Windows version: 10.0.15063 Build 15063
I record using two dongle recorders. One recorder is on one side of my desk, and one on the other. The resultant 4-bit 48K ADPCM dual-mono wav files are usually 10-12 hours in length. I record my own workday this way. I usually turn on one recorder, then the other, snap my fingers between them as a ‘slate’ so I can later time-align the tracks. My wish is to put together a stereo track with one track on Right and Left.
There is considerable time fluctuation, up to 1000-3000ms such that within 5 minutes the tracks are terribly out of sync. I then give up.
What I want to do is “anchor” a track on a point, advance to where they are out of sync, and stretch one track to sync with the other. I could do this a few times through the 12-hour track to get a “KINDA SORTA” stereo track. I am very willing to have them never perfectly line up. But I want to develop a workflow so I can have a stereo track for each day with these two autonomous mics.
I’m guessing that this is a common procedure that I’m just not familiar with–or perhaps it’s automated?
Point me to the appropriate instructions and I will play with Audacity to do this.
I don’t know of an easy solution… If the recorders are picking-up common sounds you might try [u]VocAlign[/u]. It’s not free, but I think it’s supposed to “run automatically”. (Of course, it can’t run automatically if there’s nothing in common to synchronize.)
If these are left & right channels of a stereo recording and there is common information, the common information will drift out-of-phase, and even with some time re-alignment, you could get some “phase weirdness”.
FYI - Pros use an accurate master clock and interfaces with master-clock inputs. With multiple interfaces using the same clock they are matched down to the sample. And of course, pros have multi-track in interfaces so they usually don’t need to sync multiple interfaces.
Yes, I’m aware of phase weirdness–which will be very prevalent in this recording. Let me tell you more. One microphone is embedded in the base of my travel coffee mug, and the other is in my lunchbox. When I got to meetings, one microphone goes with me to the meeting, the other stays. I will typically listen to this recording one channel at a time–noting when one microphone goes somewhere and when they come together. I understand that MUCH more sophisticated recorders would have to be used to have them sync perfectly.
Just looking for the best possible result in a challenging situation.
I don’t think we want to get involved in spying activities, and please be aware that recording people without their knowledge or consent is illegal in many countries.
Regarding the technical aspects of synchronising long tracks:
- Be aware that Audacity works with uncompressed 32-bit audio, so working with 12 hour recordings will create a huge amount of data.
- The first thing that I would do is to split the recording into more manageable size sections, say an hour each.
- Then you will need to identify points near the start and near the end for each hour section that is common to both recordings.
- In a fresh Audacity project, import an hour section of each recording, and line up the start by shifting one of the tracks with the Time Shift tool.
- Add a label at the point that lines up
- Then identify a point near the end of the hour and add a label for that point in each of the audio tracks.
- Zoom out so that you can see the (one) label near the start of the track and the two labels near the end of the tracks.
Select from the first label to first of the end labels, and look in the Selection Toolbar to see what the length is, and write down the length.
- Select from the first label to the final label, and write down that length.
- Now you need to decide which of the two audio tracks you want to stretch.
- Select in one of the audio tracks, the region from the first label to the end label for that track, and use the Change Speed effect to stretch it to the same length as the labelled section in the other track (use the time that you wrote down in the “New length” control).
Thank you. All recordings are obtained WITH CONSENT in this case.
I have found that creating a pseudo-stereo track in Goldwave, and simply snipping some from one channel or the other every hour or so is less time-intensive than the process you describe above. Usually it’s about 30-40ms, and isn’t characteristic of EITHER track. It varies. Frustrating.
Thank you very much for the procedure.
You can of course do that in Audacity, but I was answering your question, and you said that you wanted to “stretch one track to sync with the other”.