Subject:Normalize Channel Volume?

Posted by: Vocalpoint

Date:11/11/2010 1:16:44 PM

Hello,

Currently spending some time weighing both Soundforge and Wavelab 7 in trying to determine the best environment to use for my vinyl transcriptions. One thing that is crucial to the workflow is ensuring the L/R channel data is properly balanced after an LP side is transferred in.

I am wondering if Forge has any plugin or process that will quickly analyze the two channels...disply the average L/R RMS value and easily let me correct either side so the stereo file has equal weight.

Wavelab has it's Pan Normalizer which is excellent and does this exact thing. If Forge has such a thing...love to know about it....cannot find anything so far.

Cheers!

VP

Subject:RE: Normalize Channel Volume?

Reply by: musicvid10

Date:11/11/2010 1:40:31 PM

Process->Normalize->Scan Levels

You can select L/R channels independently to scan or normalize.

Subject:RE: Normalize Channel Volume?

Reply by: rraud

Date:11/11/2010 2:34:40 PM

In addition "Tools: Statistics" shows Peak, RMS. ect, ect. ect.

For instance:
Left Channel Right Channel
Cursor position (Time & Frames) 00:23:13.30 00:23:13.30
Sample value at cursor (dB) (end of sample) (end of sample)
Minimum sample position (Time & Frames) 00:21:59.26 00:21:59.26
Minimum sample value (dB) -0.786 -0.000
Maximum sample position (Time & Frames) 00:21:18.15 00:20:52.68
Maximum sample value (dB) -0.000 -0.314
RMS level (dB) -18.566 -18.863
Average value (dB) -Inf. -Inf.
Zero crossings (Hz) 1,831.74 1,840.94

Subject:RE: Normalize Channel Volume?

Reply by: Vocalpoint

Date:11/11/2010 2:43:39 PM

I am aware of the that dialog....but I doesn't show me the levels of the L and R independantly. (rraud beat me to it - Tools Stats gives me exactly that stuff) but it would be very cool to see the following in the Normalize dialog:

L: -24.25db
R: -25.25db

So I easily see that the R channel needs a 1 db boost. Right now all I see is a singular RMS value at the bottom left.

Now - are you saying that if the two channels do have a 1db difference and I dial in -23.0 under Normalize Using (Average RMS) and hit apply...this will raise the L (in my example) by 1.25db to equal -23 and will move the R up by 2.00db to get the right to -23?

Or is the singular value that I see when I click Scan levels - a composite average of both channels?

Appreciate your update.

Cheers!

VP

Subject:RE: Normalize Channel Volume?

Reply by: musicvid10

Date:11/11/2010 4:12:41 PM

Once again, by selecting a channel, one can independently scan and / or normalize the selected channel. I believe this is what you are looking for.
Normalize Audio

Subject:RE: Normalize Channel Volume?

Reply by: Vocalpoint

Date:11/11/2010 6:14:29 PM

No actually I need to be able to normalize it in relation to the other channel. In other words - bring instant balance to the entire file.

Seems like I cannot see what the exact RMS reading is for any channel - in the plugin window - without going back to Tools-Stats...but then that window takes all the focus so I cannot have both open at the same time....

Wavelab 7 wins this round...

Thanks for the help.

Cheers!

VP

Subject:RE: Normalize Channel Volume?

Reply by: musicvid10

Date:11/11/2010 7:23:29 PM

Everything you described is easy to do in Sound Forge. Good luck.

Subject:RE: Normalize Channel Volume?

Reply by: Angels

Date:11/12/2010 7:04:19 AM

I don't understand what you're asking: if your transferring vinyl the most important things are a good preamp with proper RIAA EQ, and a great A/D converter. If these are gold, you shouldn't have to re-balance anything. If you're dealing with stereo recordings, they're supposed to be unbalanced: that's what makes them stereo. A lot of older recordings by the Beatles for example are totally unbalanced: vocal on the right, piano on the left, etc... And if you're dealing with mono, then it doesn't matter. In either case it doesn't matter what software you use to record the digital stream.

Subject:RE: Normalize Channel Volume?

Reply by: rraud

Date:11/12/2010 9:16:15 AM

If it's a 'main stream' vinyl LP, I'm sure the mastering engineer and disc-cutting engineer took great pains to ensure proper left-right balance. So as Angels said, "you shouldn't have to re-balance anything."

Subject:RE: Normalize Channel Volume?

Reply by: Vocalpoint

Date:11/12/2010 10:49:18 AM

I don't understand what you're asking: if your transferring vinyl the most important things are a good preamp with proper RIAA EQ, and a great A/D converter. If these are gold, you shouldn't have to re-balance anything. If you're dealing with stereo recordings, they're supposed to be unbalanced: that's what makes them stereo. A lot of older recordings by the Beatles for example are totally unbalanced: vocal on the right, piano on the left, etc... And if you're dealing with mono, then it doesn't matter. In either case it doesn't matter what software you use to record the digital stream.

I suppose I will take this with a grain of salt. And yes - 600 dollar preamp to 1600 dollar A/D converter should suffice. But just becuase I am transcribing a 1978 album doesn't mean it was balanced correctly from a 2010 point of view. True - Revolver by the Beatles has some crazy separation going on - but if I want to "even" that out - I should be able to.

Several of my key steps of my particular method involves ensure each channel is in balance and that each song is normalized to a reasonable RMS level. Since these tracks will play in amongst all the rest of today's overblown, hypercompressed crap - I have to take some steps to at least "semi-modernize" the delivery levels.

But I do agree that messing too much with the levels is probably not a good thing. Will give SF a whirl again as I am liking it's workflow much better than WL7...which is driving me a bit nuts to be honest.

Thanks again.

VP

Subject:RE: Normalize Channel Volume?

Reply by: musicvid10

Date:11/12/2010 4:27:17 PM

As you gain more experience with leveling L/R channels, and programs against each other, you will discover that RMS levels alone have far less to do with loudness than one might expect. It's an ordinary regression that may look good on meters, but is not an accurate predictor of what we hear.

If you really want to get involved in the loudness investigation, do some research on EBU R128, which is the cutting edge of loudness normalization (without defining dynamic range). Here's a primer:
http://tech.ebu.ch/docs/techreview/trev_2010-Q3_loudness_Camerer.pdf

The Nugen Audio VisLM beta meters are a good place to start if you want to experiment on your own material.