Jump to content

Normalize tracks to eachother?


sadicus

Recommended Posts

I know what it means to 'normalize' a track, but I'm not sure what you mean by normalizing tracks ' to each other'.  A normalized track is one where the loudest peak is  moved to exactly 0 dB (or some other specified value) and the rest of the signal is scaled according to that. Thus, if you normalized every track to the same value I suppose you could say they were normalized to each other.  Somehow, I don't think that's what you mean.

If you want each track to have the same loudness then you'd need to use a loudness meter  (say like the Waves WLM Plus meter) to help you adjust the loudness  to some LUFS (Loudness Units Full Scale) value. I usually target my full mixes to -14 LUFS.

 

Link to comment
Share on other sites

I think that normalize is not  the real tool to use.

As you can read in the "Mastering" topic in this forum, this kind of result is obtained  with mastering.

The final step of mastering in fact is the use of limiter.

By limiting each song at the same volume level (brick wall), the level of each song will be identical.

Link to comment
Share on other sites

I know exactly what you are trying to do. You record vocals and now you have possibly dozens of different clips and because they were done a little bit  differently the levels of each track are all over the map. It's not always just levels, however, if the singer moves even a little bit around you might have a completely different tone due to proximity effect. That also has a bearing on the loudness and as the Sailor said this is not about peaks, it's about loudness RMS or LUFs. 

Bottom line is, because this happened you have a lot of work to do, if you want it perfect! 

Also like everything you do in a DAW, there's dozens of ways to accomplish this "repair" 

Best to download a LUFS meter, I highly recommend Youlean  https://youlean.co/youlean-loudness-meter/    Basic version is free and I highly recommend buying the full version if you are doing any of your own mastering.  

Use a Vocal Bus and send all the clips to this bus. Put the Youlean meter in the effects bin of the vocal track and keep it open on a second screen. Do not have any effects or compression active at this point. You want the raw audio. 

The trouble with normalizing is it will bring all those clips up to a very similar level but it does not address loudness, only the peak level. One little 'plosive P in there and your screwed. That becomes the peak. 

Therefore  one method would be to look for peaks in the clip and bring them under control. I do this by "tool copy" into Wave lab and I  apply hi pass EQ to all plosives. I would also do all my leveling there because the tools are always open. You can do the same in Cakewalk but the workflow is a PITA because you need to re open the EQ and gain tool over and over. 

You can also use the volume envelope which is none destructive.  

 

  • Great Idea 1
Link to comment
Share on other sites

18 hours ago, John Vere said:

Therefore  one method would be to look for peaks in the clip and bring them under control. I do this by "tool copy" into Wave lab and I  apply hi pass EQ to all plosives. I would also do all my leveling there because the tools are always open. You can do the same in Cakewalk but the workflow is a PITA because you need to re open the EQ and gain tool over and over. 

You can also use the volume envelope which is none destructive.  

 

I'm no expert but this seems to work for me. I use Melodyne 5 or iZotope RX for any plosives that I can't handle with clip automation. For everything else associated with normalization I use clip gain automation which is time consuming but tames lumps while maintaining dynamic range. I've been using clip gain automation for a long time.

  • Like 1
Link to comment
Share on other sites

For vocals, I used to split the clips at various points, then use clip gain automation to adjust the levels.

Nowadays, I use a vocal rider plugin and bounce the results to another track.  If I like the results, I'll delete or archive the original track.

  • Like 1
  • Thanks 1
Link to comment
Share on other sites

2 hours ago, msmcleod said:

For vocals, I used to split the clips at various points, then use clip gain automation to adjust the levels.

Nowadays, I use a vocal rider plugin and bounce the results to another track.  If I like the results, I'll delete or archive the original track.

First I had no clue what this "Clip gain Automation " was. Is this a fairly new feature?  It took me 10 minutes of Googling and poking around the track inspector. Geeze, talk about adding a feature that even a seasoned user like me cannot find!!  But I did find it, and as for the rest of you, it's a secret and I'm not telling.  

Yes that would certainly speed things up and only involves splitting your tracks into short clips which I do most times anyhow. So I just learned yet another feature ,, when will this all end!  

Link to comment
Share on other sites

24 minutes ago, John Vere said:

First I had no clue what this "Clip gain Automation " was. Is this a fairly new feature?  It took me 10 minutes of Googling and poking around the track inspector. Geeze, talk about adding a feature that even a seasoned user like me cannot find!!  But I did find it, and as for the rest of you, it's a secret and I'm not telling.  

Yes that would certainly speed things up and only involves splitting your tracks into short clips which I do most times anyhow. So I just learned yet another feature ,, when will this all end!  

Just checked... it's been there since at least SONAR 1 XL... so around 20 years :)

image.png.80932d403e0c21b474d3e216c7cf318e.png
 

  • Like 1
Link to comment
Share on other sites

Wow, Amazing. Cakewalk is deep!

This will sound stupid, but it's true! It's probably because when I used to hear the word "clip" I associated that with "loops".  I'm not a loop user needless to say.    As a result I avoided any feature that said "clip"   But,   I figured out a few years ago that a clip was not a loop, A loop could be a clip but any audio in a track is actually a clip.  I think of clips as being  very short as in 1 or two measures. A three minute guitar track to me was not a clip, but of course it really is.  So any feature that said "clip" was avoided by the old me.

Therefore a misinterpretation of a term ended up in me never clicking that tiny wee little icon. I will play closer attention to "clips" and embrace them from now on> 

Link to comment
Share on other sites

yet another option - Melodyne Studio can handle multiple tracks and version 5 has a level range setting you can apply to even out the performance. so this approach would be to load up the tracks and select all the events, and use the leveling range function to conform them to whatever level you need. maybe even do some cleaning and pitch/quantize whilst in there...

https://helpcenter.celemony.com/M5/doc/melodyneStudio5/en/M5tour_NewInMelodyne5?env=standAlone#chapter-5

Link to comment
Share on other sites

You can also use a loudness matching plug-in such as GainMatch, which I use and can endorse. There's also HoRNet's TheNormalizer. Either may be had for $9.

They both work by being inserted on multiple channels. The way they work is that you choose the first track you wish to use as the source, then the other instances will follow it.

Also, if you have that many tracks of the same instrument, check into iZotope Visual Mixer. It's free, and I've found it to be quite the thing for panning multiple vocal tracks.

  • Like 1
Link to comment
Share on other sites

  • 4 weeks later...
On 12/13/2020 at 2:25 PM, msmcleod said:

For vocals, I used to split the clips at various points, then use clip gain automation to adjust the levels.

Nowadays, I use a vocal rider plugin and bounce the results to another track.  If I like the results, I'll delete or archive the original track.

msmcleod... Curious what vocal rider plugin you use? I'm only familiar with the Waves Vocal Rider product.

Link to comment
Share on other sites

5 hours ago, Charles Czerny said:

msmcleod... Curious what vocal rider plugin you use? I'm only familiar with the Waves Vocal Rider product.

I use the Waves one most of the time, as I find it the easiest to use.

I've got a couple of others though:

 

Link to comment
Share on other sites

Because my vocals are most of the time very different in level, I was always very interested in "vocal levellers". I have some of them and tested most of the others (Waves Vocal Rider, TbProAudio DynaRide, TbProAudio Vocal Rider, ERA Voice Leveler, HoRNet's Normalizer, TriLeveler 2 (free), VOLA2 (free), Melodyne Leveling Tool ...).

The bottom line: I am not really impressed by vocal levellers, because if you listen, you can hear that it sounds somehow unnatural afterwards. I think the reason maybe that some parts of the vocal phrases are over-amplifed. Compressors are sometimes even better than vocal levellers (less noticeable). But if you want to use one of them I highly recommend TriLeveler 2 (by SonicAnomaly)! In my tests exactly for using it to level vocals it was one of the best ones, although it is free!

Just to put my opinion into perspective: I was testing primarily vocal levelling, not vocal riding to the instruments. This is another thing!

That's why I changed my workflow in vocal levelling:

  1. I cut the vocals in short phrases (like others above said) and fix issues (like thumps, clicks)
  2. Do clip levelling by hand in SoundForge
  3. Finally measure the LUFS of the clip to increase/decrease the overall level.

When clip gain (clip automation > gain) became visible in CbB (2020.05, IMO the best improvement of CbB) I replaced the 2nd step with it. It is very quick and convenient!

The measuring in the 3rd step I do with HOFA's 4U Meter/Fader (free). IMO it is the best tool in that area, 'cos you can simply drag the clip on the plugin for calculation. That saves a lot of time (not necessary to play the clips)!

Edited by marled
Added Melodyne
Link to comment
Share on other sites

About the clip gain. As my above post I was unaware of its use. So I have tried it but I cannot get it to behave the way everyone is describing it works. 
for me the red line behaviour is the same as a standard volume envelope. I grab in the middle and get a peak. Not a even line. Is there setting or toggle? 

Link to comment
Share on other sites

On 1/11/2021 at 5:50 AM, Bruno de Souza Lino said:

There's also Sonic Anomaly's TriLeverler2. It does exactly what Waves Vocal Rider does and it's free. I believe the term we're looking for here is "gain matching"

I'm having trouble finding a download for this. There is a post on the Cockos site that Sonic Anomaly plugins were no longer available, but maybe that was the for purchase ones.

Anyone have a "safe" download site to get TriLeverler2??

UPDATE>>>> I finally found a site with all the Sonic Anomaly plugins...

http://www.vst4free.com/index.php?dev=Sonic_Anomaly

Edited by Charles Czerny
Update
Link to comment
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...