Well I think they’re cash poor. They can attract more naive new users with features rather than fixes.
They are using ffmpeg for the engine. ffmpeg is in a constant state of “Beta” development. With the recent release of 7.x it’s really new with lots to be worked out. If Topaz updated to ffmpeg 7.x, they should go back to 6.x.
Topaz should come up with a better type of development plan, otherwise the users and reviewers will abandon the software out of frustration.
Topaz reminds me of Skylum software… Incomplete>never fully developed>then come out with a new one that you charge for.
I bought this software twice for one year each time. Now I’m sticking with my latest update to version 3. But I always enjoy reading the posts here. I think this company’s game is easy to understand. It’s a bit like gambling. The managers know very well how to get people hooked. Unfortunately, the alternatives are also too expensive so far. So the only way to change this is to stop buying the software and get the best out of the last version you bought. Constantly relearning everything because the interface is constantly changing, and unfortunately also the functions that already worked well, is simply not an option. I believe that the developers are also constantly changing.
The main problem is that Topaz has no competition. There are no FOSS AI models that are up-to-par with the Topaz models, and the other paid video AI programs are just scams that repackage free software. Topaz is the only company that develops its own models. Even the limited video AI capabilities offered by Adobe or Resolve are not even remotely close to what can be achieved with TVAI.
Essentially, you can’t go anywhere. TVAI’s only competition is older versions of itself . So people just wait around in hopes that TVAI will improve rather than jumping ship and moving to a new company as would happen in a more diverse market.
Because they keep adding new ones as they never quite manage to fix old ones. I think I manage to avoid a lot of the things people complain about most because I haven’t made much use of any of the new features added since 3.x. And it looks like there are a few others here who feel the same.
Thing is just clear, it’s a run ahead to keep the leadership
AI is all around
TVAI one of the very rare software to include AI models in video transcoding. All others systme are hosted in datacenter, deliver shit result, with ridiculous file size and unbelieavable price
To me, they do try to keep this leadership at all price, but as you said, we’re all beta tester of far from being stables releases. That point is very disapointing.
issue 1. bitrate-Constant: I wanted to output 24000 when the original video was 22045, but it ended up with 5633.
I also tried outputting the same video at 40M and 60M, but it came out at 5600.
It does still disappear if you click somewhere outside of the dialog, I’ve asked the devs to fix this - it should retain focus, stay on top, and not go away until you select create or cancel.
I think the problem with the closed beta was probably a lack of diversity in test platforms. If the vast majority of your beta testers are using more or less the same CPU/GPU combination, a lot of bugs you think you’ve fixed but haven’t really are more likely to get by.
Better yet, a good beta tester is a paid beta tester, with qualifications and hardware necessary to provide good feedback to the developer. I read some software beta testers earn up to $23 per hour for rendering the service. What incentives do Topaz beta testers have?
Yep, same problem here. After 6 hours upscaling it gets stuck at 0fps and never completes. Makes you wonder if they do any testing at all. Normally they just break the UI but this is a showstopper
I tested IRIS and PORTEUS, NYX and GAIA with 31 random files of 1 minute or less, and set the bitrate to be constant over 20M. In most cases, the result is a lower bitrate than the original.