Mastering without compression?

[Pinky]

I use very light compression on vocals to help them sit IN the music instead of on top, but use volume automation to take care of the loud and soft parts.

The degree of compression varies quite a bit for me, it's based on the vocal performance and desired vocal effect/presence. But I almost always use volume automation to get the vocal to sit *just right* in softer or louder passages, or when the singer is quieter or louder than the instrumentation. Volume automation takes the burden of getting the vocals to 'fit' out of the hands of the compressor. It's key IMO. I obviously apply volume and panning automation across other tracks as well, when needed or appropriate. Getting a good mix makes the mastering process more like polish than paint.

 

[Farview]

Surely if the peaks are too much on every beat, then that should have been fixed in the mixing stage by 'something'. Not a mastering plugin, but a normal compressor, limiter or gate combination. to work just on the offending tracks. Can you really fix something this bad after the mix is set? I suspect you can't without compromising the whole thing.

I was taking it to the ridiculous other conclusion to make the point.

The point being: if there are too many peaks, it will take forever to grab them and turn them all down.

 

[Smithers XKR]

Surely if the peaks are too much on every beat, then that should have been fixed in the mixing stage by 'something'. Not a mastering plugin, but a normal compressor, limiter or gate combination. to work just on the offending tracks. Can you really fix something this bad after the mix is set? I suspect you can't without compromising the whole thing.

I think you nailed it here Rob. Just get it right in the first place, that is why I am a fan of using pre compression. May not be right but just what I believe as old school.

 

[Smithers XKR]

Why would you do this? Surely the vocalist is key. Some are totally untrained or learned to sing badly. Those who watch the masters work a mic, but copy the look not the purpose. Some talent needs hardly any compression to be perfect and others without control need much more. Level should never be an issue nowadays. Plenty of noise free gain is normal. I never treat anything pre record. You cannot undo it afterwards. Most of my compression, if there is any at all is gentle. My music would make it horrible. I’m trying to imagine any genre where heavy compression is vital apart from death metal growling and yelling maybe?

I agree Rob. The less pre compression you need to use on vocal the better, but not all singers have the natural ability to use mike distance control so some amount of tweaking is going to be needed. I am all for natural capturing of the vocal BTW.

 

[Smithers XKR]

I was taking it to the ridiculous other conclusion to make the point.

The point being: if there are too many peaks, it will take forever to grab them and turn them all down.

Taking out the "S" and "T" high nuances in the lyric on the initial vocal recording is something I try to do, it really bugs me. But I think that is more to do with EQ capture rather than a pre compression dynamic issue. I have never been able to figure this one out apart from wrapping a sock over the mike

Any ideas?

 

[bouldersoundguy]

I do like singers who can work a mic, but the downside is that they go in and out of proximity effect. There's no one perfect solution.

If I know the performers I'm working with, I'll generally compress a little on the way in. If I don't know them, I tend to be more cautious in order to avoid overdoing it.

 

[RRuskin]

Taking out the "S" and "T" high nuances in the lyric on the initial vocal recording is something I try to do, it really bugs me. But I think that is more to do with EQ capture rather than a pre compression dynamic issue. I have never been able to figure this one out apart from wrapping a sock over the mike

Any ideas?

Popper stopper and mic placement. Often not having the mic aimed directly at the singer is the way to go.

 

[Smithers XKR]

Popper stopper and mic placement. Often not having the mic aimed directly at the singer is the way to go.

Yes I have a pop stop

 

[Massive Master]

Just echoing a bunch of stuff (and just about everything Farview said).

[A] "Mastering" isn't about "volume" - It's about translation and cohesiveness between a collection of recordings.

[2] Realistically speaking, I can't remember the last time there wasn't *some sort* of compression or limiting on a recording that's come out of here - and that includes jazz and classical. "Some sort" includes "clipping" done by the AD (some people ask me why I still use the original design Crane Song HEDD 192 and I can tell you it's because of the AD and how it handles the rough stuff in a style like no other converter I've come across) at the end of the analog chain. I can also tell you that the capture is typically done at the level that feels most appropriate for what the mixes are asking for - not what the client is expecting in the end - and then additional gain (usually via limiting) is performed post-capture.

* And I might add that I nearly *always* keep that high dynamic range version - because IMO, that's the way it's "supposed to" sound in the first place. A monkey with a limiter can make any recording just as loud as any other recording. How that recording reacts to the abuse is dependent on literally every single decision made before that step takes place.

Again, just being realistic - I'm thrilled like few others that the "loudness war" has seemed to have peaked (clipped, if you will) and levels are getting back to at least some sort of "more reasonable" (as I've said countless times, the listener never asked for this and if they knew what they were missing, they'd probably be pretty pissed about it).

That all out of the way - Are there ways to increase perceived level - substantially - without compression and limiting? Of course. I mean, in floating-bit, you could just crank the hell out of it and export to 24-bit, open the 24-bit file in float, turn it down .1dB and you're good. Heck, sometimes that actually sounds just fine (again, depending on every single decision made beforehand). But technically, in the end, you're limiting it. So just use a limiter that will do it with purpose (and style and adjustable parameters and A&R shaping, etc., etc.) instead.

And if you don't want to use any compression, limiting or various other volume-boosting techniques and you're fine with the volume that the mix is at without any consideration towards perceived and comparative volume, I'm behind you 100%. The whole "amazing" part of digital audio was that it sounded amazing and accurate at *lower* levels and had basically imperceptible noise compared to the source.

That "amazing" part lasted about ten minutes - and the volume war began.

 

[rob aylestone]

Makes sense. I probably use compression to fix things, rather than enhancing them - as in an effect so common nowadays. Popper stoppers are really handy for keeping the backs and forths to a minimum. especially with new/untrained singers. Having their lips on the mesh is great - but of course introduces the need to manually add in some pushes and pulls with faders, where they didn't do them.

 

[Smithers XKR]

Makes sense. I probably use compression to fix things, rather than enhancing them - as in an effect so common nowadays. Popper stoppers are really handy for keeping the backs and forths to a minimum. especially with new/untrained singers. Having their lips on the mesh is great - but of course introduces the need to manually add in some pushes and pulls with faders, where they didn't do them.

Very good point Rob you have a lot more experience with that than me.... but I have worked with some young very powerful young raw rock singers with a lot of dynamic range and I always found that it was useful to use a bit of pre compression/expansion after assessing their dynamic. Nothing too much, just to keep it manageble for the mix.
Tom Jones used to break valve mikes in the 60s as the engineer did not realise how loud he could sing
Of course, that was then

 

[Smithers XKR]

Just echoing a bunch of stuff (and just about everything Farview said).

[A] "Mastering" isn't about "volume" - It's about translation and cohesiveness between a collection of recordings.

[2] Realistically speaking, I can't remember the last time there wasn't *some sort* of compression or limiting on a recording that's come out of here - and that includes jazz and classical. "Some sort" includes "clipping" done by the AD (some people ask me why I still use the original design Crane Song HEDD 192 and I can tell you it's because of the AD and how it handles the rough stuff in a style like no other converter I've come across) at the end of the analog chain. I can also tell you that the capture is typically done at the level that feels most appropriate for what the mixes are asking for - not what the client is expecting in the end - and then additional gain (usually via limiting) is performed post-capture.

* And I might add that I nearly *always* keep that high dynamic range version - because IMO, that's the way it's "supposed to" sound in the first place. A monkey with a limiter can make any recording just as loud as any other recording. How that recording reacts to the abuse is dependent on literally every single decision made before that step takes place.

Again, just being realistic - I'm thrilled like few others that the "loudness war" has seemed to have peaked (clipped, if you will) and levels are getting back to at least some sort of "more reasonable" (as I've said countless times, the listener never asked for this and if they knew what they were missing, they'd probably be pretty pissed about it).

That all out of the way - Are there ways to increase perceived level - substantially - without compression and limiting? Of course. I mean, in floating-bit, you could just crank the hell out of it and export to 24-bit, open the 24-bit file in float, turn it down .1dB and you're good. Heck, sometimes that actually sounds just fine (again, depending on every single decision made beforehand). But technically, in the end, you're limiting it. So just use a limiter that will do it with purpose (and style and adjustable parameters and A&R shaping, etc., etc.) instead.

And if you don't want to use any compression, limiting or various other volume-boosting techniques and you're fine with the volume that the mix is at without any consideration towards perceived and comparative volume, I'm behind you 100%. The whole "amazing" part of digital audio was that it sounded amazing and accurate at *lower* levels and had basically imperceptible noise compared to the source.

That "amazing" part lasted about ten minutes - and the volume war began.

A lot of good points made here, albeit in a technical way
But surely a limiter will just cut the peak dynamic off rather than compression which

 

[Smithers XKR]

A lot of good points made here, albeit in a technical way
But surely a limiter will just cut the peak dynamic off rather than compression which

Will be much more musical?

 

[Smithers XKR]

Just echoing a bunch of stuff (and just about everything Farview said).

[A] "Mastering" isn't about "volume" - It's about translation and cohesiveness between a collection of recordings.

[2] Realistically speaking, I can't remember the last time there wasn't *some sort* of compression or limiting on a recording that's come out of here - and that includes jazz and classical. "Some sort" includes "clipping" done by the AD (some people ask me why I still use the original design Crane Song HEDD 192 and I can tell you it's because of the AD and how it handles the rough stuff in a style like no other converter I've come across) at the end of the analog chain. I can also tell you that the capture is typically done at the level that feels most appropriate for what the mixes are asking for - not what the client is expecting in the end - and then additional gain (usually via limiting) is performed post-capture.

* And I might add that I nearly *always* keep that high dynamic range version - because IMO, that's the way it's "supposed to" sound in the first place. A monkey with a limiter can make any recording just as loud as any other recording. How that recording reacts to the abuse is dependent on literally every single decision made before that step takes place.

Again, just being realistic - I'm thrilled like few others that the "loudness war" has seemed to have peaked (clipped, if you will) and levels are getting back to at least some sort of "more reasonable" (as I've said countless times, the listener never asked for this and if they knew what they were missing, they'd probably be pretty pissed about it).

That all out of the way - Are there ways to increase perceived level - substantially - without compression and limiting? Of course. I mean, in floating-bit, you could just crank the hell out of it and export to 24-bit, open the 24-bit file in float, turn it down .1dB and you're good. Heck, sometimes that actually sounds just fine (again, depending on every single decision made beforehand). But technically, in the end, you're limiting it. So just use a limiter that will do it with purpose (and style and adjustable parameters and A&R shaping, etc., etc.) instead.

And if you don't want to use any compression, limiting or various other volume-boosting techniques and you're fine with the volume that the mix is at without any consideration towards perceived and comparative volume, I'm behind you 100%. The whole "amazing" part of digital audio was that it sounded amazing and accurate at *lower* levels and had basically imperceptible noise compared to the source.

That "amazing" part lasted about ten minutes - and the volume war began.

I read your post again and with DAW it makes sense. I know very little regarding DAW but I understand that pre compression would not be needed

 

[Smithers XKR]

I read your post again and with DAW it makes sense. I know very little regarding DAW but I understand that pre compression would not be needed

Sorry my phone is misbehaving. But with condenser mikes then there would be some pre compression and/or expansion needed when feeding into the DAW interface in some cases to prevent overload or external noise capture at the lower dynamic levels?

 

[bouldersoundguy]

Sorry my phone is misbehaving. But with condenser mikes then there would be some pre compression and/or expansion needed when feeding into the DAW interface in some cases to prevent overload or external noise capture at the lower dynamic levels?

No, just keep the gain set low enough that it doesn't clip.

If noise is a problem, find and address the source of the noise.

 

[Smithers XKR]

No, just keep the gain set low enough that it doesn't clip.

If noise is a problem, find and address the source of the noise.

That makes sense what you say boulder, but not all condenser mikes have a flat frequency dynamic response so they will react differently according to the gain. But that can be all fixed with the DAW, so I get it what you say. The NT1a I have is quite toppy but basically the same circuitry as the NT1, so yes I understand what you say. Thanks

 

[ashcat_lt]

...measures -25dB (LUFS)...you want to get to -12dB (LUFS)...+9dB...

LOL. IDK where you learned this kinda math, but...

Anyway of course it's a good idea to try to control your peaks at the mix level, but it's not actually always very easy to do that by limiting individual tracks because those tracks all add together, and in most cases they add together differently at different times. You might have the whole band hit on the same beat, and they all add up to a certain peak level. Then some other beat they all hit together the same, and all have the same individual peak levels as before, but because of slight timing (or pitch, I suppose) differences, the waves add up differently because they have different phase relationships and things that might have been cancelling before are adding now. So you end up with your "loudest peaks" sometimes being way off from each other. If you try to squash each individual track so that the loudest of those additive peaks is at a reasonable level, you'll likely end up way overdoing it for most of the rest of the piece, and if you kind of just leave them, you end up wasting a bit of headroom. Often those really high "aberrant" peaks only actually last for a few samples and limiting or even clipping them off won't make any audible difference except to allow you to turn it up more before the red light comes on. So basically, yes try to control peaks at individual track and bus level, but you still will very often want at least a little something at the mix bus or master stage anyway, and that's not because you're doing something wrong, but just that's how wave math works.


Edit - To be fair to @keith.rogers, LUFS math isn't exactly as straightforward as RMS or peak math because of the way LUFS is sort of dynamically windowed. Changing the overall level can actually change which samples even get figured into the calculations. I tend to think that a 4db discrepancy would be fairly extreme, though.

 

[Farview]

Since you aren't working in a daw, I am assuming you don't have a mastering limiter. Using a standard compressor at a high ratio is technically limiting, however your experience will be quite different than someone using a lightning fast mastering limiter plugin. Mastering limiters are an order of magnitude faster than most outboard compressors, which makes them more transparent.

 

[LazerBeakShiek]

Limiting is heavy compression. Compression ratio's greater than 20:1 are limiter strength. Songs on the radio are most definitely limited and compressed as a standard.