Is it O.K. to Clip instead of Limit?

[MessianicDreams]

The lossy compression argument is different and not analogous because there is a definitive and demonstrable advantage in the savings of download times and storage space, and most people (including myself, frankly) are more than willing to pay the cost in relatively marginal fidelity loss for the gains in speed and capacity.

But is that still the case with the internet speed that most users can connect at now? www.speedtest.net has conducted nearly half a billion speed tests and it shows that the average European internaut is connected at over 4500kbs and the average North American internaut clocks in at just over 5000kbs. Now obviously I doubt these figures are a true reflection of the real average speeds, but but even if they're 50% off those are still fast connection speeds.

I for one will always look for the highest quality version available.

 

[SouthSIDE Glen]

Listen to any commercials lately? Same psychological principles apply ...

Have YOU listened to commercials lately? The larger the advertiser, the more likely the commercial is going to conform more closely to motion picture dynamics standards than to the standards of the old-fashoined asshole in the plaid jacket trying to sell you his lot full of used cars rescued from hurricane Katrina. Maybe it's different on some analog cable systems whose audio- and video-feed quality undergo all the quality control of a bathtub meth lab, but on digital broadcast TV the days of the over-modulated TV commercial have been pretty much relegated to 5am on Sunday Morning.

But even if that were not the case, the same principles simply don't apply to music as to commercials. When it comes to listening to a CD, the idea that bad publicity is better than no publicity at all doesn't work. The commercials that wrongly still apply the mo' louda', mo' betta' principle are no different than the music in one way, though; they both wind up the way they are because of a misinformed client, not because of a dyed-in-the-wool advertising or media engineering principle.

First, nobody cares what the music sounds like in that Joe Car Dealer commercial. As long as you know that they are located between Grand and York Road on Route 83, or that if you buy your beater by Sunday, your sex life will actually improve, the rest does not matter.

But when actually actually *listening to music*, yeah the quality of the music matters, and it's the quality of the music that will sell it to the next guy. The more they want to listen to it, the more likely they are to buy it. And RMS does NOTHING to make one want to listen to it more.

And CD RMS levels are irrelevant to making things more audible or noticeable to the public eye, because whoever is listeining to the CD they already own is going to play it at the same SPL regardless of the intrinsic RMS kevek encoded in the bits. If he want's to play it back at 100dBSPL, he'll do so by adjusting the playback volume on the dynamic CD to 2:00 and onthe pancaked CD to 10:00. Either way the guy next door or in the next car is going to get the same thing. Oh, and when was the last time that the guy next door or in the next car heard someone playing their music loud and said, "Hmm, that sounds good, I gotta get that." Maybe on bizzarro world.

But is that still the case with the internet speed that most users can connect at now?

Perhaps not from that perspective, but from the perspective of Internet loads, server loads and storage capacity loads, switching from 128k MP3 to 44.1k WAV is a 10x increase in needed capacity.

(Ironically, it took me an hour to get this post out because the HR server was overloaded and too busy to let me in ))

Then there's the mePod community, who love the idea that their little toy will hold 2000 MP3s (or whatever). Switch to WAVs and all of a sudden they run out of room after 200 songs. Even though they probably don't even intimately know 300 songs, anything less than a 2k song capacity just isn't k3wl enough.

In a few years, none of that will matter, because capacities across the board will increase, and new lossless methods that take less space than pure bitmaps will become standard probably.

G.

 

[masteringhouse]

Have YOU listened to commercials lately? The larger the advertiser, the more likely the commercial is going to conform more closely to motion picture dynamics standards than to the standards of the old-fashoined asshole in the plaid jacket trying to sell you his lot full of used cars rescued from hurricane Katrina.

www.loudcommercials.com

But even if that were not the case, the same principles simply don't apply to music as to commercials.

I don't disagree, but it's not always perceived that way by artists and others making the final decision regarding levels.

But when actually actually *listening to music*, yeah the quality of the music matters, and it's the quality of the music that will sell it to the next guy.

The song sells the music more than the quality of the audio, but I like to think the quality helps to sell it too (I wouldn't have this gig if I felt differently). Unfortunately some equate louder with better quality.

Oh, and when was the last time that the guy next door or in the next car heard someone playing their music loud and said, "Hmm, that sounds good, I gotta get that."

Probably more often than the guy who plays his music softly and can't be heard by the guy in the next car.

In a few years, none of that will matter, because capacities across the board will increase, and new lossless methods that take less space than pure bitmaps will become standard probably.

I hope that you're right, but if evolution is any indication price and convenience will prevail.

But I REALLY hope that you're right.

 

[SouthSIDE Glen]

I don't disagree, but it's not always perceived that way by artists and others making the final decision regarding levels.

Which is precisely why this debate rages on. The only reason it's old and tired and not over with yet is because the decision makers remain mostly uninformed or misinformed.

Unfortunately some equate louder with better quality.

Have you tried pumping the exact same master through two different busses with each bus set to different playback volumes, and A/Bing them as if they were different masters? Then when they pick the louder one, let them know that they were the exact same mix, that there was absolutely no difference between the two? And then asking that they remember that false volume bias in mind when they listen to the real A/B?

Now, I know that if The Q walked into my mastering sweet and said he wanted it smashed, that I wouldn't pull such a stunt. He knows what he wants. But when Joe Home Recorder comes to you and obviously doesn't know any better, don't you want to give him at least a little of the benefit of your expertise?

Probably more often than the guy who plays his music softly and can't be heard by the guy in the next car.

Ah, but there's the rub. To re-iterate, the guy playing it is going to be playing it at pretty much the same average SPL whether it's a pancake or not. Pushing the mix too hard does nothing to increase "the public eye" as long as the end user has control over the playback volume.

I hope that you're right, but if evolution is any indication price and convenience will prevail.

But I REALLY hope that you're right.

Again, where is the price or convenience advantage to high RMS? It does not sell more records, it does not sound better to most people on the other side of the speaker wires (only to the artists, who, let's face it, aren't going to pay a dime for their own CDs ), it's not cheaper or more economical to pancake the mix, and it doesn't, like the old days of AM radio, do anything to increase the number of ears it reaches.

I've said it before, I believe I'm starting to see a bit of a tide turn in many sectors of the pro music industry. It's a slow turning, but I think the fashion is starting to turn; some of the newer generation of top shelf producers are starting to discover the wonders of dynamics as a way of escaping the same ol' same ol' that they've been doing for the past 20 years. That's the positive side of me...

The cynical side of me looks at how we have totally blown the opportunity that digital technology had originally promised us by doing exactly the opposite of what advantages it did and does afford us, and wonders if the Next Big Thing meant to increase ability will simply result in a new buzz phrase that people will want to emulate for a mythical reason: "Crest Factor Free". Here's the HR forum 20 years from now: "Hey guys, I'm new to this stuff; how can I get my quantum optical holography recording to be crest factor free like the pros do? Man that latest DC voltage that the band 'Yet Another Band That Sucks' put out sounds just so AWESOME."

On a bad day I wonder, as a mixing engineer instead of a mastering engineer, what more of the sonic canvas the idiots want to take away from me. Already we have the advocates of LCR mixing wanting to take away 99.16% of the pan space, and the RMS Loudness war contingent wanting to take away about 93% of the dynamic range space.

Let's just cut to the quick; let's just go back to mono...or better yet, how about no speakers at all? Certainly no phase problems then . Or how about we just eliminate the frequencies between 500Hz and 5kHz altogether. Who needs, em? Those same clients that think louder sounds better also think that death scoop EQ sounds awesome. So let's just cut to the chase and preset our EQs to notch out the entire midrange for them.

Hmmm, you know, if we take the black keys off of the piano, it'll be a lot cheaper and easier to play....



G.

 

[masteringhouse]

Have you tried pumping the exact same master through two different busses with each bus set to different playback volumes, and A/Bing them as if they were different masters? Then when they pick the louder one, let them know that they were the exact same mix, that there was absolutely no difference between the two? And then asking that they remember that false volume bias in mind when they listen to the real A/B?

The issue isn't when they are listening in the studio, or when I explain the benefits in dynamics. The issue is when they put their CD in the car. People tend to take things at face value and go with their first reactions.

Anyway, we can debate the loudness wars over and over like your mother telling you to stand up straight and eat your peas. Sorry mom I can't hear you, my Ipod is firmly implanted in my ears listening to squashed mp3s ...

 

[SouthSIDE Glen]

The issue isn't when they are listening in the studio, or when I explain the benefits in dynamics. The issue is when they put their CD in the car. People tend to take things at face value and go with their first reactions.

I just don't get why your expertise became irrelevant just because Joe Garage Band can now afford an mBox2, especially considering it means the loss of your job before you know it if you stop defending it.

You're getting their business now, but as soon as they realize that their stuff sounds like crap no matter who masters it because crap happens to be what they *think* they like, you're ancient history my friend.

And so am I. If someone is going to make nothing but pancakes out of my mixes, I might as well just stop mixing, because I'm doing nothing but spinning my wheels.

Anyway, we can debate the loudness wars over and over like your mother telling you to stand up straight and eat your peas. Sorry mom I can't hear you, my Ipod is firmly implanted in my ears listening to squashed mp3s ...

And it's a pretty piss poor mom who stops telling them those things just because that's their reply.

Tom, you know I like and respect you, but we gotta agree to strongly disagree on this one. Just going with the flow on this one is not an option IMHO.

G.

 

[masteringhouse]

Glen you know I like and respect you as well, I think that one thing you may be failing to see is that both mastering and mixing are services not a dictatorship. It makes no difference what I prefer or educate people on. The client is the boss and pays me to take care of the technical part of what he wants to accomplish. I can preach about there being too much bass, too much compression, etc. but ultimately I follow his or her request. It's their album, not mine.

You can tell the kids not to put their fingers on the stove cause you'll get burned, eat your peas, etc. but ultimately it's their decision to do so or not. That's how they really learn. Taking away the stove, making rules, ranting or preaching until you're blue in the face, isn't going to solve the real problem. If it were the drug problems in this country would have been solved years ago.

Preaching about the loudness war reminds me of celebrities that stood behind the AIDs cause during the 80s and 90s. Sure it's a good cause, but how many people died of cancer over those years compared to AIDs? Maybe some money and celebrity could help there too? I think that many were behind it more for the "cool" factor or because it was the "thing to do" at the time more so than helping the world rid itself of disease in general.

Anyway this seems like another thread on loudness wars that as I mentioned are really old and tired. I thought that we might expand on this and talk about larger issues, but I'm going to bow out on this one.

 

[SouthSIDE Glen]

Glen you know I like and respect you as well, I think that one thing you may be failing to see is that both mastering and mixing are services not a dictatorship.

Nobody's calling it a dictatorship, it's a profession.

The difference that has crept in over the past ten years is that it used to be that we dealt with other professionals. Musicians that were usually worth recording, producers who, whether we always got along with them or not, at least usually had a game plan, and as often as not actually did know what they were doing quite well, and so forth. We may not have always agreed with them, but they usually had their reasons for asking of us what they did. And if they didn't, they at least knew who was wearing what hat and who what good at what. The artists knew what they didn't know, and if they knew that they just weren't qualified to produce, they either hired a manager or a producer to do it for them, or they at least asked our advice and understood it to be the advice of a professional.

Nowadays, not only are over half of the recording "musicians" people who couldn't hit a snare or play a Cm chord to save their lives, but they have no idea that they are that bad. They just want to play their power chords wrongly and get that stuff on disc to please their self-entitlement tendencies. Unfortunately these people also believe that they are as good at engineering and producing as they are at playing music. Unfortunately, because of their tin ears, they are absolutely right, just not in the way they think they're right.

The fact is the customer is NOT always right, and to just do what they ask no matter how silly it is without discussing it with them first is to overcharge them for our services. We need to let them know that if they just want someone who won't actually give them the benefit of their qualifications, then they shouldn't pay for it. Why should someone go to a quality ME, or mixing engineer or tracking engineer, FTM (I'm not singling you MEs out), if all they want when it's done is a stinking pile of shit? They can do that themselves.

You wanted to expand this conversation into how to improve the quality of music itself, and not just the sonic quality. Don't you see that the two are just two sides of the same coin? The only answer to either question is to improve the qualifications of the people recording it.

And since the Big Boys ain't gonna do it, and since the clients aren't going to do it on their own, all that's left is us in the middle. If we just shut up and shake the tree for Boss Paul, it ain't gonna happen.

And, based upon my experience on this board the past few years, it is not a futile struggle. The messages I have gotten, both public and private, or have read on other threads, from people who DID wind up understanding and seeing the light when it came to these things, and all because people like you and I and many others here have taken the time to explain things, have been many. Sure, you're right, most of them had to try momentarily touching that stove themselves just to test what we were preaching, but it worked. Had they touched that stove without our explanations beforehand, they would almost certainly touch it with full mitts, and done so a second or third time when it was a lot hotter, because they would have no idea why they got burned in the first place.

All it takes is one. All it takes is one artist/producer/engineering team to make it big with an album of full dynamics and suddenly all the rest of the lemmings will be copying them, and dynamics will be back in vogue again. If that one happens to be starbuck23 or ikon or Rokket or any of the many others who have heard the word, then this "old and tired" debate will have been worth every minute of it.

G.

 

[Daisy]

Are you guys still talking about this?


Christ. Give it a rest already.

 

[SonicAlbert]

Just to chime in with a personal experience:

I had an album that received quite a bit of airplay on independent stations. In at least one case the station would also stream the broadcast on the internet, and allow for it to be played back later as well.

My album was *not* mastered in the sense we are talking about here in this thread. The mix was really smooth, the dynamics were under control, everything was in it's place except that the levels weren't smashed (and it sounded good).

However, it was immediately obvious that in a playlist surrounded by other big label tunes, my album was noticeably quieter. The effect was one of lack of impact or weakness. The simple fact that it was softer than the tracks around it meant it didn't have the same impact or attention grabbing factor for the listener.

So in my opinion, an album that is intended for broadcast *must* be brought up to a level that is on a par with other tracks that it will be aired next to on a playlist.

The only issue is *how* to do that and have the music still sound great, and with what skill level the job will be done. The real problem is that so many times the leveling job is done poorly and the music suffers for it.

But again, it's not whether or not the leveling should be done, but how well it be done and in harmony with the music.

 

[benage]

It's a bit of a case of your damned if you do, damned if you don't, the only answer I can settle on is one of compromise. You say the answer is to get your levels as hot as the other tracks and still sound great unfortunately if this were possible we'd all do that, If you want it to sound good you have to accept it'll be slightly quieter than the hottest CD's around (which sound like crap) but that's not necassarily a problem. The CD you had played on the radio as you said was not mastered for loudness at all and was noticably quieter but if you push the levels up somewhat hopefully the stations compressors and limiters will do the rest and you can compete, I've even read that CD's that aren't pushed to the max can sound better on radio as the limiters respond better when there's still some dynamics to compress although I'm taking someones word for that. Plus by releasing a 'slightly' more dynamic CD your helping push the industry in the right direction and helping us all.

 

[LogicDeLuxe]

However, it was immediately obvious that in a playlist surrounded by other big label tunes, my album was noticeably quieter.

This sure would be the exception to the rule. Probably, you found one of the few stations which don't overprocess by their own. In addition to this, the DJ sucks, if he isn't able to match volumes on his own.
And still, you actually got the point, that also a soft master will be player nevertheless.

So in my opinion, an album that is intended for broadcast *must* be brought up to a level that is on a par with other tracks that it will be aired next to on a playlist.

Then, if you really must, master each CD for its purpose. That is, master a hot, but not clipping single, which includes the radio mix and one b-side. Then master a dynamic maxi (given your style is appropriate for a maxi), which includes a club mix for the DJ, a b-side and optionally the radio mix, a dub-mix, remixes, a video clip, whatever. And the album, which sure should be dynamic as well. The radio mix should be mono compatible while the other versions can be surround encoded, if it fits the music.

And if you want a hot master for the sake of one radio station, it isn't even necessary to master the regular CD hot because of this: Just burn a version specially mastered for the station and sent it in, a so called promotional CD.

 

[SouthSIDE Glen]

I had an album that received quite a bit of airplay on independent stations.
...
t was immediately obvious that in a playlist surrounded by other big label tunes, my album was noticeably quieter

I'm sorry that happened to you, Al, but I gotta agree with Logic on this one that the problem there was not the level of your mix, but rather a DJ or DJs that weren't doing their job. Think about it for a second further, from the perspective of an impartial third party listener. Which would be their most likely reaction to that: "Man, that song wasn't mastered loud enough" or "Man, the station messed up on that one"? The majority of the listening public is going to blame the radio station or DJ for *not doing what is, after all, their job* of providing the constant listening experience. This is why God invented pre-cueing .

We have a station here in Chicago (you used to hail from here, perhaps you remember WXRT?) that used to have a program called "Local Anesthetic" (I don't listen to XRT anymore, so I don't know if that proggie is still on or not.) Local Anesthetic was hosted by a friend of a friend of mine, and was a weekly program that showcased local indie recordings that were made on everything from Portastudios to pro trackings in Big Boy studios, including every range of mastering from none at all to PWM bricks. In the years that that program was on, the kind of situation you describe almost never reared it's ugly head unless the DJ admittedly glitched at the controls, which was rare, and usually only after he came back into the booth after burning a huge rope. Not only did he actually jockey the faders like a DJ is supposed to do, the program material was usually laid out in such a way whereas one end of the spectrum did not butt up against the other unless there was a commercial or a VO or something in the middle

The only issue is *how* to do that and have the music still sound great

With most forms of music - there are exceptions like forms with the suffix "-metal" - or "-core" in them, which have such high RMSs to begin with that they don't have to go very far - this is not possible unless one has a truly tin ear.

I mean come on. I have a 2000 recording here of Willie Nelson singing a blues ballad originally written in 1925. It is PUSHED. It sounded fine on the radio relative to the rest of the program, but when I went and bought my own copy, man it most definitely sounds PUSHED and PUSHED HARD. How can it not? Asking something like that to be pancaked and still sound great is like aksing Angelina Jolie to undergo Michael Jackson's plastic surgery regimen and still look great. Or, maybe a little more realistically, taking a full-color megapixel digital photograph, converting into a 255-color GIF file and asking it to still look great.

Sure a quality ME with the right gear and right knowledge of how to use it can push something harder without it looking like Michael Jackson's nose, perhaps, but with most music it's still going to sound heavily compromised once it's off the radio in into the listener's paws, no matter who does it.

G.

 

[SubLunar]

Forgive me for being retarded and reviving this debate, but what is the deal with all this clipping/maximized talk?

If you record with the audio peaked as close to the top WITHOUT clipping, isn't that the goal? Theorhetically won't the mixdown come out as loud as it's supposed to without losing signal integrity?

Clipping sounds like shit IMO. Why is there so much interest in this?

When I record anything, I have the peak levels up to like 90%. If it clips, I turn it down or completely redo it. Are people really putting up with clipping in their final product? And isn't that potentially harmful to your equipment?

What's the deal?

Am I missing something? If so, put me in my place and I'll let this discussion cease.

 

[LogicDeLuxe]

Forgive me for being retarded and reviving this debate, but what is the deal with all this clipping/maximized talk?

A good idea would be reading instead of reviving this thread. It pretty much states all the misled thoughts behind it. Also http://turnmeup.org/ and those links there are worth reading for that matter.

If you record with the audio peaked as close to the top WITHOUT clipping, isn't that the goal?

Something like that, and it won't be even a problem if there are several dB headroom left. This entire thread is not about recording, but mastering, though.

 

[NYMorningstar]

This entire thread is not about recording, but mastering, though.

That's funny, I thought it was about steroids

 

[SubLunar]

A good idea would be reading instead of reviving this thread. It pretty much states all the misled thoughts behind it. Also http://turnmeup.org/ and those links there are worth reading for that matter.Something like that, and it won't be even a problem if there are several dB headroom left. This entire thread is not about recording, but mastering, though.

Reading? What's that?

Actually I read it all and that's why I was confused. Maybe rephrasing the question would help.

Do professionals in the industry actually allow clipped audio signals in their final product? Why does this concept deserve so much attention? Clipping is avoidable and undesireable is it not?

To me (an admitted noob), the loudness war is silly. The video on the turnmeup site was common knowledge and it amazes me that people are actually ruining their audio simply for loudness sake. I do what the track calls for, whether it's quiet or peaked, but wouldn't dream of allowing it to clip.

Oh and sorry for mentioning "recording". Obviously recording has NOTHING to do with mixing/mastering. Stupid me.

 

[benage]

Hi, hope you don't mind me butting in but I have been reading this thread since I started it and someone earlier did say letting the master output clip is commonly done. The reason people do this is it's an alternative way to get your mix louder on a CD and 'compete' with the big boys. The other main way is to use a limiter which is basically chopping the top of your waveforms in your song (so you can pull up the overall level) but which adds unwanted distortion the more you do it, clipping is a very similar thing i.e chopping off the waveworms and adding distortion so I was initial trying to find out if either type of distotion was preferable and to be honest I never really got a straight answer. I have read elsewhere since limiting distortion is preferble and I have to believe thats true otherwise everyone who's bought a ultramaximizer limiter might aswell have threw there money in the bin.

 

[SouthSIDE Glen]

Reading? What's that?

Don't get me started....

Do professionals in the industry actually allow clipped audio signals in their final product?

Yes they do, and at an increasing quantity.

Why does this concept deserve so much attention? Clipping is avoidable and undesireable is it not?

Clipping, in and of itself, is undesireable if one wants quality and accuracy of reproduction of the source signal.

Clipping does add distortion which some find in a few certain circumstances to be artistically desirable. In these cases the clipping is actually being purposefully being crafted as a special audio effect. Not everybody's cup of tea, admittedly, but no less legitimate than using any other effect for sonicly artistic purposes.

Where the real controversy comes in is when clipping is used not for artistic design, but rather as one of several ways to squeeze out as much dynamic range compression as possible, with the end desire to push RMS levels and makeup gain into the >-12dBFS range - regardless of actual sonic content and regardless of artistic impression - just so that the song will be as loud and crappy sounding as the next guy's.

They (the clients) do this for three different reasons: a) they have zero confidence in the quality of their music to be able to win over the listener without having to worry about a couple of measly dB, b) they have tin ears and actually think the smashed stuff sounds better because they are newbs being tricked by and falling for the loudness bias, c) they are simply lemmings following the crowd.

G.

 

[Freudian Slip]

Reading? What's that?



Do professionals in the industry actually allow clipped audio signals in their final product? Why does this concept deserve so much attention? Clipping is avoidable and undesirable is it not?

Yes. Some systems handle clipping better than others do. You have to look at it in digital terms. It's all a bunch of numbers until it hits the end users DA converters. Samples are just that. Little dots in time along a waveform. The DA converter plays connect the dots based on pre programmed assumptions and some are kinder than others to misguided engineers.

Soooo if a tree falls in the woods............

Just play it safe and insure your recordings will sound good on all decent converters. Even without pushing to clip there are huge sacrifices in sound quality made in the pursuit of "loudness".

F.S.