Bob Katz & Compression

  • Thread starter Thread starter Bulls Hit
  • Start date Start date
B

Bulls Hit

Well-known member
So in his book Bob says he set attack time on downward compression somewhere between 50 & 300ms, usually somewhere around 100ms.

I know for some instruments you're supposed to wait so you don't ruin the character of the instrument, but isn't even 100ms a helluva long time to wait? At 300ms wouldn't it potentially be all over, so the level drops below threshold before the compressor kicks in?

My Cakewalk compressor plugin maxes out at 20ms attack - it wouldn't even get a look in
 
Isn't Bob a mastering guy? This is a completely different application.
 
Yep he's a mastering guy. Is a 300ms attack time normal in mastering?
 
You're right Bulls Hit - I just had to pull up the Cakewalk compressor and see - the attack does max out at 20ms, yikes that's a short one!

If you find one with a slower attack and you adjust the thresholds low you can play around with the attack so that the initial transient attack of the instrument is preserved while the sustained part of the sound is compressed, then you just decide what release sound you want (crunchy, pumping or smooth). I'm playing around with a vocal using 75-100 ms on the compressor attack.

You can do the same thing on full mixes - long attack times on the mastering compressor to compress the body of the song if you want to, followed by a mastering limiter where the attacks times are short to limit & control transients.
 
Bulls Hit said:
Yep he's a mastering guy. Is a 300ms attack time normal in mastering?
Click to expand...
If you are trying to just level the song out a bit, yes. If you are trying to squash the snot out of it, no.
Be careful when you take someone elses settings and try to use them. A compressor setting is very dependant on the signal you are compressing and why you are compressing it. If you read your first post, you will notice that it says to set the attack time between 50 and 300ms. 300 is six times 50. That isn't much of a guideline. That's like telling someone that their destination is somewhere between Chicago and Cincinnati.
 
100ms in mastering isn't a long time. Same with release - Anywhere between 4-8 seconds falling into the "normal" category most of the time around here...
 
Usually with mastering I am using more than one compressor or using a compressor that supports multiple attack/release times within the same box. For limiting you are going to go for shorter attack, release times (nearly instantaneous for attack, 10ms or less for release). I'ts all going to depend on what you are trying to acheive with the material. Don't just blindly apply some "rule of thumb" or you will not be optimizing the dynamics for the particular song.
 
So when you guys master, do you typically use outboard compressors/limiters etc or plugins?
 
"Typically" 90% outboard and a plug or two for tweaking. Sometimes plugs only (very rare).
 
Bulls Hit said:
So when you guys master, do you typically use outboard compressors/limiters etc or plugins?
Click to expand...

I like to use a combination. The overall compression is usually done with a combination of outboard digital and analog compressors. I will use a plug-in for some songs on an individual basis when needed, or when I need to automate compression or EQ. Noise reduction is another example of when I'll use a plug-in as I always automate this (threshold/amount) throughout the song.

I can recall the exact settings for all of the songs in a session at any point in time nearly instantaneously this way.

BTW, since digital hardware is really software anyway it doesn't much matter if it's a plug-in or not assuming it's from the same manufacturer. In fact in some cases it's better to use a plug-ins since you can chain processes at a higher bit depth than if you converted back and forth to 24 bit in order to use external digital hardware. With a digital source and using analog outboard gear you are going through at least one more D/A and A/D conversion, so an argument can be made for that as well. Usually with analog however you're going for a certain added character, so the loss in conversion is offset by what you are trying to add.
 
I use DAW plugs exclusively in my amatuer studio...anything that gets analog treatment or character is only on the front-end recording side.

masteringhouse said:
BTW, since digital hardware is really software anyway it doesn't much matter if it's a plug-in or not assuming it's from the same manufacturer. In fact in some cases it's better to use a plug-ins since you can chain processes at a higher bit depth than if you converted back and forth to 24 bit in order to use external digital hardware. With a digital source and using analog outboard gear you are going through at least one more D/A and A/D conversion, so an argument can be made for that as well. Usually with analog however you're going for a certain added character, so the loss in conversion is offset by what you are trying to add.
Click to expand...

masteringhouse - you sure said a mouthful in your post, this little snippet kind of states the reason that some of us DAW users are waiting with baited breath for native plugs of the UAD-1 caliber. In other words if software designers can write convincing dynamic processor algorithms that run in that environment then pretty soon it should spread to my station in life. I'm running a 2.6GHz P4 now and I suspect either this year or next computers should be powerful enough to pull off this kind of thing without some co-processing card - assuming that's part of the reason.

Like you alluded to plugins are already capable of double precision 64 bit [float] calculations (According to the IEEE spec - The 64-bit (double-precision) format is, a sign bit, an 11-bit exponent with a bias of 1023, and 52 bits of mantissa. With the hidden bit, normalized numbers have an effective precision of 24 and 53 bits, respectively). Voxengo plugins use this as well as a few others even though most application inserts, sends & busses like in Sonar4 and Adobe Audition operate at 32bit float internally...

Now all we need are the Weiss & UA algorithms...I'll keep my eyes peeled over at KVR. :D
 
kylen said:
In other words if software designers can write convincing dynamic processor algorithms that run in that environment then pretty soon it should spread to my station in life. I'm running a 2.6GHz P4 now and I suspect either this year or next computers should be powerful enough to pull off this kind of thing without some co-processing card - assuming that's part of the reason.
Click to expand...

Probably 2 of the main issues are latency in realtime applications like mixing and political ones (why give away the software when you can make people buy the hardware in order to use your software, helps stop piracy as well).

For mastering apps latency is less of an issue since you are only dealing with 2 tracks at a time, and most processing is done in stereo anyway.

I'm guessing that another issue with manufacturers is support for multiple operating systems and keeping up with OS changes, plug-in support for different DAWs, etc.
 
masteringhouse said:
Probably 2 of the main issues are latency in realtime applications like mixing and political ones (why give away the software when you can make people buy the hardware in order to use your software, helps stop piracy as well).
Click to expand...
True...pretty substantial dongle ! :D
 
You must log in or register to reply here.
Back
Top