All about Compressors

[jens]

Boah Jens,

also das mit dem Namensgedächtnis, da bin ich nicht sicher ob ich Dich verstanden habe. Was meinst Du? Äh ja, mein Namensgedächtnis ist recht gut. Und?

Ich kann Dir da auch leider gerade nicht wirklich folgen mit dem GCO-1 und TubeLimit. An den thread zum Thema compressors erinnere ich mich schon. Was wolltest Du mir denn jetzt sagen?

Und sicher, in den letzten postings zu schlechten plugins waren wir 100% in der Witzelzone. Ich habe auch Deine Aussage, dass meine Musik dann ja schlecht klingen müsse als innerhalb der Witzelzone gewertet, kein Problem.
Trotzdem eine willkommene Gelegenheit auf meine Website mit meiner letzten CD hinzuweisen.


Alles Gute, FRitz

oh sorry - dann hast Du so gelungen gewitzelt, daß ich der Depp war, der nicht mehr mitgekommen ist... - ich hab' daher gedacht Du könntest Dich nicht mehr daran erinnern, daß ich die beiden genannten plugins über den grünen Klee gelobt habe (und ergo meine Aussagen nicht ernst gemeint sein können)

[fritzman]

Hehe, alles cool Jens!

No prob. Alles easy. War lustich, gell?


Best wishes, FRitz

[bmanic]

Got some new sound examples from a home built compressor that works on the Vari-Mu principle. It has 4 tubes per channel driving the thing and is the first prototype of what will maybe be sold commercially in the future. It's estimated cost of production is around the 3000-4000$ mark.

Here's how it sounds at quite extreme settings (it's real ace is doing subtle mix compression but as this whole thread is about compressors at extreme settings I wanted to keep it going):

Example 1

Example 2

Example 3

Compressor Overdriven, no gain reduction, just pure tube saturation

Here's the original, un-compressed file

Cheers!
bManic

EDIT: just noticed that the level of the compressor examples are much louder than the original so please, compensate with your host while listening!

[bmanic]

Oh btw, anybody who can mimic these examples with a plugin compressor gets a chocolate cookie from me!

- bManic

[AKJ]

Hi bmanic,

this kind of heavy compression is usually not the sound I am after, but if I want it I go for blockfish, and imo it's close enough - so no need for me to self-assamble an analog comp

if you think I qualify for the chocolate cookie: how to receive it? here is the link:

www.power-xs.net/download/BlockfishExample.wav

btw: for me the original example also sounds a bit compressed

[bmanic]

Well, maybe it's just me but your example sounds like a typical plugin would.. don't get me wrong, I love blockfish but it's way of letting trough the transients and the 'resolution' of the release envelope are quite a bit worse sounding than the examples I posted. Sorry, no cookie for you but here, have a penny for the effort!

Btw. you guys can use the examples and all the other ones posted thru out this thread as good ear training for compressors. The symptoms of 'bad' compression are about the same in both the digital and analogue domain. Spiky nasty transients get trough and the release stage 'breathes' wrong. Just check out kylens examples from the TLA CL1 analogue comp for a very different flavour to the one I posted but again, the good signs are there. It's smooth and can be also punchy without being 'clumsy' or how should I say it, digitally snappy (like a nasty overload on a microphone when recording a snare drum.. similar to that sound)?

Cheers!
bManic

[AKJ]

well, to me it sounds better if some transients get through. But it's very easy to fix this with limiting. so check this:

http://www.power-xs.net/download/Blockf ... xample.wav

cheers, AKJ

I don't give up so easily to get my cookies

[bmanic]

the problem isn't about the transient's getting trough or not but rather how they sound when they get trough. I'll listen to your second example once i get home but don't be expecting no cookies!

[fgimian]

hey guys, bmatic, your samples sound great! AKJ, nice try bud, but I can easily hear the "plugin sound" I'm afraid. Not a bad sound though

The transient problem is also what I dislike about most plugin compressor ...

However, I have my own sample to share, I'll post it very soon. I think that it comes pretty close to the analog sound personally and doesn't let the transients through yet it doesn't squash them. I'm going to reach for this cookie too

Take Care
Fots

[urosh]

Nice thread this turns out to be, very educational

bmanic, i noticed issue u had with 3rd harmonic
distortion earlier in the thread, so just to put
some light on that topic :
basically this is oversimplification, but will
do for explanation. Comps and limiters determin
signal level by doing full wave rectification of
sort ( or squaring in case of rms detection ) and
they filter out that to get smooth(ish) control
signal for gain element. Now since u have finite
time constants, u cant smoothout completely so u
have residuals of peaks of waveforms in control
signal. So around wave peaks of signal u r copmressin
u have smaller gain in gain element than around zero
crossing, that is gain is signal dependent during
single waveform cycle, that is u have distortion.
Since u are doin FWR/squaring in detector, this
distortion is "squashing" both posiive and negative peaks
so u have dominant 3rd harmonic distortion ( i know
this sounds a bit confusing, one nice picture would tell
it all much better ).

And bout samples, do u notice how differently varimu
pumps room sound compared to TLA. Breathing is not
totaly signal dependent , its like comp almost has a mind
of its own. Its sound like its on edge of goin out of
controll but it behaves somehow. That is (imho) second
part of varimu mojo . TL behaves well and predictable
which u would expect from solid state gain element and
(probably) feedforward configuration. Also notice how
on Oooohdear sample tla craps ( but in a nice way )on
every kick while varimu holds on.

>>>>>>> philosophical rant ahead, ignore <<<<<<
imho, since modern production = beat detective and
sample replace everithing into oblivion, u can gues
why engeeners like to bring some life back into
track with varimus and alike
>>>>>>>>>>>>><<<<<<<<<<

And people, notice one thing on overdriven example. While
kick get squashed on every hit high end survives very
well. In digital u usually have tons of hf harshness
with tis amount of saturation. Thats what people like
bout tubes : in well designed tube device distortion
will first start to be aparent in low mid region. And
tape , amount of hf will actually decrease with more
saturation.

And finally, on drum buss compresion. One common trick
everybody mentions is parallel comperesion with some
super nasty cheap limiter on compression buss (mxr mini
limiters and talkback mic limiters in SSl desks are
legendary for this task). In virtual domain i found
classic master limter to work for this task
(thou, witout much personality).

cheerz
urosh

ps: that last block fish example is very nice, this thread
really did well in moving u people to do some great
pushin-to-the-final-limits-of-technology work

pss: imho psp master comp hands down beat every other
comp ( including uads ), just wish it can be used
on more chanells before cpu crapsout

[urosh]

Just one remark bout that beat down in last paragraph :
compression is quite a bit a matter of personal taste
and someones mixing style will determin wether he or she
likes certain comp or not. With master comp i was able for the first time to hear every parameter change without
comp being too agressive, thats it. Its like, when i fiddled round with attack on master comp i was : woah,
attack finally works the way i want. Just, dont use
its builtin limiter, it kills the sound a bit. Use pocket
limiter insead.

cheerz
urosh

[fritzman]

Hi bmanic,

that's really impressive. I don't like the long times on the very squashing examples but example #3 sounds really good. Would you do yome heavy compression with shorter times, please? I would like to hear some smashing without the pumping. Thanx a lot.
And no, the Blockfish isn't really close according to my ears. Where it absolutely breaks is the kick. Though it's definitely one of the better soft comps.


Best wishes, FRitz

P.S. Hey urosh, great insights there, thanx.

[Amberience]

I'll listen to your second example once i get home but don't be expecting no cookies!

So you made your mind up prior to listening then??

[urosh]

Notice few things : how that vari mu comp is sucking bassdrum
when it punches in, blockfish spits this undefined
block (great word for this efect bmanic, btw ) when
atack detection kicks in; how cymbals swell on varimu
samples compared to jumpy action on blockfish. And this
if I may say it, hifi quality of varimu; when bassdrum
kicks in ,comp eats it but u still hear whole basdrum
tone, unafected. Blockfish obscure distinctive features
of transients, that bd sounds like samplereplacer working
somewhere behind in signal chain, spittin some electro
bd. Also,even with much more swelling of cymbals/ambience
on that varimu example, u have room sound (in term of
tonal quality ) unafected, even enhanced a bit, where
blockfish introduce some harshness and obscure details.
Simply, u can really hear that varimu comp working
(if u have idea what i mean by that).
While this much compresion will obscure spatial cues
whatever comp u are using, point is what will sit in
mixdown and in what way. It could easily be that a bit
of harshness and detail obscuring blockfish introduces
will make u pull it up in the mix to actually hear that
room sound, but that will just put harshness more into
focus (this is just one posible scenario of course; if
something is too far in front in the mix some taming
compresion will pull it back).
Also, limiter wont help much, its character of compression in question, not transient overshoot. Notice
how big is overshoot on that first varimu example, worse
than blockfish without limiting.
>>>>>>
No wonder that everybody ranted at farchilds back in the
days they were new. Remote cutoff tube configuration
is sluggish in atack department for limiting tasks. No
wonder everibody raved with 1176 when it got out,it kills
transients like there is no tomorow. But varimu character
is something completely different....
<<<<<<<<<<<

Bmanic, u allmost get me into building one varimu for
my self. Please, pretty please with sugar on top, can
u post some more examples of that varimu working its
ass of but on some other instruments , like bass, voice,
bass+(drums-cymbals) ...

And, good people of kvr, me thinks imho this tread should
go sticky, with continuing addin examples of various
comps soft and hard. Great thing is this didnt turned
out into compwars pt.VII , people rather sat and tryed
pulling out most of their comps , some examples really
showed me at least that some comps can pull out nice
stunts with a bit of work, it was quite educational.

cheerz
urosh

ps if this tread remain on living side of flat earth
I will try to put some fatso examples next week. Plus
I have Presonus Bluemax somewhere, if its still alive
it would be very informative to compare low end analog
comp to software ones. 5000$ high end gear is sort of
no contest, me thinx

[bmanic]

I'll listen to your second example once i get home but don't be expecting no cookies!

So you made your mind up prior to listening then??

Well, kind of yes, as he named the file BlockfishBuzzMaxi or something.. which kind of gave it away as I've tried like every plugin that does any kind of dynamic alteration on the planet (including most TDM and UAD/TC ones too). So, yes, my apologies.

Btw. I listened to the clip, and still stand firm, it is light years away from the smoothness of those tube compressor clips I posted. Transients still sound bad.

@fritzman, hi! (sorry for not answering on the synthmaker forum!)

I'd post examples with shorter release if it was possible but unfortunately the compressor prototype WAS at the shortest time! I have no idea what the settings were (no labels on the compressor yet!) but it had highest ratio, threshold maximally down (except example 3) attack fastest and release fastest. I'll run some more examples trough that beast later this week or next week and will post some pictures of the compressor (provided that I get permission from the creator but I can't see any reason why not, other than because it's a rather uggly prototype ). Btw. he already has 3 people who want to buy one, two mixing enginers and a talented mastering engineer. It's an impressive kit that starts giving some slight distortion after +30dBU input, that is, it can take a helluva beating and it still sounds nice!

If any of you guys have some samples in 16bit 44,1khz .wav format then I could run it trough that compressor. I'll try to run lunchmoney's track that he sent me trough that thing and see what magic it can do on a whole mix. I'm sure he'll post the results later.

@urosh,

Thanks for the info man! Lots of stuff I need to learn still about all the technical aspects of compressors so keep em coming (Aleksey, wanna drop in on the discussion? ). You had some very good descriptions of what to listen to, regarding analogue vs digital compression. Aall the things I mentioned about plugins are also true for the Sintefex FX8000 I have sitting on my desk and the TC6000 built in compressor, those are some helluva expensive 'plugins' and they don't do compression perfectly, far from it, but they both still have a slight edge over most plugins I've ever heard.

If you have any samples that you want to hear trough the compressor then PM me or post the clips here for everybody to use.

@soundpalace,

You want the cookie? Show me da beat (and if you pull it off, show me the plug!). Btw, please make sure the filename isn't anything that gives away what you used as I'm just a mere mortal who can easily make biased decisions, like we've already seen!

Cheers!
bManic