song volume normalization for audio CD burning in Reaper. tutorial #3
There might be lots of reasons why you would like to apply song volume normalization across several items. For now let’s just imagine you like different styles of music and love listening to those in your car. Yet you don’t want to adjust volume every time a track changes from heavy metal to a lullaby.
In this tutorial I’ll show you how to even out levels of your tracks waiting to be written on a CD and introduce you to the process of manual & lossless song volume normalization using a free Digital Audio Workstation – Reaper.
Why song volume normalization is still a necessity?
Since there’s still no official standard for song loudness in the music industry, every album can be mastered and released as loud or as quiet as the artist wants it to. Although most online music listening platforms, radio stations and even music players apply some kind of song volume normalization, there’s still a strong belief from last millennium, that one should make his/her music as loud as possible for it to stand out from the “crowd”.
It’s hard do tell what world do music producers live in, but with the “stand out from the crowd trough loudness” philosophy, they are just ignoring the listeners.
Because listeners just want to enjoy the music without the need to work as a “DJ” every time we’re listening to a mix of songs from different artists.
Two most popular ways of song volume normalization.
Everyone, except people involved in music production know wish. And there are two most common ways to deal with this phenomenon:
- For the listener either to constantly “ride” volume knob…
- …or for the play back medium (music broadcaster) to apply brutal song volume normalization.
While “riding” the volume knob is annoying yet lossless (as in quality of the songs isn’t lost), the second one disfigures music so much (it evening out volume involves something more than simple level adjustment – and most often it does) that even artists themselves are sometimes having hard time recognizing their work on a radio.
Third way to even out song loudness.
I bet, you don’t want to adjust the volume every time a song changes. Also you don’t like mutilation of your favorite music tracks, that’s being done trough song volume normalization processes applied at front-end of services providing you with the music, right?
Let me guide you trough a third way of song normalizing before you burn your next mix of favorite music pieces to an audio CD. And you’ll even be able to choose whether you want it to be lossless or induce mild intervention to have a bit louder, yet relatively close-to-original result.
RMS vs. Peak or Loudness oversimplified.
Human ear perceives loudness as something that is stretched in time, thus we don’t consider one gunshot to be louder than say a Harley Davidson doing circles around us.
Same with music – while peaks (momentary volume jumps) may look same amount loud as of the other song, but something that has consistent loudness over longer period of time will appear louder to us. This is what RMS is in audio. Oversimplified. And this is what we will be looking at in order to get even perceived volume of the songs. Good luck!
TUTORIAL STARTS HERE!!!
Step #0. Download, install and launch Reaper.
As always I’ll assume you already know how to do that and already have done it. If not, head to Reaper’s download site, get the everlasting fully functioning trial, read the manual for installation instructions and once you’re done, hit the icon on your desktop or dashboard to launch it. Great job! The hardest part of this tutorial is already behind. Oh, and if you prefer to use some other DAW – that’s fine since all decent Digital Audio Workstations should do the job.
Step #1. Check project sample rate and pan law.
First let’s set project sample rate by navigating to “File” > “Project Settings…” and choosing 44100Hz in the “sample rate” drop down menu under “Project Settings” tab.
While we’re here, let’s make sure our “Pan Law” is set to +0.0(default setting) under “Advanced” tab.
Step #2. Save the project and keep saving it every few minutes.
Go to “File” and choose “Save Project”. Be sure to check both “Create subdirectory for project” as well as “Move all media into project directory” check boxes. This will ensure all files associated with and imported into your project will end up in the same folder you’ve just created by typing your project name.
Step #3. Import all songs you want to be written onto an audio CD.
Import your songs by just drag and dropping them onto Reaper’s timeline window from wherever they’re located on your machine. Reaper will ask if you want to place all the songs on the same single track. Yes, you do. Reaper also will create an empty track for you and place all your songs on it.
Step #4. Visually locate quietest song in a bunch.
Now you’ll need to locate a song that is the softest in volume even on its loudest parts. Look for waveform that is approximately at the furthest distance from edges of its media item box.
This will be our reference song for volume and we’ll do manual song volume normalization to all other songs so they match loudness of this one.
Step #5. Set master bus meter to indicate RMS level instead of Peak.
Just so we can feel more professional let’s go to master buss’es meter and make it show us RMS levels instead of Peaks so we can have a clue on what’s going on in terms of “volume over longer periods of time (300ms)”.
To do that head to “View” and hit “Mixer”. Then right click anywhere next to the master fader where column of numbers is displayed. Make sure you have all settings as in picture below.
Now we’ll be able to determine what is the RMS reading for the loudest part of our softest song we chose as reference.
Play back a short section of your reference track where the waveform is nearest to the media item’s edge. Memorize RMS meter reading (the one at the bottom of the meter) for that part with accuracy of +/- 1 dB.
Step #6. Reduce volume off louder songs manually to same RMS reading.
Now we’ll need to reduce volume of all louder songs one by one so their loudest parts have about the same RMS reading as our reference song’s. While this method might not be suitable for large amounts of tracks, but as long as we’re talking audio CD, there shouldn’t be more than 20 songs to adjust, right?
To reduce volume of the songs just place your mouse cursor on top edge of chosen song’s media item. Once your traditional “slanted arrow” becomes “two up and down facing arrows separated by a dash” you can left click and hold you mouse to drag the volume controller line down, this way reducing volume of selected media item.
After dragging this line you’ll need to reset the RMS number reading so it refreshes and shows the loudness after you’ve made the change (tiny yellow negative numbers on the bottom of green loudness bars in the picture below – marked with red rectangle). To do that just left click on the numbers indicating RMS level at the bottom of master channel’s volume meter.
Repeat dragging and resetting process until you get about the same RMS level as of your reference song. Then do the same to all other songs.
Step #7. Repeatedly micro-adjust levels of each song “by ear”.
It is possible that although RMS levels of all songs match, you will clearly HEAR that some songs are louder than the others. Reason for this being differences in frequency information (think lows / mids / highs). It is OK now to double check each song and readjust songs you feel need volume boost by 0.5 or 1dB (less is better).
Now once you’ve corrected levels of songs by ear you still might find that some songs sound louder than others despite previously performed dual stage manual song volume normalization. This is again caused by different dominating groups frequencies.
That’s another story and we’d need to dig deeper to solve it. Just so you know what’s going on – lower frequencies – think fat sub-bass line of a dub or dubstep piece – has more energy thus impact on RMS than say a soprano prima donna singing her high pitched la-la-la.
Meaning – low frequencies need to be less loud for us to interpret them being the same volume, than something pitched much higher.
Step #8. If you’re after lossless method – burn audio CD now.
If you don’t want to loose original quality of your songs – that’s it – you’re ready to burn your audio CD.
If you want some help with the process inside Reaper, feel free to jump right to the last paragraph of Step #5 of my tutorial on “how to burn audio CD using Reaper”.
But if you’re feeling techy and are okay with some qualitative sacrifices thanks to your interference by bringing up volume of ALL tracks with help of a maximizer plug-in, please proceed to the next step…
Step #9. Insert a maximizer plug-in on your Master Buss.
If you see that you have reduced your loudest song more than by 6dB, or that louder songs dominate your project and you want everything to be a bit louder, try increasing overall volume of all the songs by inserting “JS: zero crossing maximizer” plug in on your Master Buss.
To do this, click the “FX” button on your master track and in the filter field write “zero crossing maximizer” to locate the plugin. Double click the result upstairs or just select it and hit “OK”.
Step #10. Adjust maximizer’s settings.
First set its ceiling to -0.3dBFS. By doing this we’ll ensure there will be no analog distortion after the data on our audio CD is converted back to analog sound waves during playback.
Try setting threshold to about half the value of your song that has most volume reduction applied. Say if you needed to bring one song down in volume by 7.5 dB a good starting point for maximizer plug-in’s threshold would be -3.6dB to -3.8dB.
Step #11. Listen for audible distortion on loudest parts.
After you’ve adjusted threshold as instructed, play back loudest section of your quietest reference song and listen for audible unpleasant artifacts like distortion. If you hear something you don’t like, back off by 0.5dB (from -3.7dB to -3.2dB).
If you didn’t hear anything ugly at -3.7dB setting, you can try increasing volume by going deeper into negative values of maximizer’s threshold, say by increments of 0.5dB while listening for distortion on highest peaks. Once you start hearing it, back off by 0.5dB and see if distortion disappeared. If not, back off by same amount once again.
Once the audible artifacts of maximization are gone and you’re good to go. You’re welcome to jump straight to the last paragraph of Step #5 of my tutorial on “how to burn audio CD using Reaper” if you require some help with the process.
After you’ve gone trough all the steps of lossless song volume normalization process and burned your CD image, you should now be able to listen trough entire “mix-tape” without need to readjust volume after each song!
Hope you’ve found this song volume normalization tutorial useful. Obviously there are lot’s of ways to even out levels of songs. I wanted to show you one on the easier side and explain some theory behind it. If you have questions – spit ’em out and I’ll do my best to get back to you. Also let me know if you were happy with the results. Thanks for reading. Best of luck!
P.S.: I always appreciate your comments and suggestions on how to make things better!