My friend and I are doing a podcast together but live 2 1/2 hours away.
We use Audacity to record independently and sync both our start and stop time.
We both export as MP3 and have also tried as WAV, but my file is always shorter and his is always longer. All of the content is there for both but his is spaced and/or mine is compressed. We tried screen sharing to make sure all of the settings were the same and seem to be. I am unsure if this is related to bits per second or what, but somehow the final lengths and timing are far off. Makes for an editing nightmare.
Can anyone please help how we can ensure our synced recordings end up the exact same time???
That means you both “One, Two, Three, Clap!!” Another way to do that is one of you claps and the other holds the headphones up to their microphone.
Figure out how much longer one is than the other and use Audacity > Effect > Change Speed on one of them. That’s it. If you derive a percentage, the two shows will always fail sync by the same percentage no matter how long they are.
This is a cheap computer problem. The movie people spent very serious money to make sure all the recorders on a shoot run at exactly the same speed.
You should both be in 44100 sampling rate. Under certain conditions, Audacity can’t cope with different rates on two timelines.
If you get the two ends to perfectly match but it’s off in the middle, then one or both of the computers is broken and I think I hear my mum calling…
Just to be clear, chances are good you’re not both running at 44100.00. One of you may be running at 44100.01, for example. You’re experiencing that tiny difference in sampling rates. In film terms, one of your cameras is slightly under or over-cranked.
Effect > Change Speed is the least damaging of the corrections and you should be able to resolve everything OK.
The other two are Effect > Change Tempo and Effect Change Pitch. Those two have to rip the sound apart and try to put it back together with the corrections burned in. Those are much more likely to create sound damage.
Try that trick of putting the headphones up against the microphone. You can get the sync clap with no human reaction time.
I would think that Effect > Change Speed is the most correct method to use. After all, if a recording is played back at a different rate than it was recorded at (such as recording at 44100.01 and playing back at 44100.00), then both its length and its pitch will change by the same amount but in opposite directions. Effect > Change Speed will naturally correct both.
(I can’t help but think of those old Walkman cassette players with weak batteries. Everything got slower and lower pitched! Effect > Change Speed can fix that too, unless the batteries get progressively weaker to the point where the playback rate approaches 0 Hz, then fixing that becomes more challenging.)
I was able to open to separate Audacity Files - one mine and one my friend’s single track recordings from our podcast.
For this specific podcast night from weeks prior we did not start and end with claps (as we usually do) so it was a bit more difficult to get the perfect ratio… However, I did get it close enough to pass as clean transitions from beginning to end.
My track was 1 hour 54 minutes and 48 seconds
My friends track was 2 hours 4 minutes 59 seconds
If you listen to the uncompressed (without Speed Change) tracks, they are in sync and as time goes on they begin to drift further and further out of sync. This made for an editing nightmare taking 6+ hours when should have been a 3 hour or less job.
As stated above, we did not clap to start of end this specific podcast from weeks ago, but was able to attempt to time per beginning and opening conversation within fractions of a second - not the best or greatest accuracy but was able to make it work. The loosely accurate % rate from this attempt was 8.867% change.
I look forward to making a new recording and/or reviewing some of our other recordings where we did clap in and clap out.
IN SHORT, THE EFFECT > CHANGE SPEED certainly works and seems to be no distortion or quality loss.
Thank you Koz, will repost again next weekend after we make a fresh recording and use this method more accurately.
With that high of a speed change, it sounds like one of the recordings (the longer one) was actually taken at 48000 Hz instead of 44100. Speeding that up to match 44100 would be about an 8.844% change.
We did numerous screen share sessions trying to figure this out before and I am confident that both Audacities (mine and his) were recording at 44,100 and that all other visible settings were identical. We spent a lot of time trouble shooting this previously before I thought to get tech support from Audacity and found this forum with wonderful people who help - thank you!
Other than that I can not really say much more other than those are the #'s.
We should be on for this coming weekend and do the clap test and will be able to dial this in further.
We are particularly interested in you resolving this because you are using the recommended process for producing a multi-point podcast with no transmission, delay, and internet collision sound damage.
There are a handful of errors that only appear when you compare your work to somebody else’s. It would be grand to know what happened to you.
This is a four-way musical performance which looks like a live, quad-split Skype, but is actually three of them shipping independent video files to Josh (in the lower left). This is the video version of what you’re doing.
Audacity does not have an INFO service. It’s a feature request. If you ask Audacity for information about the sound file it just opened up, it will give you a blank look and shrug its shoulders. It’s up to other software to figure it out.
Right click on a Mac > Get Info and it will give you a whole page of everything you ever wanted to know about that file.
This WAV file was created on a Thursday, January 2nd, 2020 at 2:30PM (1430). It was an unusually balmy afternoon with temperatures in the 60s (15C) and moderate humidity. There was a promise of rain later.
Put a file to be inspected on your desktop. Run your browser. Type or paste this in. This is FireFox.
44100 and 48000 aren’t the only two. Those are just the two most likely to be used in normal day-to-day home operation. Higher numbers are used in studios and lower numbers are used in utility jobs such as music on-hold and Voice Response.