Hello,

So I have 2 tracks, one is lossless and another is lossy. The lossless track has incorrect runtime (silence at the beginning and at end) and the lossy track has correct runtime. Is there any way I can calculate the leading and trailing empty frames precisely from the lossy file and add them into lossless file so that the runtime of lossless file is exactly the same as lossy ?

Is there any way to accomplish this ?

Any help will be greatly appreciated.

Kind regards

 

I don't know of any way to calculate it, or what you would do with that number, but I'd try to sample align them manually by eye where the lossless tracks starts. Every DAW is capable of zooming to sample level.

 

Thank you for the response Avenger. I think machine would calculate better than my human eye can. Disheartening to know it can't be done. Not questioning your knowledge but maybe someone has some other idea ?

 

couldn't you use something like audacity and use TRUNCATE SILENCE?

 

This is not necessarily true, because in these examples you are probably dealing with additional space which is empty of any waveform information. Automatic detection methods will not be any better than you can do by zooming in and doing the edit. Personally, I would manually trim the silence. If it is that important to you that they match, I would feel better about just re-encoding the lossy example than trying to match up the waveforms and more so, the amount of any blank space.

 

yea, thought about that but it will have inconsistent silence at the beginning and at the end relative to the reference track. Thank you for the input though, appreciated.

 

I agree with @No Avenger and @clone. I know Wavelab can analyze multiple tracks and show the differences. But silence in a track is not always 'true' silence and the results can be unreliable. So I also suggest to drag the waves in normal view as close as possible first and zoom in to make fine corrections. You can play them both and make micro correction until they match exactly. The whole process will probably not take you more than a few minutes?

 

Alright, understood. I'll do that as everybody suggests the same method. I tried it on one track and got a pretty good result. Thank you for the help, highly appreciated.

 

Okay so it's decided. Thank you for the input, appreciated.

Also, is there any way to calculate the length of the track down to milliseconds ?

 

Track length is displayed down to milliseconds in every editor I've used.

But another thought came to me: Exporting to lossy formats seems to add silence at the beginning and end of a track. A few milliseconds.

 

if you need this for more precise purposes than just matching a file or two, I would look into SoX. https://sourceforge.net/projects/sox/

you can write scripts or find more advanced examples which can create excel sheets of the data, etc. like this example: https://stackoverflow.com/questions/12562023/trim-audio-files-with-sox-in-milliseconds

I haven't used this at all; but user posts on GS, etc. are posted by users of the approach. A little more complex than just using Audacity, but like most scripting options it may allow more data options.

 

Last edited: Aug 26, 2023

@krameri @clone

All done. Thanks.

 

@ceo54 Next time use Melodyne or become familiar with cross-correlation and DTW algorithm. https://github.com/andreaspung/dtw-music