Skip to main content
kaptivz62334512
kaptivz62334512
New Participant
April 24, 2019
Question

synchronization of one audio with another with variable speed

  • April 24, 2019
  • 1 reply
  • 338 views

I need to synchronise one audio track with variable playback according to another audio, which is perfectly fitting video file.

The problem is that in one audio file playback was changed to acceleration and deceleration randomly, withing 5% of original time playback of each part. It's almost not noticeable, but when add an audio to video file, it becomes unreal to see this garbage.

I have another audio file, that fits to the video perfectly. So, the question is, how can I adjust and restore the synchronisation of my audio track to video, using the one with perfect synchronisation?

How I see that, but don't know if that functionality presents in Audition:

1. I put 2 tracks into Multitrack mode

2. Using some markers I check the points or visible audio peaks that both tracks have

3. These peaks are the same for both tracks, but not synchronised. I check these audio matching points in both files.

4. Using some functionality, Audition changes one file by stretching and compressing fragments of audio to fit another audio perfectly in the points, that I marked out

Can you help me with this task? Maybe, Audition has some automatic synchronisation of two files using some key points? That question is very important for me, I appreciate if you can give an answer

    This topic has been closed for replies.

    1 reply

    SteveG_AudioMasters_
    SteveG_AudioMasters_
    Community Expert
    April 24, 2019

    Audition has an automatic speech alignment feature, and it also works quite well on music. Well up to a point it does, anyway...

    Put the audio from both files onto a track each in multitrack view, and select both of them (Ctrl+click). Now right-click and you will get a lot of options. One of them is Automatic Speech Alignment:

    Select the appropriate track as the master, and let it get on with it! The only proviso here is that because it's a memory-intensive process, it's necessary to do this in chunks of less than 2 minutes each, so it will take several goes to fix the whole thing. But when you've done that it should be easy to reassemble the results so that they align.

    kaptivz62334512
    kaptivz62334512Author
    New Participant
    April 24, 2019

    The problem is that 2 audio files don't have exactly the same content, one has longer speech parts, but this algorithm matches longer voice fragment to short original fragment. Starting points of speech matches very well, but when finishing points also matched, speech becomes unreal to understand, it plays too fast and it's unreal to hear that sound even for a minute. I wish I could manually control the alignment, and mark the points where it should work, and what it don't have to change.

    Thanks for this option, but it works too unintelligible, and the result is too bad

    Terms & ConditionsAccessibility statement
    Using the community Terms of use Privacy Cookie preferences Do not sell or share my personal information ad choicesAdChoices