Old 10-05-2011, 04:59 AM   #1
Simon J.
Human being with feelings
 
Join Date: Oct 2010
Location: Finland
Posts: 47
Default A quick question about parallel processing

Parallel processing has been a mystical buzzword for me until I watched a couple youtube videos about it.

So my question is: According to how this guy (https://www.youtube.com/watch?v=d9csv_TxAQM) explains it, putting a ReaComp in a channel and sliding the wet/dry slider to 50% is effectively parallel processing, right?
Simon J. is offline   Reply With Quote
Old 10-05-2011, 05:37 AM   #2
billybk1
Human being with feelings
 
billybk1's Avatar
 
Join Date: Jun 2006
Location: Atlanta, GA
Posts: 1,505
Default

Often times this is done using buses (Parallel Bus compression aka NY style compression). Fab has a trailer video where he briefly discusses this technique.

http://www.puremix.net/video/process...mpression.html


He has a lot of free (and more in-depth paid tutorials) which are quite good. I like his teaching style and sense of humor. He just released a great free (10) part video series (a couple of hours worth) where he tracks a song using a live band on stage (in front of hundreds of people) and then does a mix session using the same song. He descibes in detail the how, what & why as he goes through the entire process (setting up mics, preamps, etc). The video's are very well done and I learned a lot, although be aware that the sponsors of the event (AVID, Royer, Dangerous Music, Universal Audio, Great River, Mohave, Focal) are all used & promoted during the event. I guess somebody had to pay for it.
A very nice bonus is that you can download the individual session wave files and have a crack at doing the mix yourself.


The Tracking Session (6 parts):

http://puremix.net/video/othering/ev...bass-drum.html

The Mix session(4 parts):

http://puremix.net/video/othering/ev...etting-up.html

Cheers,

Billy Buck

Last edited by billybk1; 10-05-2011 at 05:46 AM.
billybk1 is offline   Reply With Quote
Old 10-05-2011, 08:11 AM   #3
mako55a
Human being with feelings
 
Join Date: Jan 2011
Location: Polska, Komorów
Posts: 228
Default

Quote:
Originally Posted by Simon J. View Post
Parallel processing has been a mystical buzzword for me until I watched a couple youtube videos about it.

So my question is: According to how this guy (https://www.youtube.com/watch?v=d9csv_TxAQM) explains it, putting a ReaComp in a channel and sliding the wet/dry slider to 50% is effectively parallel processing, right?
Yes it is and if you plan to use only compressor there is no advantage to have additional buss. Using Busses is usefull when you plan to add more plugins.

mako55a
mako55a is offline   Reply With Quote
Old 10-11-2011, 11:33 PM   #4
Tuned
Human being with feelings
 
Join Date: Nov 2010
Posts: 25
Default

Parallel compression is really a technique used for compressors with limited controls and a lot going on under the hood. With a standard compression plug-in, all parallel compression does is alter the ratio. For example if you have a 50% wet/dry mix on ReaComp and a 4:1 ratio, it is now a 2:1 ratio, since half the signal is unaffected.

In the old days if you had an LA 1176, you only had limited attack and release controls and a crapload of coloration going on under the hood. The only way to lower the ratio or lessen the coloration was parallel compression.

For the mostpart parallel compression is obsolete, as are compressors that do a bunch of stuff other than compression. I used to use the T-Racks compressor plug-in to get a fat colored sound as I compressed, now I use coloration plugs for coloration of various sorts, and ReaComp to compress, but always 100% wet.
Tuned is offline   Reply With Quote
Old 10-12-2011, 08:12 AM   #5
shemp
Human being with feelings
 
shemp's Avatar
 
Join Date: Oct 2008
Location: Texas
Posts: 3,157
Default

Quote:
Originally Posted by Tuned View Post
Parallel compression is really a technique used for compressors with limited controls and a lot going on under the hood. With a standard compression plug-in, all parallel compression does is alter the ratio. For example if you have a 50% wet/dry mix on ReaComp and a 4:1 ratio, it is now a 2:1 ratio, since half the signal is unaffected.
This isn't true. The Wet/Dry sliders effect the mix at the output after compression. It does not effect the ratio. Reducing the Wet level is still mixing in a 4:1 compressed signal with the uncompressed signal, just at a lower level.

Quote:
Originally Posted by Tuned View Post
For the mostpart parallel compression is obsolete, as are compressors that do a bunch of stuff other than compression. I used to use the T-Racks compressor plug-in to get a fat colored sound as I compressed, now I use coloration plugs for coloration of various sorts, and ReaComp to compress, but always 100% wet.
I wouldn't say it's obsolete. Modern software compressors do magical things but having a separate compressor on a separate track is more flexible IMO.
shemp is offline   Reply With Quote
Old 10-12-2011, 09:21 AM   #6
Tuned
Human being with feelings
 
Join Date: Nov 2010
Posts: 25
Default

Quote:
Originally Posted by shemp View Post
This isn't true. The Wet/Dry sliders effect the mix at the output after compression. It does not effect the ratio. Reducing the Wet level is still mixing in a 4:1 compressed signal with the uncompressed signal, just at a lower level.
Right, half of the signal is 4:1, half is 1:1. Pop quiz: what's half of 4?

Quote:
I wouldn't say it's obsolete. Modern software compressors do magical things but having a separate compressor on a separate track is more flexible IMO.
This is a matter of opinion, but I say if a compressor does something other than compression, it is not modern, it is usually a reproduction of vintage analog. Modern plug-ins say what they do and do what they say, so the engineer can do what he/she wants instead of what the plug-in designer wants. If I want compression I use a compression plug. If I want coloration I use a coloration plug. No parallel routing required. I call that obsolete.
Tuned is offline   Reply With Quote
Old 10-12-2011, 09:34 AM   #7
shemp
Human being with feelings
 
shemp's Avatar
 
Join Date: Oct 2008
Location: Texas
Posts: 3,157
Default

Quote:
Originally Posted by Tuned View Post
Right, half of the signal is 4:1, half is 1:1. Pop quiz: what's half of 4?
Still wrong. We are talking mix AFTER compression. Compressing at a 4:1 and 2:1 has DIFFERENT results on the overall signal. Mixing them into the dry signal has DIFFERENT sonic results. Please try it.
Quote:
Originally Posted by Tuned View Post
This is a matter of opinion, but I say if a compressor does something other than compression, it is not modern, it is usually a reproduction of vintage analog. Modern plug-ins say what they do and do what they say, so the engineer can do what he/she wants instead of what the plug-in designer wants. If I want compression I use a compression plug. If I want coloration I use a coloration plug. No parallel routing required. I call that obsolete.
ok dude, I don't remember saying anything about compressors not advertising what they do. I just said it's more flexible. Do whatever you want, I don't care.
shemp is offline   Reply With Quote
Old 10-12-2011, 11:18 AM   #8
shemp
Human being with feelings
 
shemp's Avatar
 
Join Date: Oct 2008
Location: Texas
Posts: 3,157
Default

ok, I'm gonna retract my argument.

I did the experiment with your numbers. A single render of a somewhat dynamic wav file with Reacomp at 4:1 50% wet visually matched (for the most part) a render of the dry file added to the same file with Reacomp at 2:1 100% wet (essentially, external parallel compression). Sonically, I could not tell the difference. However, with parallel compression, usually a larger threshold is used and it's not mixed in at exactly 50% so that was where my argument was biased. I can hear the difference when the ratio is changed, but I'm not changing the mix level to the exact proportions to mathematically figure out the new ratio, nor did I care. Point taken for future reference.
shemp is offline   Reply With Quote
Old 10-12-2011, 12:06 PM   #9
Fabian
Human being with feelings
 
Fabian's Avatar
 
Join Date: Sep 2008
Location: Sweden
Posts: 5,524
Default But it doesn't null...

I couldn't help, just had to do the experiment myself, just had to

I set up tone_generator on one track with default settings, sending to two other tracks on which I had:

1. ReaComp at 4:1, -25 dB threshold, 50% wet.
2. ReaComp at 2:1, -25 dB threshold, 100% wet.

All other settings were default, and so identical.
Inverting polarity on one of those tracks doesn't null, it only goes down to -28.3 dB. Maybe rounding errors, maybe not...

EDIT: Most probably not... since the output of track 1 (as set above) is -15.1 dB, whereas the output from track 2 is -17.2. So I'm quite convinced now that 4:1 50% wet is not exactly the same thing as 2:1 100% wet.
__________________
// MVHMF
I never always did the right thing, but all I did wasn't wrong...
Fabian is online now   Reply With Quote
Old 10-12-2011, 05:02 PM   #10
imispgh
Human being with feelings
 
Join Date: Aug 2008
Posts: 486
Default

OK so with the Wet/Dry sliders set as is - Wet at 0 and Dry at -inf it is set up to act like parallel compression with the ratio being what is stated on the Ratio slider correct?
imispgh is offline   Reply With Quote
Old 10-12-2011, 05:06 PM   #11
jacobestes
Human being with feelings
 
jacobestes's Avatar
 
Join Date: Jun 2010
Posts: 415
Default

Quote:
Originally Posted by billybk1 View Post
He just released a great free (10) part video series (a couple of hours worth) where he tracks a song using a live band on stage (in front of hundreds of people) and then does a mix session using the same song. He descibes in detail the how, what & why as he goes through the entire process (setting up mics, preamps, etc).
I'm watching these now, and just watched the part about vocals where he changes preamp, adjusts eq, and changes compression. The differences were subtle, but enough to make the previous attempt sound bad by comparison. It was really great to watch it come along. Each time I'm like, "That sounds great." And then he says, But what if we do this? And then say, "Wow, that sounds great."

Thanks for the link.
jacobestes is offline   Reply With Quote
Old 10-12-2011, 05:17 PM   #12
shemp
Human being with feelings
 
shemp's Avatar
 
Join Date: Oct 2008
Location: Texas
Posts: 3,157
Default

Quote:
Originally Posted by imispgh View Post
OK so with the Wet/Dry sliders set as is - Wet at 0 and Dry at -inf it is set up to act like parallel compression with the ratio being what is stated on the Ratio slider correct?
no, you have to mix the wet and dry for parallel compression. Usually, you set the dry level first, then set an aggressive ratio, set the threshold to somewhere below the peak of the input signal, then while listening, start the wet slider at inf and raise it slowly to hear the mix of the dry and compressed signal. Adjust the threshold as necessary for more compression.
shemp is offline   Reply With Quote
Old 10-12-2011, 07:09 PM   #13
audioguytodd
Human being with feelings
 
audioguytodd's Avatar
 
Join Date: Mar 2008
Location: Atlanta, GA
Posts: 1,054
Default

FWIW, I almost always use parallel compression on another track.

Example: Drum buss. Take a send from that track to a new track. First put ReaEQ on the new track and then put ReaComp after the ReaEQ on that track. Solo the EQ and Comp track. Set up your comp to squash the sound pretty good (4:1 or more) with the threshold low (maybe 6 to 12 db's of gain reduction). Turn the Comp track volume all the way down. Unsolo the track.

Listen to the Original drum buss and gradually raise the comp track up until you hear it give more body.

Use the ReaEQ to shape the sound of the compressed signal. Sometimes I lower quite a bit of the highs with a shelf because all I want is the 'ooomph' from the drums.

Using a send to another track leaves all the options open
__________________
Comp Specs: WIN XPSP3, Q6600, ASUSP5K, 3 GB PC6400 RAM, Focusrite Saffire, AlphaTrack, lots of plugins and hardware.
audioguytodd is offline   Reply With Quote
Old 10-13-2011, 12:12 AM   #14
Tuned
Human being with feelings
 
Join Date: Nov 2010
Posts: 25
Default

Shemp: I appreciate your work in actually testing the scenario, regardless of your conclusions. Don't see that too often

I should specify that what I've said about parallel compression only applies to *just* a standard digital compressor plug-in matched with a bare parallel channel. As soon as you put even an EQ on either side, it's a different animal.

Some use parallel techniques blending a channel with lots going on with a bare one, that's more complicated than just parallel compression. I discourage the practice because I get what I want and avoid what I don't want. I don't sorda get what I want and mix it with the bare channel to second-guess myself. If you're well practiced in a parallel technique and get good results with it, more power to you. But I discourage parallel processing in favor of learning each process and targeting their application more adeptly.

That being said, by mixing a neutral digital compressor in parallel with the same bare track 50/50, you cut the ratio in half. Everything else stays the same - attack, release, hold, knee, sidechain, everything. It's raw math. If you hear something different, there's either something wrong with your ears or your DAW.
Tuned is offline   Reply With Quote
Old 10-13-2011, 11:08 AM   #15
Fabian
Human being with feelings
 
Fabian's Avatar
 
Join Date: Sep 2008
Location: Sweden
Posts: 5,524
Default

Quote:
Originally Posted by Tuned View Post
That being said, by mixing a neutral digital compressor in parallel with the same bare track 50/50, you cut the ratio in half. Everything else stays the same - attack, release, hold, knee, sidechain, everything. It's raw math. If you hear something different, there's either something wrong with your ears or your DAW.
What you say sounds perfectly reasonable. But why don't I see that in this project?
Attached Files
File Type: rpp CompTest.RPP (5.3 KB, 60 views)
__________________
// MVHMF
I never always did the right thing, but all I did wasn't wrong...
Fabian is online now   Reply With Quote
Old 10-13-2011, 11:51 AM   #16
imispgh
Human being with feelings
 
Join Date: Aug 2008
Posts: 486
Default

Quote:
Originally Posted by shemp View Post
no, you have to mix the wet and dry for parallel compression. Usually, you set the dry level first, then set an aggressive ratio, set the threshold to somewhere below the peak of the input signal, then while listening, start the wet slider at inf and raise it slowly to hear the mix of the dry and compressed signal. Adjust the threshold as necessary for more compression.
Maybe I am asking the question wrong. I am trying to figure out if I can do it all on one track or actually have to create the second parallel track. Can I do the compression settings with ReaComp on the original track's Fx and not have to create the 2nd parallel track? Isn't this the wet/dry section of the original track's ReaComp. So I set the ratio then use the Wet/Dry to set the amount of the ratio being used in the Wet/Dry ratio which would be the same as having a compressor on a 2nd or parallel track and using the two track's volume sliders to adjust that wet/dry ratio?
imispgh is offline   Reply With Quote
Old 10-13-2011, 11:57 AM   #17
Fabian
Human being with feelings
 
Fabian's Avatar
 
Join Date: Sep 2008
Location: Sweden
Posts: 5,524
Default

Quote:
Originally Posted by imispgh View Post
Can I do the compression settings with ReaComp on the original track's Fx and not have to create the 2nd parallel track?
Yes. See the rpp I posted above. There, parallel compression is achieved on a single track with the wet/dry button that Reaper assigns all FX's. It is the little round thing top right in the FX window. Not very visible, but once you know where and what it is, it is very handy. Just press-and-drag to adjust the amount of wet and dry signal in teh output.
__________________
// MVHMF
I never always did the right thing, but all I did wasn't wrong...
Fabian is online now   Reply With Quote
Old 10-13-2011, 12:01 PM   #18
shemp
Human being with feelings
 
shemp's Avatar
 
Join Date: Oct 2008
Location: Texas
Posts: 3,157
Default

Quote:
Originally Posted by imispgh View Post
Maybe I am asking the question wrong. I am trying to figure out if I can do it all on one track or actually have to create the second parallel track. Can I do the compression settings with ReaComp on the original track's Fx and not have to create the 2nd parallel track? Isn't this the wet/dry section of the original track's ReaComp. So I set the ratio then use the Wet/Dry to set the amount of the ratio being used in the Wet/Dry ratio which would be the same as having a compressor on a 2nd or parallel track and using the two track's volume sliders to adjust that wet/dry ratio?
yes, mixing the wet and dry together on the same track is parallel compression. the dry is uncompressed, mixed with the wet compressed.

Not all compressors have wet and dry adjustments so if you used a compressor on the track without the separate wet and dry adjustments, then you would need to set up the send to a different track. Personally, I don't use ReaComp for this so I like to use the second track (plus you can add eq and whatnot as Audioguytodd mentioned (more flexible))
shemp is offline   Reply With Quote
Old 10-13-2011, 12:03 PM   #19
shemp
Human being with feelings
 
shemp's Avatar
 
Join Date: Oct 2008
Location: Texas
Posts: 3,157
Default

Quote:
Originally Posted by Fabian View Post
What you say sounds perfectly reasonable. But why don't I see that in this project?
I'd guess that the different ratios have a slight effect on time which kills null test results. The math only works perfectly under ideal conditions.
shemp is offline   Reply With Quote
Old 10-13-2011, 12:04 PM   #20
Tuned
Human being with feelings
 
Join Date: Nov 2010
Posts: 25
Default

Interesting, set the 4:1 compressor to 75% mix and it cancels, but not entirely. I checked mixing two uncompressed channels at 50% against an inverted 100%, and Reaper's wet/dry blend works fine, so I checked ReaComp's thresholds or ratios.

I set the thresholds to -24 just to work with even numbers. With the tone generator at -12, a 2:1 ratio should result in an attenuation of 6dB. It was only 4.6! In order to get -6dB you have to either set the threshold to -26.8 or set the ratio to 2.8:1.

Similarly with the 4:1 ratio, it should be -9dB, but instead it's you have to either set the threshold to -26.8 or set the ratio to 32:1! Leaving the threshold at -26.8 gave the proper -10.5 attenuation at 8:1, so ReaComp's thresholds are definitely off. Glad I've been setting it based on the attenuation and not the threshold!

This doesn't prove I'm right, only that parallel processing seems to have its place in DAW diagnostics I don't have time to mess with this more today, best try some other compressors, surely there's a freebie out there without any bells or whistles.
Tuned is offline   Reply With Quote
Old 10-13-2011, 01:23 PM   #21
imispgh
Human being with feelings
 
Join Date: Aug 2008
Posts: 486
Default

I just watched this excellent tutorial from Waves and at 34 minutes in he says the Wet/Dry control of a comp allows you to get parallel compression without using two tracks.

http://www.waves.com/lp/andrew-schep...ass/watch.html
imispgh is offline   Reply With Quote
Old 10-13-2011, 01:24 PM   #22
imispgh
Human being with feelings
 
Join Date: Aug 2008
Posts: 486
Default

Quote:
Originally Posted by shemp View Post
yes, mixing the wet and dry together on the same track is parallel compression. the dry is uncompressed, mixed with the wet compressed.

Not all compressors have wet and dry adjustments so if you used a compressor on the track without the separate wet and dry adjustments, then you would need to set up the send to a different track. Personally, I don't use ReaComp for this so I like to use the second track (plus you can add eq and whatnot as Audioguytodd mentioned (more flexible))
Excellent - thank you
imispgh is offline   Reply With Quote
Old 10-14-2011, 08:58 PM   #23
fastlanestoner
Human being with feelings
 
fastlanestoner's Avatar
 
Join Date: Jul 2010
Location: Los Angeles, CA
Posts: 248
Default

running an auxiliary send is parallel processing
__________________
Now on Reverb.com!

New Thieves & Lovers record 'Seasons' out soon!
fastlanestoner is offline   Reply With Quote
Old 10-17-2011, 12:42 AM   #24
Fabian
Human being with feelings
 
Fabian's Avatar
 
Join Date: Sep 2008
Location: Sweden
Posts: 5,524
Default

@Tuned
Sleeping on this for a few days, I think I understand it. What about this math...

We have one 2:1 compressor, 100% wet, and we have another compressor 4:1, 50% wet, else every setting is identical for the two compressors. Both compressors are fed the same input signal, and that input is above the threshold.

Let's say that the output of the ratio-2 compressor is out = in * a, where a is the attenuation (which depends on the ratio and the level of in). The output of the ratio-4 compressor, when it is 100% wet, would be out = in * a/2. However, the output signal is only 50% wet, and 50% dry (attenuation 1) so the output of the ratio-4 compressor is out = 0.5*in + 0.5*in*a/2. This is the same as out = in * (1/2 + a/4).

Thus, the output of the two compressors (for a signal above the threshold) will be equal only at a single point of attenuation, namely a = 2/3. At any other level of attenuation, they will not give the same output. Hence, parallel compression 50% wet with ratio 4:1 is not the same as 100% wet ratio 2:1 compression.
__________________
// MVHMF
I never always did the right thing, but all I did wasn't wrong...
Fabian is online now   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 12:28 PM.


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