Wow, so many bugs in the release … did anyone internally test this?
Create a Preset. Click on the “Include Preview & Export Settings” and it immediately closes the dialog box and doesn’t save export setting in the Preset.
When batch video processing, if I click on one Input source make changes to the video settings, then click on another video source I lose ALL the settings I just made for prior video source. This makes batch processing almost useless.
If 1 out of my 26 video sources doesn’t have audio it prevents me from changing Audio Mode “Copy” to “Convert”. It’s meaningless to have a UI restriction (prevent changing from Copy to Convert) since there is no audio anyway so neither Copy nor Convert would be relevant so why prevent the change?
In batch processing mode there is NO WAY to cancel all processing at once, have to go to each source one clip at a time and cancel, gets tedious with 100 plus input sources. Gets worse, if I try to “Close all Inputs” while a batch is being process, the UI just locks up and eventually get a “Not responding”.
Again for batch processing, when I select all inputs and click on the Export button (lower right) the UI doesn’t switch the tab to Export, just leaves it in Preview tab, not desirable at all, if a user performs an action that has related progress they should be presented with that process or else the user thinks NOTHING is happening and they click the Export again … and again. Action with Result … don’t leave the user wondering if anything is happening or not and make then hunt for another tab (Export) to see activity.
Sorry, but this latest version is horrible, I just can’t believe it passed any testing be it internal or beta testing. Is this a one person show/developer trying to do everything? If so that might explain why there are so many issues with UI and event messages.
I love this product when it works but so many new bugs and old bugs coming back to life. You might want to reduce the weekly deployments and focus more on a proper UAT cycle.
Bug found: Benchmark showing input resolution as undefined even though it is defined (as 1920x1080FHD in my case). Happening on both my Windows 11 PC and my MacBook Pro.
Hello,
For those who want it, I place here m configuration of 3 presets to enhance sequences from the Star Trek Voyager series. The 3 presets are to be used for enhancement. One is for denoising and sharpening. The 2nd is for scaling. And the 3rd is for overall enhancement with standard focus.
looks good, but IMO too much of the “AI upscale” look for my personal taste. I would dial up the “recover original detail” to around 50-80 to tone things down
Immediate crash if clicking on anything when program loads. I did notice that it seems to be stuck on getting license info. I ran a system restore to get back 5.02 which is working fine.
Most of the space scenes are still heavy with grain—and sometimes Proteus bakes that grain into the ships. Overall I find that acceptable because the original doesn’t look any better and those parts only last maybe two seconds at the longest.
Hahahaha. I never did the Star Trek Voyager upgrade again. This is the first time I’ve enhanced it . I’m currently in season 5. In the first phase. But first I de-interlaced all the sequences via Handbrake. Because Video AI’s deinterlacing models aren’t great.
What makes me laugh, though, are the crumpled paper backdrops painted to give the illusion of rock. Perhaps the SD source gives less of an illusion of crumpled paper, but in HD, it’s obvious .
And I know there’s a Blu-ray version out there somewhere. I’m not sure if it was only the remastered version with recreated special effects though. So I’ll stick with this version.
I am now using video proc converter AI got the lifetime student edition Man finally a piece of software that uses my decoder encoder and 3d features from my nvidia and wow the hardware converter is fenomenal I know now Topaz is dead
. But at 1m50 or 2 from the screen, the results are relatively good.
. I’m currently in season 5. In the first phase. But first I de-interlaced all the sequences via Handbrake. Because Video AI’s deinterlacing models aren’t great.
