Ok so I haven’t been able to test this yet due to waiting for my 4090 RMA to be processed, but my thoughts based on the feedback so far are:
The most voted for, long awaited and promised features such as pause and resume support, and scene detection, are still not available. IMO these should have been completed before even contemplating a UI rework.
Instead of fixing the live preview, it was removed completely.
Instead of a UI rework, the existing features of v3 could have been polished/expanded and additional requested features could have been implemented.
The v4 marketing was somewhat misleading, one of the claims made was "50% speed gain on TensorRT-compatible GPUs - ok so where is this 50% gain? What was the basis for this claim?
This update really feels like it was not well thought out and was not what most people wanted.
Hopefully a few weeks of updates can bring v4 up to a decent standard.
Disclaimer: just my opinion, feel free to agree, disagree, or flame me!
Best case scenario: The developers that program the user interface, only work on that and needed something to do. Pausing is some underlying thing that is still being worked on by other developers—and it’s just that hard to implement.
I think the fifty percent speed increase was supposed to be Nyx 2 versus 1… Didn’t they make that kind of claim when Iris 2 came out?
My over all impression is that this is VEAI 2 to TVAI 3 all over again in all the same ways.
I am sure they are refering to Iris v2 (MQ) vs Iris v1 (LQ).
I had already predicted it before the release of v4, because if you check the Program Data Model folder, you can see that most of the models except Iris v1 already using TensorRT model.
