Hello friends.
So following my first very quick test, I come back with a better analysis, having been able to test 5 different scripts.
First of all I confirm my first test (above) 0.13 is much much better.
Unfortunately I went back to FW3, the sensations remain much better!
For my tests I upgraded again to 4.0.13.
I took a 10 min video, I first played my script created by myself, in FW 4.
Then immediately after I downgraded to FW 3 and I replayed the same script.
The sensations are really better.
Then to try to understand I went back to FW 4, and I voluntarily modified my script without following the images too much but only to differentiate the sensations.
Personally, after several modifications and rereadings of this same script in FW3 then in FW4, here is what I notice:
When a move is too long over too short a duration, I voluntarily created a very tense point: 0 to 100 in 100 ms or 849 mm/s (unplayable move), so if FW 3 tries to get closer to it by remaining approximate, by playing for example 0 to 80 instead of 0 to 100, the move is played but truncated, on the other hand FW 4 knows that it will not succeed and skips this point to go directly to the next one.
Where this is most noticeable is on shorter and very close moves, for example to stay on the same crazy unplayable speed a move from 0 to 50 in 50 ms for 850 mm/s.
Same observation, FW 3 knows that he will not succeed but still tries to play something close to it, while FW 4 knowing that it is not playable skips this move.
I made many tests, going from 0 to 50, from 0 to 80 then from 0 to 100 and at different speeds from 300 mm/s to unplayable speeds at 850 mm/s to compare well.
And this is what comes out of my tests.
In summary if FW 3 knows that he cannot play a move he still tries to do something close to it, while FW 4 knows that he cannot play and directly skips the point to the next one.
Maybe I am completely wrong and mistaken, but personally this is what I was able to analyze during my tests.
This plays a lot on the feeling.
And this is confirmed by pushing the tests, on long shots not very fast for example 0 to 100 in 217 ms or 392 mm/s then FW 4 is better, more sensual, but on short and very fast shots for example 0 to 60 in 100 ms or 509 mm/s then FW 3 is much better.
All this can be in the sense of my first impressions posted here: to preserve the engine.
Personally after all my tests it is decided I return to FW 3 and I stay there for the moment.
The sensations are much better there, despite the enormous progress of FW 4.0.13.
