Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Speed-correcting vintage analog sources (Read 3280 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Speed-correcting vintage analog sources

I've got two sources for a fantastic show by one of my favorite bands that I want to blend into a "matrix" recording.  They are both in digital format now and I don't have access to the analog masters.  They both run at different speeds and I think that neither is absolutely correct.

My question concerns the best way to correct and match the pitch of both recordings before mixing them.

I will probably use Steinberg WaveLab for the task.  My proposed method is to first isolate a segment of the recording where I know what the pitch 'should' be.  Say, a guitar intro note that I know is an A.  Then I'll select that note in WaveLab and run the pitch analysis to see what the recorded pitch is.  I'll calculate the error as a ratio and use that ratio to correct the speed of the whole recording (assuming the speed is constant throughout).  I will do this for each source and then mix them together.

Is there a better or more elegant way to do this?  Feel free to suggest software, etc, I have access to some nice things through my friend's studio.  I see some pitfalls with my method, perhaps the guitar was out of tune on that song, or the note was bent, etc...

Speed-correcting vintage analog sources

Reply #1 –
If you want them to stay in synch throughout the recording you have to adjust the speed exactly, far more accurately than you could measure the pitch of one note. I would suggest comparing the times of specific events (like a snare hit) near the beginning and end of the show (this will give you the "relative" speed difference). I'm not sure of a good way to determine the "absolute" speeds, but one place to start would be if you could find some of that buzz that often comes from guitar amps. This will be a multiple of 60 Hz and is probably much more accurate than guitar notes (and you can sample for a longer duration).

I'm not sure what you mean by making a "matrix" recording, but if both (or either) of the sources really were analog you will probably run into a synch problem. Even if you exactly match a point at the beginning and the end, the variation in the middle could be significant, and even from second to second there could be quite a bit of mismatch. There may be software out there that could lock the recordings together and resample over the whole length, but I have never heard of anything like that.

Speed-correcting vintage analog sources

Reply #2 –
Quote
I'm not sure what you mean by making a "matrix" recording....[a href="index.php?act=findpost&pid=235353"][{POST_SNAPBACK}][/a]

In (legitimate) audio trading circles, a "matrix" generally refers to the mix of soundboard and audience sources, intended to provide more of an as-you-would-have-felt-it-if-you'd-been-sitting-right-there ambience, and they can be fun to work with. With modern, all-digital recordings synch issues aren't generally a problem, but I remember a few analog shows that were "fun" to mix, in a different sense of the word.

  One I fondly recall working on was the final concert of a prominent British-blues band. Both soundboard and audience recordings survive in relatively high quality, but to get an accurate synch I first broke the entire (!) soundboard into a few hundred segments - some as small as 1/3-second, depending on the apparent pitch lag - and manually identified matching points from the audience source before adjusting each segment to length. And it took a long time.

    - M.

Speed-correcting vintage analog sources

Reply #3 –
Why not just record audience mics and the soundboard mix to a single 4(or more)-track tape?

Speed-correcting vintage analog sources

Reply #4 –
M is on the money.  It's just what "they" call a blended audience and line level source.

Good points as far as getting the recordings synched to each other.  Sharp wave peaks will make a nice reference point and then I can calculate the relative speed to each other.  I will try to find some 60Hz hum in there and try that.  As far as getting the absolute pitch of the thing correct, my initial efforst at freqency analysing a known decaying note haven't worked so well.  Each channel is a different pitch!  Ack.

Speed-correcting vintage analog sources

Reply #5 –
Quote
Why not just record audience mics and the soundboard mix to a single 4(or more)-track tape?
[a href="index.php?act=findpost&pid=235391"][{POST_SNAPBACK}][/a]


If I had a time machine I would!

Speed-correcting vintage analog sources

Reply #6 –
There's a recording from the 1920s-1930s (I wish I could find the link) where a record company were comparing two new electrical recording systems (microphone and disc cutter). They set them both up and recorded the same jazz performance.

Both versions survive, and by syncing them up, a kind of stereo version has been created. However, to my ears, most of the "stereo" effect is purely due to the two recordings being not quite in sync!

As far as I know, there's no automatic way of syncing up two recordings like this. I've tried the "do it manually" process suggested by M, and it's too painfully slow to contemplate.

In situations where both sources are digital, or recorded onto video tape, it's much easier (though still not perfect). What's more, some uses of the sources require better synchronisation than others. Using them for two stereo channels requires accuracy of better than 1 sample. Adding some audience ambience from a very different sounding tape of the same event can still work OK even where the sync is a few ms off.

Cheers,
David.

Speed-correcting vintage analog sources

Reply #7 –
In order to correct the speed, multiply the sample rate by the correction factor, resample to the new sample rate with maximum quality, then reset the sample rate to the original value without processing the data. It should give you a much better quality than the standard speed correction effect.

Speed-correcting vintage analog sources

Reply #8 –
I appreciate the suggestions... I've had a go of it so far and it's miserable.  Working with just one song to start, I found a beginning event and an ending event with sharp waveforms to isolate the same segment from each recording.  Then I speed/pitch corrected one of them (Pio I will try your suggestion too), and lined them up.  They lock in at beginning and end but wander both ways in between!  Enough to be noticeable.  This could be a nightmare project, especially since the whole show is around 70 minutes long.  My real goal was to make a 5.1 mix of this.  Seems like a pipe dream now!
 

Speed-correcting vintage analog sources

Reply #9 –
Quote
There's a recording from the 1920s-1930s (I wish I could find the link) where a record company were comparing two new electrical recording systems (microphone and disc cutter). They set them both up and recorded the same jazz performance.

Both versions survive, and by syncing them up, a kind of stereo version has been created. However, to my ears, most of the "stereo" effect is purely due to the two recordings being not quite in sync!



Something a bit similar was done for th3 1956 Ellington at Newport recording for its most recent CD release.