Amplification

I have several CDs of a particular artist, many of which have been remastered and have a much louder sound than the non-remastered ones. What I would like to do is use Audacity to increase the volume of the tracks of these non-remastered CDs to match, as closely thereto as possible, the volume of the tracks from the remastered ones. I would assume I use the Amplify function to do this. If so, how would I make sure I set it right so that it matches the sound level of the remastered tracks? If Amplify is not the right way to do what I want, provided it can be done at all, how do I go about doing it? Thanks!

Steve

The Amplify effect won’t help. The tracks have been made louder by applying dynamic range compression.
See: http://en.wikipedia.org/wiki/Dynamic_range_compression
and http://en.wikipedia.org/wiki/Loudness_war

You can use Audacity’s Compressor effect to accomplish this. It will take some understanding of how the effect works, then some fiddling with the controls to get the effect you want. There is no magic formula for matching the compression of one track to another.

– Bill

Bill-

Thank you very much for your response! I checked out both articles and found them quite interesting. So I guess there’s no way to determine what level of compression the remastered tracks underwent and then just compress the non-remastered tracks to the same level? I haven’t had a chance to mess with the compression feature yet in Audacity.

What I am thinking about doing is importing two versions of the same track-one that’s remastered and one that isn’t. Fortunately, I kept the non-remastered CDs and gave them to my son. Then I could use the compression feature on the non-remastered track until the loudness of the non-remastered track was as close to the loudness of the remastered track as my ears could distinguish and then use that compression setting for the albums that have not been remastered. Does that sound like the way to go?

Once again Bill, thank you so much for your response!

Steve

Steve:
Have a look at this thread: https://forum.audacityteam.org/t/replaygain/22589/1
Steve’s ReplayGain plugin may help you with your project.
If you find it useful, or have any suggestions, please leave feedback in that thread.

– Bill

Bill:

I’ve used the ReplayGain to determine what the old db level was on the non-remastered album that came out two years before the album which I’m trying to “remaster”. I then used ReplayGain to determine what the new db level was on each track on the remastered version of the album that came out two years before the album which I’m trying to “remaster”. Every track was increased by 3. For example, Track one on the non-remastered album was at -5.5db. On the new one, that track’s level is -8.5db.

So, now I’m ready to compress. I opened up the compressor and you probably could’ve heard my eyes audibly blinked. I have no clue how to adjust the settings to result in a +3 db change. Any guidance?

Steve

You probably can’t work out the exact amount of gain that will be produced, but the following settings should give just a little over +3 dB boost:
Threshold: -12
Ratio: 2:1
Compress based on Peaks.

Leave the other settings at their defaults.

The manual page: http://manual.audacityteam.org/manual/help/manual/man/compressor.html

When I used the ReplayGain, it produced a total of 17 different db levels for the track to which i analyzed. Furthermore, at about the halfway point, ReplayGain quits. In the most recent case, 7 minutes had elapsed and there was 7 minutes to go. Is this normal?

When I applied the ReplayGain the first time, after the first db reading came up, I quit it. I’m guessing I should let it run completely for every track on both the non-remastered album tracks and the remastered album tracks on the album that came out two years prior to the one I’m wanting to remaster?

Steve

Please note that the plug-in in the “ReplayGain” topic here: ReplayGain plug-in is not “ReplayGain”.
As it says in the first post of that topic:
“This plug-in is an implementation of ReplayGain that calculates the Track Gain of the selected audio track.”

As explained here: http://wiki.hydrogenaudio.org/index.php?title=ReplayGain ReplayGain is a technique invented to achieve the same perceived playback loudness of audio files using a special algorithm (formula) to calculate perceived loudness.

Also please note that the plug-in is an experimental “work in progress”, hence there is no real documentation other than what is written in that topic.

7 minutes processing time seems extremely long for processing a track. What are you trying to do? How long are the tracks? What sort of computer are you using?

I had imported 17 tracks and even though I only had one selected, it was analyzing all of them, hence the 17 different results. I deleted all of the tracks out and reanalyzed them one by one.

I used the compressor some and on the first two tracks, to get the nonremastered track equal in db to the remastered track, I had to drop the threshold down to -3 and -2 respectively. However, on the next three tracks, when trying to go from -6.2 to -9.2, -4.3 to -7.4, and -4.3 to -7.3, I could only get to -7.8, -6.7, and -5.8 respectively, after dropping the threshold all the way down to -1. While those results would be an improvement, I’d really like to get as close as possible to the desired level. Any other adjustment I can make to get it to get those db values where I want them?

That should not happen. I don’t see how it can happen - Nyquist plug-ins can’t do that, they can only process tracks that are selected.
By default, if you don’t have a selection in Audacity and you try to apply an effect, Audacity will automatically select all audio tracks (see here: “Select all audio in project, if none selected:”. Perhaps that is what happened.

Assuming that you are using the current version of the plug-in from this topic ReplayGain plug-in (you have not confirmed or denied this), the way that the plug-in is intended to be used is that the effect is applied with the same setting to all tracks that you wish to sound the same loudness. This may be done one track at a time, multiple tracks selected at the same time, or as a “Chain” (batch) process. It is important that the same setting is used on all tracks. The “Normalize” option will then amplify each track to about the same loudness.

After applying the effect, you may use the “Amplify” effect on the tracks, but if you do so the same Amplify amount must be applied to every track.

If you use any compression effect, that should be done before applying the equal loudness plug-in effect and the settings that you use must be the same as you used for other tracks.

Steve:

laugh Now I’m very confused. I was told initially that Amplify wouldn’t help in accomplishing what I wanted to do-that I needed to use the dynamic range compression feature. Now you’re saying that between using ReplayGain and Amplify, what I’m wanting to accomplish will work.

As to which version of ReplayGain I’m using, I’m unsure.

Here’s what I want to do-I’m making a compilation of all the albums by KISS. Every album up through Crazy Nights (1987) was remastered. Hot in the Shade, released in 1989 was not. When listening to my KISS compilation on shuffle, it is quite noticeable that Hot in the Shade was not remastered as the sound goes WAY down. Since it doesn’t appear as though this album will be remastered and released anytime soon, the next best thing I can think of doing is getting it so that the sound level of those tracks are of the same approximate loudness of the rest of the remastered catalogue.

According to the ReplayGain analysis I did on the non-remastered and remastered versions of Crazy Nights, these are the values I got:

Crazy, Crazy Nights -5.3 -8.5
I’ll Fight Hell to Hold You -5.7 -8.3
Bang Bang You -4.5 -7.4
No No No -6.2 -9.2
Hell of High Water -4.3 -7.4
My Way -4.3 -7.3
When Your Walls Come Down -5.1 -8.1
Reason to Live -4.3 -7.4
Good Girl Gone Bad -3.9 -6.9
Turn on the Night -5.2 -8.2
Thief in the Night -5.9 -9.0

As you can see, the approximate difference in the db level is about 3. These are the db values I got for the Hot in the Shade tracks when I analyzed the tracks using ReplayGain:

Rise to It -4.4
King of Hearts -5.7
The Street Giveth and the Street Taketh Away -5.3
You Love Me To Hate You -5.5
Somewhere Between Heaven and Hell -4.5
Little Caesar -3.6
Boomerang -3.7
Betrayed -2.7
Hide Your Heart -5.3
Prisoner of Love -4.5
Read My Body -5.3
Love’s a Slap in the Face -4.8
Forever -4.4
Silver Spoon -4.6
Cadillac Dreams -3.5

To test the Dynamic Range Compressor, I used it as suggested on the first six tracks of the nonremastered Crazy Nights album. On Crazy, Crazy Nights, I set the Threshold on 3 and got a result of -8.5 or -8.4. To get an -8.3 on I’ll Fight Hell to Hold You, I had to drop the Threshold to 2. I failed to write down the result for Bang Bang You. On No No No, I dropped the Threshold all the way down to 1 and could get the db only up to -7.8, far away from the -9.2db value ReplayGain calculated for the remastered version of the track. On Hell or High Water, I once again set the Threshold down to 1 and only got the db level up to -6.7, whereas the db level of the remastered version of this track according to ReplayGain analysis was -7.4. On the last track I tested, My Way, I once again set the Threshold as far to the right as I could and came out with a db level of only -5.8 as opposed to the -7.3 db level on the remastered version of the track as calculated by ReplayGain.

All I’m really wanting to do here is make the tracks from Hot in the Shade sound as loud as the tracks on the rest of the KISS albums. And I’d like to do it in the simplest way possible. If I can accomplish this either way-using Dynamic Range Compression or ReplayGain/Amplify, that’s fine with me. It would seem ReplayGain/Amplify might not be the way to go as that would require importing all of the songs to get the sound level between them all equal. That’s a LOT of songs. Seems like using the DRC would be easier but so far, I haven’t been able to get the values on some of the songs to the level they need to be and I don’t know what other adjustments apart from the Threshold I can make to effectuate the result I want. So there ya have it.:

P.S. Please no judging of my love for KISS. Yes, I know they suck. I love them anyway.

Steve

The hardest part for us is often trying to work out exactly what is trying to be achieved.

You’re talking about the plug-in from This Page?
If so, to make sure that you have the correct version, delete your current copy and here’s a direct link to the version that you should have: http://forum.audacityteam.org/download/file.php?id=4685

That’s interesting and considering the “Loudness War” it’s not at all unexpected.
It’s good to see that the ReplayGain analysis agrees. The analysis indicates that the non-remastered version is about -3 dB below the remastered version.

The ReplayGain algorithm is not very accurate, even when properly implemented. The Nyquist plug-in version is an approximation of the ReplayGain algorithm, so may be even less accurate, but it does give a clear indication in this case that the non-remastered version is quieter than the remastered version.

There are several approaches that you could take.

• If your media player supports ReplayGain (the “real” ReplayGain, not the Nyquist plug-in effect) then you could use that to analyse all of your albums and adjust the playback level automatically. The advantage of this method is that it does not change the actual recording at all, it just automatically adjusts the playback level to compensate for exceptionally loud or quiet recordings. This option is only relevant if your media player supports ReplayGain (or “SoundCheck”, which is the Apple version).
• You could use the Amplify effect to lower the level of all of the remastered albums. This is easy to do because you are not needing to mess around with compressors and stuff, but if you have a lot of “loud” (remastered) albums it may be too much work to be practical.
• You could increase the loudness of the “quiet” (not-remastered) album. If it is just this one album that stands out as being too quiet, then this will be a viable option. Make sure that you keep your original recording safe in case we mess up with this. Let’s try this option.

The simplest way to make a track louder is to use the Amplify effect, however that is probably not going to work in this case because I expect that the peak level of the album is already very close to 0 dB (0 dB is the highest that you can go. Attempting to go higher will just cause “clipping” distortion. In order to be able to go “louder” without clipping, we will need to “squash down” the peaks so that there is room to amplify. This is one of the things that they commonly do in studios when they remaster old recordings, and is why many remastered recording sound louder than the originals.

I presume that the relative loudness of the different tracks on the album sounds about right?

Although the loudness analysis from the plug-in shows “No No No” to be the loudest track on the album (-6.2) and “Good Girl Gone Bad” to be the quietest (-3.9), I presume that in context they sound about right? If you agree, then what we need to do is to process the entire album with the same settings and increase the “loudness” of the entire album by about +3 dB.

Does this sound like what you want? Do I understand correctly what the job is? If so, then we can proceed, hopefully without further confusion

I presume that the relative loudness of the different tracks on the album sounds about right?

Yes.

Although the > loudness analysis > from the plug-in shows “No No No” to be the loudest track on the album (-6.2) and “Good Girl Gone Bad” to be the quietest (-3.9), I presume that in context they sound about right? If you agree, then what we need to do is to process the entire album with the same settings and increase the “loudness” of the entire album by about +3 dB.

Does this sound like what you want? Do I understand correctly what the job is? If so, then we can proceed, hopefully without further confusion >

Exactly what I want to do! I went ahead and deleted the current copy of ReplayGain and downloaded and dragged and dropped that version to the plug-in folder in Audacity.

So, what’s next?

I presume that the “Hot in the Shade” album is already in digital format on your computer. If not, you will need to do that first, with a strong preference for a lossless format such as WAV.

MP3 encoding reduces the sound quality, and processing an MP3 than re-encoding to MP3 causes additional damage. It’s the MP3 encoding that does the damage - each time you re-encode the sound quality deteriorates. Converting back to WAV does not make it better, the damage is permanent and cannot be fixed, so minimise the damage and save MP3 encoding to the very last step (if required at all) so that MP3 encoding happens as few times as possible.


Download and install the “Limiter” effect from here (direct link): http://wiki.audacityteam.org/w/images/c/c0/Limiter.zip
To install the plug-in, extract the contents of the ZIP file and place in the Audacity plug-ins folder, then restart audacity. It should now appear in the Effect menu as “Limiter…”

You will also need several GB of free disk space to do this (most of which will return after you have finished, but a lot of temporary working space is required).
Import ALL of the tracks into one Audacity project (if the recording is one continuous file then just import that. If it is one file per track, import all of the tracks and they will appear one below another. If there are multiple tracks, do not press Play without setting one track to “solo”.

Save the project to an internal hard drive that has a lot of free space (at least 10% of the drive size + 4 GB).

Select all (Ctrl+A).

Effect menu > Amplify.
Take notice of the “Amplification (dB)” amount.

• If it shows a negative amount, apply the Amplify effect (it will take a while to process)
• If it shows a positive amount more than 0.2 dB, apply the effect (strictly we should subtract this number from the “loudness increase” that we require (+ 3 dB) but we probably don’t need to bother with that unless it is greater than 1.0).
• If it shows 0.0 or 0.1, cancel the effect. The peak amplitude is high enough to go onto the next step.

If you applied the Amplify effect, Save the project.

With all tracks selected, apply the Limiter effect with the default settings:
Limit to (dB): -3 dB
Hold (ms): 10 ms
Apply Make-up Gain: Yes

Wait… wait… wait… this effect will be slow…

When it has completed (hopefully without crashing):
If you have been working on one continuous track of the whole album, you can just export it as one WAV file, or you may want to split it into tracks (see here: http://manual.audacityteam.org/manual/help/manual/man/splitting_a_recording_into_separate_tracks.html)

If you were working on multiple tracks, ensure that none of the tracks are soloed or muted, and use “File > Export multiple” to export each track as a separate file (see here: http://manual.audacityteam.org/manual/help/manual/man/export_multiple.html)