Batch processing images in Topaz AI 3.0 sees it crashing after process 50 -60 images. It does not use the Nvidia GPU as it used to. It increasingly slows down after processing the first few images. It hard uses the CPU 13 - 14% but it does bump up the power consumption. What went wrong? Every release you take one step forward 2 back and then race a new release that is unusable except if your processing 1 image at a time. This has been going on for 18 months and your product is commercially unusable.
Steps to reproduce issue:
- load a couple of hundred .jpgs or tiffs or DNG’s
- Select all images
- Press save
- Then process
- Watch the resource manager as available memory keeps growing
- Crashes back into Topaz AI and you will have fun working out which files it processed.
Topaz Photo AI [v1.x.x] on [Windows/Mac]
1 Like
Please send me your system profile information so I can check if your computer hardware is up to date.
You can securely submit your files(s) to my Dropbox using the link below.
Please be sure to send me a note here to let me know you sent something.
Hi John, systems information has been deposited in the dropbox account as requested.
Mark
Hi John,
I have posted my systems details to your dropbox account.
Regards,
Same problem here.
Sent my system configuration to the dropbox folder.
Hi John,
I have done 10Hrs testing and can report the following.
- That 3.0 is not using the GPU but rather the CPU no matter what you set it to in the preferences. It makes no difference if you us Studio or Game Drivers.
- The antimalware service in Windows 11 is constantly being triggered when Topaz AI is generating / saving images.
- That after processing 15 - 20 images Topaz AI slows down to a crawl. This can happen if you just process 15 -20 images individually without quitting Topaz or if you just run them as a batch. If you keep processing the system will crash.
Regards,
Mark
What I can add, on top of what Mark said, is that when I am processing more than 40-50 photos and I try to apply the same settings to all of them, I see a surge in RAM usage, going from 10-12GB to 64GB (on my machine that is the memory installed) and then going back to normal. When that happens, then if you try to export all the images, it will fail at some point.
Yes, I have been investigating this in detail and can report that after processing 10+ images in a batch that the link to the GPU gets dropped (stopped being used) that Topaz AI then drops to just using the CPU which then starts to increase in memory usage with each image that it processes (slowing down dramatically as it goes) until it completely fills the available physical memory at which point it crashes and restarts Topaz AI with all your images there but the AI processing settings all dropped off every image that you individually had to set. This also happens if you edit images one at a time in the time line and save them as you go. It is currently usable for processing more than a handful of images at a time. You have to exit Topaz AI after processing a handful of images and restart it to continue which clears the memory and allows the GPU to be used again. I’m not sure who’s programming / testing this but they clearly only have a 1 image testing system given this release got out. I have found an absolute host of other bugs and will not using the tool for at least 4 updates from now! Just one of these issues is that in batch processing you can see that Topaz AI runs the autopilot first before it detects your settings and then reruns! How much power is being wasted with just this one bug? Another is that if you change a settings slider in one element when making adjustments it reruns ALL of the other elements to generate a new preview image. How much power is being wasted doing this? I could keep going but don’t have the time to waste.
