Petryk

Hi I had an idea that may not have been right. I don't know what the developers of the application think about it. A nice idea would be to introduce some kind of survey, where everyone in our community could influence the direction and development of the application. So, My idea is as fallows: Someone briefly describes the functions they would expect in the program. The community then votes in favour. If a given idea reaches a large number of votes, the creators take the idea to the workshop. Next accept or for some reason reject a given function, giving the reason for rejection. What do you think about it.

In 1.9.2 there is still problems with RAM with open CL, AP use 16GB of ram ( focus merge 30 photos). When OpenCL is disabled the AP work fine and use in the same file 7GB of RAM. So still problems.

[DXCore] Found 3 adapters NVIDIA GeForce GTX 1050 Version: 27.21.14.6172 IsHardware: Yes Supports D3D12 Feature Level 12.0: Yes HardwareID: PCI\VEN_10DE&DEV_1C8D&SUBSYS_838F0000&REV_A1 LUID: 0x172F7 Intel(R) HD Graphics 630 Version: 24.20.100.6344 IsHardware: Yes Supports D3D12 Feature Level 12.0: Yes HardwareID: PCI\VEN_8086&DEV_591B&SUBSYS_838F0000&REV_04 LUID: 0x16D8A Microsoft Basic Render Driver Version: 10.0.19041.546 IsHardware: No Supports D3D12 Feature Level 12.0: Yes HardwareID: PCI\VEN_1414&DEV_008C&SUBSYS_00000000&REV_00 LUID: 0x172BD [OpenCL] Found 2 platforms: Name: Intel(R) OpenCL Vendor: Intel(R) Corporation Version: OpenCL 2.1 Name: NVIDIA CUDA Vendor: NVIDIA Corporation Version: OpenCL 1.2 CUDA 11.2.152 [OpenCL] Found 1 devices for platform Intel(R) OpenCL: Intel(R) HD Graphics 630 Vendor: Intel(R) Corporation Version: OpenCL 2.1 NEO [OpenCL] Found 2 platforms: Name: Intel(R) OpenCL Vendor: Intel(R) Corporation Version: OpenCL 2.1 Name: NVIDIA CUDA Vendor: NVIDIA Corporation Version: OpenCL 1.2 CUDA 11.2.152 [OpenCL] Found 1 devices for platform Intel(R) OpenCL: Intel(R) HD Graphics 630 Vendor: Intel(R) Corporation Version: OpenCL 2.1 NEO Installed Wintab: Failed to load Wintab32.dll Wintab: Failed to create tablet context Attempting to create Direct3D device with adapter NVIDIA GeForce GTX 1050 Setting screen DPI to 141.767441860465 (scale = 1.0) Exit

Hi all, I think I have another opencl bug. I export my work to jpg or png or any other format and have artifacts in the exported file (attachment). When I disable OpenCL, the artifacts are gone. Initially I thought it was the light settings in the file but unfortunately not. In the program itself, the artifacts are not visible in the opened file ,only after the export. There is one more thing. With hardware acceleration enabled, the program will eat up 8GB of RAM when I have the file open, if I disable OpenCL, the same file will only eat up 900MB.

Photo merge taken from 34 photos. I used a little help form the book previously purchased at Affinity shop. What do You all think. Maybe something to change ?

If you need some more information from Me: logs, system configuration, maybe some tests, write to me. I try to help to resolve this problem.

Ok I understand how this work, but still in My opinion, this function not work properly. For now i disable OpenCL. Maybe in new update this function will work fine.

I understand, but with OpenCL AF enabled, it should use the GPU core and the CPU, and it only uses the CPU and 50% of it, so with openCL enabled performance drops,it should be the other way around.

NVIDIA GTX 1050. By defoult af set renderer to NVIDIA. I got all updates on my system and updated NVIDIA studio drivers.

I also wanted to ask about it. Many free soft like "DeepSkyStacker" or "Sequator" got it. For Me and other users I think is basic function. In Affinity we can use this function only for deep sky stacking like on tutorials on affinity site. For landcapes this function is useless.

I try again and enable openCL. Next I check in task manager (Windows). With OpenCL on AF use processor but in 50% and still not use graphic card, in taks manager use 0% of graphic card procesor when I use focus merge. When openCL is disabled AF use 100% of procesor. So i think this is a bug with software not with drivers or windows.

I got the same problem. "New focus merge" with openCL enabled is very slow. I think this is not only Windows fault. Result from benchmark: