The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
Still questioning multiband compression as a "secret" Dynamics Plugins
Old 20th September 2016
  #1
Lives for gear
 
m_gant's Avatar
 

Still questioning multiband compression as a "secret"

Sorry for the lame title.

I was in a pinch yesterday to master a 14 song live recording of folk instruments & voices. 5 artists recorded live, 10 feeds (tracks). The mix came together fast.

So I pulled out my old V4 of ozone and just strapped it on the master bus in PT. I pulled up my old trusty preset CD exciter widener lol. I turned off EQ & limiter.

I was shocked. After adding about 1-2 db of GR to the 3 bands in the MB module & messing with the exciter a little, it sounded incredibly polished. I used pro-l after ozone to hit around -12 RMS.

I have to say I've abandoned ozone and MB limiters for any mastering over the past year basically because of what I read here.

But now I'm second guessing again. I actually googled MB compression while mastering. A bunch of articles turned up titled the mastering engineers secret weapon lol.

I'm definetly new to mastering, but I turn out some strong material.

The little devil on my shoulder is telling me everyone uses it & it's a secret.

Maybe I just got lucky? I'd love to know honestly who uses a MB module regularly, not to fix major problems, but as a polish for the master.
Old 20th September 2016
  #2
Lives for gear
 
Laurend's Avatar
 

A 32 bands de-expander features in my algorithm. When doing "human" mastering, I only use a MB to save bad mixes. The FLUX Alchemist is my prefered MB.
Old 20th September 2016
  #3
Lives for gear
 
Miles Flint's Avatar
 

Be in for the big MB-bashing..! :-)
I own and use 2 hardware MB comps and love them dearly because they can do things no other hardware can do, or no plugin can do. That's all what counts for me! It's a tool that can do wonders or mess things up completely, when not used carfully!
Old 20th September 2016
  #4
Gear Nut
I end up using a multiband compressor on maybe 1% of tracks.

EQ and a limiter (and sometimes a touch of single band compression) is usually all I need for the other 99%.
Old 20th September 2016
  #5
Gear Nut
Personally I find multiband shifts the overall tonal balance of the track as it plays, effectively giving you a moving target.

I understand mb can give you an instant impression that it is helping, but i genuinely think you get better results with eq and persistence!

The few tracks i might try mb on are usually the ones that sound totally wrong to begin with, and even then its very light compression. I certainly wouldn't use it on a track for 'polish'.
Old 20th September 2016
  #6
Lives for gear
 
Giuseppe Zaccaria's Avatar
 

I don't have hardware MB, but I very often or maybe always use Sonoris MB plug. I use it in a varieties of different ways like taming high freq, sit the vocal in the mix, focusing bass freq, and for a lot of other things, as said before you have to know what to do in a carefull manner, otherwise its best not use it.
best
Old 20th September 2016
  #7
Lives for gear
 
Franco's Avatar
 

Verified Member
I only own the UAD multiband compressor and have used it maybe 5 times in the last few years and it's always been to address issues on a mix that for one reason or other, can't be adjusted back at the mixing stage.

I can't say I've been "wowed" by it to the point where I've thought "Man, I will never ask anyone to make adjustments at the mixing stage, this thing is it!"

I have to also admit I've never explored this tool for creative sound-shaping as I don't create music, but have heard some interesting dynamics on some mixes that I can tell were done with the use of a multiband compressor, so my thoughts are strictly from a stereo mastering POV.

Lastly, I can't say I have any secrets. I just know in which direction to push/pull knobs and when to process and when to not (and it's come from years of working with tons of music).

EDIT: In these times, when some people are relying on robots to finalize their audio, I'd say going out of your way to actually use a multi-band processing tool is pretty damn respectable, at the very least. LOL!
Old 20th September 2016
  #8
Gear Guru
 
lucey's Avatar
 

Verified Member
Do you mean a "marketing secret" for software designers?

Never used it, never will need to. Closest thing is DeEssing, a form of narrow MBC, if you will.

The Tube Tech, Prism and Drawmer hardware units were the only ones, and not popular. Then software sprang up like weeds.

Not a secret. Just an option pushed by people who can't build digital to sound like analog.
Old 20th September 2016
  #9
Lives for gear
 
m_gant's Avatar
 

Quote:
Originally Posted by lucey View Post
Do you mean a "marketing secret" for software designers?

Never used it, never will need to. Closest thing is DeEssing, a form of narrow MBC, if you will.

The Tube Tech, Prism and Drawmer hardware units were the only ones, and not popular. Then software sprang up like weeds.

Not a secret. Just an option pushed by people who can't build digital to sound like analog.
No- not a marketing secret. A secret with held by Mastering engineers. It just makes sense to me to compress bands individually. But I digress, yesterday I was in a pinch and needed a fast solution.

I'm just confused by the anti MB for mastering. I seriously keep thinking it must be used, but people aren't giving it up. I know it sounds ridiculous. But let's face it, most people are recording themselves (hence large studios are closing and gone) and the majority are even mastering as part of their workflow from the start of mixing.

There's a thread where I was bashed called what's on your mixbuss. I chimed in and said "nothing, mastering is a separate process for me" folks were all over me. People have eq, compression, limiters on their master bus when they start a mix. The lines are blurred as far as production goes these days.
Old 20th September 2016
  #10
Lives for gear
 
sdbmastering's Avatar
 

Verified Member
Dynamic EQing is something I use regularly but multiband compression always sounds bad to me.
Old 20th September 2016
  #11
Lives for gear
Yeah maybe dynamic EQ can have fewer possible artifacts.

Top-line mastering engineers might get a higher percentage of "not awful" mixes and a lower percentage of stinkeroos. Long ago when I would do some local "mastering" back when daws were a rare commodity-- Maybe I was the only fella in town equipped to make a pre-master CD. There were a few rather awful mixes with no chance of getting a better version. Such as kick or snare ridiculously too loud, or uneven high-mid peaky vocals. The old multiband compression in Digital Performer was purt useful for such issues.
Old 21st September 2016
  #12
Lives for gear
 
m_gant's Avatar
 

Another thought...

Why do mastering chain presets in most DAW have a MB loaded? (Logic presets)

Why do products designed for mastering load MB into their presets? (ozone)
Old 21st September 2016
  #13
Motown legend
 
Bob Olhsson's Avatar
 

Verified Member
Just put two multi and compressors in series to hear why it's a bad idea. Why two? When the recording is broadcast it's almost certainly going to hit one.

Why do they show up in presets? Ignorant sales jerks who think more features will sell more software.
Old 21st September 2016
  #14
Lives for gear
 
m_gant's Avatar
 

Quote:
Originally Posted by Bob Olhsson View Post
Just put two multi and compressors in series to hear why it's a bad idea. Why two? When the recording is broadcast it's almost certainly going to hit one.

Why do they show up in presets? Ignorant sales jerks who think more features will sell more software.
Thank you. But I think a better test would be to MB, render...then apply a broadcast MB to see the damage. No?
Old 21st September 2016
  #15
Gear Guru
 
lucey's Avatar
 

Verified Member
Quote:
Originally Posted by m_gant View Post
Thank you. But I think a better test would be to MB, render...then apply a broadcast MB to see the damage. No?
Every broadcast limiter is different, all custom by the station. There is no accounting for them perfectly. And perfect translation is not the aim. There is no perfect translation.

Translation = the most musical center line for all playback.
Old 21st September 2016
  #16
Lives for gear
 
Greg Reierson's Avatar
 

Verified Member
Quote:
Originally Posted by m_gant View Post
It just makes sense to me to compress bands individually.
Compress instruments individually, but only as needed. It's pretty unusual for all of the individual elements in a frequency band to need the same amount of compression.

Quote:
There's a thread where I was bashed called what's on your mixbuss. I chimed in and said "nothing, mastering is a separate process for me" folks were all over me. People have eq, compression, limiters on their master bus when they start a mix. The lines are blurred as far as production goes these days.
People use lots of stuff on the mix bus. Whatever floats their boat, but when they think they are "mastering while mixing", what they're really doing is bus processing. Processing is a tool used in mastering but it's not really what mastering is about.
Old 21st September 2016
  #17
Lives for gear
 
m_gant's Avatar
 

Compress instruments individually, but only as needed. It's pretty unusual for all of the individual elements in a frequency band to need the same amount of compression.

Thank you for that Greg!
The light bulb just went on. This makes perfect sense if you have a balanced mix to start with.

Regarding, mixing into a "master bus", I feel it's a terrible workflow. The lines get blurred. How can you really hear what a kick compressor is doing while it's being smacked by a mastering compressor, and limiter? It's like reverse engineering a fundemental process.
Record
Mix
Master
And ideally, all steps done by different parties. Another thing of the past...
Old 21st September 2016
  #18
Lives for gear
 
SmoothTone's Avatar
 

The question is: why would a well balanced mix need multiple bands of compression across the entire spectrum to sound "polished?"
Old 21st September 2016
  #19
Lives for gear
 
JP__'s Avatar
 

Verified Member
First, I think all in all compression is quite overrated in mastering. And you should never forget that compression isnt just time coherent processing, it also can change tonality. And EQuing isnt just changing tonalty, it can also alters time behaviour.
Compression is always a compromise soundwise, especially the natural depth of a recording gets lost at first.
In the end that means the music just tells you what it needs, MB compression can be a part of it (or it can destroy everything as any other tool). The real problem isthe experiences to know what process is helpful and the ability to listen to the damage each of those processes do to the audio. So the biggest goal is to achieve the ideal compromise in every single project.
Talking tools does not help here, cause every project is different, what works here do not work on another.
Especially in the beginning I would keep my toolset as slim as possible as you simply need experiences to hear the influences of every processing. Beginners tend to overuse tools a lot.
Which leads to the old lame routine: get your monitoring as right as possible before thinking about processing in mastering. A lot of studio monitors have huge problems especially in the time domain which makes itvery difficult to hear how dynamic processing can alters the audio.
One result is that flat and deadsounding music all around, which finally leads me back to the topic.
Old 21st September 2016
  #20
Lives for gear
 
Laurend's Avatar
 

MB processors require a heavy time investment to be fully understood. Most people just don't do it because they don't really need a MB on a daily basis. Then the problem solver becomes a trouble maker.

Last edited by Laurend; 23rd September 2016 at 08:22 AM..
Old 21st September 2016
  #21
Quote:
Originally Posted by Laurend View Post
MB processors require a heavy time investment to be fully understood. Most people just do it because they don't really need a MB on a daily basis. Then the problem solver becomes a trouble maker.
Agreed! I use MB processing in my chain.
When I couldn't use the TC Electronics MD3/Brickwall anymore it took me quite some time to have a new MB-chain I could work and be satisfied with. I tried out a handful of different plugin solutions and I have now finally ended up with Fabfilters Pro-MB and DMG's Limitless.
But just as you put it; one needs time to really learn all that stuff, and of course a darn good listening environment to find out how and when these tools should be used. I never use them as any kinds of tone-shaping/character tools though. I have EQ's and single band compressors for that.
Old 22nd September 2016
  #22
Lives for gear
 

Verified Member
I use my M5000 all the time as a multi or single band unit.

It is a multi band compressor, but doesn't have to have all bands engaged.

It's just a different way to controlling frequency content other than the static EQ style.
Old 22nd September 2016
  #23
Quote:
Originally Posted by m_gant View Post
Maybe I just got lucky? I'd love to know honestly who uses a MB module regularly, not to fix major problems, but as a polish for the master.
IMHO, it's hard to generalize what people really mean with the term. There's frequency selective compression, crossover based MB, dynamic EQ, linear phase MB compression, all with very different properties.

But allow me to highlight a few generally less well known facts about the idea.

The most surprising is that MB compression in the pure sense (split, process band individually, sum) is incapable of overload protection. First, the system has no idea of the summed output level. Second, if the system bases on classic causal filtering tech, and utilizes more than 2 bands or higher order filters (higher than first order), the inherent phase shift will make it very difficult to control overshoot (and thus, compress at all) even if the output level is known.

Another central problem is the assumption that musical events in a mix can be classified into bands. Every appreciable sound starts with a relatively low fundamental frequency, to cover the full spectrum with a pattern of more or less harmonic partials. Transients in particular are always wide-band signals covering the full spectrum (preserving them is a critical aspect of serious mastering). Now what do you think is happening when these harmonic relations are crudely split apart, their balance declared pointless and their dynamics processed separately? Precious information gets lost!

Sounds are not made of bands, they are really dynamic patterns of partials, typically very wideband for the first few ms, and with a long decaying, low bandwidth phase (tonal content). MB devices are pretty dumb in the musical sense, every wideband compressor has a better idea of the incoming music than a basic MB device!

Contrary to static EQing, and by limited amounts also wideband compression, it is crucial to realize that MB compression or excessive dynamic EQing are irreversible processes. An EQing mistake can be reverted anytime, anywhere in the chain, at almost no loss. A messed up harmonic structure on the other hand can never be recovered again.

MB devices often work best in restoration or when treating material that doesn't carry much sensitive harmonic information anyway (techno, dubstep). But given proper recordings and mixes, they will tend to show the inherent problem they have with music (such as messing with the integrity of transients and harmonic structures).

I wouldn't recommend a multi-band approach for default "sweetening", except if your material really needs to be remixed (that's what I'd call restoration ).

Be aware of the static impact of traditional crossovers, too. They can't pass a transient without spreading it all over the place. This simply sounds less crisp, without even doing anything. Clearly a bottleneck for a mastering chain imho!

Last edited by FabienTDR; 23rd September 2016 at 12:05 AM.. Reason: more corrections and fixes
Old 22nd September 2016
  #24
Lives for gear
 
SmoothTone's Avatar
 

Quote:
Originally Posted by FabienTDR View Post
IMHO, it's hard to generalize what people really mean with the term. There's frequency selective compression, crossover based MB, dynamic EQ, linear phase MB compression, all with very different properties.

But allow me to highlight a few generally less well known facts about the idea.

The most surprising is that MB compression in the pure sense (split, process band individually, sum) is incapable of overload protection. First, the system has no idea of the summed output level. Second, if the system bases on classic causal filtering tech, and utilizes more than 2 bands or higher order filters (higher than first order), the inherent phase shift will make it very difficult to control overshoot (and thus, compress at all) even if the output level is known.

Another central problem is the assumption that musical events in a mix can be classified into bands. Every appreciable sound starts with a relatively low fundamental frequency, to cover the full spectrum with a pattern of more or less harminic partials. Transients in particular are always wide-band signals covering the full spectrum (preserving them is a critical aspect of serious mastering). Now what do you think is happening when these harmonic relations are crudely split apart, their balance declared pointless and their dynamics processed separately? Precious information gets lost!

Sounds are not made of bands, they are really dynamic patterns of partials, typically very wideband for the first few ms, with a long decaying bandwidth for everything else. MB devices are pretty dumb in the musical sense, every wideband compressor has a better idea of the incoming music than a basic MB device!

Contrary to static EQing, and by limited amounts also wideband compression, it is crucial to realize that MB compression or excessive dynamic EQing are irreversible processes. An EQing mistake can be reverted anytime, anywhere in the chain, at almost no loss. A messed up harmonics structure on the other hand can never be recovered again.

MB devices often work best in restoration or when treating material that doesn't carry much sensitive harmonic information anyway (techno, dubstep). But given proper recordings and mixes, they will tend to show the inherent problem they have with music (such as messing with the integrity of transients and harmonic structures).

I wouldn't recommend a multi-band approach for default "sweetening", except if your material really need to be remixed (that's what I'd call restoration ).

Be aware of the static impact of traditional crossovers, too. They can't pass a transient without spreading it all over the place. This simply sounds less crisp, without even doing anything, clearly a bottleneck for a mastering chain imho!
There's our MB compression sticky! Best explanation I've read on the topic!
Old 22nd September 2016
  #25
Lives for gear
 
Laurend's Avatar
 

Quote:
Originally Posted by FabienTDR View Post
The most surprising is that MB compression in the pure sense (split, process band individually, sum) is incapable of overload protection.
First, the system has no idea of the summed output level. Second, if the system bases on classic causal filtering tech, and utilizes more than 2 bands or higher order filters (higher than first order), the inherent phase shift will make it very difficult to control overshoot (and thus, compress at all) even if the output level is known.
That's why a limiter is required downstream. Note that compression isn't the only option.

Quote:
Originally Posted by FabienTDR View Post
Another central problem is the assumption that musical events in a mix can be classified into bands. Every appreciable sound starts with a relatively low fundamental frequency, to cover the full spectrum with a pattern of more or less harmonic partials. Transients in particular are always wide-band signals covering the full spectrum (preserving them is a critical aspect of serious mastering). Now what do you think is happening when these harmonic relations are crudely split apart, their balance declared pointless and their dynamics processed separately? Precious information gets lost!
Sounds are not made of bands, they are really dynamic patterns of partials, typically very wideband for the first few ms, and with a long decaying, low bandwidth phase (tonal content). MB devices are pretty dumb in the musical sense, every wideband compressor has a better idea of the incoming music than a basic MB device!
Steep slopes for bands separations kill transients at the signal reconstruction.
They also create unrelevent dynamic values for the bands.

Quote:
Originally Posted by FabienTDR View Post
Contrary to static EQing, and by limited amounts also wideband compression, it is crucial to realize that MB compression or excessive dynamic EQing are irreversible processes. An EQing mistake can be reverted anytime, anywhere in the chain, at almost no loss. A messed up harmonics structure on the other hand can never be recovered again.
Almost no dynamic processor can be undone

Quote:
Originally Posted by FabienTDR View Post
MB devices often work best in restoration or when treating material that doesn't carry much sensitive harmonic information anyway (techno, dubstep). But given proper recordings and mixes, they will tend to show the inherent problem they have with music (such as messing with the integrity of transients and harmonic structures).
Only the 6 dB per octave cross-overs are music friendly.

Quote:
Originally Posted by FabienTDR View Post
I wouldn't recommend a multi-band approach for default "sweetening", except if your material really needs to be remixed (that's what I'd call restoration ).
MB need to be tame.

Quote:
Originally Posted by FabienTDR View Post
Be aware of the static impact of traditional crossovers, too. They can't pass a transient without spreading it all over the place. This simply sounds less crisp, without even doing anything, clearly a bottleneck for a mastering chain imho!
This is the corner stone of the MB processors. Most processors on the market (hardware or software) just can't reproduce correctly a square wave with no gain change.
Old 22nd September 2016
  #26
Good points Laurend!
Old 23rd September 2016
  #27
Lives for gear
 

Verified Member
I still feel people are being quite binary here.

MB compression can be a totally legitimate tool to do a job.

If I have a track where for example there is over excited hi hats and a duller vocal, I can use the TC Multiband within seconds to put them into relative unity and move on with my EQ. 10% of the overall tonal processing. Not something I could with EQ without changing the already problematic vocal. That's not "restoration" that's just mastering equalizing like I do day in day out.

Would I ever use one for a kinda pseudo radio broadcasting thing? nah, sounds weird.

Tools are tools.
Old 23rd September 2016
  #28
Lives for gear
 
SmoothTone's Avatar
 

Quote:
Originally Posted by Joe_caithness View Post
I still feel people are being quite binary here.

MB compression can be a totally legitimate tool to do a job.

If I have a track where for example there is over excited hi hats and a duller vocal, I can use the TC Multiband within seconds to put them into relative unity and move on with my EQ. 10% of the overall tonal processing. Not something I could with EQ without changing the already problematic vocal. That's not "restoration" that's just mastering equalizing like I do day in day out.

Would I ever use one for a kinda pseudo radio broadcasting thing? nah, sounds weird.

Tools are tools.
Fair points Joe. I use MB/de-essers/dynamic EQ for similar tasks when needed.

Perhaps the point that doesn't get stated explicitly enough is that this type of application is more likely to be used in mastering than this weird idea that a fully loaded MB compressor (several bands across the spectrum) is somehow a staple or secret weapon for MEs.

The marketing that created this misconception seems to have been surprisingly successful.
Old 23rd September 2016
  #29
Lives for gear
 
m_gant's Avatar
 

Quote:
Originally Posted by SmoothTone View Post
Fair points Joe. I use MB/de-essers/dynamic EQ for similar tasks when needed.

Perhaps the point that doesn't get stated explicitly enough is that this type of application is more likely to be used in mastering than this weird idea that a fully loaded MB compressor (several bands across the spectrum) is somehow a staple or secret weapon for MEs.

The marketing that created this misconception seems to have been surprisingly successful.
I'd like to thank everyone for the kind wisdom shared here. And you are spot on. Buy any "mastering" bundle. Like waves, you get a MB, EQ, and limiter. Where's the broadband? How about ozone early versions? There's no broadband compression. Load up a logic "broadcast ready" master chain..yup more MB.
Old 23rd September 2016
  #30
Lives for gear
 
Silvertone's Avatar
MB is a tool. It's not used often in mastering but when necessary can be quite effective.

My secret weapon is my years of experience. To paraphrase Bob Ludwig "I can hear a track and immediately know what needs to be done" and I have several options available to do said task.

This is the only secret weapon a mastering engineer needs and it takes about 10 years or so to develop that skill set. Until you do, you are just shooting in the dark.

Have fun learning everyone... it's where the real fun is.
Topic:
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump