I am using four voice recorders to record the audio of a group interview. Three of them are Sony IC Recorder ICD-PX440 (stereo) and one is a Phillips Voice Tracker (mono). I synced the four tracks in Audacity around a clapping sound towards the middle of the recording. However, I find that whereas the three tracks coming from the Sony recorders sync perfectly, the Phillips track is out of sync as I move away from the clap. That is to say, the mono track is synced with the stereo ones around the clap, but is off by milliseconds towards the start and end of the track. Is this to do with the encoding of the audio? Can I do something to correct it in Audacity?
2 Answers
Check the sampling rate that each of the recorders are using. For example, you could be running into issues importing a 48kHz digital recording into a 44.1kHz DAW project.
-
spot on: one recording was 44.1kHz and the other 22.05 kHz. Commented Sep 28, 2016 at 0:13
-
Follow up question, Dorito: my two recorders have been syncing perfectly, but in one recording session there was heavy rain during the first 5 minutes of the intereview. Only this interval of out of sync and the rest is in perfect sync. Can background noise somehow affect playback speed? Commented Oct 3, 2016 at 21:17
Sample rate is controlled by electronic circuits. They are not perfect and will vary from one box to another and may vary depending on temperature. This means that even if they "nominally" run on the same sample rate, there will be a difference. Hence getting out of sync. So, actually the lower temperature when raining might have an effect.
The electronic circuits contains a crystal which might have a frequency stability of +/- 50 ppm (parts per million) if you are lucky. Really expensive crystal oscillators places the crystal in an oven to keep the temperature constant. Or might sync to GPS satellite signals (that are extremely accurate). You will not find those expensive solutions in audio recorders (well, there might be somewhere).
50 ppm means 2.4 samples per second at 48kHz or about 3ms per minute which is about what you can perceive. Of course if one is slow and one is fast the difference might be double that.
To solve the issue you may "resync" by cutting out parts of the slower recording at pauses or adding silence at the faster. Some programs have "elastic" tools allowing you to sort of stretch the sound. Example may be Samplitude.
