L/R Channel Imbalance

[kozikowski]

Yes, Normalize does have the DC Remover tools inside. You can apply them without applying anything else. That's not a bad idea. Koz

[waxcylinder]

Obviously if the highest thing in the show is a cat hair pop on the record, you're dead. The cat hair pop is the highest thing in the show -- almost always. Both tools will make the cat hair pop come out exactly perfect. Sorry, perrrrrfect, ignoring the real show.

But Pat is also using Brian Davies' ClickRepair and additional manual editing (I know this from another off-forum discussion) so this should take care of the cat-hair problem - in which case the Normalize effect should do the job nicely - but only if the click removal is done before the Normalize.

Rough workflow would be:
1) Capture/record
2) safety copy WAV exported
3) ClickRepair - and any manual clean-up
4) inter-track gaps clean-up
5) labelling
6) Normalize
7) multiple export of production files

This is my basic workflow - but I use Amplify at step 6 (rather than Normalize) as my set up appears to be inherently balanced.

WC

[steve]

If you wish to adjust the Left Right balance to a different position from that produced by Normalize, use the "Pan" slider on the left of the track. The pan position will be taken into account when you export.

[patcurci]

Koz:

Again, my thanks.

It's not my intention to re-engineer the LP records. I don't have the knowledge, tools, or the skill to be that foolish. I have to assume they did it "right" the first time around. My only goal is to fix the wear and tear on the records; and cure, when possible, minor defects apparent in my equipment. Then, make some CDs.

Thanks to you and the other folks on this Forum (and Audacity), I'm again enjoying my 50+ year old music collection.

Regards, Pat

[kozikowski]

<<<The pan position will be taken into account when you export.>>>

Which is precisely why I don't like using it. It's a Magic effect that sometimes appears and sometimes doesn't, Since neither the meters nor the waveforms change, how do you know when you get it right?

Koz

[steve]

how do you know when you get it right?

There are two little sensors, one each side of your head (they are also useful for securing shades, reading glasses and pencils).
Meter readings are not always reliable.
Test: Create a stereo track with a 440 Hz, 0.5 amplitude sine tone in the left channel and a 440 Hz, 0.5 amplitude square-wave in the right channel.
Adjust the levels by any means so that the left and right channels play at the same volume.

If you really need to see the levels/waveform after adjustment, select the track and from the Tracks menu > Mix and Render.

[kozikowski]

<<<If you really need to see the levels/waveform after adjustment, select the track and from the Tracks menu > Mix and Render.>>>

Two obscure tools enhancing each other. Most people can't hear the trash truck grinding day old pizza boxes outside their window much less trying to hear subtle variations in sound levels. That's a learned experience. You probably don't have $6 USD speakers on your machine, either.

One of the highly respected Systems people announced that he purchased multiple computers in a bulk offer and the vendor threw in speakers free.

Oh. Good.

Koz

[steve]

Stock answer: If you can't hear the difference then it doesn't matter.

Q. "Do I need to put songs onto my iPod at 320kbps or will 128 kbps be OK?!
A. "Can you hear the difference?"

Apart from the name, what's obscure about "Mix and Render"?
Would "Mix-down to track" be less obscure?

[DVDdoug]

<<<1. Normalize both channels separately, so that they both have 0dB peaks.>>>

You can't do that with phonograph record captures. The cat hair pops mess you up and also makes it impossible to be reliable from record to record.

Loud clicks & pops in one channel do NOT invalidate this procedure... The clicks & pops don't contribute significantly to the RMS/average, and the idea is to balance the average levels without introducing clipping...

If you "balance" without normalizing you can end-up with clipping in one channel, or you can end-up with both channels at too-low levels.

It might be simpler to match the average levels first (by reducing the louder channel) and then to normalize the stereo file (maintaining the corrected balance). I believe Audacity uses 32-bit internal processing, so you can safely reduce the volume and boost it again without loosing any resolution.

In either case, it would be best to repair/remove the clicks & pops first. These may be the only clipped (or 0dB normalized) waveforms, and it's best to normalize the music rather than the defects.

. I read the thread about the Wave Stats plugin and am not sure whether there is a "sample limit", or if the newest version of this plugin will crash Audacity.

If you can't get the RMS for both channels, it's kinda hard to make the RMS levels equal!

[billw58]

<<<1. Normalize both channels separately, so that they both have 0dB peaks.>>>

You can't do that with phonograph record captures. The cat hair pops mess you up and also makes it impossible to be reliable from record to record.

Loud clicks & pops in one channel do NOT invalidate this procedure... The clicks & pops don't contribute significantly to the RMS/average, and the idea is to balance the average levels without introducing clipping...

Indeed, but both the Audacity Normalize and Amplify effects first look through the selection to find the highest peak level, and normalize to that. Also note that Koz was responding to a suggestion to normalize for peaks.

Normalizing (or balancing) for average levels may get you closer, but the final test is your ears. Perceived loudness is a complex beast. I'm sure someone can come up with an example of a recording with the vocal mixed to the middle where the vocal would shift if the two channels were "balanced" for equal average levels.

-- Bill