Old 01-27-2009, 07:41 AM   #241
dero
Human being with feelings
 
Join Date: Jan 2009
Posts: 1
Default

great thread, thanks to all involved.

Could someone post the audio files as mp3 or .wav?

I do my internetting on a very basic pc with no audio software that didn't come preloaded.

Thanks
dero is offline   Reply With Quote
Old 01-27-2009, 08:24 AM   #242
BoxOfSnoo
Human being with feelings
 
Join Date: Apr 2006
Posts: 998
Default

Quote:
Originally Posted by yep View Post
...Goes to show why you need decent monitors! The laptop speakers wouldn't reproduce lows accurately, so I couldn't tell what was happening until I plugged the laptop into real speakers three days later.
This leads to a question. In last month's SOS, there was an interesting article how these guys recorded Chinese folk music, but got stuck with almost nothing to mix or record with. He said he used a technique to monitor without decent speakers...

He used the MDA limiter, with limiting cranked way up "to see what's ducking the mix".

I don't quite get this. Could you explain? Is it a viable technique?
__________________
Get Dropbox! (Bonus space for both of us with that link)
BoxOfSnoo is offline   Reply With Quote
Old 01-27-2009, 11:36 AM   #243
yep
Human being with feelings
 
Join Date: Aug 2006
Posts: 2,012
Default

Quote:
Originally Posted by dero View Post
great thread, thanks to all involved.

Could someone post the audio files as mp3 or .wav?

I do my internetting on a very basic pc with no audio software that didn't come preloaded.

Thanks
Just for the record, and for the benefit of any non-Reaper users who might be linking into this thread:

Reaper is the most ridiculously easy-to-demo software ever made. Takes about 40 seconds from when you click the "download" link to when you are actually recording with the full-blown unprotected software, on a moderate broadband connection. And I mean that literally. It is nothing like installing Nuendo or Sonar or that kind of stuff, where you have to set aside 2 hours to install, validate, and configure. Any examples are going to get harder to make sense of without some kind of common platform.

Even if you hate Reaper and never plan to use it for anything and have other DAW software that you love and your internet computer is a crappy piece of junk like mine, I heartily encourage you to download the little REAPER exe for the examples. If you have the bandwidth to download wav files, you have more than enough bandwidth to download reaper and my ogg sample project. Reaper is the easiest way to have a common grammar and interactive examples that everyone can use.
yep is offline   Reply With Quote
Old 01-27-2009, 12:01 PM   #244
yep
Human being with feelings
 
Join Date: Aug 2006
Posts: 2,012
Default

Quote:
Originally Posted by BoxOfSnoo View Post
...He used the MDA limiter, with limiting cranked way up "to see what's ducking the mix".

I don't quite get this. Could you explain? Is it a viable technique?
I'm just guessing, but I think he meant he was using a limiter with aggressive settings to figuratively "see" what the dynamics or low end were like because he could not trust his ability to "hear" the dynamics or the low end.

There are a few clues that a limiter could give someone in such circumstances. For one thing, limiting artifacts in the higher frequencies (that the speakers CAN reproduce) can reveal what's triggering the limiter in the frequencies that you CAN'T hear. For instance if the cymbals and vocals abruptly suck down every time there's a kick drum hit, then you might have either too much kick drum, or a kick drum that is unbalanced or overly bass-heavy, e.g. if you can hear it clearly well-balanced in the mids but if the low end is obviously causing major ducking, then the lows might be disproportionate.

Similarly he may have been using the limiter's meters and filtering controls to see the "spaces in between" the audible music, to see how the measured signal level differs from what the signal sounds like. Looking at a "limit" indicator or gain reduction meter in conjunction with an ordinary signal level meter can tell you a lot about how the compressor or limiter filters and responds to the input signal. If you already KNOW how the limiter works, then looking at those meters could theoretically tell you something about the program material in terms of how it sounds, especially in terms of how much/what aspects of the sound make it "through" the limiter or compressor and cause more of a jump in output level than they "should."

We're getting way, way ahead of the ground I've covered so far in terms of metering and technical operation, but those are ways that a knowledgeable engineer might try and chase shadows of sounds that he knows he can't actually hear. Either of them could have actually revealed to me that there was a problem with the example file I posted, but I never bothered to check anything like that.

Is it a "viable technique" for getting around the problems of bad monitors? No, not unless you consider eating dead people and tree bark a "viable technique" for camping. People in desperate and demanding circumstances must do what they must do, and some of them make it through in inspirational ways. Are you trying to be an inspirational story, or to make good recordings? (hint: the latter has a much lower rate of tragic failure).

If you need to save money, sell an instrument. Don't eat out for three months. Make your own coffee. Cancel cable. Quit drinking or smoking. But splurge on monitors. Even if they are just the cheapest monitors actually sold as "monitors" they are probably better than anything in a department store, when it comes to monitoring.
yep is offline   Reply With Quote
Old 01-27-2009, 12:16 PM   #245
stupeT
Human being with feelings
 
stupeT's Avatar
 
Join Date: Jan 2009
Location: frankonia
Posts: 1,996
Default

Yep wrote:
"unless you consider eating dead people and tree bark"

That sounds like an option to me!

Yep wrote:
"Quit drinking or smoking."

This is NOT an option for me!

Cheers
stupeT
stupeT is offline   Reply With Quote
Old 01-27-2009, 08:12 PM   #246
yep
Human being with feelings
 
Join Date: Aug 2006
Posts: 2,012
Default

Compression continued...

So how does a compressor actually work?

I'm going to start out by talking about a conventional four-control compressor, which is pretty much the norm. The four standard controls are THRESHOLD, RATIO, ATTACK, and RELEASE, or occasionally variants thereof. Makeup gain, included on virtually all compressors, is just a simple gain (volume) control that comes after the compressor and that is completely independent of the action of the compressor. I will also refer to things like "circuits," pretending that we are still in the analog realm, but the principles apply to plugins as well.

There are also simpler two-knob compressors, and more complex ones such as reacomp that actually give you control over the detection circuit, and there are also idiosyncratic things like "time constants" and so on that some compressors offer, but let's set those aside for the moment. If you want a straightforward freeware compressor to play along with then Kjaerhus classic compressor is pretty good.

Onward...
yep is offline   Reply With Quote
Old 01-27-2009, 08:40 PM   #247
yep
Human being with feelings
 
Join Date: Aug 2006
Posts: 2,012
Default

How a compressor works


Inside the compressor is a little gremlin that turns down the volume. That's it. Really. HOW and WHEN he turns down the volume is determined by the instructions you give him with the compressor controls.

THRESHOLD sets the gremlin's alarm clock. It is what tells him to wake up and start doing what he does, i.e. turning down the volume. If you set the threshold at -10dB then the gremlin just sleeps his lazy ass off, doing nothing at all until the signal level goes above that threshold. A signal that peaked at anything lower than -10dB will never wake up the gremlin and he'll never do a damn thing. (see why presets could be problematic?) But once the signal goes above the threshold, the gremlin rips off the sheets and springs into explosive action.

RATIO decides HOW MUCH the gremlin turns down the volume, and it acts completely in relation to the threshold. If the ratio is set to 2:1, and the signal goes ABOVE the THRESHOLD, then the gremlin will cut that signal in half. For example, with -10 threshold, a signal that hits -5 (which is 5dB ABOVE -10) will be turned down 2.5dB for an output of -7.5dB. Negative values can be confusing if you're not used to thinking in such terms so re-read and ask questions if you're stuck. This is important, and it does get instantly easier once you "get" it.

ATTACK is like a snooze button for the Gremlin's alarm clock. It lets the gremlin sleep in for a little while. So if the THRESHOLD is set for -10dB, and the ATTACK is set to, say, 50ms, then once the signal goes above -10dB, the gremlin will let the first 50ms pass right by while he rubs his eyes and makes coffee. An attack of zero means the gremlin will respond instantly, like a hard limiter, and will allow nothing above threshold to get through unprocessed. Any slower attack means the gremlin will allow the initial "punch" to "punch through" and will only later start to act on the body of the signal.

RELEASE is like a mandatory overtime clock for the gremlin. It tells him to keep working even after the signal has dropped below threshold. A release of zero means strict Union rules-- once the signal drops below threshold, the whistle blows, and the gremlin drops whatever he's doing and goes back to sleep. But a slower release means the gremlin keeps compressing the signal even after it has dropped below the threshold. This can lead to smoother tails and less "pumping" or "sucking" artifacts that come from unnatural and rapid gain changes.

Last edited by yep; 01-27-2009 at 10:28 PM.
yep is offline   Reply With Quote
Old 01-27-2009, 09:09 PM   #248
yep
Human being with feelings
 
Join Date: Aug 2006
Posts: 2,012
Default

So, armed with that knowledge, you could, if you want, take a second look at the example project posted above. Or better yet, you could start to mess around with your own settings and material.

Here are some things to think about:

- A compressor with a SLOW attack and a FAST release could give a very punchy, lurchy sound, as the compression lets the initial attack through and then clamps down on the "body" of the note, bringing it down in level, and then lets go as soon as the note starts to decay. This would actually INCREASE the dynamics in the track, and would probably require a limiter on the output after makeup gain was applied.

- A compressor with VERY SLOW release times could overlap the release into the next note, compressing the initial attack even further, leading to a time-dragging feel.

- A compressor with a high threshold and a heavy ratio will flatten out the peaks of the notes, but will leave the body and decay unaffected.

- A compressor with a very low threshold will compress the entire sound, and will make the attack and body blend into the decay, ambience, and noise of the track.

If you "tune" the compressor by setting the threshold low and the ratio high so that it catches every note, you can adjust the attack and decay times so that gain reduction "bounces" along with each note in a way that complements the natural dynamics of the track. Then you can back off the threshold or ratio to get more natural sound.

If you instead "tune" the compressor by setting a slowish attack and realease time, and then tweaking the threshold and ratio to get the right kind of pumping and breathing, you can then adjust the attack and release so that the the impact and decay sound natural and well-balanced.

Practicing both approaches will quickly give you an ear for the subtle ways that compression affects the sound, and you will be able to achieve the best results by tweaking everything in tandem. But remember that certain settings can have opposite effects-- with a longer release time, lowering the threshold could cause the release to overlap into the next note, killng your attacks. With a slower attack, increasing the ratio and lowering the threshold for heavier compression could actually produce MORE dynamic swing. And so on.

Every control is interactive, and every control depends on what is going on in the signal. Presets such as "rock bass" or "vocals" are basically completely meaningless. They might as well be labeled "random 1" and "random 2" when it comes to compression. The tempo and source material could make appropriate settings for one song have a completely opposite effect on another song with a different singer.
yep is offline   Reply With Quote
Old 01-27-2009, 09:41 PM   #249
guitartb
Human being with feelings
 
Join Date: Mar 2008
Posts: 14
Default

This is pure gold, yep. It's the missing link to reducing my ongoing compressor learning curve.

"If you "tune" the compressor by setting the threshold low and the ratio high so that it catches every note, you can adjust the attack and decay times so that gain reduction "bounces" along with each note in a way that complements the natural dynamics of the track. Then you can back off the threshold or ratio to get more natural sound.

If you instead "tune" the compressor by setting a slowish attack and realease time, and then tweaking the threshold and ratio to get the right kind of pumping and breathing, you can then adjust the attack and release so that the the impact and decay sound natural and well-balanced."
guitartb is offline   Reply With Quote
Old 01-28-2009, 01:04 AM   #250
LOSER
Human being with feelings
 
Join Date: May 2006
Posts: 2,373
Default

Quote:
Originally Posted by BoxOfSnoo View Post
This leads to a question. In last month's SOS, there was an interesting article how these guys recorded Chinese folk music, but got stuck with almost nothing to mix or record with. He said he used a technique to monitor without decent speakers...

He used the MDA limiter, with limiting cranked way up "to see what's ducking the mix".

I don't quite get this. Could you explain? Is it a viable technique?
I know a similar technique that uses a clipper. The basic idea is that you just slam the clipper (on the master buss) until the mix starts distorting, now everything just start distorting equally (= mix is balanced no major problems), when something starts distorting way before the other instruments, e.g. the bass gets all boomy before the kick or vocals start distorting, the bass is too loud or has too high peaks in its spectrum, etc...
Is this technique usable ... if you know what you're doing everything is usable, but if you don't you will just make things even worse.
LOSER is offline   Reply With Quote
Old 01-28-2009, 02:44 AM   #251
stupeT
Human being with feelings
 
stupeT's Avatar
 
Join Date: Jan 2009
Location: frankonia
Posts: 1,996
Default

Quote:
Originally Posted by LOSER View Post
I know a similar technique that uses a clipper. The basic idea is that you just slam the clipper (on the master buss) until the mix starts distorting, now everything just start distorting equally (= mix is balanced no major problems), when something starts distorting way before the other instruments, e.g. the bass gets all boomy before the kick or vocals start distorting, the bass is too loud or has too high peaks in its spectrum, etc...
Well. That causes one problem in my brain.

Actually I am only using my ears to judge a mix, so I'm not experienced to this technique which is also true for looking at analyzers on the master bus.

But I always thought a "balanced" spectrum is closer to the pink noise idea (-3dB per octave, or something like that) and that technique from above will only work on white noise like spectrum (equal in all octaves).

I think it could kindof peak vs. RMS difference which I do not get here.... hmm... the clipper is working on peaks, right?... hmm...

Cheers
stupeT
stupeT is offline   Reply With Quote
Old 01-28-2009, 09:37 AM   #252
DerMetzgermeister
Human being with feelings
 
DerMetzgermeister's Avatar
 
Join Date: Feb 2008
Posts: 1,400
Default

Quote:
Originally Posted by yep View Post
How a compressor works


Inside the compressor is a little gremlin that turns down the volume. That's it. Really.
I knew it!
__________________
"Everybody here is going to eventually die with hard drives full of half finished relics." viscofisy
DerMetzgermeister is online now   Reply With Quote
Old 01-28-2009, 10:19 AM   #253
LOSER
Human being with feelings
 
Join Date: May 2006
Posts: 2,373
Default

Quote:
Originally Posted by stupeT View Post
Well. That causes one problem in my brain.

Actually I am only using my ears to judge a mix, so I'm not experienced to this technique which is also true for looking at analyzers on the master bus.

But I always thought a "balanced" spectrum is closer to the pink noise idea (-3dB per octave, or something like that) and that technique from above will only work on white noise like spectrum (equal in all octaves).

I think it could kindof peak vs. RMS difference which I do not get here.... hmm... the clipper is working on peaks, right?... hmm...

Cheers
stupeT
I don't use that technique and therefor also don't adores it. But I once met a guy that swears by it (well not really, but he used it on occasions). Regarding the pink vs. white noise, the clipper is clipping the peaks, but not the frequency spectrum, so imagine a 1000Hz wave on top of a 50Hz wave, the 1000Hz wave gets "lifted" up and down by the 50Hz wave, while the 50Hz wave isn't really affected by the 1000Hz wave, because it has a much shorter wavelength (more details on this would go too far into the territory of amplitude modulation), so when the clipper chops off the peaks it chops off parts of the audio that both belong to the 50Hz wave AND the 1000Hz wave (because the 50Hz wave elevates the 1000Hz wave).
So whether this works or not can be disputed.
But let's not clutter up Yeps excellent thread.
LOSER is offline   Reply With Quote
Old 01-28-2009, 04:36 PM   #254
fOZf8
Human being with feelings
 
Join Date: Jan 2009
Posts: 1
Default

good thread...easy to summate if one needs a quick resolution as well....


IF you know what you like, ie, the sound you want to make, everything else is logical trouble shooting.

sure, it took me about 10 years to figure everything else out, but the last 3 have been easy as pie because of it. Do it yourself, don't read manuals, and you'll no so much more from it. It will become INNATE to yourself. I can spend weeks mixing a track, but in a pinch I can dial a track in thats half there in 30 minutes. Everything else I do to it otherwise is just preferential fun!

oh yeah, some reading is good, true, but don't approach ANYTHING about mastering and mixing UNTIL you've read Yamaha's Sound Reinforcement Handbook!
fOZf8 is offline   Reply With Quote
Old 01-28-2009, 07:36 PM   #255
yep
Human being with feelings
 
Join Date: Aug 2006
Posts: 2,012
Default

So let's talk about some guidelines for where to set these settings...

THRESHOLD approaches:

- set the threshold just above the "average" signal level if you just want transparent-ish peak compression, like a limiter.

- set the threshold deeper, below the "average" signal level but well above the noise foor if you want to actually modulate the sound or performance dynamics.

(I cannot give numbers, because it depends totally on what your signal is doing. Look at the meters.)

RATIO approaches:

- Any ratio above say 10:1 is basically acting like a limiter-- there will be VERY little dynamic variation above the threshold with these settings, EXCEPT as you allow via the "attack" window, or force via the "release." Ask if this is not making sense.

- Ratios of 2:1 or 3:1 will be very gentle compression, basically inaudible as processing effects, just giving a slight evening out of the signal levels.

- Ratios of around 4:1~8:1 will offer medium compression with some pumping

- As said above, ratio is totally dependent on the threshold

Attack and release later.
yep is offline   Reply With Quote
Old 01-29-2009, 08:12 AM   #256
shemp
Human being with feelings
 
shemp's Avatar
 
Join Date: Oct 2008
Location: Texas
Posts: 3,157
Default

ok, two questions for me:

1. Does a limiter compress? Meaning, I sometimes use the Kjaerhus classic Limiter and I *think* I can hear some compression but there are no threshold and ratio settings on it. Please explain?

2. Please explain the 2 knob compressors. Is it more of a pre-set threshold/ratio/attack/release in one knob?

thanks!!!!!!!!!
shemp is offline   Reply With Quote
Old 01-29-2009, 08:50 AM   #257
matey
Human being with feelings
 
matey's Avatar
 
Join Date: Apr 2008
Location: Civitavecchia (Italy)
Posts: 575
Default

shemp, about your first Q. I could say that a limiter acts like a compressor with an extreme amount of fixed ratio setting.
Actually you can count on the ceiling level <=0db, and the threshold value.

the more the threshold value decreases, the harder the average amplitude of the signal will be squashed to the wall of the given ceiling of 0db or less...

exemplum: if you push your face gently to a window people outside can see your noise (peaks) and, as you push your face harder, every part of your face will be clear in a bi-dimensional (distorted) way.... ah ah ah .. this is good...sorry..
matey is offline   Reply With Quote
Old 01-29-2009, 08:55 AM   #258
stupeT
Human being with feelings
 
stupeT's Avatar
 
Join Date: Jan 2009
Location: frankonia
Posts: 1,996
Default

Quote:
Originally Posted by matey View Post
exemplum: if you push your face gently to a window people outside can see your noise (peaks) and, as you push your face harder, every part of your face will be clear in a bi-dimensional (distorted) way.... ah ah ah .. this is good...sorry..
OK!

Which do you think looks nicer?!?
stupeT is offline   Reply With Quote
Old 01-29-2009, 08:57 AM   #259
Masarin
Human being with feelings
 
Masarin's Avatar
 
Join Date: Dec 2007
Location: Malmö, Sweden
Posts: 369
Default

Quote:
Originally Posted by matey View Post
exemplum: if you push your face gently to a window people outside can see your noise (peaks) and, as you push your face harder, every part of your face will be clear in a bi-dimensional (distorted) way.... ah ah ah .. this is good...sorry..
Simply brilliant! That was the best/funniest analogy I've ever heard.

/Jens
__________________
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Skodak sounds @ http://skodak.bandcamp.com
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Masarin is offline   Reply With Quote
Old 01-29-2009, 09:20 AM   #260
yep
Human being with feelings
 
Join Date: Aug 2006
Posts: 2,012
Default

Quote:
Originally Posted by shemp View Post
ok, two questions for me:

1. Does a limiter compress? Meaning, I sometimes use the Kjaerhus classic Limiter and I *think* I can hear some compression but there are no threshold and ratio settings on it. Please explain?

2. Please explain the 2 knob compressors. Is it more of a pre-set threshold/ratio/attack/release in one knob?

thanks!!!!!!!!!
1. "Limiter" is a bit of a fuzzy term. A pure, unadulterated brickwall instant limiter would be a clipper. I.e. it would simply clip the top off anything that exceeded the limit, like digital clipping. And this approach can actually be very transparent for short overs.

But most "limiters" on the market are actually very high- or infinite-ratio compressors with a fast or instantaneous attack and carefully-tuned release curves designed to have as little sonic impact as possible without actually squaring off the tops of the wave forms. How the designer approaches the release is what determines the sound and response of the limiter.

Digital look-ahead limiters actually slightly delay the output signal, which allows them to start compressing BEFORE the signal reaches threshold, which in turn allows them to modulate the very top of the waveform in ways that keep a microscopic smidgen of level variation, allowing extremely heavy limiting without the kind of obvious harmonic distortion that would come from a conventional instantaneous attack.

2. Yeah, exactly. For example, in optical compressors, the signal is passed through an LED or lightbulb that varies in brightness according to the signal strength. This in turn fires on a photovoltaic element of some sort (like a solar cell) that modulates the signal (i.e. reduces the gain) according the intensity of the light. Besause the light element does not respond instantly and has a certain delay before it achieves full brightness and another delay as it goes dark, there is a sort of built-in attack/release that varies according to the intensity of the light.

By selecting a "just so" combination of light source and photocell, a designer might achieve a continuously-variable response that becomes faster and slower according signal intensity and speed of change that sounds musical and natural at a variety of compression settings and on a variety of material. The designer might not need to add any additional attack and release delays. And a simple control to adjust the relative voltage sent to the light source could control whether it generally responded more quickly or more slowly.

Please note that there are also very fast-response, four-knob optical compressors, and slow-response two-knob VCA compressors. The optical type is just a little easier to visualize the operation of, I think, so that's the example I used.

You could also have 3-knob or 8-knob compressors, depending on how the designer decided to approach it. The famous LA-2A is basically a one-knob compressor plus gain (no wonder people like it!), as is the old Ross guitar compressor. More controls have been added over the years to make compressors more versatile for different kinds of signal and specific technical or creative goals.
yep is offline   Reply With Quote
Old 01-29-2009, 09:30 AM   #261
Tedwood
Human being with feelings
 
Tedwood's Avatar
 
Join Date: Feb 2008
Location: South Coast UK
Posts: 14,303
Default

Quote:
Originally Posted by matey View Post
exemplum: if you push your face gently to a window people outside can see your noise (peaks) and, as you push your face harder, .

Don't you mean nose peaks?

Yes, best analogy for a compressor I have heard - I am truly glad you didn't do the bare arse and photo copier one as my imagination can only stand so much
__________________
The grass is greener where it rains
Tedwood is offline   Reply With Quote
Old 01-29-2009, 09:43 AM   #262
stupeT
Human being with feelings
 
stupeT's Avatar
 
Join Date: Jan 2009
Location: frankonia
Posts: 1,996
Default

Yep,

can you talk about the feedback compressor design and what it does to the sound?

Cheers
stupeT
stupeT is offline   Reply With Quote
Old 01-29-2009, 10:16 AM   #263
matey
Human being with feelings
 
matey's Avatar
 
Join Date: Apr 2008
Location: Civitavecchia (Italy)
Posts: 575
Default

Quote:
Originally Posted by Tedwood View Post
Don't you mean nose peaks?

Yes, best analogy for a compressor I have heard - I am truly glad you didn't do the bare arse and photo copier one as my imagination can only stand so much
..err... yes, NOSE works way better than noise, even tho my nose is often noisy... whatever....my bad english...
matey is offline   Reply With Quote
Old 01-29-2009, 11:06 AM   #264
yep
Human being with feelings
 
Join Date: Aug 2006
Posts: 2,012
Default

Quote:
Originally Posted by stupeT View Post
Yep,

can you talk about the feedback compressor design and what it does to the sound?

Cheers
stupeT
In most modern technical compressors, the design is feedforward through a sidechain. If you take the opto compressor example above, it would work this way:

The signal comes into the compressor, and is split off into two seperate circuits. The main signal is fed right into the gain-modulated compressor circuit for processing, and a seperate "side chain" is fed to the LED or light bulb. This way, the plain unprocessed signal, complete with dynamics intact, is used to TRIGGER the compression that happens in the main compression circuit. That is feed-forward, and when you hear talk of side-chaining, it just means the ability to feed some other signal into the compressor's sidechain, so that for example you could use kick drum hits to trigger compression on a bassline to "lock" the two instruments together.

Feedback designs are actually much simpler. The signal only passes through the compressor once, and the level-detection circuit uses the output of the compressor. This is less precise, but some people like the slower, squishier sound for some kinds of applications. The sonic differences might not be very pronounced until you get into fairly heavy compression settings, but try it both ways if your compressor has a switch.

For technical compression such as targeted control of peaks, feedforward is usually better.
yep is offline   Reply With Quote
Old 01-29-2009, 06:30 PM   #265
yep
Human being with feelings
 
Join Date: Aug 2006
Posts: 2,012
Default

Aside:

The acoustics thread that I referenced at the very beginning of this thread has a lot less hits than this one does. I really meant what I said-- studio acoustics is an absolute basic. Anybody who is following this thread who has not read through the acoustics thread is missing a gigantic part of this stuff.
yep is offline   Reply With Quote
Old 01-29-2009, 09:17 PM   #266
mamm7215
Human being with feelings
 
Join Date: Jan 2009
Posts: 23
Default

Or let's put it this way, if you execute every great tip in this thread to learn how to NOT make your recordings sound like ass, and you DON'T treat your room, they'll STILL sound like ass...
mamm7215 is offline   Reply With Quote
Old 01-30-2009, 03:25 AM   #267
stupeT
Human being with feelings
 
stupeT's Avatar
 
Join Date: Jan 2009
Location: frankonia
Posts: 1,996
Default

Quote:
Originally Posted by yep View Post
Aside:

The acoustics thread that I referenced at the very beginning of this thread has a lot less hits than this one does. I really meant what I said-- studio acoustics is an absolute basic. Anybody who is following this thread who has not read through the acoustics thread is missing a gigantic part of this stuff.

Can somebody please post that link again? Cant find it anymore
stupeT is offline   Reply With Quote
Old 01-30-2009, 07:52 AM   #268
yep
Human being with feelings
 
Join Date: Aug 2006
Posts: 2,012
Default

Quote:
Originally Posted by stupeT View Post
Can somebody please post that link again? Cant find it anymore
You can find it here, very top thread:

http://forum.cockos.com/forumdisplay...aysprune=&f=29
yep is offline   Reply With Quote
Old 01-31-2009, 10:30 PM   #269
Smurf
Human being with feelings
 
Smurf's Avatar
 
Join Date: Jan 2006
Posts: 2,173
Default

New yep PDF is up!

http://www.filesavr.com/whydoyourrec...-1-09thread268

Enjoy!
__________________
Yep's First 3 Years in PDF's
HP Z600 w/3GHz 12 Core, 48GB Memory, nVidia Quadro 5800, 240GB SSD OS drive, 3 480GB SSD Sample/Storage drives, 18TB External Storage, Dual 27" Monitors
Smurf is offline   Reply With Quote
Old 02-01-2009, 05:20 AM   #270
spikemullings
Human being with feelings
 
spikemullings's Avatar
 
Join Date: Aug 2007
Location: East London
Posts: 1,697
Default

Quote:
Originally Posted by Smurf View Post
Excellent, Yep has his Boswell!

Thank you so much for this thread.
__________________
Cloth-Eared Hobbyist
http://www.reverbnation.com/spikemullings
spikemullings is offline   Reply With Quote
Old 02-02-2009, 03:58 AM   #271
yhertogh
Human being with feelings
 
yhertogh's Avatar
 
Join Date: Nov 2006
Location: Belgium
Posts: 1,461
Default

Yep,

first...you have a way with words man...it's amazing how you can make things so clear with using plain words, rather than the typical behaviour you see where 'audio experts' use very scientific terms to bury their lack of knowledge under. Thanks again! awesome work!


Quote:
Originally Posted by yep View Post
If you "tune" the compressor by setting the threshold low and the ratio high so that it catches every note, you can adjust the attack and decay times so that gain reduction "bounces" along with each note in a way that complements the natural dynamics of the track. Then you can back off the threshold or ratio to get more natural sound.
This is a very similar phylosophy as the one in Michael Paul Stavrou's book 'Mixing with your mind'. Come to think of it, your and his views on things are very alike i find. He takes this a step further to explain that compressors are like a safe. You crack the dials one by one. He uses the ARRT acronym for this, first you crack attack, then Release, then Ratio, and then Threshold.

First you set ratio to the maximum value, put release to the lowest/fastest setting, and lower the threshold so that the entire signal is compressed. Ignore the horrible pumping you hear, but focus only on the 'beginning' of the sound i.e. you focus on the attack button only. With this button you can create e.g. the thickness (or thinness) of a sound. If the material is e.g. a snaredrum you can almost 'tune' the size of the stick the drummer is using. The attack affects the size of the hit.

Once you are happy with that, leave attack alone and adjust release. Release essentially controls the groove, the volume envelope over time. Try to set it as slow as possible while still hearing a nice groove.

Then leave attack and release alone and adjust the ratio (which was at its max). You can think of the ratio as a sort of lens. High ratios the sound will be firm, but small. Lower ratios the sound will be bigger but softer (also less controlled). Lower the ratio until you loose your above created groove, then increase it again to get the groove back.

Then adjust the threshold so that some sound still gets uncompressed so that the compressor comes to rest 'in special moments' as stav puts it.

I hope this helps some people, and the above is not at all my invention. I just wanted to post this as i believe it is in the same vein as Yep's other comments in this thread AND it surely helped me to finally understand how a compressor works.

Yves
yhertogh is offline   Reply With Quote
Old 02-02-2009, 08:43 AM   #272
Smurf
Human being with feelings
 
Smurf's Avatar
 
Join Date: Jan 2006
Posts: 2,173
Default

Quote:
Originally Posted by spikemullings View Post
Excellent, Yep has his Boswell!
??????????
__________________
Yep's First 3 Years in PDF's
HP Z600 w/3GHz 12 Core, 48GB Memory, nVidia Quadro 5800, 240GB SSD OS drive, 3 480GB SSD Sample/Storage drives, 18TB External Storage, Dual 27" Monitors
Smurf is offline   Reply With Quote
Old 02-02-2009, 02:01 PM   #273
spikemullings
Human being with feelings
 
spikemullings's Avatar
 
Join Date: Aug 2007
Location: East London
Posts: 1,697
Default

Sorry smurf, its just me being a pretentious arse

James Boswell is known for his biography of Samuel Johnson but also for his journals chronicling their travels together, to the extent that his name has become a term denoting companion and (as in this context) chronicler.

Bet you wish you'd never asked now
__________________
Cloth-Eared Hobbyist
http://www.reverbnation.com/spikemullings
spikemullings is offline   Reply With Quote
Old 02-02-2009, 04:10 PM   #274
yep
Human being with feelings
 
Join Date: Aug 2006
Posts: 2,012
Default

A little more on compressor controls...

I left off describing attack and release controls because I was trying to think of a good, easy way to get started with them, but yhertogh's synopsis does a pretty good job.

These things have to be adjusted by ear, but having good meters helps give feedback to what you are doing. The recent REACOMP review at ProRec cited in the main Reaper forum actually gives a great overview of reacomp's controls for experienced users:

http://www.prorec.com/Articles/tabid...-A-Review.aspx

However, I'm not sure I would recommend REACOMP as a first compressor for a beginner, because the controls are so powerful and so inter-related. The bottom half of the control panel in particular is really advanced stuff, allowing you to design your own detection circuit. And unless you either already understand compression AND frequency in a pretty detailed way, or are extremely patient, it could be hard to make sense of.

But I do recommend reading the review. Even if it seems a little overwhelming, there is a huge amount of two-steps-forward-one-step-back to learning audio, and having some exposure to advanced concepts helps as your understanding grows into it.
yep is offline   Reply With Quote
Old 02-02-2009, 04:44 PM   #275
yep
Human being with feelings
 
Join Date: Aug 2006
Posts: 2,012
Default

Usually, compression (and almost all effects) should be adjusted with the whole mix playing, i.e. not by soloing one instrument at a time. Very often, what sits well and punches through a mix well is very different from what sounds ideal as a solo instrument.

That said, there are at least two, and more often three or four distinct stages to making a record. When you're only tracking one instrument at a time, it is obviously impossible to evaluate the sounds you're capturing in context. And for that matter, even during mixdown, it's impossible to compare any single element in the context of the whole, finished mix, because the mix is not finished until you have adjusted all the different elements.

I don't want to go too far into mixing approaches yet, because the stuff that we are talking about still has very real implications at the tracking and "pre-mix" stage, even if you track without effects.

For most engineers, there is a stage in-between straight tracking and full-blown mixing where you are doing some basic cleaning up and sound-sculpting just to get the tracks knocked into shape before you settle into the real task of mixing. I'm going to call this "pre-mixing." The specific boundaries between tracking, pre-mixing, and mixing can be a little blurry, but virtually every professional engineer does these as more or less separate stages.

Pre-mixing is all the processing that you do to a track before you actually sit down to mix them all together. In the analog days, the division was usually pretty straightforward-- anything you did to the signal BEFORE you recorded it to tape was "pre-mixing," and the realities of tape saturation, hiss, limited access to finite numbers of outboard effects, and tape's natural frequency alterations kind of forced you to get clean, clear, punchy, airy, warm tracks of reasonable signal strength if you wanted to have good tracks to mix with. Analog mixing consoles typically have eq and dynamics controls as well as effects returns for just this purpose (known collectively as "channel strips"). You would do obvious cleanup and intrinsic effects at the tracking stage, and set aside the real work of mixing for later.

In a commercial kitchen, this would be similar to the work done by "prep cooks"-- picking out wilted lettuce, sifting flour, making stocks and broths, chopping vegetables, trimming meat, making sauces and marinades, cutting loins into steaks of the right thickness and so on. Nothing immediately edible comes out of it, it's just getting the ingredients into shape so that the line cook can focus on cooking.

In a studio, the idea is to get tracks that not only sound good but that will be easy to mix without getting bogged down in humdrum technicalities. And this process is even more critical to be aware of in the DAW age where it is all too easy to just record everything to an infinite number of tracks with an infinite number of available processors and then have a gigantic mess of ingredients to pick through and manage while you're trying to actually cook.
yep is offline   Reply With Quote
Old 02-02-2009, 05:09 PM   #276
yep
Human being with feelings
 
Join Date: Aug 2006
Posts: 2,012
Default

In the example project that I posted above, both versions were over-processed deliberately to illustrate the ways that compression can radically alter the "feel" of a track. You can use a compressor to chop a track into short staccato hits or to flatten it into a gently pulsating pad. You can make it pump and suck in an off-time, funky way or you can lock into an exaggerated syncopation. The compressor's detection circuit combined with how your gremlin handles attack and release can make for some pretty drastic changes, to the point where it sounds like there is a whole different player or instrument.

One of the biggest things that trips up beginners is finding that "sweet spot" of how far to go in the pre-mix versus what decisions to leave for mixing. There is a tendency to either leave every possible decision for mixdown, or to "mix" each instrument one at a time and end up with a collection of tracks that all sound big, hype, thumpy, punchy, and so on, and that are impossible to fit together.

Have you ever tried to make your own sauces or soups without a recipe? If so, you have probably had the experience of making something that tastes absolutely perfect when you dip your spoon into the pot and taste it, only to find that it is way too heavy and over-powering when you actually sit down to eat a whole plate of it. A half-teaspoon on the tip of your tongue is very different from a whole meal of mouthful after mouthful. This is the culinary equivalent level-matched listening. If you make a roux with some cooked fat, flour, sugar, and salt together it might taste fantastic on the tip of your tongue, but try and eat a whole bowl of it and you'll be vomiting in two spoonfuls.

Pre-mixing is the art of making tracks that are clean, consistent, noise-free, well-balanced, and appropriately dynamic, so that they are easy to work with come mixdown.
yep is offline   Reply With Quote
Old 02-02-2009, 05:16 PM   #277
yep
Human being with feelings
 
Join Date: Aug 2006
Posts: 2,012
Default

I would encourage beginning mixers to get into the habit of saving pre-mixes as a separate, rendered project. For example, you track all your instruments, save the project as "minimum rage" or whatever, then go through each track and clean up and polish each track with mild eq, compression, gating, and any obvious effects such as intrinsic delays or guitar effects, and save. Then render each track with those effects embedded, and then save that as "minimum rage pre-mix."

Then use that project to do your actual mixing. If you have to go back, so be it. It might take a little trial-and-error, but it much easier and more intuitive to mix a project with cleaned-up, committed sound than it is to try and cook while sorting wilted lettuce and making stocks and so on.
yep is offline   Reply With Quote
Old 02-02-2009, 06:02 PM   #278
Smurf
Human being with feelings
 
Smurf's Avatar
 
Join Date: Jan 2006
Posts: 2,173
Default

That's OK spikemullings, I am glad I asked, it is always nice to know what your being called! But all I could think of was the Boswell from "Charlies Angels" and did not have a clue how that fit in! LOL :cool
__________________
Yep's First 3 Years in PDF's
HP Z600 w/3GHz 12 Core, 48GB Memory, nVidia Quadro 5800, 240GB SSD OS drive, 3 480GB SSD Sample/Storage drives, 18TB External Storage, Dual 27" Monitors
Smurf is offline   Reply With Quote
Old 02-02-2009, 06:07 PM   #279
yep
Human being with feelings
 
Join Date: Aug 2006
Posts: 2,012
Default

Bringing this all back to compression, it is absolutely standard operating procedure to use more than one instance of compression on every track. And compression does NOT automatically mean killing dynamics-- compression can actually make a track MORE dynamic.

Unless you're doing live broadcast work, there is no reason to use compression as an automated volume control to adjust the differences between loud and quiet passages. Fader automation is much easier and much more flexible these days. Use automation to even out the overall performance, and compression to affect the sound and the sense of intensity and performance vibe.

One of the reasons why I'm talking about compression early on, before getting into eq or reverb or even tracking specific instruments is that compression occurs naturally in all sorts of analog processes, and some of the best compression does not even use a compressor. If you listen to some older recordings of rock and roll, there is a great effect where the singer gets louder and more emotional, and the recording saturates and overloads, giving a terrific "effect" of loudness and emotional intensity, without much change in volume. The Temptations' "Ain't to Proud to Beg" is a great example, as are a lot of John Lennon's vocals. There is an explosive, analog "fire" on the intense syllables without actually varying the signal level all that much.

In recent years, there has been a kind of divergence, where cleaner, poppier, more "mainstream" records have avoided this kind of overload sound in favor of "cleaner" look-ahead compression and limiting, and where more "heavy" rock and metal records have tried to get that "overload" sound on every note of every instrument.

I'm not here to tell you what kind of sound you should go for, but there is a lot of potential to use the sonic illusions available to you to really make certain sounds "explode" out of the speakers with saturation and creative/intense compression effects. And having that kind of textural variation makes it possible get recordings that are fairly hot without becoming the constant white-noise earache of modern loudness-race stuff.

Stuff like old Rolling Stones or Velvet Underground has a very "analog" sound that sounds full-bodied and satisfying, even when quiet, and without degenerating into white-noisy fizz and "ringing phone" effects. By contrast, the latest Guns N Roses record sounds somehow too clean and un-ballsy in spite of being a very "hot" record. It somehow never seems to be at the right volume-- no matter how you adjust the volume control, it either seems too loud or not loud enough, which is a sad departure from Appetite for Destruction, which is a record that sounds exactly the way its' supposed to (for good or for ill).

There is perhaps no better example of what compression is capable of than the snare on Simon and Garfunkel's studio recording of "The Boxer." That giant explosion that somehow sounds like a gunshot or a bullwhip without overpowering or even sounding artificial against the soft, delicate vocal harmonies is a perfect illustration of how careful dynamics control (plus reverb) can give massive creative power to the studio engineer, and maybe even make a megahit from a single effect.

Compression is a big part of what makes a record sound "right" at a variety of playback volumes. It's not about making things sound louder or quieter so much as making them sound proportionate and "right" in a dynamic sense. It is the closest that a mix engineer gets to actually playing an instrument, because it affects the sound in exactly the same ways that a really good singer or player does-- it alters the texture and tone of the sound in real-time, dynamic ways.
yep is offline   Reply With Quote
Old 02-02-2009, 06:11 PM   #280
yep
Human being with feelings
 
Join Date: Aug 2006
Posts: 2,012
Default

Quote:
Originally Posted by Smurf View Post
...Boswell from "Charlies Angels"...
I like that one better! (cuz it makes me a hot spy chick instead of a crippled, impoverished curmudgeon)

Although I actually tried the link to read my own posts, and it didn't work. Is it a broken link or my broken computer?
yep is offline   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -7. The time now is 11:26 AM.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2019, vBulletin Solutions Inc.