Issues with Posterize Time effect and decimals (maybe)?

[Ignatz Johnson Higham]

So I've got a comp and I want to have it on 2s or rather I want to halve the frame rate. I could do this by changing the comp frame rate and such, but I still want it in the original frame rate, just each frame to hold for 2 frames.

So I stick a posterize time effect on it and enter the frame rate as half the original. In this case the original is 24.976 and so halved is 11.988. Ae displays this as 12 as I assume it rounds it for the ui. But it's not accurately posterizing time.

I've done a test where I have a comp with a frame counter, duplicated it and halved the frame rate using three methods. Now if I'm right these three methods should all return the same result, but they don't. Please someone tell me where I've gone wrong or is the Posterize Time effect doddgy?

Here are my three methods and a screen shot to explain:

A - Posterize time with an expresion that halves the frame rate

B - Time Remap with an expression that halves the frame rate

C - individual toggle hold keyframes applied to time remap - the I don't trust AE method

Shouldn't these all return the same result, please someone tell me why?

[Mylenium]

Such a simple test is not really telling us much. In order to figure out whether there are any quantization or rounding issues you would have to have the timecode effect show some decimals and you would have to show us the result after at least a few cycles, not right at the beginning. Anything is possible of course but I would think that outside specific considerations for comp vs. source framerate and how AE samples stuff based on shutter phase and duration there should not be much of a difference indeed. That may in fact worth be looking into. If the shutter has some odd offset, it may simply sample the result at a different time wheres the expressions and time-remapping work on the actual time of the parent comp.

Mylenium

[Ignatz Johnson Higham]

Ok, here's a more detailed test. It's one precomp with three time displays. The only effects that behave as expected are highlighted in red. Plus my AE test file. Please let me know where I've gone wrong as to my mind these effects should all return the same result.

[Mylenium]

I messed around with this stuff and the interesting observation would have to be that this seems to be an API bug or in the code of the plug-in itself. Inputting 1/frameDuration*0.5 returns correct results, but not on frame 0, while hardwiring the values always gives wrong results. And here's the kicker: I ran this on CS6 here at home since my access to spiffy machines and newer versions is severely limited due to my health condition these days, but this problem seems to exist forever. Never noted before, as apparently here in PAL land we don't deal that much with fractional framerates.

Mylenium

[Ignatz Johnson Higham]

Great, I'd only notecied this issue recently as I usually work at 25fps and it hadn't been a problem. But I've been working on a job where the client want's it in 24.976, and the animation director want's it on twos.

So how do I bring this to the attention of the Adobe Devs?