Getting Volumes Equal

[grovergpace]

Everyone says it's "ears", the only way to get all the tracks on a project
to be of equal volume. I'll agree that this is a big factor in the process
but I'm still believing that there should be some other helps too.
I record on a Tascam DA38 then mix the tracks through a Mackie
mixer to Adobe Audition 1.5. I use the normalize function. When I
get 10 or 12 songs burned to a CD they all have varying volume levels.
I'm using professional soundtracks and adding one vocal.
Any ideas?

[DT Chris]

After you have completed the mix that you feel you like, you save down into one track and open it up in Edit View.

Here you can do all your 'mastering' of the song before you output it to the format desired.

Read about Audio: Home Recording Studio (Music) / Recording Software

This is the stage where you can clean up the final mix, compress, equalize and make it uniform volume with other tracks in your project.

How to do so in detail is beyond my scope to teach at this point, but Audition has Dynamics Processing ability, as well as Hard Limiting, which will boost the signal a decibel range, while compressing it on the top end so that it doesn't.

I'll let your research more and learn how to do it.

[grovergpace]

I use normalize and the other features in Audition 1.5, but normalize
does not correct the different volumes at all. I do master in the EV, but it's
all trial and error and guess work. I've been using Audition 1.5 and before
that Cool Edit, and so far I've not found a feature that will do this correctly.
They all have an effect on the volumes but nothing levels them equally.

[DT Chris]

Originally Posted by grovergpace

I do master in the EV, but it's
all trial and error and guess work.

Welcome to sound engineering. Investigate, read, learn, experiment to see what happens.

Originally Posted by grovergpace

I've been using Audition 1.5 and before
that Cool Edit, and so far I've not found a feature that will do this correctly.
They all have an effect on the volumes but nothing levels them equally.

Nothing?
Normalize
The Normalize effect amplifies the entire file or selection equally. For example, if the original audio reaches a loud peak of 80% and a quiet low of 20%, normalizing to 100% amplifies the loud peak to 100% and the quiet low to 40%

As the Audition 1.5 Help file displays, you will not get what you want using Normalize.

You need to use Gain/Compression techniques to achieve an overall volume mix for the mixdown track, and get rid of extreme peaks and valleys in the mix.

Check out Effects->Amplitude->Dynamic Processing (which allows you to use several presets you can try out)

Dynamic Processing (Effect/Filter)
The Dynamics Processing effect varies the output level of a waveform based on its input level. You can use this effect to limit or compress dynamic range, producing a consistent level of perceived loudness. You can also expand or gate the signal so that low-level signals are reduced in level, increasing perceived dynamic range, or eliminating signals with noise that falls below a specific threshold.

Also Effects->Amplitude->Hard Limiting which you can use after you have generally compressed and compensated for the mix volume. This feature allows you to boost the overall volume and compresses it hard at the end of a project to get you as loud as possible without clipping.

Hard Limiting
The Hard Limiting effect drastically attenuates audio that rises above a defined threshold, leaving audio below the threshold unaffected. This effect is particularly useful for increasing perceived volume because you can amplify audio beyond the digital maximum, 0 dbFS, and you can lower areas that would otherwise be clipped.

I am not an expert on these applications, I'm pretty much in the learning stage myself.



Above image shows the state of the wave file.

1) No process shows the mixdown before any processing

2) Some Processing shows my inexperienced application of gain/compression filters to eliminate some of the peaks and valleys in the mix

3) Final Master Limit would have me boosting the meat of the mix up to -1 db or less in the mix, with some compression, and without clipping out too much.

The point I am making is, through technique, your final waveform will show your levels. I still have a lot of peaks and valleys due to my inexperience, but in general, I can make all my recordings the same volume level by boosting the DB and compressing appropriately.

[grovergpace]

OK, thanks
I'm going to print your reply and study your suggestions

[earwax]

Never normalize. Ever. It has no positive function.

In Audition, after you have reduced the song to a stereo track, in the Edit window, click on Window, Amplitude Statistics. Look at "average RMS Power" That will tell you the overall average loudness. Have a target in mind like -14 or whatever you like.

Then close that window and click "effects, amplitude, hard limiting" and in the box that says "boost input by" but the number of dbs you are short of your target, like if your song is at -20, put in "6". Run that then check amplitude statistics again. If you do this with all your songs, they will sound somewhat equal, but use your ears. A sparce tune will sound different at -14 than a dense one.

[grovergpace]

Thanks, I'll try that now.
Why do you say never normalize?

[earwax]

I highly recommend Bob Katz's book "Mastering Audio." He covers this and many other important points in detail. Briefly, normalizing does nothing to improve the sound, it only degrades it and adds distortion. It is not the way to increase loudness because it works on the basis of peak levels while the perception of loudness is a product of RMS levels.

[earwax]

Originally Posted by DT Chris

I still have a lot of peaks and valleys due to my inexperience, ...

DT Chris has provided a very good and detail explanation here. I only take issue with his self criticism (quoted above).

What's wrong with peaks and valleys. That is dynamic range. Most of the bad sounding records I hear have a waveform that looks like a bunch of mesas crammed together, all peak no valley. Look at the wave forms of music you like and I'll bet you see peaks and valleys.

Now, some of the mixes you hear that are all close to peak but still sound good didn't get there by boosting the overall level. They got there through meticulously filling the spaces with material from different tracks, like bass goes down, organ goes up, like a bunch of pistons all firing at different times. It can be very tedious to mix that way, but that's one of the reasons some of the big hit records take hours and hours to mix. You can't short cut that process by just "maximizing" the mix.

[DT Chris]

Originally Posted by earwax

In Audition, after you have reduced the song to a stereo track, in the Edit window, click on Window, Amplitude Statistics. Look at "average RMS Power" That will tell you the overall average loudness. Have a target in mind like -14 or whatever you like.

Then close that window and click "effects, amplitude, hard limiting" and in the box that says "boost input by" but the number of dbs you are short of your target, like if your song is at -20, put in "6". Run that then check amplitude statistics again. If you do this with all your songs, they will sound somewhat equal, but use your ears. A sparse tune will sound different at -14 than a dense one.

Interesting. I've never tried that.

Originally Posted by earwax

DT Chris has provided a very good and detail explanation here. I only take issue with his self criticism (quoted above).

What's wrong with peaks and valleys. That is dynamic range. Most of the bad sounding records I hear have a waveform that looks like a bunch of mesas crammed together, all peak no valley. Look at the wave forms of music you like and I'll bet you see peaks and valleys.

Now, some of the mixes you hear that are all close to peak but still sound good didn't get there by boosting the overall level. They got there through meticulously filling the spaces with material from different tracks, like bass goes down, organ goes up, like a bunch of pistons all firing at different times. It can be very tedious to mix that way, but that's one of the reasons some of the big hit records take hours and hours to mix. You can't short cut that process by just "maximizing" the mix.

That is true. Thanks for that