About clipping

First off, I'm an analog dude, I'm 63 years old and grew up with vinyl, even 78 rpm bakelite records when I was very young.

And tape, my dad owned a Philips portable tapedeck he'd bought in 1958.

With this tapedeck and a keramic mic I made my first recordings when I was 14-15 years old.

Anyway, I found it all out the hard way with stumble and fall and the last five or six years I've learned tons about recording from the forums, from guys like George Massenburg, Bob Ohlson, Fletcher, the list is very long and of course Bruce Swedien.

I know you can hit tape very hard and sometimes I do this on purpose in order to get that special tape compression and that edge of tape saturation.

I also know from experience that some analog consoles sound better when you push them near its limits.

I only use digital for mastering and mixing to and still a 1/4" tape sounds better in my experience than even 24/96, but that's another discussion.

In the beginning of my digital era I've had some problems with CD factories because there were some minor clips on my DAT masters and I've always thought that digital clipping is not done.

I have two sons who are professional computer dudes, ICT guys that is and I've had some discussion with them about the subject.

When a digital recording device clips severely it's no longer capable to translate the wave to sound and as a result you'll hear a crack or something like that. This is what I've been told and this is what I hear when there's a digital clip, a nasty sound.

This afternoon I was in the studio and I've played the MP3 of 'Earthsong' once more. And I was thinking about the hassle that happened when John Sayers posted the waves of that song.

First off, it sounds wonderful, but I was thinking, how did it sound before it got pushed, 'cause it reads really loud on the VU meters compared to even CD's from after 2000 and it clips very little, but it does. Maybe there's some headroom in Adobe Audition because I don't hear much from it.

I've put a brick wall limiter over it, just limiting at -02, no compression and then there's no red light anymore, but I don't hear any difference.

So I think that the original mix of it, before it got compressed must have had a sound from heaven.

My point is: Bruce says he put the clips there for an artistic reason and Craig says that he uses clipping on purpose as well. But I've always been told that digital clipping is absolutely not done.

What are your thoughts about that?

I am not sure, but i think they were not talking about digital clipping. The guy who posted the waveform was talking about "over compression" that clips the waveform. But you are right i was wondering about the same thing, because ive heard digital clipping and it hasnt never been pleasant.

Yeah... I couldn't help but feel a lot of different issues were getting conflated there, saturation/clipping, over-compression/competitive loudness, and digital clipping...

I try to listen with my ears -- but I've seen a lot of great sounding tracks that show fairly extreme effects of tape or compression saturation -- to the extent where the natural ballistics of the wave envelope are CLEARLY quite affected, clipping, if you will.

But that's a LONG way from the pro forma squashing we see on a lot of pop records today.

Even using my WinAmp player, I could see that the finale of the track in question was mighty saturated -- but it still sounded clean and defined -- and the overall impression of the track was that it had a huge dynamic range. Listening to the intro -- which, with much musical logic, had a much lower average level -- there was no sense that it was sonically compromised by lack of defnition or other ills associated with "too-low" levels.

There's much food for thought in this recording. And comparing it with the released version was also really interesting.

Big thanks to Bruce for sharing. I wish the dialog had gone differently but I learned a lot, and I'm walking away with a lot to think about, as I have from a number of his threads.

When you look at a WAV file and see a horizontal bar where the dynamics used to be, you can't really tell (at that scale) whether it's clipped or compressed or both.

If you zoom in to see an individual cycle (or alternation), and the top is flat, then you know it's clipped.

If it is not flat, it has been compressed but is not clipped.

I'm just trying to get my head around what the benifit to super compression is. Is loudness the only thing it's good for? I'm thinking the industry knows what it's doing. The listening public isn't as keen on whats going on when they hear a song. Especially when 80% or so listen in thier cars, Mp3 players and boom boxes. Heck, I swear my Ford truck radio has a built in compressor! When I crank it i can hear it pumping! Now if I get this right, compression is to prevent clipping. Lots of compression will help a boombox get to louder levels without clipping( distorting). Same would go for an MP3 player or a computer.

So i'm thinking that a huge commercial release has to sound as good as it can, on all listening devices. Being that sound quality on a large number of devices used nowadays is crap anyhow, the benifts must outweigh the problems created. Only people that are into this even come close to knowng what they are hearing.

Now if I get this right, radio and television compress everything to get uniform sound levels. So no matter what the level on the song or artist is, thier systems will adjust to thier set limits. So if you've got the new Red Hot Chilli Peppers with massive comp'ing, thier systems will work less. On the other hand an old Jimi Hendrix recording, thier systems will increase to bring things up to level. I may not be correct about that but after thinking about it for quite awile it makes sense to me.

I guess i'm just offering a different point of view here. In a way you can call it similar to alot of things that progress. I remember in the mid 70's or so, they started coming out with emmision control devices on cars. Well, to a 15 year old gearhead, we ripped that stuff out because it made no sence and hurt performance. After a few years you figure out, they are dropping octane levels in the gas so it's not burning efficient anymore. Ripping the emmission stuff out did nothing for performance.

Sorry for that ramble but my point is that the way people listen to music is much different than it was 10 years ago. The major lables and radio stations are well aware of the trends in mastering. I'd have to have faith that the people that are running the machine know what they are doing. And if we can still have guys like Bruce recording things as incredibly as he's done over the years, i'm sure even the pickiest of audio conesuers will still love what they are hearing. And speaking as a musician, any way my music can sound better to more people is cool with me. If I really want to hear dynamics, I'll pull out some Robert Johnson records!! (just kidding!)I'd honestly like to know more of what the bennifits are vs. the bad side. Rob

I use digital clipping only on very short transients, like under 10 ms (and more like 5 ms). This adds a short burst of harmonics that can do amazing things to something like an analog drum machine kick. But for any sustained passage, digital clipping is NOT advised and doesn't sound remotely like the distortion that you get from hitting analog tape really hard.

It was not my intention to discuss loudness, but digital clipping.

I do know what digital clipping sounds like and looks like and it's bad.

But why do engineers accept clipping when there's such things as brick wall limiters?

seriously? wow...

And how do you make sure the clipping will last just the time you want?

I hear many people are experimenting with the introduction of extra jitter through the use of external clocking (where it's not actually necessary).

You might want to check that out.

I'll say one thing... rules are there for a reason... but it doesn't mean they have to be applied to everything in every situation.

As my old 8th grad English teacher (well, she wasn't old then... she was a maximum babe but I digress) used to say, Most of the fun of breaking rules is knowing you're breaking them. And THAT requires knowing what the rules are in the first place.

At least I THINK that's what she said. Frankly my mind was on other stuff in that class. And those thoughts were DEFINITELY breaking the rules... Yum. It warms my... uh... heart just thinking about it.

So should I stop slammin' the levels on the DAT real hard?

It's trade secret time...shhhh...

I edit the track or sample in a digital audio editor, highlight the first 5-10 ms of the attack, and boost the gain sufficiently to clip.

Just remember, that "punch" switch on Kurzweil keyboards was a direct result of my analysis of what makes a Minimoog have punch I'm way into this kind of stuff.

Because when used the way I use it, everything except for a few transients remains totally unprocessed. A Brickwall limiter has more of a "sound" to my ears.

Okay, here's another secret: With drum samples, a lot of times I draw spikes in the first few milliseconds of the attack. It REALLY makes drums cut through on a track without having to boost the volume.

I never said I was normal, BTW.

Why be normal?

Rick

Thank goodness you aren't "normal"

Craig....

"Normal" is anathema to creativity IMHO.