Music Into Noise: The Destructive Use Of Dynamic Range Compression part 2

It seems to my ears like the sound check feature on iTunes compensates for differing loudness to some extent (as it's supposed to). For instance, it makes the peaks from an Esquivel disc mastered in 1994 sound louder than tracks from Sleater-Kinney's The Woods. It's not quite the same as a built-in compressor, but it would theoretically make it possible for music to be mastered dynamically while allowing listeners to have relatively consistent overall volume among different tracks.

But maybe mere consistency of volume isn't what people actually want, if they do prefer the overtly harsh sound.

Since this thread is about both kinds of compression, a good example is something like an early Who single. Kit Lambert and Pete Townshend would have compressed the hell out of it in the production stage as an artistic choice for impact, punch, sound good on a transistor radio etc.

Afterward, it would be mastered for 45, LP, comp LP, early CD, remastered CD, digitally remastered audiophile vinyl, SACD, original analog tape proper digital remaster/remix, surround sound HD DVD-A.

All I want is the full dynamic range high resolution reproduction of what the band heard in the playback the day they sent their final mix off. Or maybe I'm looking for an exact reproduction of what the 45 sounded like on a normal record player.

I don't know how anybody could even determine if Radiohead sounds bad or not, it's a completely manufactured music. The only thing that could sound wrong is if someone at the factory messed up what the creators sent out. Or if the label destroyed the sound during a format shift. Meaning, unless I hear a band member complaining about a failure, I assume every sound on an OK Computer CD including compression artifacts is an artistic choice and only right or wrong as a matter of taste.

A low bit-rate after the fact reproduction, on the other hand, diminishes the intent, and I think it is too bad that many people don't know the difference or care. Ipods will play higher resolution music either through itunes or third party software, and can still hold and play more music at one time than any one can listen to in a reasonable time period. You bought it though, so use it as you like.

OK Computer is supposed to have a flat shiny modernistic plastic sheen over the whole sound. That's kind of the point.

Indeed it is. Sadly a lot of people copied it without realising the semiotics of it.

In Rainbows though I would say is simply too loud, too bright, mixed with elements in odd proportions. To me it's a pretty standard example of what's wrong with modern mainstream rock sound. There's delicacy inherent in many of the tunes, melodies, and arrangements, but it's really not there in the actual physical qualities of the sound.

Johnny Greenwood has been very keen to point out at various points that he's not an audiophile, that they're not a band into high fidelity sound. This is not surprising.

Not sure if anyone cares, but I know Mouthy enjoyed 'Open Season'. Encouraging news on BSP's 'Man of Aran' soundtrack:

http://yanbsp.wordpress.com/2009/05/20/further-notes-that-i-don’t-mind-if-anyone-reads-or-don’t-reads/

"Basically turn up your volume. Because it sounds quieter but is in fact louder."

http://www.nytimes.com/2009/06/28/arts/music/28pare.html

Current R&B is “crushed” for radio play, Maxwell said: “It’s made to have no peaks, no valleys. And we wanted, like, landscape. I wanted the music to sound like, ‘Oh my god, I’m looking at a vista, like I see a mountain over here, and water, and some clouds.’ “ As the album was about to be mastered, Maxwell decided that it had already grown too compressed — its ups and downs leveled out — and sent it back for a remix.

!!

nick do know retribution gospel choir? (sparhawk from low's new band)

anyway that record is BIZARRE sounding...i'd be curious if you've heard it

http://i21.photobucket.com/albums/b276/oliver8bit/comparison_powerless.jpg

I did a needledrop recording the 2xLP vinyl "Embryonic" and at left is the vinyl and at right is the CD. The vinyl sounds much better to me, and as you can see above is less compressed. If you're curious you can find my rip on the d3m0n01d.

Were you trying to post waveforms there?

The new retribution is mixed by a guy who has also mixed britney spears, vanessa hudgens, avril lavigne. I think Alan is friends with the guy and thought it would be fun to have a more glossy amped up sound as RGC is his more 'rock' project. fwiw i think it kind of sounds like boston in places - and i mean that in a very good way, it's a really fun lp although the mix probably goes a little too far in the compressed direction for my liking.

No mention of this followup article, here?

hey adam thanks, i love embyronic but frankly find the CD tiring to listen to....will definitely buy it on vinyl based on your post, was kinda on the fence about it

"Embryonic"'s vinyl mix should be released on CD for maximum sound quality in every possible way.

(x-x-post) That's a very encouraging article.

There's a DVD audio release of Embryonic as well but I haven't checked it out. I wonder if the compression/volume is different on that as well. After reading that article (and thanks so much for posting it!) I suppose not. I do know that mastering for vinyl you must take volume into account to prevent the needle from jumping during overwhelming passages.

Embryonic on vinyl not only sounds right but _feels_ right, like it's some lost artifact of experimental 1970s prog.

That is a striking difference, in terms of all that peak info being lost on the digital. Just so I am understanding this correctly, you do set a recording level when ripping the vinyl, right? I can see the differences in the peaks there, so they are obvious, but it seems like the *absolute* volume levels aren't comparable since presumably you are adjusting the level when it goes into the computer. But clearly the peak info is lost in the digital in this case. I think sometimes people post wave forms from vinyl vs. CD here without understanding that you have to set the peak at close to the same level to get a look at the differences in compression.

Yeah you're totally right there. Thing is I plugged the audio out straight from my turntable's RCA jack into the line in on my computer and had the volume set at 100%. Anyways I just loaded the files up and amplified the vinyl WAV by 190% and the top volume looks close to the CD version. Here's the new comparison:

http://i21.photobucket.com/albums/b276/oliver8bit/comparison.jpg

Yeah, that is a huge difference in dynamic range there, need to get this one on vinyl!

lolllllllllllllllll:

http://blog.winamp.com/2010/01/19/how-to-make-your-mp3-sound-like-vinyl/

http://www.blogcdn.com/blog.winamp.com/media/2010/01/izotope-vinyl.jpg

LOL (Would want to fiddle with the year button)

"Of course, in this case sound quality means making it sound realistically bad, and we spent significant time tuning models and samples and developing 64-bit signal processing for noise shaping, distortion, EQ and dithering."

Holy crap, this is the George Bush of software applications. We are slowly sinking into the sea.

"Crackling Noises OK. Do Not Correct!"

um, dithering anyone?

it's kinda funny how much people are amazed when they come over and i put on a record that's new or in good condition and there's hardly any noise or crackles and stuff...

they should have marketed that winamp thing as *izotope noize*. for budding merzbows.

that response article is interesting. quick question, does mp3gain do the same job as replaygain? is one better than the other?

now you can make your own Gucci Mane 78s

no wax cylinder option no credability

that response article is interesting. quick question, does mp3gain do the same job as replaygain? is one better than the other? ― NI, Tuesday, 16 February 2010 17:26 (20 minutes ago)

iirc, there are differences between the two- for instance, in terms of where the gain information is written (At least last time I read about it). There are posts about it on hydrogen audio forums and in the FAQ.

ok, looks like I mis-spoke and should have checked before I posted. Anyhow, 'replaygain' seems to be a generic term for the process, whereas mp3gain and the 'replay gain' scanner in foobar, and specific implementations. As for how they can differ- browse this

ARE specific implementations

I don't think the peaks being visually more prominent in the vinyl-to-digital transfer necessarily means that the overall dynamic range is greater. Any time you transfer from vinyl, even if the source for the vinyl master is basically the original CD master with the volume turned down, you'll see peaks that look more "natural."

Waveform Plots Considered Harmful

excellent link, thanks! I didn't know any of that stuff.

lol at this:

Audio engineers use an "RMS" value that is conceptually about the same as ReplayGain, and shares its faults. However, it is an acceptable figure of merit proving that Iggy Pop is a crappy producer.

The-Dream, "Fancy"

http://i47.tinypic.com/xp77zm.jpg

i feel it's only fair as a patented annoying vinyl fan to give a great big shout-out to all the people who made AMAZING sounding CDs that I heard last year. digital rules! seriously, i heard so much great electronic and electro-acoustic music on cd last year. there are people who really know what they are doing when they make music for CD. the problem is a lot of them are making music that not a lot of people hear.

Dynamic range day is on my 'stag do'.

http://productionadvice.co.uk/dynamic-range-day/

<3 the "fancy" crescendo

I like the way the Phoenix record sounds

i still don't think this is a widespread thing at all tbh - i mean if you only listen to radio hits or songs built for radio play, maybe, but who does that? sometimes you want beyoncé, sometimes you want maxwell.

No, it's widespread.

Not all records are necessarily totally brickwalled and clipped to unlistenability, but the vast, vast majority are considerably louder than they need to be, with flatter dynamics. it's a very subtle thing but it does make a difference.

The Phoenix record does sound really good, but it could be quieter and more dynamic and I think that would make it sound even better.

If you use something like Scott walker's The Drift, or Kate Bush's Aerial, as a sonic benchmark, pretty much everything else sounds louder and less detailed than them, and there's no reason for it. Even radio-focused stuff like Beyonce doesn't need to be louder than Kate Bush; radio stations run everything through massive compression anyway.

I know I'm banging head against brick wall here.

nick as far as i know you don't listen to the artists i'm referring to anyway - how would you know whether erykah badu, maxwell, sade or any of the techno/dubstep/funky/bobbins i listen to are over-compressed?!

and yeah how would i know whether phoenix or whoever are over-compressed but i don't really give a shit about them so

Well I may not have listened to the latest badu, maxwell, or sade records, but I do listen to quite a bit of techno and have been known to dabble in dubstep a little; plus you said "this isn't widespread" not "badu, maxwell, and sade aren't overcompressed" so I assumed you were talking about a wider range of artists than that!

That waveform for the The-Dream track does look absolutely fine though.

well, do you think the techno/dubstep you listen to is overcompressed? b/c in the main, i don't, or at least i don't hear it in those genres, in the slightest, the way i do in eg an alexandra burke radio hit.

i basically mean "a large proportion of all the artists i listen to" and just gave those three as particularly well-known examples in lieu of actually going through my last.fm profile and listing everyone whose music i think doesn't sound overcompressed.

the more generally electronic (inc. rock guitar) the music/genre sonically the more likely it is to be overcompressed perhaps. i'd be surprised if hard dance music played in clubs was not also affected by this. more organic/trad productions are less likely to be affected.

I think a lot of electronic music that I've listened to since 2001, say, of all different stripes, from Burial to The Field to Nathan Fake (especially him) to Vitalic to The Knife to Lindstrom to Four Tet, even (many of these I absolutely adore) are all louder and less dynamic than they need to be, in pure absolutest terms. Some of the time this doesn't bother me that much (The Field), some times it means I get rid of records cos I can't stand them (Nathan Fake). I think the new Knife album is WAY better than Silent Shout, for instance (Fever ray is also way better than Silent Shout, though still a little too loud ion absolutist terms).

Tell you what isn't, though, and what I wish a lot of other electronic records sounded like in terms of mastering levels; the DJ Sprinkles album. Fucking amazing sounding record.

It's not that The Field is ruined by being overcompressed, for instance; just that I think it could be a LOT better.

the dj sprinkles record sounds really nice but i'm not hearing the "fucking amazing wow wow wow" compared with any other given deep house (either in terms of quality or sound tbh)

the knife's monochrome flatness is a kinda integral part of their aesthetic surely? can't imagine anything about their music isn't a deliberate decision.

listening to the field is like being smothered in cotton wool but i ascribe this to his limited sonic palette and apparent ignorance of this thing called bass.

everything i've heard by lindstrøm has been a+ and spot on when it comes to production.

I think the new Knife album is WAY better than Silent Shout, for instance (Fever ray is also way better than Silent Shout, though still a little too loud ion absolutist terms)

do you just mean from a loudness/compression pov? because the only reason i can think why they may have toned that down since Silent Shout is because they're less dance-orientated (which is interesting). haven't actually heard "Tomorrow, In A Year" yet tho