Quality music recording corrects intonation?

I suppose this could be a CS thread, but I’m starting it in GQ instead because it’s more technical than artistic.

Years ago I used to record my conducting teacher’s orchestra. While they sounded quite professional to me live, the recordings always sounded horrible. I’ve always chalked this up to either my ears being more objective with the recordings, or my microphone being crap. At any rate, the recordings sounded far worse than any commercial recordings of top-notch orchestras I had. The exceptions were some very old recordings, such as from the 1920s. These sounded pretty bad, too.

Fast forward to the past decade. My own orchestra was recorded pretty extensively. The recordings sounded pretty faithful, but I was surprised at how sloppy and out of tune the music was. I got a CD of one performance and couldn’t listen to more than 10 minutes of it. It was the sloppiest crap I ever had the misfortune of listening to.

Until last Thursday. I finally got a DVD of the same performance. I was pleasantly surprised at how tight we sounded. Everything was dead on. Everything was in tune.

So what happened? Is there some sort of remixing process that cleans up the crap? Or is there some element in the original recording process that exagerates mistakes, and can later be corrected?

How did we get so much better?

There is some kind of post-processing software now that can pull out-of-tune notes into the nearest correct intonation. Not sure how it work or if it works on multiple instruments recorded on the same track. I heard a singer talking about this in a radio interview and she hated the way it sounds when used on singers (the interviewer was surprised that she could tell, but she could) and refused to use it herself.

Here is one example, though it doesn’t sound like it would work in the case you described:

Product Description

Hailed as a “Holy Grail of recording” by Recording magazine, Auto-Tune is a multi-platform plug-in that corrects intonation problems in vocals or solo instruments, in real time, without distortion or artifacts, while preserving all of the expressive nuance of the original performance š with audio quality so pristine that the only difference between what goes in and what comes out is the intonation.

That doesn’t sound like it. I shouldn’t say that on the DVD we sounded perfect, just far better. Egregious intonation mistakes were still there, just not so, um, abrasive.

And yeah, it sounds like that product wouldn’t handle a live recording of an orchestra plus soloists plus chorus plus dialog plus applause very gracefully.

This can only be applied to a single instrument at a time. You simply cannot fix the sharp 3rd horn part in a Beethoven symphony in post-production.

As for the OP, maybe some good microphones and low noise just gave you the impression that things were better overall. Or, perhaps the orchestra simply played better that time. But I can’t see how any technology can actually make a whole group play with better intonation.

I wonder if there was a different mix used for the DVD audio. If the audio-only recording you have was done using multiple microphones over different sections of the orchestra, you’d be much more able to here individual mistakes and rough patches. Now, if the DVD used those mics in combination with ambient or hall microphones, or only hall mics, there will be a huge amount of difference. Hearing the orchestra only as a composite and not the sum of its individual parts does a lot in the way of “correcting” small errors, as you’d be less likely to hear them in the first place.

A good illustration of this occurs when you play within an orchestra. As I’m sure you know, you can hear what’s going on in your immediate vicinity very well. But some rhythmic approximations or instances of bad tuning that you hear will be covered up by other sounds and won’t be heard at all in the audience.

Just speculatin’.

I don’t know how many mics were used, but I imagine 1 or 2 on the orchestra and 2 or 3 on the cast (this was a light opera). Possibly another ambient mic, but I doubt it. I kind of doubt that more than 3-4 mics were used overall, and I’m quite certain that the DVD and the CD didn’t use separate mic’ing systems. There was only one audio source recorded. (I’m not sure how many tracks. The disks were in stereo, though.)

Both the CD and DVD were the same performance, I’m sure of it. Although we had seven performances, certain stupid mistakes were made only once, and that’s what was recorded.

I’m thinking it has something to do with EQ. The DVD version sounds fuller, not as tinny. And the engineers had a good year longer to work with than than the CD, which was run off in maybe just a few days.