Recording With Compression........opinions please ??

[Bobbsy]

Yup. When you combine stand alone analogue gear with digital then all bets are off. The 0dBu=-18dBFS mantra is only a guide and certainly not all analogue gear can handle a +18dBu signal (while others are happy up to +20 or even +22). However, I'd like to think that anyone getting up to that complexity of installation would also understand how to calibrate things and decide on maximum levels.

Alas, based on various post and misinformation, there are all sorts of people convinced that their basic 2 channel interface sounds better (in an audiophile sense) at lower levels like an external mixer might.

 

[Farview]

Yup. When you combine stand alone analogue gear with digital then all bets are off. The 0dBFS=-18dBu mantra is only a guide and certainly not all analogue gear can handle a +18dBu signal (while others are happy up to +20 or even +22). However, I'd like to think that anyone getting up to that complexity of installation would also understand how to calibrate things and decide on maximum levels.

A $50 Mackie mixer feeding an M-Audio interface...best not push it. Hell, the old Mackie 2408's didn't like to be pushed. But that was because they weren't at +4 line level. They were somewhere between +4 and -10, so feeding a +4 input was a stretch in the first place, feeding it enough signal to almost clip the converters could sound pretty bad.

Alas, based on various post and misinformation, there are all sorts of people convinced that their basic 2 channel interface sounds better (in an audiophile sense) at lower levels like an external mixer might.

I still think it might, depending on the source and the cheapness of the unit. How much headroom could anything the size of your hand with two mic preamps, headphone amp, converters, IO, all running off a little wallwart really have? The fidelity is a little suspect to begin with, much less if you push it a bit...

But, obviously, that is a case by case thing.

 

[Mo Facta]

+18dBu is very low headroom compared to high end and vintage designs. So is +22dBu. Gear that exhibits these kind of headroom figures came as a result of manufacturers cutting costs and using cheaper components yet still pushing their stuff into the "pro" market, i.e. Mackie. If a pro audio unit can't handle +18dBu it's not very good, imo and you'll struggle to get a clean signal out of it when gain staging with other conceivably "better" pro audio gear. This is because low-headroom gear made with cheap components often saturate many dB before clip point. So, you may think you're getting 16 or 18 dB of headroom (i.e. clean audio) but what you're actually getting is more like 10 or 12.

Bob Katz reckons it's better to use the -10dBv reference in Pro Audio (especially in the prosumer market) these days because it adds around 12dB of headroom and avoids saturating cheap circuits. Of course, the trade off is a higher noise floor but you'll at least get a clean, distortion-less signal.

Cheers

 

[Bobbsy]

I'd like to know what you consider "high end and vintage designs". I've been kicking around with pro audio gear since the early 1970s and don't remember anything that got much better than that and a lot that got far worse. The Nagra IV-S, for example, long the mainstay of the field recording industry, had a headroom of about 10dB once you corrected for the weird scale they used on their modulometer.

I find the opposite to you--it's in the last decade or two that analogue electronics have improved enough to take the sort of 18-22dB headroom we generally enjoy now. Indeed it's probably down to "cheaper components" but the move to ICs etc. often means cheaper can also be more reliable and flexible.

 

[Farview]

And it all depends on the signal being sent through. transients at +18 aren't generally a big deal, but having the bulk of the signal up there Will suck hard.

 

[Ethan Winer]

The only reason for that is so that you don't run your analog circuits out of headoom on the way to the converters. The sound difference between recording hot and recording at line level happens on the analog side, not the digital side. You are, of course, correct that the digital side doesn't sound any different at any reasonable (and most unreasonable) level.

Most competent analog circuits are perfectly clean right up to the point of hard clipping. It's true that with transformer gear (or tubes that use less negative feedback), distortion can creep up a bit as you approach clipping. But it's nothing like analog tape where distortion rises evenly above 0 VU or whatever. With modern solid state gear it's not a problem at all. So the obvious approach is to calibrate your mixer / preamp and sound card such that clipping never occurs at any level below 0 dBFS. This is easy to do!

--Ethan

 

[philbo]

Exactly. Further, there's no reason to record both compressed and not compressed versions onto two tracks. As I already explained, if you patch the compressor into the playback path rather than the recording path, you hear exactly the same sound and get exactly the same result, but you still have a chance to change things later if needed.

--Ethan

An excellent point, Ethan. My comments, though in reply to the use of compression, were somewhat pointed toward the need to commit to what sounds good, rather than indulge in fence-sitting till the music suffers due to indecision. It is not a technology problem, it is more like a neurosis. I've watched it develop over the last 3 decades, as DAWs have replaced tape machines.

For an extreme example, I know a guy who refuses to print audio from a recorded midi take because "I might want to change the patch some day" -- even though the midi would still be there...) So, literally, nothing ever gets finished, it just lies there in a spaghetti-pile of potentialities. While this guy is an amazing musician and composer, his producing is a bit of a weak spot.

 

[Bobbsy]

An excellent point, Ethan. My comments, though in reply to the use of compression, were somewhat pointed toward the need to commit to what sounds good, rather than indulge in fence-sitting till the music suffers due to indecision. It is not a technology problem, it is more like a neurosis. I've watched it develop over the last 3 decades, as DAWs have replaced tape machines.

The problem is that there seems to be two different philosophies about what compression is for.

Some are looking for a "compression sound" and using it as an effect like EQ or Reverb. For this purpose, there MIGHT be an argument to record with the compression (though I'm not sure I buy your indecision argument--I take thousands of decisions per mix and don't see why it should be any more difficult to decide on compression after the fact than it is for, say, EQ).

The other philosophy is that compression is a technical tool purely to control dynamic range. I fall into this category and make it my goal that the compression itself is completely inaudible as an effect and simply makes sure all parts of a track are clear in the mix. For this use, applying the compression can ONLY happen when you can see how it sits in the mix.

I'm not saying either viewpoint is right or wrong. Like so many other things to do with recording a mixing there are just different ways of doing things and both are equally valid.

 

[Farview]

Some of it is genre specific. Certain styles need to compress things in order to sound right, while others should only use compression when something is dynamically out of place. (and it's done almost apologetically)

Since I do al lot if hard rock and metal, I tend to need to compress things just to get the expected sound that fits the style. When I do jazz and classical, I try to use as little compression as I can get away within tend not to compress on the way in. Im also blessed with a couple really nice sounding compressors that allow me to compress the snot out of something without it sounding bad.

 

[Farview]

Most competent analog circuits are perfectly clean right up to the point of hard clipping. It's true that with transformer gear (or tubes that use less negative feedback), distortion can creep up a bit as you approach clipping. But it's nothing like analog tape where distortion rises evenly above 0 VU or whatever. With modern solid state gear it's not a problem at all. So the obvious approach is to calibrate your mixer / preamp and sound card such that clipping never occurs at any level below 0 dBFS. This is easy to do!

--Ethan

Im not sure most of the people on this board have converters that can be calibrated. I don't, and Im using the upper end of prosumer converters. (motu 24io, 2408)

 

[GONZO-X]

The problem is that there seems to be two different philosophies about what compression is for.

Some are looking for a "compression sound" and using it as an effect like EQ or Reverb.

exactly what i said on my very first post.

if you know how to use a compressor properly, you will know that you can 'color' the sound.....
change the dynamics....
mold the signal to do just what you want it to, through the use of a nice compressor.

i can't think of any professional mixers that do NOT use a compressor on input, on certain sound sources.

there are two different philosophies.

that's the problem with discussing all of this, some folks just don't get it.

either they have never had any real experience using a compressor as a color device, or they just cannot grasp the concept.

it can smooth off peaks at the same time as creating a 'specific sound', a color of sound, a richness of tone from the compression effect, a 'squishing' of the tone by artful use of the attack and release.....


it's compression 101.

 

[bouldersoundguy]

Im not sure most of the people on this board have converters that can be calibrated. I don't, and Im using the upper end of prosumer converters. (motu 24io, 2408)

Users don't need to recalibrate their converters, they just need to know what 0vu translates to in dBFS on their particular converters.

 

[Ethan Winer]

My comments, though in reply to the use of compression, were somewhat pointed toward the need to commit to what sounds good, rather than indulge in fence-sitting till the music suffers due to indecision.

You can still "commit" when patching a compressor into the playback path. Just don't ever change the settings!

It is not a technology problem, it is more like a neurosis.

Yes, and the better solution is psychotherapy rather than forcing yourself to work one way or another. I don't have this problem, though I certainly change my mind sometimes about a mix. I'm sure everyone who prefers to "commit" has at one time or another wished he could change something after it was too late. Either that or they're lying.

For an extreme example, I know a guy who refuses to print audio from a recorded midi take because "I might want to change the patch some day"

That's not extreme! I never render software synths as audio because it just wastes disk space. And being able to change a MIDI patch or filter position later when you can hear everything in context is just as useful as being able to change the amount of compression or the ratio etc. I honestly don't understand why anyone would prefer to tie their own hands behind their back when it's not needed.

So, literally, nothing ever gets finished, it just lies there in a spaghetti-pile of potentialities.

Again this is a psychology problem, not a technology problem. But hey, we all get better, and more confident over time.

--Ethan

 

[Ethan Winer]

Users don't need to recalibrate their converters, they just need to know what 0vu translates to in dBFS on their particular converters.

Exactly. Before the Focusrite Scarlett 8i6, my main sound card was a Delta 66. The Delta series have a control panel that lets you set the nominal input and output levels individually for -10, +4, and an in-between value they call "Consumer." I set both to +4 because I wanted to send and receive enough signal to be well above the residual noise of my mixer. I sent to the sound card from the Insert point of my Mackie 1202, and I measured that as +5 dBm output to hit 0 dBFS in my recording software. So the only "calibrating" I did was set the input level switches to +4, which is what most sound cards default to anyway.

--Ethan

 

[jaynm26]

it can smooth off peaks at the same time as creating a 'specific sound', a color of sound, a richness of tone from the compression effect, a 'squishing' of the tone by artful use of the attack and release.....


it's compression 101.

I think thats what people are missing is that it can be used for different things. People hear compression they automatically think LOUD. When thinking of it as a tracking tool they think too LOUD & permanent. GONZO, I mean for the people that are not professional you could clearly see the misconception and disconnect in which people would stray away from it. But! Most of the professionals here (you, Ethan, Boulder, etc..)
#1 Track well at the correct level
#2 Understand the many uses of a compressor that its not just a mixing or leveling device
#3 Proly have multiple hardware comp's to select the different tonal colors & understand this type of use
#4 Know how to use it for that specific purpose with correct tracking technique and settings

For the people who do not have the extensive knowledge on the use as you guys do..or the people who always tracked without them and still generate a good sound its easy for them to shy away from the technique or just reject in entirety.

 

[jaynm26]

Also GONZO is right I have never not seen an external comp in the recording chain.

A professional trackng chain most of the times consist of
Mic->Pre Amp->Comp (for whatever reason GONZO stated)->External A/D Converter (sometimes)->DAW Converters->PC

 

[Mo Facta]

I'd like to know what you consider "high end and vintage designs". I've been kicking around with pro audio gear since the early 1970s and don't remember anything that got much better than that and a lot that got far worse. The Nagra IV-S, for example, long the mainstay of the field recording industry, had a headroom of about 10dB once you corrected for the weird scale they used on their modulometer.

I find the opposite to you--it's in the last decade or two that analogue electronics have improved enough to take the sort of 18-22dB headroom we generally enjoy now. Indeed it's probably down to "cheaper components" but the move to ICs etc. often means cheaper can also be more reliable and flexible.

Well as you said, a Nagra is a field recorder. It's not the same as high end studio gear, as great a recorder as it is. And was it calibrated to +4dBu/0VU?

My TL Audio C-1 has a max output of +27dBu. A Manley Vari-Mu has a max output of +30dBu (26dB of headroom!). Tonelux 500 series units have max outputs of+24dBu. Generally you find high end units with specs like this because electrical engineers know that once you saturate a circuit, you get distortion, often many dB below clip point. Whether or not that distortion is "musical" is an entirely different debate and depends on the design, components, etc. Using high-headroom components minimizes the effects of distortion when signals increase and supply cleaner signals within a broader dynamic range. This is also how mastering engineers get a clean sound when clipping masters after processing - by using high headroom processors going back into converters calibrated several dB lower.

Cheers

 

[Bobbsy]

Well as you said, a Nagra is a field recorder. It's not the same as high end studio gear, as great a recorder as it is. And was it calibrated to +4dBu/0VU?

Hey! Don't bad mouth the beloved Nagra! In terms of build quality and sound quality it was streets ahead of most "high end studio gear". In terms of price (if you consider price per channel) it's several streets ahead!

The meter (which they called a "modulometer") was strangely calibrated to -8 on the meter=0VU=+4dBu, hence my comment about correcting for the weird scale. I guess when you're Swiss you can do things your own way!

 

[iqi616]

#2 Understand the many uses of a compressor that its not just a mixing or leveling device

+1......

 

[Mo Facta]

Hey! Don't bad mouth the beloved Nagra! In terms of build quality and sound quality it was streets ahead of most "high end studio gear". In terms of price (if you consider price per channel) it's several streets ahead!

The meter (which they called a "modulometer") was strangely calibrated to -8 on the meter=0VU=+4dBu, hence my comment about correcting for the weird scale. I guess when you're Swiss you can do things your own way!

Ha ha, Bobbsy, I'm not badmouthing it. I know how great they are, they're fantastic. I know the sound quality is great. I'm talking strictly from a headroom perspective here.

Cheers