Question re: Bobby Owsinski Compression Tutorial

steakmusic

steakmusic

New member
I've been checking out Brian Lee White and Bobby Owsinki's tutorials on Lynda.com, which are really comprehensive and clear, but I'm a little perplexed by Bobby Owsinki's compression technique.

He seems to employ what I think is referred to as the "safecracking" method of setting compression parameters. His method is to start with the longest attack setting possible, the shortest release setting possible, and then dial the two parameters in from there.

What confuses me is that when he talks about using a compressor on a snare track to "help even out the dynamics of the snare hits," he uses an attack of 125 ms and a release of 294 ms. If you're trying to control the dynamics of a snare track, which contains transient material of varying intensity, then it seems logical to me that you would want quite a fast attack time, in the neighborhood of perhaps 10-20 ms or less. With an attack time of 125 ms, I would think that the transient portion of the waveform has already long since passed by the time the compressor starts reducing gain.

So my question is: if you're trying to even out transients, why would you use a long attack time on your compressor? If anything, I think you'd just be clamping down on the "sustain" portion of the snare hits, making them even quieter relative to the onset of the snare hits.

Obviously Bobby Owsinski is a total pro, and he has at least 100,000 hours more experience than I do in the realm of mixing (I'm just a peon who started learning this stuff several years ago), but I just want to try to understand the logic behind this compression technique that he's detailing.

Thanks!
 
The main parameters your missing is the ratio settings he had and where was the threshold. If his ratio was 4:1 or even 6:1 then there would be no need for a fast attack a medium attack and slow release as what was shown would suffice. Also if his threshold was low the same would apply.

And the "safecracking" statement made... when setting a compressor, a long to no attack and setting a fast release is quite normal. Again the thing your missing is where Bobby had his ratio & threshold those parameters are vital.
 
steakmusic said:
So my question is: if you're trying to even out transients, why would you use a long attack time on your compressor? If anything, I think you'd just be clamping down on the "sustain" portion of the snare hits, making them even quieter relative to the onset of the snare hits.
Click to expand...
100% depending on the compressor just to "even out transients" you do not have an attack time of 0.1mm to accomplish this. I mean your not limiting you just want control. In some instance (of course adding ratio & threshold in the equation AND...Where the actual signal is hitting) having a medium attack. You also leaving out the release in this statement if you have a slow release, Hello! This is evening out transients cause your holding them. There is no reason to have a fast attack and fast release that would result in pumping. So if your relaying on the release parameter to do most of the work by holding the transients there is no need for a fast attack just a moderate one for "control" would work fine. Again whats more important and missing is the ratio & threshold level. For example

Im trying to control transients
Thresh - -10dB so I am REALLy crushing the signal with this thresh setting this will def control the transients if I do nothing else from here.
Ratio - (This is where compression really occurs) 4:1 here every 4dB of the signal it get turns down to 1dB so a lot of compression
Release - 250m, A long release time so I am holding the transients
Attack - Well no need to set a fast attack now, Im crushing the transients with thresh, Im compressing heavily with ratio and, Im holding the transients with a long release. A fast attack here and I might as well scrap the comp all together and just get the L2 out.
 
When setting a compressor from start

1.) Set all parameters to 0 to where no parameter is effecting the audio, so no effect
2.) Set release to the fastest setting (cause you dont want any interference as your about to set the thresh)
3.) Next thing you adjust is threshhold level to just kiss the signal first, once you get 1-2dB of gain reduction stop!
4.) Next adjust attack, might have to readjust release again you want the release & attack to be in sync with the rate of the signal (BPM, flow, etc.)
5.) Set Ratio cause this is where the compression really happens

(there is knee as well which is adjusted as the approach to compression hard to soft knee, soft knee is less notable hard knee more noticeable.)
 
jaynm26 said:
Attack - Well no need to set a fast attack now, I'm crushing the transients with thresh, Im compressing heavily with ratio and, Im holding the transients with a long release. A fast attack here and I might as well scrap the comp all together and just get the L2 out.
Click to expand...

Thanks so much for the detailed responses! Yeah, just as you suspected, I failed to pay closer attention to the threshold and ratio settings that Bobby was using on the DigiRack Compressor/Limiter, and they were indeed low (for threshold) and high (for ratio). The snare track looked like it was peaking at -10 dBFS before any processing, and he set his compressor threshold way down at -33.9 dBFS. Additionally the ratio was nearly 6:1. This was yielding 3-6 dBs of gain reduction for the snare hits.

So, as you said, in spite of the "slow" attack time, the compressor was already being triggered quick enough to deal with those transients, due to the low threshold he was using.

This was kind of a foreign approach to compression for me, because whenever I've dealt with transient material, I've always had the philosophy of fast attack+fast release, using a higher threshold relative to where the audio signal is peaking. So it's good to now know another method of evening out drum dynamics. I'm going to have to start experimenting more with using lower thresholds (coupled with slower attack and release times).

Thanks again!
 
^ This is good tho know you know more about compression now. Compression is proly the most difficult tool to get down. You tackling the single most difficult thing so kudos to you. Member dont stop reading and researching thats how you grow. Im 10 yrs in and I am STILL a student to the arts and it will never change. So keep doin what you doin man you will be a great engineer just cause the fact your not doing what 90% of young engineers do which is...Go buy a bunch of equipment spend thousands dont know how to use any of it then find us here are another forum and try and get the entire arts in a few post :wtf: . Just cant happen gotta study put in work.
 
Absolutely, I hear you--I had a brief period of hubris a little less than a year back when I thought, "oh...OK, I totally get compression now." Obviously I didn't hold on to that mind set for too long!

Compression really does seem to be the least intuitive of all the elements of mixing, but asking questions and having hands on experience helps slowly but surely. Still lots to learn about multi-band compression, creative side-chaining, compressing for tone/effect, multi-stage compression, etc. etc.
 
steakmusic said:
Absolutely, I hear you--I had a brief period of hubris a little less than a year back when I thought, "oh...OK, I totally get compression now." Obviously I didn't hold on to that mind set for too long!

Compression really does seem to be the least intuitive of all the elements of mixing, but asking questions and having hands on experience helps slowly but surely. Still lots to learn about multi-band compression, creative side-chaining, compressing for tone/effect, multi-stage compression, etc. etc.
Click to expand...
I mean not to get into but just cause u stated

Multi-band - aint nutting but 4 or 5 compressors you got the basics down you can handle this one (you will rarely use this one)

Side-Chaining - just use same technique and set the key in put or other signal to control the compressor either raise or lower the compressor based on the incoming signal. Again you got the basics so half the battle is won.

Compressor for tone - This is the one that will take time cause you have to listen to the character each comp as you go thru life as an engineer this is the most fun one lol.

Multi-Stage Comp - you got this as well set the 1st comp heavy set the 2nd comp light simple so the first does all the heavy lifting and the 2nd one does peak control. (This will rarely be used as well unless you get into mastering then it might occur more often)
 
Excellent--thanks again!

Yep, I've had at least one somewhat successful use of the Waves C4 multi-band for some problematic low-mids on an acoustic guitar, and I've felt pretty good about my use of side-chaining (more for utility, and clearing things out of the way to make space in the mix--I haven't yet really explored the Flying Lotus/electronica/dance use of side-chain compression for creative effect).

Multi-stage compression feels a little intimidating to me, though it probably shouldn't. I like the guideline of using one compressor to do the heavy lifting and another for peak control.
 
^ Yea with multi-band comp just get one and stick with it for life C4 is a good choice I use C4 and LinMB in mastering if I have to use it. I rarely use MBC tho. Your side chain technique is correct thats what most use it for. I dont use Side-Chain at all I automate either way works tho. Flying Lotus is not using any new or unknown type/technique of compression he just uses side chain on *multiple* samples and elements in the mix. The elements that he uses the side-chain comp on he drives the comp to EXTREME measure to PUMP the heck out of stuff nothing special or special type compression. There is no such thing a NEW style of compression cause its all based on 1950 hardware. All multi-stage compression is one heavy one light no reason to be afraid.

Lets say drum bus and you have a comp to give it the color/harmonic sound/character its punchy and thick and your really pushing it cause your pushing so hard the peaks are causing the signal to clip (and you dont wanna pull out the L2 lol). You add a second comp with some light settings: Medium Attack, Medium to slow release high threshold something just below the clip level -1 dbFS or -2 dBFS a low very ratio 2:1 or even 1.5:1 like 1 to 1&1/2 of gain reduction proly a hard knee. Soft knee on the 1st comp and your good.
 
steakmusic said:
So my question is: if you're trying to even out transients, why would you use a long attack time on your compressor? If anything, I think you'd just be clamping down on the "sustain" portion of the snare hits, making them even quieter relative to the onset of the snare hits.
Click to expand...

This "even out transients" thing has always confused me. It feels like it's something that's only become a "thing" since the advent of the visual element of the DAW. Using your ears to set a compressor will assist you more in coming to a realisation than any other method. Practice on a hardware compressor if you can. There are no visual distractions there.

FWIW, I never compress individual drums. The day I gave it up was a great day, to tell you the truth. The problem I see with compressing, say, just the snare, is that it makes it sound disconnected from the rest of the kit. It gives it a false dynamic that just doesn't gel with the kit as a whole. Sometimes you can get away with compressing just the kick but that too doesn't always feel right to me.

I prefer to bus all the drums to a group and compress the group as a whole with a DEDICATED drum compressor like an 1176 (UAD), or a 160VU (Steinberg Vintage Series) or even Bootsy's ThrillseekerLA works great. You can strike a good balance and tailor the detector of the compressor by playing with the track levels, i.e. how hard each element hits the threshold, and use a sidechain to filter the input if you want more kick to come through without the compressor reacting. I've been using the ThrillseekerLA a lot lately and think it stands up with the best.

So just try compressing the drum group sometime. You can get all the smack you want out of the snare and the rest of the kit will breathe with it, instead of sound like it was superimposed on top of it. Generally you want to use a medium attack like 15 to 25ms, and a fast release. You want the compressor to get in and out of there as quick as possible without distortion artifacts or cymbal pumping.

Cheers :)
 
I'm late in responding, but thanks again guys. Great advice. I'm going to try using compression on just the drum bus rather than compression on the snare, kick, and toms respectively.

Since all of my drums are programmed (mostly using BFD2), I have a good handle on dynamics in the first place, just by virtue of the fact that I can go in and tweak velocity levels to my heart's content before even printing the drums. So I should definitely be able to benefit from drum bus compression for tonal rather than corrective purposes.

Guess I could also try sequencing drums through a compressor on the drum bus, though I'm a little more hesitant to try that for now!
 
steakmusic said:
I'm late in responding, but thanks again guys. Great advice. I'm going to try using compression on just the drum bus rather than compression on the snare, kick, and toms respectively.

Since all of my drums are programmed (mostly using BFD2), I have a good handle on dynamics in the first place, just by virtue of the fact that I can go in and tweak velocity levels to my heart's content before even printing the drums. So I should definitely be able to benefit from drum bus compression for tonal rather than corrective purposes.

Guess I could also try sequencing drums through a compressor on the drum bus, though I'm a little more hesitant to try that for now!
Click to expand...
Try everything so you know if anythings works.
 
You must log in or register to reply here.
Back
Top