Nudge bug is in 5.1.1 too

[ripple FX]

I think I can explain a bug that I first discovered in 5.1 that still exists in 5.1.1

1.Open a new 44.1KHz session
2.Set frame rate to 29.97 FPS
3.Set Main ruler to Time Code and a second ruler to Samples
4.Set Main counter to Time Code and Sub counter to Samples
5.Set Nudge to 1 Second (00:00:01:00:00)
6.Open 1 Audio Track
7.Zoom in as far as possible
8.Press the "+" key to nudge forward 5 seconds
9.Notice the cursor is now 1 sample early
10.Nudge another 10 seconds and it's 2 samples early
Although it's a very small error (1 sample every 10 seconds) it is not correct.

What I think is happening is: PT uses samples as its' internal timeline so 1 second at 44.1KHz = 44100 samples. But at a frame rate of 29.97 FPS, 1 second = 44144.1 samples. PT nudge appears to be calculating 1 second as 44144.0 samples and therefore accumulating a 1 sample error every 10 seconds. PT does seem to nudge properly in 1 frame and 10 frame increments, so it's resolution must be smaller than whole samples.

I'm a musician, not a mathematician, but I think this is a fixable bug, no?

[ripple FX]

Yeah, I've noticed this too. I wonder if Digi Tech Support is aware of it? Hopefully they are now. I'm sure someone from Digi will respond soon. They're pretty busy right now.

[NK]

Yup.,
Set the sample rate to 48K and this goes away.
44,100 is a harder number to use than 48,000 mathematically I hear.

[ripple FX]

It also nudges correctly at 1 frame and 10 frame values. These also require fractions of samples to correctly calculate. It's only when the nudge value is 1 second that the error occurs.

[ripple FX]

bumpitty-do-dah

[DigiTechSupt]

-I have reported this behavior as a bug but have no idea as to why this happens (or when to expect a fix).

-All I know is that it is specific to the 29.97 frame rate. 30FPS does not exhibit this sample offset when comparing timelines.

Brent

[TommyT]

Dito..When I Spot regions to a new TC they are off..For example If I have a 30 second commercial at 01:00 hour TC with a 2beep at 58sec and get a revised picture where the spot starts a few EVEN minutes later..I spot all the regions with the 2beep by adding a few even minutes to the even start time in the spot window...But when I zoom in , the moved regions now start a frame late...And the more I repeate this process the later the regions become, First 1 frame then 2 and so on....This only happens when I have Digial Picture on line (miro DC30+) ..Yes I've checked the session and the Pic are at 29.97 NDF.I've reported this a year ago and got a "not able to reproduce" response. [img]images/icons/confused.gif[/img]

[dcornutt]

Good Job Ripple!

Now THATS how to report a bug.

I want to thank you for posting a clear and directed list of steps to reproducing this bug. It makes it clear and easy to reproduce this behavior and most certainly has made the difference in getting attention to it.

Thanks!

dcornutt

[ripple FX]

Brent, thanks for getting this info to those who should know.

dcornutt, thanks for your comments. I've been troubled by this bug for some time, but I know if it's not reproducable, it's pretty hard to diagnose.

Tommy T, I'm not sure if your problem is related to mine. I'll see if I can reprodue it. For a long time, I've noticed little slivers of regions that shouldn't be there after nudging and spotting, by-products of the slight inaccuracies we're talking about. But it wasn't until now that I was able to take the time to see the error clearly.

Digi has a superb piece of software. I'm sure a programmer there is going to say, "D'oh! I can't believe we didn't see that!"