parallel compression in mastering

[BitShift]

hi,
i would love to know how do you use parallel compression in mastering. Tried it out in some projects and i think it works fine. I used the strip pro for the parallel channel.
I think the best would be to have the normal mixdown plus a groupexport of the drum & bass.

Would you do it also with just the final mixdown without any extra export??

[Cellotron]

Originally Posted by BitShift

hi,
i would love to know how do you use parallel compression in mastering. Tried it out in some projects and i think it works fine. I used the strip pro for the parallel channel.
I think the best would be to have the normal mixdown plus a groupexport of the drum & bass.

Would you do it also with just the final mixdown without any extra export??

Amount of times I've used parallel compression when mastering tracks in the past year: 0.

Amount of times the year before that: once.

It works way better for things like the drum bus (where even then it's not always best thing to do) than it ever does for the entire program imho. Obviously OMMV!!

Anyway - there's some hardware that has "blend" controls between dry and processed signals such as the Safesound Dynamic Toolbox and Altomoda Unicomp - and there are some insert controllers such as the Manley Backbone that can do this as well.

Best regards,
Steve Berson

[Ben B]

I think parallel compression works badly just as often as it works well. I think it's an easy trap to fall into to believe that a certain technique is generically "good" or "bad," in any aspect of the recording process. Having said that, lately I've been into the "less is more" approach in mastering. Great mixes don't really need very much work at all in mastering. When I add parallel compression to a mix that already sounds good, it tends to make it more congested sounding for me. When I do use parallel compression in mastering, I usually find myself filtering the parallel signal to work on a specific area of the spectrum. There have been lots of times when I've regretted doing this, however. Sometimes parallel compression provides a type of instant gratification that sounds good initially, but when listening to the mixes later on I realize that some of the "space" that used to exist in the mixes has been altered, not for the better. For me, the challenge is in identifying the circumstances in which parallel compression really is the best tool. While there certainly have been times when it has been beneficial, there have been just as many times when I've regretted doing it.

-Ben B

[Jerry Tubb]

Originally Posted by Cellotron

Amount of times I've used parallel compression when mastering tracks in the past year: 0.

Same here.

So many myths about what actually happens in mastering get propagated somehow...

1. multiband compression

2. parallel compression

3. widening

4. adding reverb

When in reality it's most often experience with basic EQ and light Compression...

JT

[Silver Sonya]

Originally Posted by Jerry Tubb

Same here.

So many myths about what actually happens in mastering get propagated somehow...

1. multiband compression

2. parallel compression

3. widening

4. adding reverb

When in reality it's most often experience with basic EQ and light Compression...

JT

SSSSHHHHHHHHHH!!!!!!

Don't tell anyone this. Judicious moderation is boring.

To everybody reading this: WE'RE WIZARDS! WE SPRINKLE FAIRY DUST! WE MAKE YOUR MUSIC MAGICALLY MAGICAL MAGIC!!!!!

[masteringhouse]

It all depends on what you are trying to achieve. Sometimes a little filtered parallel comp or parallel MBC can be used to pull out elements of a mix if it needs it. It can also be helpful when you are asked to make a louder master but would rather not kill the transients as much.

I think that parallel comp in mastering has gotten undeserved bad rep because according to some guidelines on how to set it up some may be a little heavy handed on the compression side. Often just a little goes a long way as in straight compression. Blending in compression with low ratios, medium attack, and only a slight amount of reduction can often be the ticket over applying it to the mix overall.

[bgrotto]

Originally Posted by Silver Sonya

SSSSHHHHHHHHHH!!!!!!

Don't tell anyone this. Judicious moderation is boring.

To everybody reading this: WE'RE WIZARDS! WE SPRINKLE FAIRY DUST! WE MAKE YOUR MUSIC MAGICALLY MAGICAL MAGIC!!!!!

[dcollins]

Originally Posted by Jerry Tubb

Same here.

So many myths about what actually happens in mastering get propagated somehow...

1. multiband compression

2. parallel compression

3. widening

4. adding reverb

When in reality it's most often experience with basic EQ and light Compression...

That sounds boring. Can't you jazz it up a little?


DC

[masteringhouse]

What's the best fairy dust under $100?

This looks promising anyone try it yet?

Celtic Shopping Center: Fairy Dust

[Darius van H]

The only time i use parallel compression is in the Ammunition plugin in Sequoia - it has a "mix" control as well as separate M/S sections - very useful little pluggy.

[masteringhouse]

I wonder sometimes if people would have as many issues with parallel comp if it was presented in a different way. Rather than "parallel comp" let's use an inverse concept termed "parallel source injection".

To use PSI, compress the way that you normally would (nobody has a problem with this) and try summing a given amount of the source track from the point before the compressor (possibly delayed slightly to avoid latency or phase issues) back into your compressed signal.

That way you're adding good to "better" rather than adding nasty compression to your source. Assuming that the purpose of this technique is to add back some element of the source in order to undo the negative artifacts of the original compression (e.g. lost transients, space, etc.).

Additionally the source track being summed might be filtered, which is a slightly different approach than filtering the compressed signal and adding that as in PMBC, or FPC (filtered parallel compression).

Ok, now pass the fairy dust ...

[illacov]

Is the congestion coming because of an increase in amplitude of undesirable freqs??
That were attenuated by so many db to get them into a good spot and then if you parallel compress it you've in essence increased their amplitude??

Interesting...

I always experienced it from the tracking and mixing side of things to be honest. On drum overheads, vocals, bass etc...I think it really depends on the compressor.

I've heard some sick mixes with the Manely VariMus in parallel on the 2 buss.

As far as mastering I've only personally heard it on the masters of my band's first album but I believe that one was the mastering edition, with the hipass sidechain?

Whenever I've tried to do fake masters, I always find better success with very very light parallel compression with a Fairchild emulation (1/2 db to 1 db of GR), like say 30 raw / 70 compressed and saturation for volume/heft. So far it translates pretty good but I still prefer a mastering engineer's results to mine.

I just do those fakies for my mix clientele who want to hear it loud.



Peace
Illumination

[wado1942]

What's the best fairy dust under $100?

I'll have a guy they call "Itchy" meet you tomorrow night by the corner street light. Come alone.

[finetuner]

Originally Posted by masteringhouse

...To use PSI, compress the way that you normally would (nobody has a problem with this) and try summing a given amount of the source track from the point before the compressor (possibly delayed slightly to avoid latency or phase issues) back into your compressed signal.

That way you're adding good to "better" rather than adding nasty compression to your source...

I can see the beauty of this concept.
But doesn't it result in higher peak add up, leaving more 'ugly' work for the digital limiter instead?
Whereas parallel comp is more like 'filling up the gaps' with less peak build up?

[masteringhouse]

Originally Posted by finetuner

I can see the beauty of this concept.
But doesn't it result in higher peak add up, leaving more 'ugly' work for the digital limiter instead?
Whereas parallel comp is more like 'filling up the gaps' with less peak build up?

Hey Peter!

That's the perception, but parallel comp is adding in the source component in the same way as PSI. It's all just a question of how you setup the comp and the mixture of source to compressed. That's really been my main point in the previous posts. Other than the difference of filtering the source versus the compressed signal PSI and PC are six in one, half of a dozen in the other.

For one to say that they don't like PC (in my book) it simply means that that don't like mixing a source signal back into the chain after compression.

[romanski]

Hi to everyone,

did not want to carry off the topic but thought it'd be appropriate if i put my question in here. i was searching the web about parallel compression and found this post on some forum:

What Is Parallel Compression?

Parallel compression is the mixing together of two differently compressed signals coming from the same source. One compressor will compress lightly (or not at all) and the other compressor will have more gain reduction. The heavily compressed signal is usually mixed under (lower in level than) the uncompressed/lightly-compressed signal. The advantage is that it brings up lower level signals in the mix while leaving the highest signals alone. It can raise the average loudness of the sound and yet obtain a sound that is quite natural.

Parallel compression is used both in mastering and in mixing. In mastering it is usually fastest to import the same audio file into two tracks of your DAW or editing program and then processing them differently before mixing them together. In mixing, you would need to setup two submix aux channels and apply compression differently in each channel before mixing their outputs.

What is Motown Compression?

Motown compression is the use of parallel compression with equalization added before the heavy compression, usually done during mixing. The equalization is added to accent certain frequencies needed on an instrument (or an entire mix). Originally it was used for vocals with increased energy at 8 kHZ but with minimal distortion on the tape or disc due to the accented highs - 8 kHz was accented to get the mixes sounding better over AM radio. The technique however can be used on any instrument in the mix by using a frequency key to that instrument.

the question is what exactly do you guys use: motown or parallel? i used to use parallel compression myself and just tried to put eq before the comp..sound a bit different but more pleasant to my ear. any thoughts?

[Superdisc]

Parallel compression depends a lot on how you mix the "stems" There is no formula and until you use your ears you can think that it:

1. results in higher peak add up
2. makes it more congested sounding
3. works way better for things like the drum bus

The truth is you may experience this or not depending on how you mix the stems.

Blend controls are boring and could work I guess, but using parallel compression is not always mixing an uncompressed stem with a compressed stem. It, just as easily, could be (and often if for me) two stems compressed or processed differently.

There absolutely is no "formula" to set it up and the different blends are endless.

[Superdisc]

Originally Posted by romanski

Hi to everyone,

did not want to carry off the topic but thought it'd be appropriate if i put my question in here. i was searching the web about parallel compression and found this post on some forum:

What Is Parallel Compression?

Parallel compression is the mixing together of two differently compressed signals coming from the same source. One compressor will compress lightly (or not at all) and the other compressor will have more gain reduction. The heavily compressed signal is usually mixed under (lower in level than) the uncompressed/lightly-compressed signal. The advantage is that it brings up lower level signals in the mix while leaving the highest signals alone. It can raise the average loudness of the sound and yet obtain a sound that is quite natural.

Parallel compression is used both in mastering and in mixing. In mastering it is usually fastest to import the same audio file into two tracks of your DAW or editing program and then processing them differently before mixing them together. In mixing, you would need to setup two submix aux channels and apply compression differently in each channel before mixing their outputs.

What is Motown Compression?

Motown compression is the use of parallel compression with equalization added before the heavy compression, usually done during mixing. The equalization is added to accent certain frequencies needed on an instrument (or an entire mix). Originally it was used for vocals with increased energy at 8 kHZ but with minimal distortion on the tape or disc due to the accented highs - 8 kHz was accented to get the mixes sounding better over AM radio. The technique however can be used on any instrument in the mix by using a frequency key to that instrument.

the question is what exactly do you guys use: motown or parallel? i used to use parallel compression myself and just tried to put eq before the comp..sound a bit different but more pleasant to my ear. any thoughts?

Motown compression and New York compression are different variations of parallel compression. The vocal sound of Motown was famous and the above description is accurate. However, 95%+ of the compression used in mixing those old Motown three track masters was parallel compression.

I know the author of those quotes rather well :-)

[masteringhouse]

In addition to parallel comp, Motown comp, NY comp, upward comp, there is side chain comp:
SIDE CHAIN COMPRESSION

They are all essentially the same technique regardless if you filter one or the other signal or process one or both. We should just call all of this "parallel processing" as opposed to "serial processing".

Has anyone come up with names for parallel expansion or other forms of parallel processing? Why does compression seem to be the main focus for this?

[romanski]

Originally Posted by Superdisc

Motown compression and New York compression are different variations of parallel compression. The vocal sound of Motown was famous and the above description is accurate. However, 95%+ of the compression used in mixing those old Motown three track masters was parallel compression.

I know the author of those quotes rather well :-)

thanx for the quick replythumbsup

i dont know what about the vocal 'cause i tried it on drums and it gives absolutely different effect. mmmmmaybe its more solid...but as usuall..need to use my ears i guess))

[Caput]

Originally Posted by Ben B

I think parallel compression works badly just as often as it works well. Sometimes parallel compression provides a type of instant gratification that sounds good initially, but when listening to the mixes later on I realize that some of the "space" that used to exist in the mixes has been altered, not for the better.

-Ben B

This is very true. I have sometime found that ME working with parallel comp say to me - see how the transient is still there! But they didn't notice the new over-crowded sound.

SB

[Jerry Tubb]

Originally Posted by dcollins

That sounds boring. Can't you jazz it up a little?

DC

Perhaps a Finalizer in Wizard mode would do the trick!

Pass the fairy dust please.

JT

[finetuner]

Originally Posted by masteringhouse

Hey Peter!

That's the perception, but parallel comp is adding in the source component in the same way as PSI. It's all just a question of how you setup the comp and the mixture of source to compressed. That's really been my main point in the previous posts. Other than the difference of filtering the source versus the compressed signal PSI and PC are six in one, half of a dozen in the other.

For one to say that they don't like PC (in my book) it simply means that that don't like mixing a source signal back into the chain after compression.

Hi Tom!

Re-reading your post fresh in the morning i see that you didn't literally mean to suggest a different concept but rather describe the process in another way right?

As it's clear that one can mix processed to source or processed 1 to processed 2 in any ratio to taste.

I'm not afraid of Parallel Compression either and combined with some analog compression in series it allows to keep the PC mixed in subtly and/or - indeed - with a lower ratio. Less fingerprints by adding small amounts of processing divided over a few tools.

O, btw i DO use PSI on everything :

[Lagerfeldt]

I think parallel compression is a great way of enhancing a mix while maintaining punch.

The Flux Syrah does great parallel compression, also in combination with the de-expansion feature. Since this can be entirely threshold independent you don't get any nasty sideeffects such as pumping.

Parallel compression in the Flux Solera can lift information in an interesting way, especially in the high frequencies.

BTW, there's a 50% off sale at Flux at the moment: Flux:: sound and picture development

[spasm_dtc]

Originally Posted by masteringhouse

It all depends on what you are trying to achieve.

That's really all there is to say about it in my opinion... That and "how you want to do it". Any engineer will use the techniques he's most comfortable with to achieve a given goal - the rest is blahblah (no offence anyone) in my opinion, your ears are the only judge here.

[masteringhouse]

Originally Posted by finetuner

O, btw i DO use PSI on everything :

Haven't had an opportunity to hear them yet, they look gorgeous!

Originally Posted by Jerry Tubb

Perhaps a Finalizer in Wizard mode would do the trick!

Pass the fairy dust please.

JT

Isn't that called Isengard compression?

[Superdisc]

[QUOTE=masteringhouse;4861031]
They are all essentially the same technique regardless if you filter one or the other signal or process one or both. We should just call all of this "parallel processing" as opposed to "serial processing".
QUOTE]

So right

It illustrates my point - many ways to skin the same cat. I usually call it parallel processing rather than parallel compression.

The unfortunate thing, however, is that the particular technique described often did more damage than good, if the instructions in the article are followed. Too many times people come up with "formulas" for processing that people try to use on everything - that doesn't work (understatement).

I don't like the name because the term "Sidechain" is used otherwise much more often - so that use of the term would be confusing to many.

[masteringhouse]

Bob,

There often seems to be a lot of "wizardry" that hides under terminology or names in audio production for the purposes of making a process or technique more proprietary. For example "separations vs stems" or "harmonic balance vs matching FFT curves".

Saying that "I've used parallel processing with a compressor and filter on one side" pretty much demystifies the process thereby making it useless as a marketing tool.

[Jerry Tubb]

Originally Posted by masteringhouse

Isn't that called Isengard compression?

Whatever it takes to fend off the Orcs! Thou shall not pass!

If Elfin (mbpc) dust is workin' for you... so be it.

Cheers - JT

[flatfinger]

Since we have some real expertise here today , I thought it a good time to chime in with a related question that's been buggin me latley .

Some comps have the "Range " adjustment ( ya , comps , not gates!!) For instance , the aforementioned flux pure comp has one .
basically on the transfer function after the knee , things are as expected for ahile , but just like some old analog gear , the ability to lower gain craps out so the bigger transients get through . The response in non- linear ABOVE knee.

My question is this.. PC is supposed to be nice cause it is'nt stomping the transients .. so what about comps like this one
Cytomic | Sound Music Software............

It has a range knob to allow the gain reduction to be reduced to a set maximum...
The manual just says that it is an "alternative " to PC ( and this comp also has a dry/wet mix control too !!!! ...So it seems that the biggest signal values are preserved ( which would seem to be leaving some work for a needed limiter next in the chain !)



Are these two methods pretty much equivalent ?????






Cheers