APHEX TWIN - SYRO

[Frank Poole]

Someone leak this ffs.

[MassfreeKid]

lol @ complaining about imperfect sound quality on an Aphex Twin release. It's like everything pre-1997 gets... listened to... with rose-colored, uh... earholes?

 

Go listen to that fucked up distortion about 3:30 into Next Heap With a couple hundred times as a sort of mantra to allow yourself to move on.

 

That's the impression I get. Everything he did before was perfect, but this one, OMG!!! Clipping!!!!

[SmackSmash]

 

I think people are only showing the waveform because it backs up what we've heard in the track. This was my experience (probably similar to some others here):

 

I listened to it when it premiered on Radio 1, and recorded it too. I thought it sounded kinda overcompressed but I figured that it was just the transmission (radio is notorious for this kind of stuff). I went back to my recording, saw the clipping (limiting, maximising, whatever... it was audible) and figured that I'd recorded at too high a level. I waited for it to come on again on Radio 6 and ended up with the same results. "Oh well, that's radio I guess". Then I saw that bleep had released it I was really excited to hear it in it's proper, shimmering, full, beautiful Tuss-like glory. I grabbed the 24bit wav and was disappointed to find out that it sounded just the same. Now of course it's nothing major: it's not had the shit smashed out of it or anything, but compared to what I was expecting it really didn't sound that good. It doesn't hinder my enjoyment of the track, but I thought it was weird considering that what we've had from him in the past decade or so has been so clean and, dare I say, 'perfect'. When I saw so many people on here talking about how incredible it sounds I just felt like I had to give an opinion on this. It's all subjective anyway, but after hearing the live version so many times I was really looking forward to hearing that bass pulse ripple around my room and fill the air with crystal clear definition, and this version just didn't quite hit the spot. Anyway, I think the vinyl version will be much better so I'm looking forward to that even more now.

 

INB4 "expectations", "audio snob", "RDJ knows better than you" - I'm starting to remember why I don't post here too often.

Now to the real question: what is your setup? What do you listen to this album with? This determines a huge portion of what you wrote.

 

 

So far I've tried it on a bunch of things: Adam A7's, Sennheiser HD650's, car stereo, a mates' expensive setup (not sure what it is, but I've had some amazing listening experiences on it. Listened to Burial and The Microphones in the same session and it was awesome), Sennheiser in-ear headphones, and a couple of Logitech PC speaker setups. Not the highest end gear list I know, but the same issues were apparent on every bit of gear so I don't think that's it. I'd be very surprised if it sounded better on anything else. Before everyone gangs up on me I want you to know that I'm totally nitpicking here. All of this stuff is subjective so I really hope this can remain a discussion and not turn in to a flaming match.

[nature]

When did so many people stop enjoying music?

[SmackSmash]

When did so many people stop enjoying music?

 

I love this track, and I think that's why I care enough to say this. It's a bit like this; you can love your wife but still think she could lose a couple of pounds. Of course you're never meant to actually say that

[Guest trananhhung]

This looks like neogaf when kids complain about 30 fps.

[YELLOW]

Speaking of questions interviewers should be asking...

 

 

STEINVORD

Edited September 8, 2014 by YELLOW

[MisterE]

i would agree that something being squashed flat, whether its actually clipping or just limited to death, it's going to sound like shit. but looking at that waveform, it doesn't look close to the high end of extreme. it sounded ok to me. maybe i would have liked it a bit more if there were a few db more of dynamics for the drums to punch out above the rest of the mix, but it is what it is and there's a lot to like about it.

 

maybe the vinyl got a different mastering process with less limiting, maybe not. im not too bothered tho because it just didn't sound that extreme to me.

[awepittance]

some of you guys need to stop using the word clipping so interchangeably with limiting, or just assuming that when you look at a track that has been limited, that what you're seeing is clipping. just about every track that gets released on a professional level is going to have some limiting somewhere, and of course pretty much all or most of the mainstream stuff on big labels uses a lot of it. but if any time a limiter was used meant that something clipped, then pretty much every single piece of music you listen to is clipped. they aren't the same thing. they are two totally different things. yes they are related in some ways, but not the same.

 

in fact, the basic idea behind limiting is to PREVENT clipping.

 

when you look at a waveform like that, all you can say is 'the dynamics have been reduced' and the peaks are being stopped from going beyond a point around 0db. that can be accomplished a few ways, limiting being one of them, but yeah sometimes mastering guys will intentionally do things like clip their A/D converters to lob maybe the top dB or so off of the peaks. they'd probably do that in combination with limiting. some converters are known to supposedly sound good when clipping like that. there are different types of clipping or at least different results. some is 'softer' (bit more rounded edges), some is just perfectly flat.

 

a 16bit wav cant actually have anything going over 0db recorded in that wav. which is why this whole thing goes on.

 

but when you see a waveform like that with the whole track on display, all you know is that there is almost surely limiting, and the signal may or may not have also been clipped in the mastering process. limiting is compression which means the volume level gets dropped as the signal is about to exceed the limit, so limiting prevents clipping, and if you zoom in horizontally on a track you limit with something like waves l2, even with extreme limiting, you will see that the waveform goes up close to 0db (or wherever you set the limit) but then curves back down, but it always stays 'rounded', and looks somewhat like the rest of the waveform. up close it still looks kind of 'natural', which is due to some crazy algorithms in how those limiters work. obviously you can still hear them affecting the sound, but it's NOT the same thing as clipping. if you zoom in on a clipped signal, its going to be like a flat line going across where that peak tried to go above a point and coudln't due to limitations of digital systems. so instead it just stays 'pegged' at that maximum allowed point, and you see a flat line. that causes distortion. yeah extreme limiting causes distortion too, but the signal is typically kept rounded and more natural looking with the much used big name limiters out there. theres a pretty huge difference between that and a signal just being completely chopped off. i think something like l2 tries to kind of 'mirror' the waveform back down below the limit, to retain it's basic structure. with clipping, that musical information that was happening where the signal tried to cross that point and got flattened, that information is just gone. its totally gone forever. and its sounds a lot different.

 

just showing the entire track zoomed out horizontally like that, you have absolutely no idea that there is significant 'clipping' on that track. zoom in, look for flat lines where peaks used to be, then talk about clipping.

 

limiting ≠ clipping

yeah i think some of the other users were incorrectly calling just extreme limiting = clipping, but the track in all versions indeed actually clips and hits the red multiple times (worse in the 16-bit version than in the 24-bit version) check it out for yourself. I've seen this before but I always assumed it was from a poor mastering engineer, but surely there must be a reason to allow a digital master to literally clip in the final file (doesn't just seem like an A/D convert clipping type of trick)? maybe you can explain it. Im not saying it detracts from the track or it sounds worse as a result, in fact i'm surprised it actually sounds fine given the clipping. So my question is more of a technical curiosity, I love the track wholeheartedly

 

(edit: of course whoever mastered this used a lot of limiting too, but for some reason decided not to set it as a -.01 but instead let it actually hit 0)

Edited September 8, 2014 by John Ehrlichman

[MisterE]

@JE

well i havent actually looked at it myself because i don't have a wav of it yet, but i dont know if i would outright call just going to 0db clipping. it became a standard to stop just below that but it was supposed to be because some cd players had problems with signals that went all the way up to 0db (maybe you said something like that already? cant remember). but i wouldn't call something 'clipped' unless i could actually see the top of the peaks literally 'clipped' off, and there would be something like a straight/flat line going across, until the signal comes back down on the other side of the peak.

 

so the thing about, if the signal is just going up to 0, but say maybe thats just because of the wave being normalized to 0db or limited at 0db, but its not actually having its tops chopped off, i would just say they limited at 0db instead of the usual practice of stopping just before that, which isn't a hardfast rule but more of a standard. if its actually clipped, and im not an expert on this, but they do that because clipping allows them to get the signal louder, just like limiting does, but they both have different results. a limiter has to recover, determined by the release time, so the level is still being held down for a bit even after the signal goes below the limit/threshold. so stuff thats just right after that limited peak will now be quieter than it was before (can be heard as what some people call 'pumping'), unless you use a super fast release. but super fast releases lead to other problems/distortions.

 

on the other hand, if you clip the tops of the peaks, there is no release time, so as soon as the signal comes back below the point of clipping, its pretty much untouched. the peak is gone, but everything else is the same. no pumping or any of those other side effects of limiters to worry about. some mastering guys found that different converters clip the signal in different sounding ways, some less harsh than others, with some supposedly even sounding a bit 'nice' if just a db or so is clipped off this way.

 

so, i think what mastering guys do now, is multiple passes of various methods of gradually chipping away at the dynamic range, squashing it a few dbs at a time. they'll maybe use a regular compressor first, then some clipping of peaks then some limiting. or maybe theyd do the limiting first then the clipping, im not sure if there would be a 'best' order to do it in. but each step would give them maybe a db or two or 3 of peak reduction. which adds up and supposedly sounds better than just using one tool to do it all in one go. which they will tell you is why taking L2 to your mix for 6+ db of peak reduction won't sound as good as what they do.

 

anyway there are lots of sometimes interesting discussions about intentionally clipping masters using a/d converters over at gearslutz, with some of the posts being actual mastering engineers.

Edited September 8, 2014 by MisterE

[perunamuusi]

You fucking people

[vertsk8er419]

page 181: tl;dr

[awepittance]

@JE

well i havent actually looked at it myself because i don't have a wav of it yet, but i dont know if i would outright call just going to 0db clipping. it became a standard to stop just below that but it was supposed to be because some cd players had problems with signals that went all the way up to 0db (maybe you said something like that already? cant remember). but i wouldn't call something 'clipped' unless i could actually see the top of the peaks literally 'clipped' off, and there would be something like a straight/flat line going across, until the signal comes back down on the other side of the peak.

 

so the thing about, if the signal is just going up to 0, but say maybe thats just because of the wave being normalized to 0db or limited at 0db, but its not actually having its tops chopped off, i would just say they limited at 0db instead of the usual practice of stopping just before that, which isn't a hardfast rule but more of a standard. if its actually clipped, and im not an expert on this, but they do that because clipping allows them to get the signal louder, just like limiting does, but they both have different results. a limiter has to recover, determined by the release time, so the level is still being held down for a bit even after the signal goes below the limit/threshold. so stuff thats just right after that limited peak will now be quieter than it was before (can be heard as what some people call 'pumping'), unless you use a super fast release. but super fast releases lead to other problems/distortions.

 

on the other hand, if you clip the tops of the peaks, there is no release time, so as soon as the signal comes back below the point of clipping, its pretty much untouched. the peak is gone, but everything else is the same. no pumping or any of those other side effects of limiters to worry about. some mastering guys found that different converters clip the signal in different sounding ways, some less harsh than others, with some supposedly even sounding a bit 'nice' if just a db or so is clipped off this way.

 

so, i think what mastering guys do now, is multiple passes of various methods of gradually chipping away at the dynamic range, squashing it a few dbs at a time. they'll maybe use a regular compressor first, then some clipping of peaks then some limiting. or maybe theyd do the limiting first then the clipping, im not sure if there would be a 'best' order to do it in. but each step would give them maybe a db or two or 3 of peak reduction. which adds up and supposedly sounds better than just using one tool to do it all in one go. which they will tell you is why taking L2 to your mix for 6+ db of peak reduction won't sound as good as what they do.

 

anyway there are lots of sometimes interesting discussions about intentionally clipping masters using a/d converters over at gearslutz, with some of the posts being actual mastering engineers.

while others might roll their eyes at a post like this, extremely informative for me. Yeah it could just be that it's going all the way up to 0db. All I know is that when I normalised a bunch of my own tracks to 0db instead of say right around -.01 the CD player in my car would actually make loud scratchy bursts during some of the percussive parts of 3-4 of the tracks. Most of the tracks actually sounded totally fine, and It wasn't until i double checked my masters did I realize they were all normalised to 0db (but seemingly only having an audible effect on a handful of the tracks only on one system, sounded fine on other systems)

[randomAmateur]

I think people are only showing the waveform because it backs up what we've heard in the track. This was my experience (probably similar to some others here):

 

I listened to it when it premiered on Radio 1, and recorded it too. I thought it sounded kinda overcompressed but I figured that it was just the transmission (radio is notorious for this kind of stuff). I went back to my recording, saw the clipping (limiting, maximising, whatever... it was audible) and figured that I'd recorded at too high a level. I waited for it to come on again on Radio 6 and ended up with the same results. "Oh well, that's radio I guess". Then I saw that bleep had released it I was really excited to hear it in it's proper, shimmering, full, beautiful Tuss-like glory. I grabbed the 24bit wav and was disappointed to find out that it sounded just the same. Now of course it's nothing major: it's not had the shit smashed out of it or anything, but compared to what I was expecting it really didn't sound that good. It doesn't hinder my enjoyment of the track, but I thought it was weird considering that what we've had from him in the past decade or so has been so clean and, dare I say, 'perfect'. When I saw so many people on here talking about how incredible it sounds I just felt like I had to give an opinion on this. It's all subjective anyway, but after hearing the live version so many times I was really looking forward to hearing that bass pulse ripple around my room and fill the air with crystal clear definition, and this version just didn't quite hit the spot. Anyway, I think the vinyl version will be much better so I'm looking forward to that even more now.

 

INB4 "expectations", "audio snob", "RDJ knows better than you" - I'm starting to remember why I don't post here too often.

btw I really like the track and

I had the same feeling - it somehow felt like a DJ MIX, I'm not saying Tiesto here, NOPE! The important thing is still the atmosphere, the melodies and the rhythm, it get's better with every listen due to the dynamics? Looking back at it today I don't feel like it has no dynamics, it was just weird on the first listen to me...

Edited September 9, 2014 by randomAmateur

[MisterE]

i think what you could say is that in that case, your cd player was probably clipping the signal as it tried to play it back, which comes down to the cd player and its own limitations. but the recording you were playing wasn't itself clipped (unless it was, which it would then sound clipped anywhere if that was the case)

[Consul]

Speaking of questions interviewers should be asking...

 

 

STEINVORD

 

Finally someone asked.

[Guest skibby]

my first impression on hearing minipops was that it possessed a lot of bass. one thing about slamming a track with a lot of bass is that it tends to really reduce the other frequencies because they are all in a sense, riding along the bass frequencies. so as everything approaches 0db, the mids and highs have no place left to go, so they get reduced, resulting in audible squishing. this isnt necessarily a mistake, it could be a creative decision.

 

i would have brought the bass down a touch on mixdown, but i bought my copy, and i will probably buy more copies for friends.

[randomAmateur]

next AFX album is going to be mono

http://www.youtube.com/watch?v=qY0ia6GaGaA

[Guest skibby]

 

Speaking of questions interviewers should be asking...

 

 

STEINVORD

 

Finally someone asked.

 

 

while we're dreaming i hope this one is on syro, extended, b side, i dont care, its one of my favourite jams!

 

[Hugh Mughnus]

pretty good fake huh?

 

watmmer made that

[Guest skibby]

pretty good fake huh?

 

watmmer made that

 

did you see him or her push the render button?

[whylessness]

youu-u-uu u-u-u

[Klopjob]

 

So far I've tried it on a bunch of things: Adam A7's, Sennheiser HD650's

 

 

Beautiful taste

[Hugh Mughnus]

 

pretty good fake huh?

 

watmmer made that

 

did you see him or her push the render button?

 

 

Nope but what does that matter?

 

Did you see richard push the render button on any of his releases? How do you know they weren't all made by his cat who is a visitor from another galaxy that only landed here to make us listen to strange sounds?

[Hugh Mughnus]

isn't minipops the one was mastered by someone else? the same person who did the recent warp compilation with the autechre track that people were saying much the same about?

 

interesting if minipops was originally destined for that compilation and then he decided to go all out and just release an album

that would be interesting. where did you read that?