Vocal Compression - Real Time vs. Track Processing

  • Thread starter Thread starter mark4man
  • Start date Start date
mark4man

mark4man

MoonMix Studios
Crew,

One of our vocalists (of the female Cool Jazz variety), at times, sings whispery/quietly; & drifts away from the microphone (when she closes her eyes.)

It is, needless to say, very difficult to get good levels in these circumstances.

[It is my understanding that this (vocalists undisciplined in the recording booth) was one of the main reasons for the development of dynamic compression.]

Aside from that, I have always had to apply compression to her tracks...& normally do so after the fact (after the track is recorded.) My DAW is SONAR XL 2.2; & my audio editing/pre-mastering software is WaveLab 4.0...&; between the two, I can always seem find a plug-in which proves very effective for the track in question.

But...

I have also been contemplating using compression during the actual recording; & thought I might post an inquiry among the brains here, to determine which method is actually better...during the recording, in real time, or afterward, as post processing (keeping in mind that it is software dynamic processing I would be using.)

And also...if I go with real time...I could use some advice as to typical setups (i.e., do I also need limiting, what would be effective plug-ins & settings, etc.) I'm really curious as to which methods are used by our members, along with the various results.

Thanks very much,

mark4man,
 
It's not possible to use software to apply destructive compression during recording. As long as it's software, you have to do it after recording.

You can always use Input Monitoring, but the recording itself will still be without compression.



You can of course use compression during recording, but you need a hardware compressor.

What's the best? I think you need to find that out yourself. I hate applying effects desctructively to tracks, so the only effect I record is distortion/fuzz/overdrive when I play guitar, but that's just me... :)
 
An FMR Really Nice Compressor (RNC) would be just about right for that purpose. Set on Super Nice mode. Almost perfectly transparent. Cheap too.
 
moskus said:
It's not possible to use software to apply destructive compression during recording. As long as it's software, you have to do it after recording.
Click to expand...

Actually Protools can do this. You can bring the input into a aux bus, apply the plug in and then route to a writable track.

Not Sonar however or most of the rest of the packages out there.
 
Middleman said:
Actually Protools can do this. You can bring the input into a aux bus, apply the plug in and then route to a writable track.

Not Sonar however or most of the rest of the packages out there.
Click to expand...

SAWStudio (all versions) can do this too, you can route your input device to any input channel on the virtual console, and use the built in EQ, noise gate, compressor, and any plugins that are inserted into the pre fader FX patchpoint as well, the record point is taken between the pre fader FX patchpoint and the aux sends.

You can also record that input channel onto any track you want in the multi-track, so if you're doing multiple takes, you don't have to keep setting up different input channels, you just have to use the track adjuster on the record panel to change the destination track for the recording.

I used to use Sonar 2.2 XL, and in your case, you'll be needing a hardware compressor to use while tracking, or, just set the recording level so that her loudest singing won't clip going in, and then do all of the compresson after the fact.

Of course, using a compressor in software cannot help you to keep from driving your AD converters into clipping, but at 24 bit, you can keep the peaks of your input level nicely below the danger zone, and not really have to worry about boosing the noise floor too much, unless you are doing some heavy compressing and extreme makeup gain while tracking, which wouldn't be advisable in any case.

When I have to deal with this type of thing, I sometimes get better results by recording uncompressed, and then copy the wave to a seperate track, archive the original, and then go in and manually edit the waveform to be more consistant in overall volume, or even chop the track's quieter passages out, and put them on a seperate track, and automate the two tracks using volume envelopes to acheive the desired results.

You could also just make her keep recording the song over and over, until she get a handle on how to do it better. LOL!!
 
Last edited:
Middleman said:
Actually Protools can do this. You can bring the input into a aux bus, apply the plug in and then route to a writable track.
Click to expand...
But the point is lost. One reason to use compression is (as you know) to get as hot input signal as possible before the AD-converter. Clipping occurs in the converter and not in the software, so it doesen't matter if you apply it during or after recording (if you're using software compressors). :)
 
moskus, StevenLindsey, Middleman & Strryder...

Thanks, all.

Have never used hardware processing during live recording; & thought perhaps there may be a way to bring her (the vocalist's) mike level up; & at the same time apply some sort of limiting so as to avoid clipping...not so much as to have to worry about changing the dynamics of the vocal recording...but just to insure the signal doesn't spike overall.


Of course, using a compressor in software cannot help you to keep from driving your AD converters into clipping, but at 24 bit, you can keep the peaks of your input level nicely below the danger zone, and not really have to worry about boosing the noise floor too much, unless you are doing some heavy compressing and extreme makeup gain while tracking, which wouldn't be advisable in any case.
Click to expand...
Strryder,

Please explain the "but at 24 bit" statement. This goes back to the old "0dB = 0dB = 0dB" argument...with which I have a bit of difficulty conceptualizing how gain can be higher or lower at varying resolutions (unless you're simply referring to cleaner signal at lower volume...which I believe you are, since you go on to explain deeper noise floor...I'll shut up, now.)


When I have to deal with this type of thing, I sometimes get better results by recording uncompressed, and then copy the wave to a seperate track, archive the original, and then go in and manually edit the waveform to be more consistant in overall volume, or even chop the track's quieter passages out, and put them on a seperate track, and automate the two tracks using volume envelopes to acheive the desired results.
Click to expand...
I go through similar steps myself...to get the track just right, level wise; & at the same time preserve it's original integrity (which was where I was going on the limiter question.) I wind up with a series of comped tracks for every voice & instrument, with tons of carefully cut & automated clips. It's kind of like "manual" or "home-made" processing...but always woth the effort.


It's not possible to use software to apply destructive compression during recording. As long as it's software, you have to do it after recording.

You can always use Input Monitoring, but the recording itself will still be without compression.
Click to expand...
moskus,

I know...but I was hoping. I was anticipating somehow getting a limiting effect out of a plug-in.

I'll probably start looking into one of the hardware units mentioned.

Thanks again, everyone,

mark4man


BTW - moskus...what ever became of the collaboration everyone was busy on?
 
moskus said:
But the point is lost. One reason to use compression is (as you know) to get as hot input signal as possible before the AD-converter. Clipping occurs in the converter and not in the software, so it doesen't matter if you apply it during or after recording (if you're using software compressors). :)
Click to expand...


No denying your point. Limiting prior to AD conversion is the goal. At best, Protools and SAWStudio are adding color and maybe some control to dynamics but if clipping has occured its too late.
 
Last edited:
Just got an FMR Audio Really Nice Compressor

I too was struggling with the problem of getting the levels as high as possible while tracking, but trying to avoid clipping.

You might finally get that perfect take, then find that it just clips in one spot. Unless you can get a perfect overdub, the take is ruined.

So I broke down and got a RNC. Everyone on these forums gives it a great review. I'll let you know what I think when I have had a chance to try it. You can usually find it on Ebay for about $175 US although I confess I bought mine retail.
 
But the problem might also be the signal-chain. A good signal-chain will add much less noise, and low-level recordings will sound much better. A compressor won't make up for poor mic.amplifiers, etc... ;)
 
Hi,

From my experience I tried compressing vocal tracks after recording(software compressor), and I also tried compressing the signal while i was recording (hardware compressor). To be honest i prefer the second option as I saved a lot of time on the mix. that's my opinion

Lapieuvre
 
jamie_drum said:
IYou can usually find it on Ebay for about $175 US although I confess I bought mine retail.
Click to expand...

thats how much it costs retail unless you got jerked...
 
lapieuvre said:
To be honest i prefer the second option as I saved a lot of time on the mix. that's my opinion
Click to expand...
A mix shouldn't be done fast. It should be done right! :D

What if you screw up the compression on the recording? ;)
 
"What if you screw up the compression on the recording?"

I track again with correct compression... :o/

seriously i compress very little just to flatten those exagerated peaks that bad signers do.

It can avoid clipping too!

Lapieuvre
 
I agree that the best way to do it for vocals is to use a limiter set to catch peaks and very little or no compression on the way in. Moskus is correct in relation to the need for hardware compression as it occurs prior to the A/D conversion. Run it back through the same box or use a software compressor after the take is perfect.

:) Q.
 
You must log in or register to reply here.
Back
Top