What's new

Hyper Compression: Am I doing it wrong?

Piano Pete

Senior Member
Hi guys,

Per my other thread, I have been hitting the books and practicing a lot of mixing. Something that I noticed is that when I hyper compress anything, especially if there are a ton of tracks going, I seem to spend a lot of time fighting for clarity. As I have been reviewing my exports back to back, I have been noticing a trend: the works that are not hyper compressed are balanced but not freakishly loud, no odd buildups. The works that are hyper compressed tend to have a large build up in the lowmids-mids that are muddying everything up. So I am thinking following:

1) The distortion caused by hyper compression tends to build up around these muddying frequencies.

2) This is a reflection that I mix with my speakers tuned to 75dB, spotting at 80, and at the higher dB I am noticing something I should've fixed when it was mixed lower. (Even if I crank my speakers up to check the dynamic mixes they tend to be fine).

or

3) I am not actually producing a clean mix to begin with regardless of my speakers' level.

Thoughts?
 
Hi guys,

Per my other thread, I have been hitting the books and practicing a lot of mixing. Something that I noticed is that when I hyper compress anything, especially if there are a ton of tracks going, I seem to spend a lot of time fighting for clarity. As I have been reviewing my exports back to back, I have been noticing a trend: the works that are not hyper compressed are balanced but not freakishly loud, no odd buildups. The works that are hyper compressed tend to have a large build up in the lowmids-mids that are muddying everything up. So I am thinking following:

1) The distortion caused by hyper compression tends to build up around these muddying frequencies.

2) This is a reflection that I mix with my speakers tuned to 75dB, spotting at 80, and at the higher dB I am noticing something I should've fixed when it was mixed lower. (Even if I crank my speakers up to check the dynamic mixes they tend to be fine).

or

3) I am not actually producing a clean mix to begin with regardless of my speakers' level.

Thoughts?

It would be easier to hear with an example.

Hyper-compression makes sense if you wanna bring back the tail of percussive sounds or short strings. Its almost more a sound design tool, then a mixing technique.

Mix wise, I would only compress busses ever so slightly to glue them together.

Regarding your low mids build up, I would use a multiband compressors to even out the signal in that area.
 
The reason I am trying the whole hyper compression thing, is that some people commented that I need to be able to competed with "loud" tracks, but I have this aversion to slamming stuff into a limiter. Since I have been messing around with it, I have been noticing more issues.

Thank you for the MB advice, that is actually what I had used on the stereo mix when doing my side by sides when I noticed the spike in clarity.

The other thing I was wondering is if it simply comes down to that each project has to be mixed, and tweaked, differently when trying to just pump volume out versus when not. I have noticed I reach for more MB compression and M/S EQ's to tame the loud tracks than my dynamic mixes.
 
First things first, why are you 'hyper compressing', what do you define as hyper compression and what stage and what elements in your mix are you applying it to?
 
I am trying to get better at mixing loud, dense mixes. Anything that I have been shipping for cash has been pretty dynamic, and I always leave headroom for mastering. Typically, I hand the stems off to someone who enjoys mixing as much as I do composing, so I want to get better at mixing in general. Up until now, I have only been doing some small contract work, and I have never been asked, "could it be louder," but I want to be able to handle "loud" tracks in case I ever need to.

I have always been under the impression that hyper compression is just maxing out the loudest a track can go before distortion kicks in-- trying to shove the rms as high as possible while retaining some dynamic movement. I have been given conflicting information about how to best achieve this, either through small compression and limiting throughout the mix or by just putting a single limiter on the master. I have had mixed results with both, although I have achieved better clarity when applying small amounts across the entire mix.

My mixing routine has so far been: Get a good balanced mix first, remove any wonky buildups and accentuate what I want to pop out (make sure everything is heard), and then lately, slowly cranking up the overall rms to dip my toes into the "loudness wars." The last step is where I have ran into the most problems with clarity and build ups of wonky frequencies that previously did not exist.
 
If it's mostly loudness you're concerned with that can be surely achieved in the mastering stage. A well balanced mix with enough headroom should be able to be pushed up in that "loudness war"-area, if so desired. Limiters, Multibands and dedicated mastering plugins like the FG-X from Slate or the Ozone suite can really be helpful in that regard.
Unless you're producing a dance track I wouldn't compress the final mix too much, it just kills dynamics.
 
It just hit me that I am trying to battle mixing/mastering at the same time, although I have been doing it in the right order. It is still bothering me that when I try to, in reality, master a track, I get that build up and lose clarity. I hope that isnt a reflection of an unbalanced mix, although I am pretty paranoid about testing anything I make on my focals, phone, and a pair of standard dell speakers.
 
Right, yeah it's probably best to seperate mixing and mastering, so when you're satisfied with your mix export it as .wav with around -3 to -6db and then do the mastering in a seperate project. It's hard to say without an example as to what might be the reason for your build ups. Could be arrangement, mix, EQ etc. basically a lot that might not have anything to do with the mastering process directly.
 
It just hit me that I am trying to battle mixing/mastering at the same time, although I have been doing it in the right order. It is still bothering me that when I try to, in reality, master a track, I get that build up and lose clarity. I hope that isnt a reflection of an unbalanced mix, although I am pretty paranoid about testing anything I make on my focals, phone, and a pair of standard dell speakers.
somebody else said before that loudness is something you archive at the mastering stage, so, since I agree I'll not comment on this further. but there is one crucial thing that influences the result of the last stage: the internal dynamic of your mix. You'll only get to a loud and distortion free master if you make sure that your mix has no significant peaks in the first place. in my opinion all this comes down to levels and those directly reflect what you actually hear. I had a real eye opener a couple of days ago, mixing in a really good room. i always thought my room is good (or at least sufficient) but what I suddenly was able to hear in the other room made me understand a couple of things about levels. i've only changed levels and EQed out bumps and my mix is now louder and less cluttered (> more clarity).
 
What genre are you working in, and what would you consider to be a comparable reference track with a sound you're trying to aim for? That would help with some specific advice.

The thing about doing 'hyper compression' is that to do it well it's a very high maintenance procedure. All your sounds, edits, recordings, dynamics, tonal balances, glitches etc have to be sorted out because that's all the crap that'll be brought out by smashing stuff. The chances are you'll have to be very interventionist with the automation too, to make the dynamics work and create space as elements come in and out and develop over a piece.

What others are saying about mastering being the cherry on top is true- your mix needs to be very solid and 'loud' sounding all by itself. If you can't throw an L2 on there at the end and get the overall level in the ballpark yourself without artefacts, then that means you need to work more on the mix.
 
Thanks for the comments guys.

Im away from my setup right now, but I think what I am going to do is check my mixes with the increased gain to see what sticks out, to see what I am missing when mixing, before I hit the mastering stage. This is probably the stimulus for a lot of my issues. Additionally, I need to separate the mixing and mastering stages. For some reason, I have completely ignored this since trying to learn to mix louder tracks.

Super loud tracks aside, what I truly wish to gain from this endeavor is to be able to produce nice, clean mixes that are at comfortable levels without me having to rely on other engineers as much. A big wake up call was a piece of "epic" trailer music that I was writing that I could not wrangle for the life of me (another forum post). I handed it off to the engineer which was all fine and dandy, but I feel horrible that I could not at least get it to a presentable level of quality on my own. I come from the concert world, pencil and paper is comfortable to me, and it is terrifying to me that orchestration that I know works all of a sudden becomes a pile of mud in the digital environment. Unfortunately, I tend to write pretty dense, which is something I have been learning to taper the more I have been asked to work digitally.

Scoremixer, most of my music tends to be more traditional in instrumentation, although I do work quite a bit with synths now, which is where the "loudness" wars thing stemmed from when talking to a buddy of mine. I have had more issues trying to mix samples than live recordings-- I think I spend more time trying to remove shrill tones and such from sampled strings and brass. For reference for the trailer piece, I believe I used some Two-steps From Hell stuff. I cannot remember which pieces I looked at as I don't really listen to this genre, but I know they have amazingly clear mixes that translate well across playback devices. For my standard affairs, I have just recently been referencing my mixes against Howard Shore's, Desplat's, Miklós Rózsa's, and some of Jerry Goldsmith's finished works. I know these are not hyper compressed, but they are well done--as one would expect from professional productions.

Unfortunately, I do not have any recordings or anything on my person that I can upload.
 
Gotta hear to say anything intelligent, but...

- If the mix sounds good before you "hypercompress" it, well, you're over-compressing! Sort of like when Larry Holmes was commenting on a boxing match: "If his left hand is hurt, he's smart to be hitting him with his right." :)

- If everything is all the way up, nothing is all the way up. In general, sustained notes sound louder than short ones, so longer snare and kick sounds (if you're using them) will make the whole thing sound louder. Tune the snare higher if the tempo seems to slow down as a result.

(I learned about short and long notes from playing recorder, which has almost no dynamic control, so you have to resort to tricks. That was way before I started playing with recording toys.)

- Compressors really do change the tonal balance. It's more common to use EQ before compression, but not always.
 
Gotta hear to say anything intelligent, but...

- If the mix sounds good before you "hypercompress" it, well, you're over-compressing! Sort of like when Larry Holmes was commenting on a boxing match: "If his left hand is hurt, he's smart to be hitting him with his right." :)

- If everything is all the way up, nothing is all the way up. In general, sustained notes sound louder than short ones, so longer snare and kick sounds (if you're using them) will make the whole thing sound louder. Tune the snare higher if the tempo seems to slow down as a result.

(I learned about short and long notes from playing recorder, which has almost no dynamic control, so you have to resort to tricks. That was way before I started playing with recording toys.)

- Compressors really do change the tonal balance. It's more common to use EQ before compression, but not always.
Haha, thanks for the words of wisdom nick!
 
I like to pretend.

What I didn't mention is that you get sustain on percussive sounds (e.g. snare and bd) with fast attack-slow release settings.
 
Are you sure you're not referring to parallel compression? The only technique I know of where you'd 'hyper' compress anything is where you only mix a small amount back in to gain some RMS power... Even then it's not common to do it to a mix. A drum buss? Sure, (1176 All buttons, Distressor Nuke etc... Totally common in metal or heavy rock...) Even then you do it because you're after specific sound.

But a whole mix? No, not common. Limiters, and/or a combination of limiters and soft clippers and saturation are the only way to create the perceived volume people tend to chase when beating a track to death. You can parallel compress into a limiting chain, but even then less is more. The limiter's going to be doing the lions share of heavy lifting and get you significantly more perceived volume than compression will. Too much compression will just flatten your track no matter what genre.
 
Last edited:
Whilst compression and certainly multi-band compression could come a long way, you might be asking yourself the wrong questions. Certainly, loudness is a thing these days. But there's loudness and there's noise. You want loud? Don't look at your level meter, listen to your mix.

Mixing (and composing as well, for that matter) is a matter of making choices. What is, at any given moment, the most important element in the mix? Is it the drums? The string staccatos? The blaring of the horns? The choir that comes in?

Good mixes are good mixes because of the choices made within that mix. There's this stupid trend these days to just pile everything up into one big cluster of sounds in which nothing stands out. It's not a wall of sound, it's a wall of noise. More and more mixing engineers should lay off the compressors and first look at what choices they have made and wether they are good. And so should composers too!

Not everything can be as important as everything else. Because then, nothing is important. What do you want to say with your composition? Make a list of priorities. Drums first, then strings, then horns? The other way around? Two elements that are equally important clash sonically with each other? Make a choice. Kill your darlings, is what they say. Having elements not clashing with each other creates space. Space in which every element can be loud, without piling up on each other. Remember, however, that only one element can be loudest.

Giving elements space makes music sound much louder than overcompression. Want your drums to really stand out among the crowd? Want people to say 'whoa, do you hear these drums, those things are huge!'?

Give them space. Or give any other element in the mix space. You say dense composition? I hear too much elements all fighting for space at the same time (This is not a critique on your composing skills, by the way).

I know this can be irritating, because you don't want to go back to your composition and change things. At the least, learn from this. Don't fix it in the mix, build your composition up with heavy loudness in mind. And then, only then, can compression, parallel compression, multiband compression, EQing get you to where you want. Many engineers say that a well recorded song can become great in the mixing stage, and a badly recorded song can only go so much in that direction. The same goes for the composition! A well done composition can become great in mixing, a bad composition can only go so far.
 
Last edited:
Whilst compression and certainly multi-band compression could come a long way, you might be asking yourself the wrong questions. Certainly, loudness is a thing these days. But there's loudness and there's noise. You want loud? Don't look at your level meter, listen to your mix.

Mixing (and composing as well, for that matter) is a matter of making choices. What is, at any given moment, the most important element in the mix? Is it the drums? The string staccatos? The blaring of the horns? The choir that comes in?

Good mixes are good mixes because of the choices made within that mix. There's this stupid trend these days to just pile everything up into one big cluster of sounds in which nothing stands out. It's not a wall of sound, it's a wall of noise. More and more mixing engineers should lay off the compressors and first look at what choices they have made and wether they are good. And so should composers too!

Not everything can be as important as everything else. Because then, nothing is important. What do you want to say with your composition? Make a list of priorities. Drums first, then strings, then horns? The other way around? Two elements that are equally important clash sonically with each other? Make a choice. Kill your darlings, is what they say. Having elements not clashing with each other creates space. Space in which every element can be loud, without piling up on each other. Remember, however, that only one element can be loudest.

Giving elements space makes music sound much louder than overcompression. Want your drums to really stand out among the crowd? Want people to say 'whoa, do you hear these drums, those things are huge!'?

Give them space. Or give any other element in the mix space. You say dense composition? I hear too much elements all fighting for space at the same time (This is not a critique on your composing skills, by the way).

I know this can be irritating, because you don't want to go back to your composition and change things. At the least, learn from this. Don't fix it in the mix, build your composition up with heavy loudness in mind. And then, only then, can compression, parallel compression, multiband compression, EQing get you to where you want. Many engineers say that a well recorded song can become great in the mixing stage, and a badly recorded song can only go so much in that direction. The same goes for the composition! A well done composition can become great in mixing, a bad composition can only go so far.

Thank you for the comments. I completely understand about giving elements their space in the mix, and I do try to take care and consider this during the orchestration stage, even with it being contrapuntal. No love lost over your comments, it needs to be said and considered!

Im currently in the middle of recording for several projects, and I have been noticing another trend. When working with samples towards the goal of a loud track, the clarity is so-so. (Haven't been able to mess with any samples since I left.) Currently, from grabbing the stems from the performers to experiment during my off time, I seem to have an easier time achieving balance with the live recordings. I am sure this is just coincidence, or a reflection I am actually learning something, but it is a night and day difference. :\ The sessions have been a mix of ensemble recording and individual dubs.
 
Top Bottom