fractorium/Source/Ember/Iterator.h

593 lines
21 KiB
C
Raw Permalink Normal View History

#pragma once
#include "Ember.h"
/// <summary>
/// Iterator and derived classes.
/// </summary>
#define CHOOSE_XFORM_GRAIN 16384//The size of xform random selection buffer. Multiply by the (number of non-final xforms present + 1) if xaos is used.
#define CHOOSE_XFORM_GRAIN_M1 16383//All 1s, so it's logically and-able.
namespace EmberNs
{
#define ITERATORUSINGS \
using Iterator<T>::NextXformFromIndex; \
using Iterator<T>::DoFinalXform; \
using Iterator<T>::DoBadVals;
0.4.1.5 Beta 11/28/2014 --User Changes Remove limit on the number of xforms allowable on the GPU. This was previously 21. Show actual strips count to be used in parens outside of user specified strips count on final render dialog. Allow for adjustment of iteration depth and fuse count per ember and save/read these values with the xml. Iteration optimizations on both CPU and GPU. Automatically adjust default quality spinner value when using CPU/GPU to 10/30, respectively. --Bug Fixes Fix severe randomization bug with OpenCL. Fix undo list off by one error when doing a new edit anywhere but the end of the undo list. Make integer variation parameters use 4 decimal places in the variations list like all the others. New build of the latest Qt to fix scroll bar drawing bug. Prevent grid from showing as much when pressing control to increase a spinner's increment speed. Still shows sometimes, but better than before. --Code Changes Pass count and fuse to iterator as a structure now to allow for passing more params in the future. Slightly different grid/block logic when running DE filtering on the GPU. Attempt a different way of doing DE, but #define out because it ended up not being faster. Restructure some things to allow for a variable length xforms buffer to be passed to the GPU. Add sub batch size and fuse count as ember members, and remove them from the renderer classes. Remove m_LastPass from Renderer. It should have been removed with passes. Pass seeds as a buffer to the OpenCL iteration kernel, rather than a single seed that gets modified. Slight optimization on CPU accum. Use case statement instead of if/else for xform chosing in OpenCL for a 2% speedup on params with large numbers of xforms. Add SizeOf() wrapper around sizeof(vec[0]) * vec.size(). Remove LogScaleSum() functions from the CPU and GPU because they're no longer used since passes were removed. Make some OpenCLWrapper getters const. Better ogranize RendererCL methods that return grid dimensions.
2014-11-28 04:37:51 -05:00
template <typename T, typename bucketT> class Renderer;
template <typename T>
struct IterParams
{
size_t m_Count;
size_t m_Skip;
//T m_OneColDiv2;
//T m_OneRowDiv2;
};
/// <summary>
/// Iterator base class.
/// Iterating is one loop level outside of the inner xform application loop so it's still very important
/// to take every optimization possible here.
/// The original had many temporary assignments in order to feed the output of the current iteration
/// into the input of the next iteration. All unneccessary temporary assignments are eliminated by simply using i and i + 1
/// as the input and output indices on the samples array passed to Xform.Apply().
/// Note that the samples array is assigned to while fusing. Although this technically doesn't make sense
/// since values computed during fusing get thrown out, it doesn't matter because it will get overwritten
/// in the actual loop below it since the index counter is reset to zero when fusing is complete.
/// Flam3 needlessly computed the final xform on each fuse iteration only to throw it away. It's omitted here as an optimization.
/// Rather than place many conditionals inside the iteration loop, they are broken into separate classes depending
/// on what's contained in the ember's xforms.
/// The biggest difference is whether xaos is present or not it requires extra work when picking
/// the next random xform to use. Further, each of those is broken into two loops, one for embers with a final xform
/// and one without.
/// Last, the fuse loop and real loop are separated and duplicated to omit the conditional check for fuse inside the real loop.
/// Although this makes this file about four times as verbose as it would normally be, it does lead to performance improvements.
/// Template argument expected to be float or double.
/// </summary>
template <typename T>
class EMBER_API Iterator
{
public:
/// <summary>
--User changes -Remove the option --intpalette to format the palette in the xml as ints. If they are not hex formatted, then they should always be float. This option was pointless. -Cleanup some options text for the command line programs. -Allow for dragging around flames in the library tab. This is useful for setting up the order of an animation. -Make the opening of large files in Fractorium much more efficient when not-appending. -Make the opening of large files in all EmberRender and EmberAnimate more efficient. -Better error reporting when opening files. --Bug fixes -Get rid of leftover artifacts that would appear on preview thumbnails when either switching SP/DP or re-rendering previews. -Filename extension was not being appended on Linux when saving as Xml, thus making it impossible to drag that file back in becase drop is filtered on extension. --Code changes -Move GCC compiler spec to C++14. Building with 5.3 now on linux. -Use inline member data initializers. -Make a #define for static for use in Utils.h to make things a little cleaner. -Make various functions able to take arbitrary collections as their parameters rather than just vectors. -Make library collection a list rather than vector. This alleviates the need to re-sync pointers whenever the collection changes. -Subclass QTreeWidget for the library tree. Two new files added for this. -Remove all usage of #ifdef ROW_ONLY_DE in DEOpenCLKernelCreator, it was never used. -Add move constructor and assignment operator to EmberFile. -Add the ability to use a pointer to outside memory in the renderer for the vector of Ember<T>. -Make a lot more functions const where they should be.
2016-04-03 21:55:12 -04:00
/// Empty constructor and virtual destructor so proper derived class destructors get called.
/// </summary>
--User changes -Remove the option --intpalette to format the palette in the xml as ints. If they are not hex formatted, then they should always be float. This option was pointless. -Cleanup some options text for the command line programs. -Allow for dragging around flames in the library tab. This is useful for setting up the order of an animation. -Make the opening of large files in Fractorium much more efficient when not-appending. -Make the opening of large files in all EmberRender and EmberAnimate more efficient. -Better error reporting when opening files. --Bug fixes -Get rid of leftover artifacts that would appear on preview thumbnails when either switching SP/DP or re-rendering previews. -Filename extension was not being appended on Linux when saving as Xml, thus making it impossible to drag that file back in becase drop is filtered on extension. --Code changes -Move GCC compiler spec to C++14. Building with 5.3 now on linux. -Use inline member data initializers. -Make a #define for static for use in Utils.h to make things a little cleaner. -Make various functions able to take arbitrary collections as their parameters rather than just vectors. -Make library collection a list rather than vector. This alleviates the need to re-sync pointers whenever the collection changes. -Subclass QTreeWidget for the library tree. Two new files added for this. -Remove all usage of #ifdef ROW_ONLY_DE in DEOpenCLKernelCreator, it was never used. -Add move constructor and assignment operator to EmberFile. -Add the ability to use a pointer to outside memory in the renderer for the vector of Ember<T>. -Make a lot more functions const where they should be.
2016-04-03 21:55:12 -04:00
Iterator() = default;
virtual ~Iterator() = default;
Iterator(const Iterator<T>& iter) = delete;
/// <summary>
/// Accessors.
/// </summary>
const byte* XformDistributions() const { return m_XformDistributions.empty() ? nullptr : m_XformDistributions.data(); }
size_t XformDistributionsSize() const { return m_XformDistributions.size(); }
/// <summary>
/// Virtual empty iteration function that will be overidden in derived iterator classes.
/// </summary>
/// <param name="ember">The ember whose xforms will be applied</param>
/// <param name="count">The number of iterations to do</param>
/// <param name="skip">The number of times to fuse</param>
/// <param name="samples">The buffer to store the output points</param>
/// <param name="rand">The random context to use</param>
/// <returns>The number of bad values</returns>
0.4.1.5 Beta 11/28/2014 --User Changes Remove limit on the number of xforms allowable on the GPU. This was previously 21. Show actual strips count to be used in parens outside of user specified strips count on final render dialog. Allow for adjustment of iteration depth and fuse count per ember and save/read these values with the xml. Iteration optimizations on both CPU and GPU. Automatically adjust default quality spinner value when using CPU/GPU to 10/30, respectively. --Bug Fixes Fix severe randomization bug with OpenCL. Fix undo list off by one error when doing a new edit anywhere but the end of the undo list. Make integer variation parameters use 4 decimal places in the variations list like all the others. New build of the latest Qt to fix scroll bar drawing bug. Prevent grid from showing as much when pressing control to increase a spinner's increment speed. Still shows sometimes, but better than before. --Code Changes Pass count and fuse to iterator as a structure now to allow for passing more params in the future. Slightly different grid/block logic when running DE filtering on the GPU. Attempt a different way of doing DE, but #define out because it ended up not being faster. Restructure some things to allow for a variable length xforms buffer to be passed to the GPU. Add sub batch size and fuse count as ember members, and remove them from the renderer classes. Remove m_LastPass from Renderer. It should have been removed with passes. Pass seeds as a buffer to the OpenCL iteration kernel, rather than a single seed that gets modified. Slight optimization on CPU accum. Use case statement instead of if/else for xform chosing in OpenCL for a 2% speedup on params with large numbers of xforms. Add SizeOf() wrapper around sizeof(vec[0]) * vec.size(). Remove LogScaleSum() functions from the CPU and GPU because they're no longer used since passes were removed. Make some OpenCLWrapper getters const. Better ogranize RendererCL methods that return grid dimensions.
2014-11-28 04:37:51 -05:00
virtual size_t Iterate(Ember<T>& ember, IterParams<T>& params, Point<T>* samples, QTIsaac<ISAAC_SIZE, ISAAC_INT>& rand) { return 0; }
2014-09-10 01:41:26 -04:00
/// <summary>
/// Initialize the xform selection vector by normalizing the weights of all xforms and
/// setting the corresponding percentage of elements in the vector to each xform's index in its
/// parent ember.
/// Note that this method of looking up and index in a vector is how flam3 did it and is about 10%
/// faster than using a while loop to check a random number against a normalized weight.
/// Also, the ember used to initialize this must be the same ember, unchanged, used to iterate.
/// If one is passed to this function, its parameters are changed and then it's passed to Iterate(),
/// the behavior is undefined.
/// </summary>
/// <param name="ember">The ember whose xforms will be used to populate the distribution vector</param>
/// <returns>True if success, else false.</returns>
bool InitDistributions(Ember<T>& ember)
{
0.4.1.3 Beta 10/14/2014 --User Changes Size is no longer fixed to the window size. Size scaling is done differently in the final render dialog. This fixes several bugs. Remove Xml saving size from settings and options dialog, it no longer applies. Final render can be broken into strips. Set default save path to the desktop if none is found in the settings file. Set default output size to 1920x1080 if none is found in the settings file. --Bug Fixes Better memory size reporting in final render dialog. --Code Changes Migrate to C++11, Qt 5.3.1, and Visual Studio 2013. Change most instances of unsigned int to size_t, and int to intmax_t. Add m_OrigPixPerUnit and m_ScaleType to Ember for scaling purposes. Replace some sprintf_s() calls in XmlToEmber with ostringstream. Move more non-templated members into RendererBase. Add CopyVec() overload that takes a per element function pointer. Add vector Memset(). Replace '&' with '+' instead of "&amp;" in XmlToEmber for much faster parsing. Break strips rendering out into EmberCommon and call from EmberRender and Fractorium. Make AddAndWriteBuffer() just call WriteBuffer(). Make AddAndWriteImage() delete the existing image first before replacing it. Add SetOutputTexture() to RendererCL to support making new textures in response to resize events. Remove multiple return statements in RendererCL, and replace with a bool that tracks results. Add ToDouble(), MakeEnd(), ToString() and Exists() wrappers in Fractorium. Add Size() wrapper in EmberFile. Make QString function arguments const QString&, and string with const string&. Make ShowCritical() wrapper for invoking a message box from another thread. Add combo box to TwoButtonWidget and rename.
2014-10-14 11:53:15 -04:00
size_t i;
size_t distribCount = ember.XaosPresent() ? ember.XformCount() + 1 : 1;
auto xforms = ember.Xforms();
if (m_XformDistributions.size() < CHOOSE_XFORM_GRAIN * distribCount)
m_XformDistributions.resize(CHOOSE_XFORM_GRAIN * distribCount);
if (m_XformDistributions.size() < CHOOSE_XFORM_GRAIN * distribCount)
return false;
0.4.1.3 Beta 10/14/2014 --User Changes Size is no longer fixed to the window size. Size scaling is done differently in the final render dialog. This fixes several bugs. Remove Xml saving size from settings and options dialog, it no longer applies. Final render can be broken into strips. Set default save path to the desktop if none is found in the settings file. Set default output size to 1920x1080 if none is found in the settings file. --Bug Fixes Better memory size reporting in final render dialog. --Code Changes Migrate to C++11, Qt 5.3.1, and Visual Studio 2013. Change most instances of unsigned int to size_t, and int to intmax_t. Add m_OrigPixPerUnit and m_ScaleType to Ember for scaling purposes. Replace some sprintf_s() calls in XmlToEmber with ostringstream. Move more non-templated members into RendererBase. Add CopyVec() overload that takes a per element function pointer. Add vector Memset(). Replace '&' with '+' instead of "&amp;" in XmlToEmber for much faster parsing. Break strips rendering out into EmberCommon and call from EmberRender and Fractorium. Make AddAndWriteBuffer() just call WriteBuffer(). Make AddAndWriteImage() delete the existing image first before replacing it. Add SetOutputTexture() to RendererCL to support making new textures in response to resize events. Remove multiple return statements in RendererCL, and replace with a bool that tracks results. Add ToDouble(), MakeEnd(), ToString() and Exists() wrappers in Fractorium. Add Size() wrapper in EmberFile. Make QString function arguments const QString&, and string with const string&. Make ShowCritical() wrapper for invoking a message box from another thread. Add combo box to TwoButtonWidget and rename.
2014-10-14 11:53:15 -04:00
for (size_t distrib = 0; distrib < distribCount; distrib++)
{
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
double totalDensity = 0;
//First find the total densities of all xforms.
for (i = 0; i < ember.XformCount(); i++)
{
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
double d = xforms[i].m_Weight;
2014-09-10 01:41:26 -04:00
if (distrib > 0)
d *= xforms[distrib - 1].Xaos(i);
totalDensity += d;
}
//Original returned false if all were 0, but it's allowed here
//which will just end up setting all elements to 0 which means
//only the first xform will get used.
//Calculate how much of a fraction of a the total density each element represents.
0.4.1.3 Beta 10/14/2014 --User Changes Size is no longer fixed to the window size. Size scaling is done differently in the final render dialog. This fixes several bugs. Remove Xml saving size from settings and options dialog, it no longer applies. Final render can be broken into strips. Set default save path to the desktop if none is found in the settings file. Set default output size to 1920x1080 if none is found in the settings file. --Bug Fixes Better memory size reporting in final render dialog. --Code Changes Migrate to C++11, Qt 5.3.1, and Visual Studio 2013. Change most instances of unsigned int to size_t, and int to intmax_t. Add m_OrigPixPerUnit and m_ScaleType to Ember for scaling purposes. Replace some sprintf_s() calls in XmlToEmber with ostringstream. Move more non-templated members into RendererBase. Add CopyVec() overload that takes a per element function pointer. Add vector Memset(). Replace '&' with '+' instead of "&amp;" in XmlToEmber for much faster parsing. Break strips rendering out into EmberCommon and call from EmberRender and Fractorium. Make AddAndWriteBuffer() just call WriteBuffer(). Make AddAndWriteImage() delete the existing image first before replacing it. Add SetOutputTexture() to RendererCL to support making new textures in response to resize events. Remove multiple return statements in RendererCL, and replace with a bool that tracks results. Add ToDouble(), MakeEnd(), ToString() and Exists() wrappers in Fractorium. Add Size() wrapper in EmberFile. Make QString function arguments const QString&, and string with const string&. Make ShowCritical() wrapper for invoking a message box from another thread. Add combo box to TwoButtonWidget and rename.
2014-10-14 11:53:15 -04:00
size_t j = 0;
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
//These must be double, else roundoff error will prevent the last element of m_XformDistributions from being set.
double tempDensity = 0, currentDensityLimit = 0, densityPerElement = totalDensity / CHOOSE_XFORM_GRAIN;
//Assign xform indices in order to each element of m_XformDistributions.
//The number of elements assigned a given index is proportional to that xform's
//density relative to the sum of all densities.
for (i = 0; i < ember.XformCount(); i++)
{
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
double temp = xforms[i].m_Weight;
2014-09-10 01:41:26 -04:00
if (distrib > 0)
temp *= xforms[distrib - 1].Xaos(i);
2014-09-10 01:41:26 -04:00
currentDensityLimit += temp;
2014-09-10 01:41:26 -04:00
//Populate points corresponding to this xform's weight/density.
//Also check that j is within the bounds of the distribution array just to be safe in the case of a rounding error.
while (tempDensity < currentDensityLimit && j < CHOOSE_XFORM_GRAIN)
{
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
#ifdef _DEBUG
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
//Ensure distribution contains no out of bounds indices.
if (byte(i) >= ember.XformCount())
throw "Out of bounds xform index in selection distribution.";
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
#endif
2016-12-05 22:04:33 -05:00
//cout << "offset = " << j << ", xform = " << i << ", running sum = " << tempDensity << "\n";
m_XformDistributions[(distrib * CHOOSE_XFORM_GRAIN) + j] = byte(i);
tempDensity += densityPerElement;
j++;
}
}
//If probability was zero, then nothing was filled in, so make all zero.
//If it was non zero but for some reason didn't fill all elements, then just make the remaining
//elements have the index of the last xform.
byte val = j ? byte(i - 1) : 0;
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
for (; j < CHOOSE_XFORM_GRAIN; j++)//Make absolutely sure they are set to a valid value.
m_XformDistributions[(distrib * CHOOSE_XFORM_GRAIN) + j] = val;
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
//Flam3 did this, which gives the same result.
//T t = xforms[0].m_Weight;
//
//if (distrib > 0)
// t *= xforms[distrib - 1].Xaos(0);
//
//T r = 0;
//
//for (i = 0; i < CHOOSE_XFORM_GRAIN; i++)
//{
// while (r >= t)
// {
// j++;
//
// if (distrib > 0)
// t += xforms[j].m_Weight * xforms[distrib - 1].Xaos(j);
// else
// t += xforms[j].m_Weight;
// }
//
// m_XformDistributions[(distrib * CHOOSE_XFORM_GRAIN) + i] = j;
// r += densityPerElement;
//}
}
return true;
}
2014-09-10 01:41:26 -04:00
protected:
/// <summary>
/// When iterating, if the computed location of the point is either very close to zero, or very close to infinity,
/// it's considered a bad value. In that case, a new random input point is fed into a new randomly chosen xform. This
/// process is repeated up to 5 times until a good value is computed. If after 5 tries, a good value is not found, then
/// the coordinates of the output point are just set to a random number between -1 and 1.
/// </summary>
/// <param name="xforms">The xforms array</param>
/// <param name="badVals">The counter for the total number of bad values this sub batch</param>
/// <param name="point">The point which initially had the bad values and which will store the newly computed values</param>
/// <param name="rand">The random context this iterator is using</param>
/// <returns>True if a good value was computed within 5 tries, else false</returns>
0.4.1.3 Beta 10/14/2014 --User Changes Size is no longer fixed to the window size. Size scaling is done differently in the final render dialog. This fixes several bugs. Remove Xml saving size from settings and options dialog, it no longer applies. Final render can be broken into strips. Set default save path to the desktop if none is found in the settings file. Set default output size to 1920x1080 if none is found in the settings file. --Bug Fixes Better memory size reporting in final render dialog. --Code Changes Migrate to C++11, Qt 5.3.1, and Visual Studio 2013. Change most instances of unsigned int to size_t, and int to intmax_t. Add m_OrigPixPerUnit and m_ScaleType to Ember for scaling purposes. Replace some sprintf_s() calls in XmlToEmber with ostringstream. Move more non-templated members into RendererBase. Add CopyVec() overload that takes a per element function pointer. Add vector Memset(). Replace '&' with '+' instead of "&amp;" in XmlToEmber for much faster parsing. Break strips rendering out into EmberCommon and call from EmberRender and Fractorium. Make AddAndWriteBuffer() just call WriteBuffer(). Make AddAndWriteImage() delete the existing image first before replacing it. Add SetOutputTexture() to RendererCL to support making new textures in response to resize events. Remove multiple return statements in RendererCL, and replace with a bool that tracks results. Add ToDouble(), MakeEnd(), ToString() and Exists() wrappers in Fractorium. Add Size() wrapper in EmberFile. Make QString function arguments const QString&, and string with const string&. Make ShowCritical() wrapper for invoking a message box from another thread. Add combo box to TwoButtonWidget and rename.
2014-10-14 11:53:15 -04:00
inline bool DoBadVals(Xform<T>* xforms, size_t& badVals, Point<T>* point, QTIsaac<ISAAC_SIZE, ISAAC_INT>& rand)
{
0.4.1.3 Beta 10/14/2014 --User Changes Size is no longer fixed to the window size. Size scaling is done differently in the final render dialog. This fixes several bugs. Remove Xml saving size from settings and options dialog, it no longer applies. Final render can be broken into strips. Set default save path to the desktop if none is found in the settings file. Set default output size to 1920x1080 if none is found in the settings file. --Bug Fixes Better memory size reporting in final render dialog. --Code Changes Migrate to C++11, Qt 5.3.1, and Visual Studio 2013. Change most instances of unsigned int to size_t, and int to intmax_t. Add m_OrigPixPerUnit and m_ScaleType to Ember for scaling purposes. Replace some sprintf_s() calls in XmlToEmber with ostringstream. Move more non-templated members into RendererBase. Add CopyVec() overload that takes a per element function pointer. Add vector Memset(). Replace '&' with '+' instead of "&amp;" in XmlToEmber for much faster parsing. Break strips rendering out into EmberCommon and call from EmberRender and Fractorium. Make AddAndWriteBuffer() just call WriteBuffer(). Make AddAndWriteImage() delete the existing image first before replacing it. Add SetOutputTexture() to RendererCL to support making new textures in response to resize events. Remove multiple return statements in RendererCL, and replace with a bool that tracks results. Add ToDouble(), MakeEnd(), ToString() and Exists() wrappers in Fractorium. Add Size() wrapper in EmberFile. Make QString function arguments const QString&, and string with const string&. Make ShowCritical() wrapper for invoking a message box from another thread. Add combo box to TwoButtonWidget and rename.
2014-10-14 11:53:15 -04:00
size_t xformIndex, consec = 0;
Point<T> firstBadPoint;
while (consec < 5)
{
consec++;
badVals++;
firstBadPoint.m_X = rand.Frand11<T>();//Re-randomize points, but keep the computed color and viz.
firstBadPoint.m_Y = rand.Frand11<T>();
firstBadPoint.m_Z = 0;
firstBadPoint.m_ColorX = point->m_ColorX;
firstBadPoint.m_Opacity = point->m_Opacity;
xformIndex = NextXformFromIndex(rand.Rand());
if (!xforms[xformIndex].Apply(&firstBadPoint, point, rand))
return true;
}
//After 5 tries, nothing worked, so just assign random values between -1 and 1.
if (consec == 5)
{
point->m_X = rand.Frand11<T>();
point->m_Y = rand.Frand11<T>();
point->m_Z = 0;
}
return false;
}
/// <summary>
/// Apply the final xform.
/// Note that as stated in the paper, the output of the final xform is not fed back into the next iteration.
/// Rather, only the value computed from the randomly chosen xform is. However, the output of the final xform
/// is still saved in the output samples buffer and accumulated to the histogram later.
/// </summary>
/// <param name="ember">The ember being iterated</param>
/// <param name="tempPoint">The input point</param>
/// <param name="sample">The output point</param>
/// <param name="rand">The random context to use.</param>
inline void DoFinalXform(Ember<T>& ember, Point<T>& tempPoint, Point<T>* sample, QTIsaac<ISAAC_SIZE, ISAAC_INT>& rand)
{
if (IsClose<T>(ember.FinalXform()->m_Opacity, 1) || rand.Frand01<T>() < ember.FinalXform()->m_Opacity)
{
T tempOpacity = tempPoint.m_Opacity;
ember.NonConstFinalXform()->Apply(&tempPoint, sample, rand);
sample->m_Opacity = tempOpacity;
}
else
{
*sample = tempPoint;
}
}
/// <summary>
/// Retrieve an element in the distributions vector between 0 and CHOOSE_XFORM_GRAIN which will
2014-09-10 01:41:26 -04:00
/// contain the index of the next xform to use. When xaos is prsent, the offset is the index in
/// the ember of the previous xform used when.
/// </summary>
/// <param name="index">The index to retrieve</param>
/// <param name="distribOffset">When xaos is prsent, the index of the previous xform used. Default: 0 (xaos not present).</param>
/// <returns></returns>
0.4.1.3 Beta 10/14/2014 --User Changes Size is no longer fixed to the window size. Size scaling is done differently in the final render dialog. This fixes several bugs. Remove Xml saving size from settings and options dialog, it no longer applies. Final render can be broken into strips. Set default save path to the desktop if none is found in the settings file. Set default output size to 1920x1080 if none is found in the settings file. --Bug Fixes Better memory size reporting in final render dialog. --Code Changes Migrate to C++11, Qt 5.3.1, and Visual Studio 2013. Change most instances of unsigned int to size_t, and int to intmax_t. Add m_OrigPixPerUnit and m_ScaleType to Ember for scaling purposes. Replace some sprintf_s() calls in XmlToEmber with ostringstream. Move more non-templated members into RendererBase. Add CopyVec() overload that takes a per element function pointer. Add vector Memset(). Replace '&' with '+' instead of "&amp;" in XmlToEmber for much faster parsing. Break strips rendering out into EmberCommon and call from EmberRender and Fractorium. Make AddAndWriteBuffer() just call WriteBuffer(). Make AddAndWriteImage() delete the existing image first before replacing it. Add SetOutputTexture() to RendererCL to support making new textures in response to resize events. Remove multiple return statements in RendererCL, and replace with a bool that tracks results. Add ToDouble(), MakeEnd(), ToString() and Exists() wrappers in Fractorium. Add Size() wrapper in EmberFile. Make QString function arguments const QString&, and string with const string&. Make ShowCritical() wrapper for invoking a message box from another thread. Add combo box to TwoButtonWidget and rename.
2014-10-14 11:53:15 -04:00
size_t NextXformFromIndex(size_t index, size_t distribOffset = 0)
{
return size_t(m_XformDistributions[(index & CHOOSE_XFORM_GRAIN_M1) + (CHOOSE_XFORM_GRAIN * distribOffset)]);
}
vector<byte> m_XformDistributions;
};
/// <summary>
/// Derived iterator class for embers whose xforms do not use xaos.
/// </summary>
template <typename T>
class EMBER_API StandardIterator : public Iterator<T>
{
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
ITERATORUSINGS
public:
/// <summary>
/// Empty constructor.
/// </summary>
StandardIterator()
{
}
/// <summary>
/// Overridden virtual function which iterates an ember a given number of times and does not use xaos.
/// </summary>
/// <param name="ember">The ember whose xforms will be applied</param>
/// <param name="count">The number of iterations to do</param>
/// <param name="skip">The number of times to fuse</param>
/// <param name="samples">The buffer to store the output points</param>
/// <param name="rand">The random context to use</param>
/// <returns>The number of bad values</returns>
0.4.1.5 Beta 11/28/2014 --User Changes Remove limit on the number of xforms allowable on the GPU. This was previously 21. Show actual strips count to be used in parens outside of user specified strips count on final render dialog. Allow for adjustment of iteration depth and fuse count per ember and save/read these values with the xml. Iteration optimizations on both CPU and GPU. Automatically adjust default quality spinner value when using CPU/GPU to 10/30, respectively. --Bug Fixes Fix severe randomization bug with OpenCL. Fix undo list off by one error when doing a new edit anywhere but the end of the undo list. Make integer variation parameters use 4 decimal places in the variations list like all the others. New build of the latest Qt to fix scroll bar drawing bug. Prevent grid from showing as much when pressing control to increase a spinner's increment speed. Still shows sometimes, but better than before. --Code Changes Pass count and fuse to iterator as a structure now to allow for passing more params in the future. Slightly different grid/block logic when running DE filtering on the GPU. Attempt a different way of doing DE, but #define out because it ended up not being faster. Restructure some things to allow for a variable length xforms buffer to be passed to the GPU. Add sub batch size and fuse count as ember members, and remove them from the renderer classes. Remove m_LastPass from Renderer. It should have been removed with passes. Pass seeds as a buffer to the OpenCL iteration kernel, rather than a single seed that gets modified. Slight optimization on CPU accum. Use case statement instead of if/else for xform chosing in OpenCL for a 2% speedup on params with large numbers of xforms. Add SizeOf() wrapper around sizeof(vec[0]) * vec.size(). Remove LogScaleSum() functions from the CPU and GPU because they're no longer used since passes were removed. Make some OpenCLWrapper getters const. Better ogranize RendererCL methods that return grid dimensions.
2014-11-28 04:37:51 -05:00
virtual size_t Iterate(Ember<T>& ember, IterParams<T>& params, Point<T>* samples, QTIsaac<ISAAC_SIZE, ISAAC_INT>& rand) override
{
0.4.1.3 Beta 10/14/2014 --User Changes Size is no longer fixed to the window size. Size scaling is done differently in the final render dialog. This fixes several bugs. Remove Xml saving size from settings and options dialog, it no longer applies. Final render can be broken into strips. Set default save path to the desktop if none is found in the settings file. Set default output size to 1920x1080 if none is found in the settings file. --Bug Fixes Better memory size reporting in final render dialog. --Code Changes Migrate to C++11, Qt 5.3.1, and Visual Studio 2013. Change most instances of unsigned int to size_t, and int to intmax_t. Add m_OrigPixPerUnit and m_ScaleType to Ember for scaling purposes. Replace some sprintf_s() calls in XmlToEmber with ostringstream. Move more non-templated members into RendererBase. Add CopyVec() overload that takes a per element function pointer. Add vector Memset(). Replace '&' with '+' instead of "&amp;" in XmlToEmber for much faster parsing. Break strips rendering out into EmberCommon and call from EmberRender and Fractorium. Make AddAndWriteBuffer() just call WriteBuffer(). Make AddAndWriteImage() delete the existing image first before replacing it. Add SetOutputTexture() to RendererCL to support making new textures in response to resize events. Remove multiple return statements in RendererCL, and replace with a bool that tracks results. Add ToDouble(), MakeEnd(), ToString() and Exists() wrappers in Fractorium. Add Size() wrapper in EmberFile. Make QString function arguments const QString&, and string with const string&. Make ShowCritical() wrapper for invoking a message box from another thread. Add combo box to TwoButtonWidget and rename.
2014-10-14 11:53:15 -04:00
size_t i, badVals = 0;
Point<T> tempPoint, p1;
auto xforms = ember.NonConstXforms();
if (ember.ProjBits())//No xaos, 3D.
{
if (ember.UseFinalXform())//No xaos, 3D, final.
{
p1 = samples[0];
0.4.1.5 Beta 11/28/2014 --User Changes Remove limit on the number of xforms allowable on the GPU. This was previously 21. Show actual strips count to be used in parens outside of user specified strips count on final render dialog. Allow for adjustment of iteration depth and fuse count per ember and save/read these values with the xml. Iteration optimizations on both CPU and GPU. Automatically adjust default quality spinner value when using CPU/GPU to 10/30, respectively. --Bug Fixes Fix severe randomization bug with OpenCL. Fix undo list off by one error when doing a new edit anywhere but the end of the undo list. Make integer variation parameters use 4 decimal places in the variations list like all the others. New build of the latest Qt to fix scroll bar drawing bug. Prevent grid from showing as much when pressing control to increase a spinner's increment speed. Still shows sometimes, but better than before. --Code Changes Pass count and fuse to iterator as a structure now to allow for passing more params in the future. Slightly different grid/block logic when running DE filtering on the GPU. Attempt a different way of doing DE, but #define out because it ended up not being faster. Restructure some things to allow for a variable length xforms buffer to be passed to the GPU. Add sub batch size and fuse count as ember members, and remove them from the renderer classes. Remove m_LastPass from Renderer. It should have been removed with passes. Pass seeds as a buffer to the OpenCL iteration kernel, rather than a single seed that gets modified. Slight optimization on CPU accum. Use case statement instead of if/else for xform chosing in OpenCL for a 2% speedup on params with large numbers of xforms. Add SizeOf() wrapper around sizeof(vec[0]) * vec.size(). Remove LogScaleSum() functions from the CPU and GPU because they're no longer used since passes were removed. Make some OpenCLWrapper getters const. Better ogranize RendererCL methods that return grid dimensions.
2014-11-28 04:37:51 -05:00
for (i = 0; i < params.m_Skip; i++)//Fuse.
{
if (xforms[NextXformFromIndex(rand.Rand())].Apply(&p1, &p1, rand))
DoBadVals(xforms, badVals, &p1, rand);
}
DoFinalXform(ember, p1, samples, rand);//Apply to last fuse point and store as the first element in samples.
ember.Proj(samples[0], rand);
0.4.1.5 Beta 11/28/2014 --User Changes Remove limit on the number of xforms allowable on the GPU. This was previously 21. Show actual strips count to be used in parens outside of user specified strips count on final render dialog. Allow for adjustment of iteration depth and fuse count per ember and save/read these values with the xml. Iteration optimizations on both CPU and GPU. Automatically adjust default quality spinner value when using CPU/GPU to 10/30, respectively. --Bug Fixes Fix severe randomization bug with OpenCL. Fix undo list off by one error when doing a new edit anywhere but the end of the undo list. Make integer variation parameters use 4 decimal places in the variations list like all the others. New build of the latest Qt to fix scroll bar drawing bug. Prevent grid from showing as much when pressing control to increase a spinner's increment speed. Still shows sometimes, but better than before. --Code Changes Pass count and fuse to iterator as a structure now to allow for passing more params in the future. Slightly different grid/block logic when running DE filtering on the GPU. Attempt a different way of doing DE, but #define out because it ended up not being faster. Restructure some things to allow for a variable length xforms buffer to be passed to the GPU. Add sub batch size and fuse count as ember members, and remove them from the renderer classes. Remove m_LastPass from Renderer. It should have been removed with passes. Pass seeds as a buffer to the OpenCL iteration kernel, rather than a single seed that gets modified. Slight optimization on CPU accum. Use case statement instead of if/else for xform chosing in OpenCL for a 2% speedup on params with large numbers of xforms. Add SizeOf() wrapper around sizeof(vec[0]) * vec.size(). Remove LogScaleSum() functions from the CPU and GPU because they're no longer used since passes were removed. Make some OpenCLWrapper getters const. Better ogranize RendererCL methods that return grid dimensions.
2014-11-28 04:37:51 -05:00
for (i = 1; i < params.m_Count; i++)//Real loop.
{
if (xforms[NextXformFromIndex(rand.Rand())].Apply(&p1, &p1, rand))
DoBadVals(xforms, badVals, &p1, rand);
DoFinalXform(ember, p1, samples + i, rand);
ember.Proj(samples[i], rand);
}
}
else//No xaos, 3D, no final.
{
p1 = samples[0];
0.4.1.5 Beta 11/28/2014 --User Changes Remove limit on the number of xforms allowable on the GPU. This was previously 21. Show actual strips count to be used in parens outside of user specified strips count on final render dialog. Allow for adjustment of iteration depth and fuse count per ember and save/read these values with the xml. Iteration optimizations on both CPU and GPU. Automatically adjust default quality spinner value when using CPU/GPU to 10/30, respectively. --Bug Fixes Fix severe randomization bug with OpenCL. Fix undo list off by one error when doing a new edit anywhere but the end of the undo list. Make integer variation parameters use 4 decimal places in the variations list like all the others. New build of the latest Qt to fix scroll bar drawing bug. Prevent grid from showing as much when pressing control to increase a spinner's increment speed. Still shows sometimes, but better than before. --Code Changes Pass count and fuse to iterator as a structure now to allow for passing more params in the future. Slightly different grid/block logic when running DE filtering on the GPU. Attempt a different way of doing DE, but #define out because it ended up not being faster. Restructure some things to allow for a variable length xforms buffer to be passed to the GPU. Add sub batch size and fuse count as ember members, and remove them from the renderer classes. Remove m_LastPass from Renderer. It should have been removed with passes. Pass seeds as a buffer to the OpenCL iteration kernel, rather than a single seed that gets modified. Slight optimization on CPU accum. Use case statement instead of if/else for xform chosing in OpenCL for a 2% speedup on params with large numbers of xforms. Add SizeOf() wrapper around sizeof(vec[0]) * vec.size(). Remove LogScaleSum() functions from the CPU and GPU because they're no longer used since passes were removed. Make some OpenCLWrapper getters const. Better ogranize RendererCL methods that return grid dimensions.
2014-11-28 04:37:51 -05:00
for (i = 0; i < params.m_Skip; i++)//Fuse.
{
if (xforms[NextXformFromIndex(rand.Rand())].Apply(&p1, &p1, rand))
DoBadVals(xforms, badVals, &p1, rand);
}
samples[0] = p1;
ember.Proj(samples[0], rand);
0.4.1.5 Beta 11/28/2014 --User Changes Remove limit on the number of xforms allowable on the GPU. This was previously 21. Show actual strips count to be used in parens outside of user specified strips count on final render dialog. Allow for adjustment of iteration depth and fuse count per ember and save/read these values with the xml. Iteration optimizations on both CPU and GPU. Automatically adjust default quality spinner value when using CPU/GPU to 10/30, respectively. --Bug Fixes Fix severe randomization bug with OpenCL. Fix undo list off by one error when doing a new edit anywhere but the end of the undo list. Make integer variation parameters use 4 decimal places in the variations list like all the others. New build of the latest Qt to fix scroll bar drawing bug. Prevent grid from showing as much when pressing control to increase a spinner's increment speed. Still shows sometimes, but better than before. --Code Changes Pass count and fuse to iterator as a structure now to allow for passing more params in the future. Slightly different grid/block logic when running DE filtering on the GPU. Attempt a different way of doing DE, but #define out because it ended up not being faster. Restructure some things to allow for a variable length xforms buffer to be passed to the GPU. Add sub batch size and fuse count as ember members, and remove them from the renderer classes. Remove m_LastPass from Renderer. It should have been removed with passes. Pass seeds as a buffer to the OpenCL iteration kernel, rather than a single seed that gets modified. Slight optimization on CPU accum. Use case statement instead of if/else for xform chosing in OpenCL for a 2% speedup on params with large numbers of xforms. Add SizeOf() wrapper around sizeof(vec[0]) * vec.size(). Remove LogScaleSum() functions from the CPU and GPU because they're no longer used since passes were removed. Make some OpenCLWrapper getters const. Better ogranize RendererCL methods that return grid dimensions.
2014-11-28 04:37:51 -05:00
for (i = 1; i < params.m_Count; i++)//Real loop.
{
if (xforms[NextXformFromIndex(rand.Rand())].Apply(&p1, &samples[i], rand))
DoBadVals(xforms, badVals, samples + i, rand);
p1 = samples[i];
ember.Proj(samples[i], rand);
}
}
}
else//No xaos, no 3D.
{
if (ember.UseFinalXform())//No xaos, no 3D, final.
{
p1 = samples[0];
0.4.1.5 Beta 11/28/2014 --User Changes Remove limit on the number of xforms allowable on the GPU. This was previously 21. Show actual strips count to be used in parens outside of user specified strips count on final render dialog. Allow for adjustment of iteration depth and fuse count per ember and save/read these values with the xml. Iteration optimizations on both CPU and GPU. Automatically adjust default quality spinner value when using CPU/GPU to 10/30, respectively. --Bug Fixes Fix severe randomization bug with OpenCL. Fix undo list off by one error when doing a new edit anywhere but the end of the undo list. Make integer variation parameters use 4 decimal places in the variations list like all the others. New build of the latest Qt to fix scroll bar drawing bug. Prevent grid from showing as much when pressing control to increase a spinner's increment speed. Still shows sometimes, but better than before. --Code Changes Pass count and fuse to iterator as a structure now to allow for passing more params in the future. Slightly different grid/block logic when running DE filtering on the GPU. Attempt a different way of doing DE, but #define out because it ended up not being faster. Restructure some things to allow for a variable length xforms buffer to be passed to the GPU. Add sub batch size and fuse count as ember members, and remove them from the renderer classes. Remove m_LastPass from Renderer. It should have been removed with passes. Pass seeds as a buffer to the OpenCL iteration kernel, rather than a single seed that gets modified. Slight optimization on CPU accum. Use case statement instead of if/else for xform chosing in OpenCL for a 2% speedup on params with large numbers of xforms. Add SizeOf() wrapper around sizeof(vec[0]) * vec.size(). Remove LogScaleSum() functions from the CPU and GPU because they're no longer used since passes were removed. Make some OpenCLWrapper getters const. Better ogranize RendererCL methods that return grid dimensions.
2014-11-28 04:37:51 -05:00
for (i = 0; i < params.m_Skip; i++)//Fuse.
{
if (xforms[NextXformFromIndex(rand.Rand())].Apply(&p1, &p1, rand))
DoBadVals(xforms, badVals, &p1, rand);
}
DoFinalXform(ember, p1, samples, rand);//Apply to last fuse point and store as the first element in samples.
2014-09-10 01:41:26 -04:00
0.4.1.5 Beta 11/28/2014 --User Changes Remove limit on the number of xforms allowable on the GPU. This was previously 21. Show actual strips count to be used in parens outside of user specified strips count on final render dialog. Allow for adjustment of iteration depth and fuse count per ember and save/read these values with the xml. Iteration optimizations on both CPU and GPU. Automatically adjust default quality spinner value when using CPU/GPU to 10/30, respectively. --Bug Fixes Fix severe randomization bug with OpenCL. Fix undo list off by one error when doing a new edit anywhere but the end of the undo list. Make integer variation parameters use 4 decimal places in the variations list like all the others. New build of the latest Qt to fix scroll bar drawing bug. Prevent grid from showing as much when pressing control to increase a spinner's increment speed. Still shows sometimes, but better than before. --Code Changes Pass count and fuse to iterator as a structure now to allow for passing more params in the future. Slightly different grid/block logic when running DE filtering on the GPU. Attempt a different way of doing DE, but #define out because it ended up not being faster. Restructure some things to allow for a variable length xforms buffer to be passed to the GPU. Add sub batch size and fuse count as ember members, and remove them from the renderer classes. Remove m_LastPass from Renderer. It should have been removed with passes. Pass seeds as a buffer to the OpenCL iteration kernel, rather than a single seed that gets modified. Slight optimization on CPU accum. Use case statement instead of if/else for xform chosing in OpenCL for a 2% speedup on params with large numbers of xforms. Add SizeOf() wrapper around sizeof(vec[0]) * vec.size(). Remove LogScaleSum() functions from the CPU and GPU because they're no longer used since passes were removed. Make some OpenCLWrapper getters const. Better ogranize RendererCL methods that return grid dimensions.
2014-11-28 04:37:51 -05:00
for (i = 1; i < params.m_Count; i++)//Real loop.
{
if (xforms[NextXformFromIndex(rand.Rand())].Apply(&p1, &p1, rand))//Feed the resulting value of applying the randomly selected xform back into the next iter, and not the result of applying the final xform.
DoBadVals(xforms, badVals, &p1, rand);
DoFinalXform(ember, p1, samples + i, rand);
}
}
else//No xaos, no 3D, no final.
{
p1 = samples[0];
0.4.1.5 Beta 11/28/2014 --User Changes Remove limit on the number of xforms allowable on the GPU. This was previously 21. Show actual strips count to be used in parens outside of user specified strips count on final render dialog. Allow for adjustment of iteration depth and fuse count per ember and save/read these values with the xml. Iteration optimizations on both CPU and GPU. Automatically adjust default quality spinner value when using CPU/GPU to 10/30, respectively. --Bug Fixes Fix severe randomization bug with OpenCL. Fix undo list off by one error when doing a new edit anywhere but the end of the undo list. Make integer variation parameters use 4 decimal places in the variations list like all the others. New build of the latest Qt to fix scroll bar drawing bug. Prevent grid from showing as much when pressing control to increase a spinner's increment speed. Still shows sometimes, but better than before. --Code Changes Pass count and fuse to iterator as a structure now to allow for passing more params in the future. Slightly different grid/block logic when running DE filtering on the GPU. Attempt a different way of doing DE, but #define out because it ended up not being faster. Restructure some things to allow for a variable length xforms buffer to be passed to the GPU. Add sub batch size and fuse count as ember members, and remove them from the renderer classes. Remove m_LastPass from Renderer. It should have been removed with passes. Pass seeds as a buffer to the OpenCL iteration kernel, rather than a single seed that gets modified. Slight optimization on CPU accum. Use case statement instead of if/else for xform chosing in OpenCL for a 2% speedup on params with large numbers of xforms. Add SizeOf() wrapper around sizeof(vec[0]) * vec.size(). Remove LogScaleSum() functions from the CPU and GPU because they're no longer used since passes were removed. Make some OpenCLWrapper getters const. Better ogranize RendererCL methods that return grid dimensions.
2014-11-28 04:37:51 -05:00
for (i = 0; i < params.m_Skip; i++)//Fuse.
{
if (xforms[NextXformFromIndex(rand.Rand())].Apply(&p1, &p1, rand))
DoBadVals(xforms, badVals, &p1, rand);
}
samples[0] = p1;
0.4.1.5 Beta 11/28/2014 --User Changes Remove limit on the number of xforms allowable on the GPU. This was previously 21. Show actual strips count to be used in parens outside of user specified strips count on final render dialog. Allow for adjustment of iteration depth and fuse count per ember and save/read these values with the xml. Iteration optimizations on both CPU and GPU. Automatically adjust default quality spinner value when using CPU/GPU to 10/30, respectively. --Bug Fixes Fix severe randomization bug with OpenCL. Fix undo list off by one error when doing a new edit anywhere but the end of the undo list. Make integer variation parameters use 4 decimal places in the variations list like all the others. New build of the latest Qt to fix scroll bar drawing bug. Prevent grid from showing as much when pressing control to increase a spinner's increment speed. Still shows sometimes, but better than before. --Code Changes Pass count and fuse to iterator as a structure now to allow for passing more params in the future. Slightly different grid/block logic when running DE filtering on the GPU. Attempt a different way of doing DE, but #define out because it ended up not being faster. Restructure some things to allow for a variable length xforms buffer to be passed to the GPU. Add sub batch size and fuse count as ember members, and remove them from the renderer classes. Remove m_LastPass from Renderer. It should have been removed with passes. Pass seeds as a buffer to the OpenCL iteration kernel, rather than a single seed that gets modified. Slight optimization on CPU accum. Use case statement instead of if/else for xform chosing in OpenCL for a 2% speedup on params with large numbers of xforms. Add SizeOf() wrapper around sizeof(vec[0]) * vec.size(). Remove LogScaleSum() functions from the CPU and GPU because they're no longer used since passes were removed. Make some OpenCLWrapper getters const. Better ogranize RendererCL methods that return grid dimensions.
2014-11-28 04:37:51 -05:00
for (i = 0; i < params.m_Count - 1; i++)//Real loop.
{
if (xforms[NextXformFromIndex(rand.Rand())].Apply(samples + i, samples + i + 1, rand))
DoBadVals(xforms, badVals, samples + i + 1, rand);
0.4.1.5 Beta 11/28/2014 --User Changes Remove limit on the number of xforms allowable on the GPU. This was previously 21. Show actual strips count to be used in parens outside of user specified strips count on final render dialog. Allow for adjustment of iteration depth and fuse count per ember and save/read these values with the xml. Iteration optimizations on both CPU and GPU. Automatically adjust default quality spinner value when using CPU/GPU to 10/30, respectively. --Bug Fixes Fix severe randomization bug with OpenCL. Fix undo list off by one error when doing a new edit anywhere but the end of the undo list. Make integer variation parameters use 4 decimal places in the variations list like all the others. New build of the latest Qt to fix scroll bar drawing bug. Prevent grid from showing as much when pressing control to increase a spinner's increment speed. Still shows sometimes, but better than before. --Code Changes Pass count and fuse to iterator as a structure now to allow for passing more params in the future. Slightly different grid/block logic when running DE filtering on the GPU. Attempt a different way of doing DE, but #define out because it ended up not being faster. Restructure some things to allow for a variable length xforms buffer to be passed to the GPU. Add sub batch size and fuse count as ember members, and remove them from the renderer classes. Remove m_LastPass from Renderer. It should have been removed with passes. Pass seeds as a buffer to the OpenCL iteration kernel, rather than a single seed that gets modified. Slight optimization on CPU accum. Use case statement instead of if/else for xform chosing in OpenCL for a 2% speedup on params with large numbers of xforms. Add SizeOf() wrapper around sizeof(vec[0]) * vec.size(). Remove LogScaleSum() functions from the CPU and GPU because they're no longer used since passes were removed. Make some OpenCLWrapper getters const. Better ogranize RendererCL methods that return grid dimensions.
2014-11-28 04:37:51 -05:00
}
}
}
return badVals;
}
};
/// <summary>
/// Derived iterator class for embers whose xforms use xaos.
/// </summary>
template <typename T>
class EMBER_API XaosIterator : public Iterator<T>
{
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
ITERATORUSINGS
public:
/// <summary>
/// Empty constructor.
/// </summary>
XaosIterator()
{
}
/// <summary>
/// Handler for bad values similar to the one in the base class, except it takes the last xform used
/// as a parameter and saves the xform used back out because this iterator is meant to be used with xaos.
/// </summary>
/// <param name="xforms">The xforms array</param>
/// <param name="xformIndex">Index of the last used xform before calling this function</param>
/// <param name="lastXformUsed">The saved index of the last xform used within this function</param>
/// <param name="badVals">The counter for the total number of bad values this sub batch</param>
/// <param name="point">The point which initially had the bad values and which will store the newly computed values</param>
/// <param name="rand">The random context this iterator is using</param>
/// <returns>True if a good value was computed within 5 tries, else false</returns>
0.4.1.3 Beta 10/14/2014 --User Changes Size is no longer fixed to the window size. Size scaling is done differently in the final render dialog. This fixes several bugs. Remove Xml saving size from settings and options dialog, it no longer applies. Final render can be broken into strips. Set default save path to the desktop if none is found in the settings file. Set default output size to 1920x1080 if none is found in the settings file. --Bug Fixes Better memory size reporting in final render dialog. --Code Changes Migrate to C++11, Qt 5.3.1, and Visual Studio 2013. Change most instances of unsigned int to size_t, and int to intmax_t. Add m_OrigPixPerUnit and m_ScaleType to Ember for scaling purposes. Replace some sprintf_s() calls in XmlToEmber with ostringstream. Move more non-templated members into RendererBase. Add CopyVec() overload that takes a per element function pointer. Add vector Memset(). Replace '&' with '+' instead of "&amp;" in XmlToEmber for much faster parsing. Break strips rendering out into EmberCommon and call from EmberRender and Fractorium. Make AddAndWriteBuffer() just call WriteBuffer(). Make AddAndWriteImage() delete the existing image first before replacing it. Add SetOutputTexture() to RendererCL to support making new textures in response to resize events. Remove multiple return statements in RendererCL, and replace with a bool that tracks results. Add ToDouble(), MakeEnd(), ToString() and Exists() wrappers in Fractorium. Add Size() wrapper in EmberFile. Make QString function arguments const QString&, and string with const string&. Make ShowCritical() wrapper for invoking a message box from another thread. Add combo box to TwoButtonWidget and rename.
2014-10-14 11:53:15 -04:00
inline bool DoBadVals(Xform<T>* xforms, size_t& xformIndex, size_t lastXformUsed, size_t& badVals, Point<T>* point, QTIsaac<ISAAC_SIZE, ISAAC_INT>& rand)
{
0.4.1.3 Beta 10/14/2014 --User Changes Size is no longer fixed to the window size. Size scaling is done differently in the final render dialog. This fixes several bugs. Remove Xml saving size from settings and options dialog, it no longer applies. Final render can be broken into strips. Set default save path to the desktop if none is found in the settings file. Set default output size to 1920x1080 if none is found in the settings file. --Bug Fixes Better memory size reporting in final render dialog. --Code Changes Migrate to C++11, Qt 5.3.1, and Visual Studio 2013. Change most instances of unsigned int to size_t, and int to intmax_t. Add m_OrigPixPerUnit and m_ScaleType to Ember for scaling purposes. Replace some sprintf_s() calls in XmlToEmber with ostringstream. Move more non-templated members into RendererBase. Add CopyVec() overload that takes a per element function pointer. Add vector Memset(). Replace '&' with '+' instead of "&amp;" in XmlToEmber for much faster parsing. Break strips rendering out into EmberCommon and call from EmberRender and Fractorium. Make AddAndWriteBuffer() just call WriteBuffer(). Make AddAndWriteImage() delete the existing image first before replacing it. Add SetOutputTexture() to RendererCL to support making new textures in response to resize events. Remove multiple return statements in RendererCL, and replace with a bool that tracks results. Add ToDouble(), MakeEnd(), ToString() and Exists() wrappers in Fractorium. Add Size() wrapper in EmberFile. Make QString function arguments const QString&, and string with const string&. Make ShowCritical() wrapper for invoking a message box from another thread. Add combo box to TwoButtonWidget and rename.
2014-10-14 11:53:15 -04:00
size_t consec = 0;
Point<T> firstBadPoint;
2014-09-10 01:41:26 -04:00
while (consec < 5)
{
consec++;
badVals++;
firstBadPoint.m_X = rand.Frand11<T>();//Re-randomize points, but keep the computed color and viz.
firstBadPoint.m_Y = rand.Frand11<T>();
firstBadPoint.m_Z = 0;
firstBadPoint.m_ColorX = point->m_ColorX;
firstBadPoint.m_Opacity = point->m_Opacity;
xformIndex = NextXformFromIndex(rand.Rand(), lastXformUsed);
if (!xforms[xformIndex].Apply(&firstBadPoint, point, rand))
return true;
}
//After 5 tries, nothing worked, so just assign random.
if (consec == 5)
{
point->m_X = rand.Frand11<T>();
point->m_Y = rand.Frand11<T>();
point->m_Z = 0;
}
return false;
}
/// <summary>
/// Overridden virtual function which iterates an ember a given number of times and uses xaos.
/// </summary>
/// <param name="ember">The ember whose xforms will be applied</param>
/// <param name="count">The number of iterations to do</param>
/// <param name="skip">The number of times to fuse</param>
/// <param name="samples">The buffer to store the output points</param>
/// <param name="rand">The random context to use</param>
/// <returns>The number of bad values</returns>
0.4.1.5 Beta 11/28/2014 --User Changes Remove limit on the number of xforms allowable on the GPU. This was previously 21. Show actual strips count to be used in parens outside of user specified strips count on final render dialog. Allow for adjustment of iteration depth and fuse count per ember and save/read these values with the xml. Iteration optimizations on both CPU and GPU. Automatically adjust default quality spinner value when using CPU/GPU to 10/30, respectively. --Bug Fixes Fix severe randomization bug with OpenCL. Fix undo list off by one error when doing a new edit anywhere but the end of the undo list. Make integer variation parameters use 4 decimal places in the variations list like all the others. New build of the latest Qt to fix scroll bar drawing bug. Prevent grid from showing as much when pressing control to increase a spinner's increment speed. Still shows sometimes, but better than before. --Code Changes Pass count and fuse to iterator as a structure now to allow for passing more params in the future. Slightly different grid/block logic when running DE filtering on the GPU. Attempt a different way of doing DE, but #define out because it ended up not being faster. Restructure some things to allow for a variable length xforms buffer to be passed to the GPU. Add sub batch size and fuse count as ember members, and remove them from the renderer classes. Remove m_LastPass from Renderer. It should have been removed with passes. Pass seeds as a buffer to the OpenCL iteration kernel, rather than a single seed that gets modified. Slight optimization on CPU accum. Use case statement instead of if/else for xform chosing in OpenCL for a 2% speedup on params with large numbers of xforms. Add SizeOf() wrapper around sizeof(vec[0]) * vec.size(). Remove LogScaleSum() functions from the CPU and GPU because they're no longer used since passes were removed. Make some OpenCLWrapper getters const. Better ogranize RendererCL methods that return grid dimensions.
2014-11-28 04:37:51 -05:00
virtual size_t Iterate(Ember<T>& ember, IterParams<T>& params, Point<T>* samples, QTIsaac<ISAAC_SIZE, ISAAC_INT>& rand) override
{
0.4.1.3 Beta 10/14/2014 --User Changes Size is no longer fixed to the window size. Size scaling is done differently in the final render dialog. This fixes several bugs. Remove Xml saving size from settings and options dialog, it no longer applies. Final render can be broken into strips. Set default save path to the desktop if none is found in the settings file. Set default output size to 1920x1080 if none is found in the settings file. --Bug Fixes Better memory size reporting in final render dialog. --Code Changes Migrate to C++11, Qt 5.3.1, and Visual Studio 2013. Change most instances of unsigned int to size_t, and int to intmax_t. Add m_OrigPixPerUnit and m_ScaleType to Ember for scaling purposes. Replace some sprintf_s() calls in XmlToEmber with ostringstream. Move more non-templated members into RendererBase. Add CopyVec() overload that takes a per element function pointer. Add vector Memset(). Replace '&' with '+' instead of "&amp;" in XmlToEmber for much faster parsing. Break strips rendering out into EmberCommon and call from EmberRender and Fractorium. Make AddAndWriteBuffer() just call WriteBuffer(). Make AddAndWriteImage() delete the existing image first before replacing it. Add SetOutputTexture() to RendererCL to support making new textures in response to resize events. Remove multiple return statements in RendererCL, and replace with a bool that tracks results. Add ToDouble(), MakeEnd(), ToString() and Exists() wrappers in Fractorium. Add Size() wrapper in EmberFile. Make QString function arguments const QString&, and string with const string&. Make ShowCritical() wrapper for invoking a message box from another thread. Add combo box to TwoButtonWidget and rename.
2014-10-14 11:53:15 -04:00
size_t i, xformIndex;
size_t lastXformUsed = 0;
size_t badVals = 0;
Point<T> tempPoint, p1;
auto xforms = ember.NonConstXforms();
if (ember.ProjBits())//Xaos, 3D.
{
if (ember.UseFinalXform())//Xaos, 3D, final.
{
p1 = samples[0];
0.4.1.5 Beta 11/28/2014 --User Changes Remove limit on the number of xforms allowable on the GPU. This was previously 21. Show actual strips count to be used in parens outside of user specified strips count on final render dialog. Allow for adjustment of iteration depth and fuse count per ember and save/read these values with the xml. Iteration optimizations on both CPU and GPU. Automatically adjust default quality spinner value when using CPU/GPU to 10/30, respectively. --Bug Fixes Fix severe randomization bug with OpenCL. Fix undo list off by one error when doing a new edit anywhere but the end of the undo list. Make integer variation parameters use 4 decimal places in the variations list like all the others. New build of the latest Qt to fix scroll bar drawing bug. Prevent grid from showing as much when pressing control to increase a spinner's increment speed. Still shows sometimes, but better than before. --Code Changes Pass count and fuse to iterator as a structure now to allow for passing more params in the future. Slightly different grid/block logic when running DE filtering on the GPU. Attempt a different way of doing DE, but #define out because it ended up not being faster. Restructure some things to allow for a variable length xforms buffer to be passed to the GPU. Add sub batch size and fuse count as ember members, and remove them from the renderer classes. Remove m_LastPass from Renderer. It should have been removed with passes. Pass seeds as a buffer to the OpenCL iteration kernel, rather than a single seed that gets modified. Slight optimization on CPU accum. Use case statement instead of if/else for xform chosing in OpenCL for a 2% speedup on params with large numbers of xforms. Add SizeOf() wrapper around sizeof(vec[0]) * vec.size(). Remove LogScaleSum() functions from the CPU and GPU because they're no longer used since passes were removed. Make some OpenCLWrapper getters const. Better ogranize RendererCL methods that return grid dimensions.
2014-11-28 04:37:51 -05:00
for (i = 0; i < params.m_Skip; i++)//Fuse.
{
xformIndex = NextXformFromIndex(rand.Rand(), lastXformUsed);
if (xforms[xformIndex].Apply(&p1, &p1, rand))
DoBadVals(xforms, xformIndex, lastXformUsed, badVals, &p1, rand);
lastXformUsed = xformIndex + 1;//Store the last used transform.
}
DoFinalXform(ember, p1, samples, rand);//Apply to last fuse point and store as the first element in samples.
ember.Proj(samples[0], rand);
2014-09-10 01:41:26 -04:00
0.4.1.5 Beta 11/28/2014 --User Changes Remove limit on the number of xforms allowable on the GPU. This was previously 21. Show actual strips count to be used in parens outside of user specified strips count on final render dialog. Allow for adjustment of iteration depth and fuse count per ember and save/read these values with the xml. Iteration optimizations on both CPU and GPU. Automatically adjust default quality spinner value when using CPU/GPU to 10/30, respectively. --Bug Fixes Fix severe randomization bug with OpenCL. Fix undo list off by one error when doing a new edit anywhere but the end of the undo list. Make integer variation parameters use 4 decimal places in the variations list like all the others. New build of the latest Qt to fix scroll bar drawing bug. Prevent grid from showing as much when pressing control to increase a spinner's increment speed. Still shows sometimes, but better than before. --Code Changes Pass count and fuse to iterator as a structure now to allow for passing more params in the future. Slightly different grid/block logic when running DE filtering on the GPU. Attempt a different way of doing DE, but #define out because it ended up not being faster. Restructure some things to allow for a variable length xforms buffer to be passed to the GPU. Add sub batch size and fuse count as ember members, and remove them from the renderer classes. Remove m_LastPass from Renderer. It should have been removed with passes. Pass seeds as a buffer to the OpenCL iteration kernel, rather than a single seed that gets modified. Slight optimization on CPU accum. Use case statement instead of if/else for xform chosing in OpenCL for a 2% speedup on params with large numbers of xforms. Add SizeOf() wrapper around sizeof(vec[0]) * vec.size(). Remove LogScaleSum() functions from the CPU and GPU because they're no longer used since passes were removed. Make some OpenCLWrapper getters const. Better ogranize RendererCL methods that return grid dimensions.
2014-11-28 04:37:51 -05:00
for (i = 1; i < params.m_Count; i++)//Real loop.
{
xformIndex = NextXformFromIndex(rand.Rand(), lastXformUsed);
2014-09-10 01:41:26 -04:00
if (xforms[xformIndex].Apply(&p1, &p1, rand))//Feed the resulting value of applying the randomly selected xform back into the next iter, and not the result of applying the final xform.
DoBadVals(xforms, xformIndex, lastXformUsed, badVals, &p1, rand);
DoFinalXform(ember, p1, samples + i, rand);
ember.Proj(samples[i], rand);
lastXformUsed = xformIndex + 1;//Store the last used transform.
}
}
else//Xaos, 3D, no final.
{
p1 = samples[0];
0.4.1.5 Beta 11/28/2014 --User Changes Remove limit on the number of xforms allowable on the GPU. This was previously 21. Show actual strips count to be used in parens outside of user specified strips count on final render dialog. Allow for adjustment of iteration depth and fuse count per ember and save/read these values with the xml. Iteration optimizations on both CPU and GPU. Automatically adjust default quality spinner value when using CPU/GPU to 10/30, respectively. --Bug Fixes Fix severe randomization bug with OpenCL. Fix undo list off by one error when doing a new edit anywhere but the end of the undo list. Make integer variation parameters use 4 decimal places in the variations list like all the others. New build of the latest Qt to fix scroll bar drawing bug. Prevent grid from showing as much when pressing control to increase a spinner's increment speed. Still shows sometimes, but better than before. --Code Changes Pass count and fuse to iterator as a structure now to allow for passing more params in the future. Slightly different grid/block logic when running DE filtering on the GPU. Attempt a different way of doing DE, but #define out because it ended up not being faster. Restructure some things to allow for a variable length xforms buffer to be passed to the GPU. Add sub batch size and fuse count as ember members, and remove them from the renderer classes. Remove m_LastPass from Renderer. It should have been removed with passes. Pass seeds as a buffer to the OpenCL iteration kernel, rather than a single seed that gets modified. Slight optimization on CPU accum. Use case statement instead of if/else for xform chosing in OpenCL for a 2% speedup on params with large numbers of xforms. Add SizeOf() wrapper around sizeof(vec[0]) * vec.size(). Remove LogScaleSum() functions from the CPU and GPU because they're no longer used since passes were removed. Make some OpenCLWrapper getters const. Better ogranize RendererCL methods that return grid dimensions.
2014-11-28 04:37:51 -05:00
for (i = 0; i < params.m_Skip; i++)//Fuse.
{
xformIndex = NextXformFromIndex(rand.Rand(), lastXformUsed);
if (xforms[xformIndex].Apply(&p1, &p1, rand))
DoBadVals(xforms, xformIndex, lastXformUsed, badVals, &p1, rand);
lastXformUsed = xformIndex + 1;//Store the last used transform.
}
samples[0] = p1;
ember.Proj(samples[0], rand);
0.4.1.5 Beta 11/28/2014 --User Changes Remove limit on the number of xforms allowable on the GPU. This was previously 21. Show actual strips count to be used in parens outside of user specified strips count on final render dialog. Allow for adjustment of iteration depth and fuse count per ember and save/read these values with the xml. Iteration optimizations on both CPU and GPU. Automatically adjust default quality spinner value when using CPU/GPU to 10/30, respectively. --Bug Fixes Fix severe randomization bug with OpenCL. Fix undo list off by one error when doing a new edit anywhere but the end of the undo list. Make integer variation parameters use 4 decimal places in the variations list like all the others. New build of the latest Qt to fix scroll bar drawing bug. Prevent grid from showing as much when pressing control to increase a spinner's increment speed. Still shows sometimes, but better than before. --Code Changes Pass count and fuse to iterator as a structure now to allow for passing more params in the future. Slightly different grid/block logic when running DE filtering on the GPU. Attempt a different way of doing DE, but #define out because it ended up not being faster. Restructure some things to allow for a variable length xforms buffer to be passed to the GPU. Add sub batch size and fuse count as ember members, and remove them from the renderer classes. Remove m_LastPass from Renderer. It should have been removed with passes. Pass seeds as a buffer to the OpenCL iteration kernel, rather than a single seed that gets modified. Slight optimization on CPU accum. Use case statement instead of if/else for xform chosing in OpenCL for a 2% speedup on params with large numbers of xforms. Add SizeOf() wrapper around sizeof(vec[0]) * vec.size(). Remove LogScaleSum() functions from the CPU and GPU because they're no longer used since passes were removed. Make some OpenCLWrapper getters const. Better ogranize RendererCL methods that return grid dimensions.
2014-11-28 04:37:51 -05:00
for (i = 1; i < params.m_Count; i++)//Real loop.
{
xformIndex = NextXformFromIndex(rand.Rand(), lastXformUsed);
if (xforms[xformIndex].Apply(&p1, &p1, rand))
DoBadVals(xforms, xformIndex, lastXformUsed, badVals, &p1, rand);
samples[i] = p1;
ember.Proj(samples[i], rand);
lastXformUsed = xformIndex + 1;//Store the last used transform.
}
}
}
else//Xaos, no 3D.
{
if (ember.UseFinalXform())//Xaos, no 3D, final.
{
p1 = samples[0];
0.4.1.5 Beta 11/28/2014 --User Changes Remove limit on the number of xforms allowable on the GPU. This was previously 21. Show actual strips count to be used in parens outside of user specified strips count on final render dialog. Allow for adjustment of iteration depth and fuse count per ember and save/read these values with the xml. Iteration optimizations on both CPU and GPU. Automatically adjust default quality spinner value when using CPU/GPU to 10/30, respectively. --Bug Fixes Fix severe randomization bug with OpenCL. Fix undo list off by one error when doing a new edit anywhere but the end of the undo list. Make integer variation parameters use 4 decimal places in the variations list like all the others. New build of the latest Qt to fix scroll bar drawing bug. Prevent grid from showing as much when pressing control to increase a spinner's increment speed. Still shows sometimes, but better than before. --Code Changes Pass count and fuse to iterator as a structure now to allow for passing more params in the future. Slightly different grid/block logic when running DE filtering on the GPU. Attempt a different way of doing DE, but #define out because it ended up not being faster. Restructure some things to allow for a variable length xforms buffer to be passed to the GPU. Add sub batch size and fuse count as ember members, and remove them from the renderer classes. Remove m_LastPass from Renderer. It should have been removed with passes. Pass seeds as a buffer to the OpenCL iteration kernel, rather than a single seed that gets modified. Slight optimization on CPU accum. Use case statement instead of if/else for xform chosing in OpenCL for a 2% speedup on params with large numbers of xforms. Add SizeOf() wrapper around sizeof(vec[0]) * vec.size(). Remove LogScaleSum() functions from the CPU and GPU because they're no longer used since passes were removed. Make some OpenCLWrapper getters const. Better ogranize RendererCL methods that return grid dimensions.
2014-11-28 04:37:51 -05:00
for (i = 0; i < params.m_Skip; i++)//Fuse.
{
xformIndex = NextXformFromIndex(rand.Rand(), lastXformUsed);
2014-09-10 01:41:26 -04:00
if (xforms[xformIndex].Apply(&p1, &p1, rand))
DoBadVals(xforms, xformIndex, lastXformUsed, badVals, &p1, rand);
lastXformUsed = xformIndex + 1;//Store the last used transform.
}
DoFinalXform(ember, p1, samples, rand);//Apply to last fuse point and store as the first element in samples.
0.4.1.5 Beta 11/28/2014 --User Changes Remove limit on the number of xforms allowable on the GPU. This was previously 21. Show actual strips count to be used in parens outside of user specified strips count on final render dialog. Allow for adjustment of iteration depth and fuse count per ember and save/read these values with the xml. Iteration optimizations on both CPU and GPU. Automatically adjust default quality spinner value when using CPU/GPU to 10/30, respectively. --Bug Fixes Fix severe randomization bug with OpenCL. Fix undo list off by one error when doing a new edit anywhere but the end of the undo list. Make integer variation parameters use 4 decimal places in the variations list like all the others. New build of the latest Qt to fix scroll bar drawing bug. Prevent grid from showing as much when pressing control to increase a spinner's increment speed. Still shows sometimes, but better than before. --Code Changes Pass count and fuse to iterator as a structure now to allow for passing more params in the future. Slightly different grid/block logic when running DE filtering on the GPU. Attempt a different way of doing DE, but #define out because it ended up not being faster. Restructure some things to allow for a variable length xforms buffer to be passed to the GPU. Add sub batch size and fuse count as ember members, and remove them from the renderer classes. Remove m_LastPass from Renderer. It should have been removed with passes. Pass seeds as a buffer to the OpenCL iteration kernel, rather than a single seed that gets modified. Slight optimization on CPU accum. Use case statement instead of if/else for xform chosing in OpenCL for a 2% speedup on params with large numbers of xforms. Add SizeOf() wrapper around sizeof(vec[0]) * vec.size(). Remove LogScaleSum() functions from the CPU and GPU because they're no longer used since passes were removed. Make some OpenCLWrapper getters const. Better ogranize RendererCL methods that return grid dimensions.
2014-11-28 04:37:51 -05:00
for (i = 1; i < params.m_Count; i++)//Real loop.
{
xformIndex = NextXformFromIndex(rand.Rand(), lastXformUsed);
2014-09-10 01:41:26 -04:00
if (xforms[xformIndex].Apply(&p1, &p1, rand))//Feed the resulting value of applying the randomly selected xform back into the next iter, and not the result of applying the final xform.
DoBadVals(xforms, xformIndex, lastXformUsed, badVals, &p1, rand);
DoFinalXform(ember, p1, samples + i, rand);
lastXformUsed = xformIndex + 1;//Store the last used transform.
}
}
else//Xaos, no 3D, no final.
{
p1 = samples[0];
0.4.1.5 Beta 11/28/2014 --User Changes Remove limit on the number of xforms allowable on the GPU. This was previously 21. Show actual strips count to be used in parens outside of user specified strips count on final render dialog. Allow for adjustment of iteration depth and fuse count per ember and save/read these values with the xml. Iteration optimizations on both CPU and GPU. Automatically adjust default quality spinner value when using CPU/GPU to 10/30, respectively. --Bug Fixes Fix severe randomization bug with OpenCL. Fix undo list off by one error when doing a new edit anywhere but the end of the undo list. Make integer variation parameters use 4 decimal places in the variations list like all the others. New build of the latest Qt to fix scroll bar drawing bug. Prevent grid from showing as much when pressing control to increase a spinner's increment speed. Still shows sometimes, but better than before. --Code Changes Pass count and fuse to iterator as a structure now to allow for passing more params in the future. Slightly different grid/block logic when running DE filtering on the GPU. Attempt a different way of doing DE, but #define out because it ended up not being faster. Restructure some things to allow for a variable length xforms buffer to be passed to the GPU. Add sub batch size and fuse count as ember members, and remove them from the renderer classes. Remove m_LastPass from Renderer. It should have been removed with passes. Pass seeds as a buffer to the OpenCL iteration kernel, rather than a single seed that gets modified. Slight optimization on CPU accum. Use case statement instead of if/else for xform chosing in OpenCL for a 2% speedup on params with large numbers of xforms. Add SizeOf() wrapper around sizeof(vec[0]) * vec.size(). Remove LogScaleSum() functions from the CPU and GPU because they're no longer used since passes were removed. Make some OpenCLWrapper getters const. Better ogranize RendererCL methods that return grid dimensions.
2014-11-28 04:37:51 -05:00
for (i = 0; i < params.m_Skip; i++)//Fuse.
{
xformIndex = NextXformFromIndex(rand.Rand(), lastXformUsed);
2014-09-10 01:41:26 -04:00
if (xforms[xformIndex].Apply(&p1, &p1, rand))
DoBadVals(xforms, xformIndex, lastXformUsed, badVals, &p1, rand);
lastXformUsed = xformIndex + 1;//Store the last used transform.
}
samples[0] = p1;
0.4.1.5 Beta 11/28/2014 --User Changes Remove limit on the number of xforms allowable on the GPU. This was previously 21. Show actual strips count to be used in parens outside of user specified strips count on final render dialog. Allow for adjustment of iteration depth and fuse count per ember and save/read these values with the xml. Iteration optimizations on both CPU and GPU. Automatically adjust default quality spinner value when using CPU/GPU to 10/30, respectively. --Bug Fixes Fix severe randomization bug with OpenCL. Fix undo list off by one error when doing a new edit anywhere but the end of the undo list. Make integer variation parameters use 4 decimal places in the variations list like all the others. New build of the latest Qt to fix scroll bar drawing bug. Prevent grid from showing as much when pressing control to increase a spinner's increment speed. Still shows sometimes, but better than before. --Code Changes Pass count and fuse to iterator as a structure now to allow for passing more params in the future. Slightly different grid/block logic when running DE filtering on the GPU. Attempt a different way of doing DE, but #define out because it ended up not being faster. Restructure some things to allow for a variable length xforms buffer to be passed to the GPU. Add sub batch size and fuse count as ember members, and remove them from the renderer classes. Remove m_LastPass from Renderer. It should have been removed with passes. Pass seeds as a buffer to the OpenCL iteration kernel, rather than a single seed that gets modified. Slight optimization on CPU accum. Use case statement instead of if/else for xform chosing in OpenCL for a 2% speedup on params with large numbers of xforms. Add SizeOf() wrapper around sizeof(vec[0]) * vec.size(). Remove LogScaleSum() functions from the CPU and GPU because they're no longer used since passes were removed. Make some OpenCLWrapper getters const. Better ogranize RendererCL methods that return grid dimensions.
2014-11-28 04:37:51 -05:00
for (i = 0; i < params.m_Count - 1; i++)//Real loop.
{
xformIndex = NextXformFromIndex(rand.Rand(), lastXformUsed);
2014-09-10 01:41:26 -04:00
if (xforms[xformIndex].Apply(samples + i, samples + i + 1, rand))
DoBadVals(xforms, xformIndex, lastXformUsed, badVals, samples + i + 1, rand);
lastXformUsed = xformIndex + 1;//Store the last used transform.
}
}
}
return badVals;
}
};
2014-09-10 01:41:26 -04:00
}