Increasing volume through multiple amplify and normalize effects

[snowbird]

My normal method for increasing the sound output/volume of a track is to first manually "amplify -3" any of the outliers, amplify the whole track as much as possible without clipping, and then start over with reducing the peaks manually. I would usually only do this to about 5 or 10 peaks before I find the whole thing can be amplified enough without clipping.

I have recently started doing something similar except with normalizing the whole track after reducing the outliers.

My questions...

1. Is this a stupid method and I'm missing out on a better way to do this?
2. Does repeating the amplify/normalize effect somehow degrade the track (for example as if I was saving an mp3 and repeatedly opening it for editing)?
3. Is there anything wrong with using amplify and normalize on the same track?

I admit that I don't really understand the difference between amplify and normalize, other than that normalize increases the output of a track and also increases the volume of the quiet parts more. But if that's the case I don't understand why repeatedly normalizing the same track doesn't eventually make everything the same volume.

Thanks!

[kozikowski]

You might like RMS Normalize rather than "straight" Normalize.

viewtopic.php?p=325078#p325078

This was designed to satisfy the audiobook requirement of maintaining a stiff standard for volume (roughly RMS) which normal Normalize won't do since both it and Amplify work on sound peaks, not volume. Sound peaks do not correlate well with loudness.

Follow that with the soft Limiter to tame peaks and avoid overload (or audible problems) and you're done.

Screen Shot 2019-07-25 at 2.16.32.png (18.39 KiB) Viewed 481 times

Please note that sound inside Audacity does not overload in spite of what View > Show Clipping tells you. So it's perfectly valid to have a clipping warning between RMS Normalize and Limiter (Soft). Just don't forget to apply both.

That's not the full audiobook protocol. The complete suite has a rumble filter before everything because many home microphones have low pitch trash that can throw off filters, effects and corrections and most voices don't miss suppression of very low pitch sounds. This is the full suite if that's what you're doing.

https://wiki.audacityteam.org/wiki/Audiobook_Mastering

Feel free to mess with the settings and see what the tools do. Please note there are no compressors in the suite even though it would seem this is a natural place for them. Soft Limiter is a version of compressor in this application.

Koz

[kozikowski]

There is another variation on this. Chris' Compressor was designed to mimic broadcast compressors and processors so Chris could listen to opera in the car.

https://theaudacitytopodcast.com/chriss ... -audacity/

I change the first number, Compress Ratio from the default 0.5 to 0.77. Mixed shows and performances (musical podcasts) come out nearly identical to the local radio station.

It's completely self-contained. Apply the effect and walk away.

Koz

[DVDdoug]

1. Is this a stupid method and I'm missing out on a better way to do this?

If you get the results you want it's not stupid! But selecting part of the file and amplifying or normalizing can create sudden, unnatural sounding jumps in volume.

Making volume adjustments like this manually by-ear if often the best way. But it's usually better to use the Envelope Tool to fade-up and/or fade-down so there are no sudden jumps. (That's more time consuming).

2. Does repeating the amplify/normalize effect somehow degrade the track

No. Audacity uses floating-point internally which is very precise and for all practical purposes has no upper or lower limit. If you save-as WAV you can get rounding errors but you can probably do it 100 times without audible damage (as long as the adjustments are not too extreme). Volume adjustments are not considered lossy, although in some cases they are mathematically imperfect and not perfectly reversable.

(for example as if I was saving an mp3 and repeatedly opening it for editing)?

Right. Every time you export to MP3, you go through another generation of lossy compression and the damage does accumulate. If you are going to save the file and them come-back and do more edits, you can save the Audacity project (which retains floating point), or export to WAV. Floating-point WAV is "best", but 16-bit or 24-bit WAV is OK too.

If you want an MP3, export ONCE as the last step. (If the original was MP3, just try to minimize the number of times it's compressed/decompressed/compressed.

3. Is there anything wrong with using amplify and normalize on the same track?

No,

I admit that I don't really understand the difference between amplify and normalize

They are mathematically the same. Amplify has an "extra feature" that allows you to normalize to a desired peak level, and this is exactly what Normalize does. And, the Normalize effect offers a couple of "extra features" such as adjusting the left & right channels independently and removing DC offset.

other than that normalize increases the output of a track and also increases the volume of the quiet parts more

No, they both make ONE adjustment to the whole file (or whole selection). Normalizing to 0dB does the same thing as the Default Amplify setting. And, running either one twice with those settings does nothing the 2nd time.

Of course if you amplify (attenuate) BY -3dB (not TO -3dB) twice you have changed the volume by -6dB.

[snowbird]

Thank you!! That is exactly what I was after. You even manged to answer some questions I didn't even know I had.

Screen Shot 2019-07-25 at 2.16.32.png
Koz

I'm just starting to use macros... Could those two commands with those setting be put into a single macro. I'm assuming they can since it sounds like there is really nothing to be done between them.

It's not for audio books exactly, but close enough.

Of course if you amplify (attenuate) BY -3dB (not TO -3dB) twice you have changed the volume by -6dB.

Thank you both for taking the time to write such detailed explanations. The support on this forum is amazing.

[snowbird]

There is another variation on this. Chris' Compressor was designed to mimic broadcast compressors and processors so Chris could listen to opera in the car.

I've been using this effect with great success. Thank you so much.

I have one problem, though. If there is a long (relatively) silent part, it gets amplified really high. It even effects the parts that are not silent on either side of the silence. And any silence at the beginning or end of the selection is also amplified.

I work around this by making sure to paste any bit of sound at the beginning and end, and to remove any large silent sections. But is there a better solution?

[kozikowski]

Chris doesn't like falling off the beginning or end of the show, so pasting sacrificial "stuff" on the ends and then cutting it off later solves that. It's not likely to be fixed because Chris reached end-of-life.

Like a broadcast compressor, if the show goes silent for too long, it will think there's something wrong and go "diving for sound." Very few entertainment shows have long stretches of silence, so you just have to deal with it.

But is there a better solution?

What's "better?" I don't think we ever found out what the shows were. What are they? If you have special needs, then the generic entertainment tools aren't going to work for you. Describe what you're doing in detail rather than a list of tools that don't work right.

Koz

[snowbird]

But is there a better solution?
What's "better?" I don't think we ever found out what the shows were. What are they? If you have special needs, then the generic entertainment tools aren't going to work for you. Describe what you're doing in detail rather than a list of tools that don't work right.

Sorry, I wasn't trying to be critical. I just thought that perhaps there was a setting I didn't understand or just something else I was missing. (I actually don't understand any of the settings, but I'm just learning as I go.) I'm really very happy with the effect and appreciate the person who created it as well as you for telling me about it.

I'm just editing people talking. We are amateurs, so our microphone skills aren't that great. I'm just trying to get a suitable even volume and except for the noted issue, this effect is great. The show I am doing is made up of segments that I edit individually and then piece together. The actual final product doesn't contain large gaps. But this particular effect does seem to really boos1 the volume at the beginning and the end in a way that I didn't understand the usefulness of. I'm happy to just paste something in before I apply it, though.

Thanks for your help!