Compression Doesn't Mean Bad Sound

Nightfall

I do agree that the loudness war has had an adverse affect on music but it certainly doesn't mean that a highly compressed album is automatically garbage sounding, am I alone in this?

Tool "10,000 Days" sounds pretty damn compressed to me but still sounds REALLY quite good. Soundstage width is fantastic on this album for whatever reason.

nooshinjohn

I think you would be alone in this, but I would not be surprised at all if X offers you his support.

Nightfall

DSkip wrote: »

Compression is there, but it's a different compression than most albums.

Maybe this is why I am drawing this conclusion. What do you mean by "different than most"?

DSkip wrote: »

It's still not good for it to be there though.

Fully agree.

mhardy6647

Virtually all commercial recordings use compression, limiting, "imaginative" use of EQ, artificial reverb, double-tracking (of vocals), etc. -- and always have (even in the all-analog, pre "AutoTune" days). And, of course, virtually all pop music recordings are not stereo in any true sense of the word -- they're multiple-track, multi-mike mono recordings artificially mixed and sweetened to produce some completely artificial sense of three-dimensional acoustic space.

There are exceptions (mostly audiophile recordings, some small-combo studio and live jazz and some - particularly older - "classical" music recordings, large and small) to all of the proceeding, of course, but by and large -- it's all just part of the game.

I realize the topic is relatively extreme processing/dynamic range limiting, but if one stops and thinks about how real (or not) a recording sounds... it's actually kind of amazing that so many recordings can sound as good as they do in a home environment. It's a credit to the skills, knowledge and experience of the engineer(s) and the producer.

nooshinjohn

I now have music that used to sound pretty damn good that now sounds like complete crap thanks to the upgrades I have made in my gear...

BlueFox

nooshinjohn wrote: »

I now have music that used to sound pretty damn good that now sounds like complete crap thanks to the upgrades I have made in my gear...

That is impossible since all gear sounds the same if the specs are the same. Actually, you are just imaging there is a difference. Humans are inherently stupid and delusional, and are easily fooled by the upgrade myth. You are easily manipulated by the manufacturer’s snake oil marketing into to thinking your system sounds better. Only certain intellects are capable of not falling for these audiophool myths. You need to be sent to the audio reeducation camp, and have your mind wiped clean of this nonsense of better sound from better gear, cables, and clean power.

Sorry, just trying to poke fun at the two current motor mouth posters who know everything.

Nightfall

nooshinjohn wrote: »

I now have music that used to sound pretty damn good that now sounds like complete crap thanks to the upgrades I have made in my gear...

I just can't imagine 10,000 Days sounding like complete crap no matter the level of system used to play it back.

I could be very wrong though.

tonyb

BlueFox wrote: »

...sent to the audio reeducation camp,

Ha...now that's sig material there, I dig it.

mhardy6647

There are good components that are much more tolerant/forgiving of poor source material.

Lowther drivers = unforgiving
QUAD ESL-57 electrostatics = forgiving

QUAD ESL-57s are on many folks' list of the most significant hifi products ever made. They are not poorer loudspeakers for the property of making even not-so-hot recordings sound good.

cnh

I am partial to rigs that are forgiving. Probably one reason why I spend so much time with old vintage gear where even a cheap Sony CDP from the late '80s early '90s sounds just fine. Of course, it goes without saying that these systems are NOT the most resolving. But having been in a number of live venues, I dare say that all THAT resolution is often artificial, not found, or even heard live in many cases.

Just how many times has a live performance driven you out of the concert hall screaming "my ears are bleeding"?

mhardy6647

Horses for courses, as they say.

The SE 2A3 Altec Duplex system sounds great (relative to my expectations) with most any source component and with most any recording in any format that I've tried. That said, the Duplexes are hardly forgiving -- they are in fact quite revealing of things like microphone technique. There's a reason tha they were ubiquitous monitors in studios and control rooms literally for decades. The nice thing, though, is that I find that I can listen as deep or as shallow as I wish.

That said, it's handy to keep things around like a fairly pedestrian class AB push-pull solid state amplifier & "low maintenance" loudspeakers just to listen to stuff. The original Polk Monitor series will fit that spec quite nicely, as would the classic British box speakers upon which they were modeled (KEF, Rogers, Chartwell, and a couple of others). The vintage ads L-710 is one of my favorite smooth, musical and forgiving loudpeakers. Everyone should have a pair handy.

steveinaz

Compression that results in clipping, is the worst variety. I'd rather listen to dynamic range limited music, than clipped (overdriven). This may explain why many prefer vinyl to CD; and the kicker is, when CD's are "squashed" they have far less dynamic range than vinyl! Go figure.

CoolJazz

Here is a little back ground information for those interested in dynamic compression. (As opposed to data compression).

The first difference to know is the various terms applied to dynamic compression…automatic gain control (AGC), limiting and then finally, clipping. The single biggest difference between all of these is speed with which the compression is applied. Otherwise, you're saying the same thing.

First with AGC it’s speed that it reacts at is generally regarded as a “slow hand on the pot”. Something starts to get louder and the hand gently and in a controlled fashion will reduce the level. Then at some point if it gets low enough in level, it slowly turns back up the level.

With limiting, it’s much faster. Individual elements in the music can cause it to react quickly. Normally the circuitry is such it’s used in only a downward direction. Or it reduces the loud to maintain a more consistent peak level. This stage is where the term compression is most appropriate.

Then with clipping, which is much easier to associate from the term, a threshold is set that anytime something exceeds this, it’s simply removed. Think of it like a zener diode set to trigger at 4.5 volts. 4.45 it ideally does nothing. At anything over the precise value of 4.5 though, it simply doesn’t pass. It’s gone. Initially, the clip was only for a “if all else fails” and was to prevent gross distortion or illegal operation with too much level. Think of a dropped microphone or somebody bumping a mic with an instrument.

So the way for decades audio has been processed, these elements were stacked one after the other. Slowest first to fastest last. Individual equipment came about that combined all of these elements into one box. A manufacturer might configure for more of one step and then rely on less of another. Over time, the adjustability to the end user became more and more available to allow the user to tailor to their chosen results. All along, more and more adjustments became available too. Speed for the attack or the reaction time to louder or softer material being presented to a stage. Release time or how long it stays at the changed level expecting more of the same.

Poor implementation of the controls caused things like a holes punched in material. Think of a dummer steadily, one handed, slowly popping a drum beat. If set wrong, each pop on the drum might change the mixed level and cause other instruments to be reduced in level for a moment at each hit. So a sustained piano note would get dropped by each beat and then come back up, to only on the next beat get whacked back down and rhythmically repeated.

Then digital processing came along. First implementations allowed more precise control over each setting. You could set an attack to 7.6 milliseconds and if that sounded good in a particular situation, it could be repeated in the future…rather than trying to remember to set one of many adjustments to about “3:30” or so. Actually digitizing the audio and getting as good quality as more traditional analog took a few generations to get decent. Early attempts at all digital just didn’t sound good!

But eventually what is possible in digital has become very capable. Capable of good and also very capable of poor. What is called “look ahead” is possible so that a stage knows what level is coming and how repetitive it is and can therefore adapt in a predicted fashion to do the job and do it more transparently. As hot poker to the eye as it sounds, clipping is also adaptable and smart and more removal of level than would ever look to be possible can be done without terrible artifacts. For instance, a single transient that’s stopped at some level doesn’t have to cause a lower level for everything else and can be removed without even being heard.

Steve you would be shocked at how much clipping isn’t necessarily terribly audible or offensive. Some of the grungier sonics can come from too aggressive use of the earlier stages being used too deeply in many cases and may not be from clipping at all.

Production processes of music have picked up on these modern units and have obviously used them extensively. Some better than others. But just as a guitarist uses distortion as part of the fabric of the sound, some groups and producers use them as part of the song signature. Then in other cases, someone with a heavy hand and a belief of his target audience being low-fi consumers listening on cheap equipment, only pursues the loudest possible result so the song “hits harder” in the first notes and stands out.

As being brought up in the original post, some dynamic compression can yield some things that are pleasing to listen to. Level through a song can be consistent. Some individually recorded instruments maybe lacking in eq due to mic use or room placement and be slightly corrected after the fact to fit into the song fabric better. Level can be held to some degree and still allow what seems to be some pretty good dynamics through too. It's all in how the combination of adjustments have been done.

On the whole, some of the gross distortions heard are part of intended, some are powerful tools that are cheaper than ever and packaged in easy to implement designs. Some can be well used and some is horrible.

This drawn out wordy information hasn’t even touched on all the additional points in the processing chain where a variety of eq can be used either. Or the fact that every stage can and usually is done in multiple bands. And the frequency and the slope of the crossover between those bands may be adjustable. Or the fact that each and every stage has tons of ability to set where it turns on or is "gated". It’s very touchy and capable equipment and sometimes very good results can be obtained with a good touch and ear and pretty often some less good product is manufactured too!

Perhaps more than you wanted to know...

CJ