Yet another question about Compression.

Posted: Thu Jan 10, 2008 7:08 pm

If using a compressor reduces the frequencies, maybe it should not be used for music if you want to keep the dynamics. However, for radio it is fine. For instance Howard Stern is always saying 'give me more more compression' to the engineer. For him, it means that he does not have to talk as loud, since he is doing it for five hours.
For singing, the singer should learn to move back as he or she raises the voice. Some frequencies are only present when a voice gets louder, the overtones from the false vocal cords. Tuvan throat singers make use of these folds to sing multiple frequencies at once.
If you got the chance to record them, a compressor would be counter productive.
Radio stations compete with each other on the dial by using compressors, they want people to stop on the loudest station. It is insanity.
Sirius does not need to do that, which is great.

Posted: Tue Feb 26, 2008 7:11 pm

oh wow my brain hurts . greatful for the info, and explained in laymen's terms--absolutly wonderful. Exclamation

Posted: Wed Mar 19, 2008 11:34 am

Wow. A post for the ages, Shotgun! I love your colorful analogies.

And for the sake of further understanding of the less technical nature, I've been sharing this link as much as possible. Something for every engineer to think about.

http://www.rollingstone.com/news/story/17777619/the_death_of_high_fidelity

Posted: Wed Mar 19, 2008 11:40 am

BluesDimeBag wrote:

If using a compressor reduces the frequencies, maybe it should not be used for music if you want to keep the dynamics. However, for radio it is fine. For instance Howard Stern is always saying 'give me more more compression' to the engineer. For him, it means that he does not have to talk as loud, since he is doing it for five hours.
For singing, the singer should learn to move back as he or she raises the voice. Some frequencies are only present when a voice gets louder, the overtones from the false vocal cords. Tuvan throat singers make use of these folds to sing multiple frequencies at once.
If you got the chance to record them, a compressor would be counter productive.
Radio stations compete with each other on the dial by using compressors, they want people to stop on the loudest station. It is insanity.
Sirius does not need to do that, which is great.

I can't begin to tell you how much I disagree with this point of view. Radio compression is necessary because they have limited bandwidth and HAVE to compress the signal in order to get it out there. Sure mic technique is important, but damn. If you don't have something USEFUL to say, go say it at harmonycentral. Howard Stern? Really?

Posted: Sun Jul 06, 2008 9:46 pm

i suggest getting a copy of the recording engineers handbook

Posted: Sun Jul 06, 2008 9:54 pm

I smell something whiffy, some sort of packaged meat or meat-based product?

Posted: Mon Jul 28, 2008 1:47 pm

Ok...bluesDimeBag has demonstrated a "fundamental misunderstanding" of compression. He understands that compression is used in radio, but doesn't know what, how, or why, and so, let's help him, shall we?

...radio stations DO use compression, on ALL of their signal. Why? it has to do with two things:

First: the law. The radio station has a license to broadcast on a certain frequency(center), and at a certain maximum power. If they go over their rated power, then they get fined. If they go significantly under their rated power, then they are broadcasting to a significantly smaller area, thus a smaller listener base.

Second: perception. As you flip the dial, you are drawn to the stations that have a clear, strong signal. Most certainly, the "seek" button on your car stereo is convinced that the clear, strong signal is the one to stop on.

Ok, I lied...Third: listener environment. MOST radio listeners are in what we would characterize as a "disadvantaged" environment for listening...in a car...or, the radio is on as background...like at work.

In each of these situations, the radio station compressor helps to do something to the program material...it CHANGES the relationship between the "average amplitude" and the "peak amplitude". Specifically, by lowering the peak amplitude (i.e. 'compressing' it), it changes the standard deviation from the average amplitude, makes it smaller.

So, for problem 1...the station can bring the average level up, without the peaks going over their licensed power rating. For problem 2...the higher average volume makes the station "sound louder" than the other stations around it on the dial. For problem 3, the signal remains intelligible against the background noise in the listener's environment. Also, because the average sound level is higher, the nominal signal-to-noise ratio is improved, for the transmission chain.

When Howie Stern calls for more compression on his mic, that's a different thing. His mic technique is to get very close to the mic, and take advantage of the PROXIMITY EFFECT that boosts the bass on his voice, and to do so, he has to speak fairly softly. This does two things....first, it makes him sound rich and full...and second, it makes his delivery more "close" or "intimate". Using a mic this way results in a lot of variation, however, and the compressor is needed to smooth things out.

dwoz

Posted: Mon Jul 28, 2008 2:45 pm

So...shotgun has impeccably addressed the "what" of compression...what is it?

Now, let's get started on the "why".

Why use compression, within a mix?

Two main reasons.

First...to bind, or "glue" mix elements together.

Second...to separate mix elements from each other.

Third...to fix deficiencies in the mix element.

Number three is easy to understand. You have a bass part or singer with a fair amount of amplitude variation, where the desired effect is to have less variation. Applying compression will help to smooth out those variations. Or, you want to add some "body" to the part...a common technique is to take a parallel copy of the part, run that through a compressor, hitting quite hard, and bring that parallel track up along with the untouched track. (might need time-alignment). Such a technique leaves the peaks intact, while also raising the apparent "body" of the track.

There's a hundred more techniques like that.

For the FIRST point..."gluing" parts together...people will often buss several parts together onto an aux buss, and apply the compressor to the buss. This makes it so the compressor is responding to the overall level of both(multiple) inputs, and is compressing all inputs based on the most dominant (or "peaky" ) input. Net result is to tend to put all the elements included on the buss in the same sonic space.

The MOST SECRET and amazing use of a compressor, is to SEPARATE mix elements.

Let's go there.

Sidebar: you're looking at a photograph, a very nice photo, well composed, well exposed, well printed, well lit. In a gallery.

You notice something about it. The photograph has a characteristic "depth of field". The elements in front, close to the lens, are in sharp, crisp focus. The highlights are stark, the shadows dense. As you move "farther away" into the background, you note that the focus gets less sharp, and the highlights and shadows tend more toward grey. This is also an atmospheric effect, not just a lens effect. The air itself diffuses distant elements.

Now, surprisingly enough, the exact same principle applies to DEPTH OF FIELD in a music mix.

mix elements that are up front are detailed, crisp, with high dynamic range.

Mix elements that are away in back are softer, less transients, and less dynamic range.

Try an example? ok. Take a song that you're working on, and take two lead vocal takes. the best two...take 3 and take 11. bring them both up, pan them at 9:00 and 3:00 (i.e. a little away from each other to either side). level match them as best you can. Now, on take 3, apply a compressor, where you use a LONG attack, and a fast(ish) release. Go for a 10 db threshold, with a 2:1 ratio. Now, over on Take 11, set a VERY FAST attack...like, 0. set a slow release. set the same threshold and ratio.

Now, listen. I will bet you ANYTHING, that the fast attack take sounds farther back, and the slow attack take sounds in your face. pan them back to center, on top of each other. Hear the separation? Now, bypass the comps, and listen...hear the separation go away? Now, treat the "background" take with a very tiny amount of delay or just a TINY touch of reverb.

Wow.

You're used to using WAY more reverb than that, to get that "depth" thing going...

To recap....dynamics processing can affect the perception of DEPTH OF FIELD. things that have less transient attack are perceived as being farther away. Things that have a little high end rolled off sound farther away. Add it all together, and you have a soundstage.

dwoz

Posted: Mon Jul 28, 2008 11:05 pm

Mr Dwoz, thanx for yet another succinct and seemingly effortless post on music production realities.

Good to see ya around these parts.

Thanx for sharing.

Readership.....of course hes 100% correct.

Posted: Thu Jul 31, 2008 11:20 pm

Dwoz / Shotgun,
Thanks for those nuggets of wisdom. Your depth of field explanation was a real eye opener for me. I think it is really cool that there is a few of you experienced guys out there willing to share their experience with us.

This may be outside the scope of this thread, but I often wonder about the different compressor types like the FETs or Vari-Mu’s, etc. Do they have a sonic character that pairs them with certain instrument types or task. People are always talking about a Fairchild on the 2-buss, SSL on the drums, or a LA2A on the vocals. Rules of thumb or nostalgia?

I do realize that the bottom line is run what ya brung and use your ears creatively to get the job at hand done quickly with the best possible quality.

Posted: Tue Aug 05, 2008 3:02 pm

There's an encyclopedia brittanica's worth of material to cover in that question, so I'll just address a couple of general points.

First off...a question. Why do you use different mics on different sources? Why not just pick one very good mic, and use a lot of those?

The answer is a little more complex than "certain mics are better-suited to certain sources", and as simple as "because they sound different".

Think about it. A certain microphone has it's own particular "sonic signature". Even high-priced measurement mics. So, for a couple or three tracks, no big deal...but let's say you use the same mic on 14 different sources, all combining into one mix. the problem is, you start multiplying that sonic signature in your mix! it isn't just freq. response either. Transient response, how "fast" the mic capsule and electronics are...whether there's a lot of hysteresis (or not).

And when you get that "sonic signature" buildup, you can find yourself struggling against it.

The same phenomenon happens with mic pre's ...one of the reasons for "boutique" pre's, is to break up the monotony of using only the console pre's...not because they sound particularly BETTER, but more because they just don't sound exactly the same! and next to the summing buss, the mic pre's are a major component of a console's "sound".

There's times that you WANT the same sonic signature...like in stereo pair situations...but...

anyway, compressors are different than most other pieces of audio gear, because they seem to impart more "self-sonic-signature" to the programme than most other types of signal processors. In fact, you'll hear anecdotal stories of how so-and-so runs his mix through an 1176 just to impart some of that mojo...with compressor settings that won't do anything.

Because of this, it's a good idea to mix them up a bit, brand-for-brand, type-for-type.

Consider it like garlic (I'll digress into the 7th deadly sin of forum posting...using a cooking analogy). A little garlic is fantastic, but keep adding teaspoonfuls, over and over, and well....that's a bad analogy, because you simply CANNOT have "too much garlic".

Let's pick "oatmeal" instead. great stuff. can't get enough, really. But enough "oatmeal" in a mix is DEFINITELY too much oatmeal. You should never have "enough" oatmeal.

Unfortunately, the final answer to your question is, "It's mostly about taste". Generally speaking, units with a lot of "personal color" are ill-suited to be mix buss compressors. But, in a given situation, that may be just the thing.

It's decidedly less of an issue with software (plugin) comps...but the issue doesn't go away. Especially if you're using an "emulation" plugin...one that's TRYING to sound like the real hardware. Most often, they don't really achieve that, but they often DO achieve coloration.

dwoz

Posted: Thu Aug 07, 2008 10:51 am

Thank you, Dwoz. That makes good sense to me.

Posted: Tue Sep 02, 2008 12:42 am

New to the forum, and nearly as new to recording.
Like many newbies, compression was almost taboo to me -
didn't understand it well enough to use it w/o fear of doing damage,
so just leave it alone.
This thread has added to knowledge I've garnered from experience and other sources, particularly Shotgun's post(s).

Here's my add, as someone working from the other end of the experience spectrum:
Worry more about the "why" than the "how" or "what".
Those will follow naturally from the first.
Use your ears.

I had a tendency to just slap preset software compressors on tracks, certain the manufacturers knew better than me. While these presets may get you in the general neighborhood, city, or state, that's all they do.
I like to think of compression as fire - an essential tool that can do amazing things, but if you go beyond your skill set, you end up burning down your own home.

And sometimes you don't need a compressor on an indivd'l track. If you have good mic choice/placement and preamp choice/settings, then EQ, panning, and reverb or delay *may* be all you need to make it sit right in the mix. In my early projects, a lot of my rough, relatively untouched mixes sounded better (albeit quieter and well, "rougher") than the suped up versions.

As a musician turned audio engineer, my main focus is on capture rather than editing. Or tracking rather than mixing. I prefer to get as good of a track as possible and let that do the talking. None of that focus applies to mixdown, of course.

AND, if you use a DAW, avoid compression during tracking - compression is a destructive process, and you can't undo what you put in the DAW. If it sounds messed up, you can always return to the original tracks in the DAW - assuming they're not messed up.

I may be way off, and have reiterated a lot of older points. But these are some guidelines that have helped me learn to use compression more appropriately.

Posted: Tue Sep 02, 2008 6:34 pm

Keep learning like that and you'll go far.

Goanyersel!

Posted: Thu Sep 11, 2008 10:04 am

Shotgun, that article deserves a grammy (or the journalist equivalent).

I've been struggling with this very issue. Thanks a stack and thanks to all the contributors who take the time to pass on insight to the helpless souls like me.

This topic sponsored by: