We had to improvise our equipment on our last podcasting session since one of us was remote. That means we ended up recording one of us in audacity and two on a Zoom H1 recorder. The problem is we didn’t sync up our bitrates so audacity recorded at 44100 hz and the H1 recorded at 48000 hz. Obviously no clap syncing will make these tracks line up now as they play out at different rates.
Changing the rate of one to match the other does not work (again, I find out that I’m better at math then at not screwing stuff up in the first place) Simply changing the bitrate to a mathematical equivalent doesn’t seem to work as different sections will alternate between being early, on time, late, and back again for some reason. I tried the change speed effect but I couldn’t figure out the right percentage to adjust it or even if that would fix the problem.
Does anyone have any suggestions of how I can get these two to mesh? I’m not worried about pitch changes because THOSE I can correct in post but because of how little I know about bitrates the time problem is vexing me.
I’m surprised Audacity messed up the Time between these two files. Most video editors (48000) “know” how to automatically and completely in the background use both sample rates.
Desperation Method (there may be other ways to do this).
Assume you want everything in 44100. Open the 48000 show in Audacity by itself. The duration of the show should be correct. Change the default sample rate with the number on the bottom left of the window. Export a WAV sound file.
Totally wasn’t audacity’s fault. One of us was recording on a hardware voice recorder. I’ll give that a shot but will that have a different effect than just changing the bitrate in the track?
Well what could the other problem be? I assume it’s bitrate because its the only variance between the track and it seems like the time just plays out slower on the second track. Simply changing the bitrate on the track doesn’t do anything.
First, you are talking about sample rate (kHz) which is the number of samples per second. Bitrate (kbps) is the number of bits per second. Bitrate is commonly associated with compressed files as an indication of the amount of compression (although we can calculate the bitrate of uncompressed files).
You may just have to experiment with speed changes until you get the timing right. 48,000/44,100 is 1.088 (about a 9% difference), but I doubt that’s your problem and I doubt you’re that far off.
If you record and play back at the same sample rate, the tempo (and pitch) will be correct. Or you can mix two different sample rates together and the software will re-sample to the selected project rate and there should be no timing/tempo/pitch problems.
That means we ended up recording one of us in audacity and two on a Zoom H1 recorder.
Every device has its own clock (oscillator) and no clock is perfect. Two different devices will never match perfectly and eventually they will drift-apart, although they are usually close-enough for the duration of a song and sometimes close enough for the duration of a concert. If you’re using a regular consumer soundcard, I would suspect that the soundcard has a bigger error than the Zoom. Sometimes, musicians will play a backing track through their soundcard while recording a 2nd track through an external interface or a USB “podcast” microphone, and when they mix the tracks they won’t sync.
FYI - Pros use a super-accurate (expensive) master clock and interfaces with master clock inputs, so everything is synchronized to the exact sample count.
as different sections will alternate between being early, on time, late, and back again for some reason.
If the timing is not consistent, you may have to split the recording into sections and sync the sections.
If it’s not extreme, variation in timing are perfectly normal. Effect > Change Speed. I think one of the variations is put in the durations of the two tracks and the effect will figure everything else out.
The people who do four and five different independent locations have to do this all the time.
We recorded separately and the tracks seem to be playing at different speeds so duration based solutions are not helpful at the moment, unfortunately.
The problem with percentage based solutions is it seems that the audio is not consistently shifting away at a constant rate but fluctuating. I can match up two words perfectly then in a 15 second clip the 2nd track will be a bit ahead, then it will resync, then be behind. So it doesn’t appear just to be a length or speed problem.
We need to start somewhere, so Media INFO is a grand place. Do that.
I don’t have same duration tracks.
You won’t if one recorder was running at 44100 and the other was running at 44101 or some other fraction off. The second track will play with your machine “assuming” it was correctly recorded. It wasn’t so your 20 minute show appears as 19.75 minutes. Different duration. As near as I can tell, Audacity doesn’t care about 44100 and 48000.
The problem with percentage based solutions
The tool has provision for typing in the two durations and it will try and resolve them. No percentages.
the audio is not consistently shifting away at a constant rate but fluctuating.
And that’s the phrase that pays. I think you just killed your show. You should notify the owner not to make any more presentations with that recorder.
I would try correcting the whole show anyway and see maybe the short variations and wanderings aren’t that far off. If they wander back and forth around sync and not by much, you may be able to get away with it. If quiet portions of the “wrong” voice are included in the wandering segment, then you’re stuck. This is why the Big Kids do total isolation between voices in a complicated show.