Topaz Video AI Beta 5.2.3.1.b

Topaz Video AI Video AI Beta
1

Hi everyone,

Today we have a new beta release for Video AI.

5.2.3.1.b

Changelog:

  • Fixed network share locations causing an app crash on export on some systems
  • Fixed certain Radeon cards showing the AV1 export option when not supported in hardware
  • Updated minimum window size to avoid UI elements overlapping
  • Improved login flow for Standard+Pro license owners
5 Likes
2

Fixed network share locations causing an app crash on export on some systems

God bless it

2 Likes
3

Any chance of an updated Linux Alpha/Beta sometime soon?

2 Likes
4

Great. Thanks. :blush:

5

Getting some major unstability issues patched with this one, you’re on the good way, but goal is yet far.
Giving some feedback on this release, but it is common with all since introducing “Live rendering”

  • “Live rendering” is a bug nest. As soon as you use it, it generates a crash by a way or another when you touch to any button, cursor or value.
  • “Live render” is at least 2x longer to render 1 single frame than using 5 frames rendering in preview mode. This is not a normal behaviour.
  • Rendering frames/seconds is rock stable whatever you stress the app. “Live render” absolutely needs to fullfill operations else it crashes. (app end + reload)

Improvments

  • Keep the zoom + position of the video it was positionned to, previously you start a new render in A/B or slide : It is anoying as hell to rezoom and find the sweet spot you wanted to analyse.

  • Remove from workspace things that are fully useless, as an example, the upper part, that you touch at least 10s in an jour of job. This space would be SO USEFULL on a wide screen.
    (picture below, upper part with TOPAZ, project aso)

    Waste of space
    Waste of space2560Ă—1039 262 KB

  • Add preset name beside model name. When you deals with tons of tries to fine tune the parameters of the model, you can’t remember all the time at what point you were, this is insane.

  • Model names are not enough, we often fine tweak parameters inside a same model / combo model
    Adding the recorded preset help you to go forward in your tuning.
    This is called dichotomous search and is the fastest way you achieve our goal (in a situation like this one)

    My Preset
    My Preset1395Ă—78 48.9 KB

2 Likes