Skip to main content
Known Participant
August 27, 2023
Question

20 milliseconds delay

  • August 27, 2023
  • 2 replies
  • 7185 views

Hi, friends. I use Premiere to synchronize audio and video (.avi) and I love this software, but there's a thing bothering me. I noted the sometimes a frame suffers a time variation of 20 milliseconds. The difference is always 20 milliseconds! For example: Before to repair the audio synchronization, I select a frame that begins at 00:03:11:43200. After I synchronize that and several others parts, I return to "frame 00:03:11:43200" in order to see it again, but, now, the same frame begins at 00:03:11:41200! I would like to know why it happens and if is there a way to avoid that variation, since it destroys the confidence on lipsync. Should I use mp4 intead avi?

[url=https://ibb.co/3fR3hyL][img]https://i.ibb.co/3fR3hyL/Adobe-Premiere-Sy.jpg[/img][/url]

Thank you very much in advance.

This topic has been closed for replies.

2 replies

Community Expert
August 27, 2023

It looks like you are working with Audio units, there is a good chance Premiere is going to the nearest viseo frame.

Known Participant
August 27, 2023

I don't understand it. When I revisit  the same frame several times as I continue fixing other parts, it can return to normal or no. I would like to know a way to stop it, since I can't be sure about the timing. If I set some audio 5 miliseconds after the mouth opens, for example, and the same frame now starts 20 miliseconds earlier, sound will be produced 15 miliseconds before the mouth opening. And, sometimes, it returns to original timing when I revisit it. I have no confidence to do the job anymore. Do you have any solution?

Community Expert
August 30, 2023

Hi, Remote Index. Thank you for supporting me.

Originally, I worked based on "audio samples". When I select "show audio time units", looks timeline is more sensitive. For example, I can listen exactly when the following speech "contaminarion noise" begins.

I never changed to "miliseconds" before open this thread, but it looks still more sensitive than "audio samples".

Talking about this scene, for example, girl's mouth opened at 00:04:38:09340. "following speech noise contamination" begins at 00:04:38:10333 and audio wave begins around 00:04:38:12628. I was totally satisfied with this arrangement. Adjusting based on "audio samples" is wonderful. I really don't understand why the time on video is changing.

And, after I linked audio and video as R Neil Haugen said, but this event turned to happen again. There was an alteration of "20-units of audio sample". The same happened with miliseconds metrics. Without this disturbing fact, I think Premiere is wonderful for audio sync.

https://ibb.co/pxX02w9

[Using miliseconds metrics, girl's mouth opened at 00:04:38.194. "following speech noise contamination" begins around 00:04:38.206 and audio wave begins at 00:04:38.262. ]

When I say "following speech noise contamination", I mean the noise from following wave sound that we can hear before it.

When I open the software again, mouth opens at 00:04:38:07338 in video. It means "following speech noise contamination" and audio wave will spend more time to be produced.


I've been a video editor for 15 years and an audio mixer for 30 years before that and I only use audio samples when I am syncing two audi tracks together. When syncing video to audio I stick with frame timecode. The spec for broadcast audio sync used to be plus or minus one frame. Modern digital transmission is often worse than that. I agree you nay be seeing some issues but sticking to frame timecode will probably solve it.

Known Participant
August 27, 2023