Frame interpolation and duplicates removal appears to work great now. I’ve only tried Apollo v8, but on a couple of different samples I went from 25 fps sources with random duplicates throughout to both 23.976 fps and 24 fps outputs and both came out nice and smooth. The internal preview playback is still jerky. I had to open the preview in an external player to verify that the interpolation was actually working correctly.
Thank you for fixing the one side with no changes was reprocessing each time. Is that the “No longer reprocessing overlapping preview areas.” in the description above?
And a minor thing - If you zoom in with mouse wheel in preview area. The zoom size drop-down continues to say “Fit”, then if you click it to Fit, it doesn’t work.
In all sincerity and honesty, I really don’t know what to say anymore. I agree with what PaulM said in the general discussion in his post RIP Topaz. There is no longer any doubt about that. I’m sorry to be realistic, it’s getting worse and worse. I don’t know if it’s because of the developer team who are incompetent, while respecting these people behind the software, I’m not here to insult them, I’m just saying the reality of things, either that’s the direction which is bad… But there’s really nothing going right… And I’ll say it again, Topaz will lose customers if this continues. Gigapixel AI for example had huge potential, many of us here love this software. Why did you suddenly stop it at version 6.3.3? Topaz Photo AI will never exceed Gigapixel, in fact it is worse than GPAI, and many do not like this software. In short, this is not the subject of this post, this concerns TVEAI. Really I say to the team with sincerity, version 3 was not really finished, why are you quickly launching version 4 which is full of bugs, crashes, and lots of things that are wrong… I hope that Adobe Res Up will overtake you because it’s truly horrible. It really feels like Topaz has really completely regressed. as I say one last time, is it the fault of the developer team or the management but nothing is going well there.
When i switch Between the different models in the preview the looping stopps and this message appears, so i cant semless switch between models to compare. Looping is enabled while switching the models. Video is freezing for about 1 second while switching model. Dont know if its possible that its playing without stopping while switching :
But if I switch between the Preview and the Original, I get different Zoom-Factors for the Preview! So I can’t compare the Original with the Upscaled Video!
Moving along the timeline is painfully slow again. Click a new place on timeline and wait 5-10s before it shows up in preview, and that is no processing or anything.
Same video I have been using for awhile to test this, Apollo v8 is still doing blocks on high action scenes.
Original
And 50fps
It is not as bad as it was last version looks like, but still bad enough.
I think this is the worst Video you could give an AI.
There are no really predictable objects in here.
This kind of video will be maybe possible in the far far future.
I couldn´t even imagine how i would calculate here in betweens…
With this video i wouldn´t expect too much at the moment…while the AI isn´t able yet to do easier jobs.
Give it some years…and also this would go…
With this kind of video…try different AI´s…the most problem here is that the AI doesn´t know the algorithm that is the base of this video is made of…maybe you can think how this could look…but for an AI it has to think of objects and how they could move. This is much too complex.
Not necessarily true. It is QUITE old hat to be able to look at 2+ frames before and after to guess what a frame in the middle should be, this has been around for awhile used to analyze frames to better help noise/temporal noise reduction, I used to use that in megui. There is no way someone can tell me that AI, with how powerful it is now, cannot take a few frames before and after and THEN use AI matching patterns, and if none found just best guess placement between frames. It should easily be able to see there are no squares in that.
Now… that said this is WHY I provided this, because you do not optimize and modify for the BEST case scenarios, you do it for the worst ones because you know if you get the worst down, the best are easy. This is, of course, sans other models which AI is learning, and it very much should have a path of easy reproduction if their standard models do not apply by analyzing multiple frames. I could be wrong 100% on how this works, but I already do this type of thing with our AI bots we have deployed at work, if our trained responses and models do not suffice, they take the context given and try to extrapolate from that.
edit Additionally, I would think it should not be a monumental task to include an artifact hunter addition that essentially views each frames ±5 from the frame it is on to see if that frame contains artifacts, either self generated or from compression, and rebuild as necessary. I say ±5 because a scene change in the middle could throw that off, and I cannot imagine anything that has 2 scene changes in 5 frames. look at 3 frames beforelook at 3 frames afternotice single frame artifacts not in the other original framesrebuild those frames using a frame-to-frame model and not a learned model something like that would make sense to me. Obviously standard is training models, and the more training the better it can “guess” the data, but if you have lines pre and post that frame that move a tad, it should be easy to replicate the middle frame of that. My case IS one of the worst case, so if they get it to work on that, it should be fine for all other light flashes, explosions, electricity jumping scenes, etc. Make it an option for this, make sure all knows this will be for high action/movement movies or scenes and that it will dramatically slow down processing due to the frame compares, but for high speed movies or scenes this could be invaluable…
Not sure if anyone else on this thread is using Linux, I am on Pop!_OS 22.04 and this version is failing to load due to missing library:
TopazVideoAIBETA.desktop[11216]: /opt/TopazVideoAIBETA/bin/Topaz Video AI BETA: error while loading shared libraries: libcloudprocessing.so: cannot open shared object file: No such file or directory
I could not find that file at all, is this supposed to be supplied by the debian package or is there a dependency that I am missing?
I hope that Adobe Res Up will overtake you because it’s truly horrible. It really feels like Topaz has really completely regressed. as I say one last time, is it the fault of the developer team or the management but nothing is going well there.
