Cliping Converters

[AnalogBob]

There's alot of talk about clipping converters...here's a question I have. Would there be any difference between cliping DA/AD.. in the Daw if I have the software push up past 0 would the Da converter not represent that in the same way as if I were to push an AD converter past 0? So in effect if I pushed my daw and recorded a shreded DA would that not have the same effect as shredding AD?

[SnakeCained]

What rescue techniques can you use on clipped recordings?

I have a guy bringing me some stuff that was recorded into PT way too hot and it need cleaning.

Thanks.

[AnalogBob]

No rescue techniques, and I wouldn't advocate trying it.

There is just alot of talk in this months thread about mastering engineers going hard into nice AD i.e Lavry or Prism or the "Sterling" thing. And I wondered if you could get the same effect by pushing the DA into the red and recording that instead meanwhile lowering the volume on your input...(provided your DA was a good one.) In theory it would seem to make sense that whatever you could represent with AD you could represent with DA and vice versa..maybee there is somthing unique in the design of certain converters that allow then to be put past 0 and it still sounds decent but if you had the like DA of that converter I don't see why you couldn't do the same thing from the other end and achieve a simliar result.

[pingu]

Originally Posted by AnalogBob

No rescue techniques, and I wouldn't advocate trying it.

There is just alot of talk in this months thread about mastering engineers going hard into nice AD i.e Lavry or Prism or the "Sterling" thing. And I wondered if you could get the same effect by pushing the DA into the red and recording that instead meanwhile lowering the volume on your input...(provided your DA was a good one.) In theory it would seem to make sense that whatever you could represent with AD you could represent with DA and vice versa..maybee there is somthing unique in the design of certain converters that allow then to be put past 0 and it still sounds decent but if you had the like DA of that converter I don't see why you couldn't do the same thing from the other end and achieve a simliar result.

Good point.

[grandmasters]

Usually Mastering Engineers use a DA to feed an analog EQ and Compressor. I doubt you'd want to send a clipped signal into that.

[brendondp]

Originally Posted by grandmasters

Usually Mastering Engineers use a DA to feed an analog EQ and Compressor. I doubt you'd want to send a clipped signal into that.

Hey Noah,

Great to have you here.

Just wanna say "thanks" for all the Broken Social Scene work you've done. It's appreciated.

Cheers,

bdp

[robot gigante]

Wouldn't your software 'clip' before your D/A does?

I don't think that I've tried this intentionally, though I have got quite a few mixes to master where they hit the master bus in Pro Tools waay too hard- sounds godawful!

It certainly didn't sound like, say, clipping the inputs of the Hedd a bit (not that I'm a huge fan of clipping btw).

[Kiwiburger]

When the pro's talk about clipping their converters, you need to look at the converters they are talking about. Lavry Blue AD has both analog and digital limiting built in, which most cheaper converters don't have. They also have a special clipping feature which adds 6dB boost digitally - but with an algorithmn that compresses the top top 6dB to emulate tape.

So from what I can see, they've discovered digital tape emulation ...

I like Voxengo Analog Flux Tape Bus - it uses saturation and convolution of real tape samples.

If you use 24bits fixed, you can experiment with actual clipping by exporting a mix with overs, and when you re-import it, it's clipped. Doesn't work if you use 32bit floating. Of course, what you are probably hearing through your 24bit D/A will be clipping like crazy, but that's a different thing. Very different. Because you can 'unclip' a 32bit float file that appears to be clipping your D/A, simply by applying a gain reduction. But fixed bits makes it permanent.

I like Voxengo Elephant2 as a classy limiter, but it can also do clipping, which is handy because you can use lower thresholds than 0dB.

There is a place for skilfull use of actual clipping.

[neilwight]

[QUOTE=Kiwiburger]

So from what I can see, they've discovered digital tape emulation ...
QUOTE]

except that no one i know uses this analog saturation thing on their lavrys. response isnt predictable enough and it isnt that transparent though that is the point of this function as i read it.

just hitting it hard with a signal forcing it to clip sounds much better and it can take alot before it breaks up.
try this with lesser converters, apogee incl and you just get cracking audio almost immediately

given the hugely critical ear that most MEs turn to whatever they are doing and the thought that goes into each decision, to just have turned on sat mode on the AD, and thought we'd discovered the holy grail.......c'mon


ref clipping DAs from analog bob,
the one issue with your theory is that you will need to capture this clipped DA signal back into the DAW via an AD or you will not see any gains in level being physically passed on to the material. as such you arent leaving an AD out the equation anyways. theres also issues surrounding getting a hot signal. when clipping the AD you have the huge amounts of excellent gain and headroom on the analog outboard to drive the signal. this isnt available with your clipping DA theory so you would see very little practical gain from this stage when compared to doing it the other way.

with good converters the software would clip before the DA, though this is often exploited to get some extra gain (only around a db or so) and not use a limiter. limiting can often soften the sound too much by dulling attacks etc and while clipping DAW adds distortion, so does limiting anyways so atleast impact is kept to a minimum.

with poorer DAs these will also crack at anything slightly above 0dbfs, making it very hard to judge what is happening from where and being able to make sound judgements on what you are doing.

while it would obviously be better if none of this were requested and audio went a few DBs quieter, gaining tremendously in other ways, it doesnt change the fact that this isnt what most labels/clients want.

beyond a point, anything to gain volume is a trade off with quality and all an ME can do is attempt to control and limit impact.

[dave-G]

I just want to stop and point out my awe at the fact that this forum's populace is usually in lockstep condemnation of the clipping distortion and other artifacts from the much lamented quest for loudness in CD releases these days ... Do a search for the phrase "loudness wars" and see how many me-toos and how many nose-pinched claims of hearing clipping and distortion you'll find.. And yet, this impromptu forum pops up, and guys like Chris Athens show up, representing a wealth of knowledge, stories and experiences, and the denizens are lining up to discuss the secrets of how to do clipping, to get things "loud".

To this I'd say there is no free lunch ... I think Chris has tried to make that point as well.. And the guys that are doing a better job of loudness-izing records aren't getting free lunch either, but are just experienced enough to make good decisions on what account they can put that particular lunch's 'expenses' on ... (so to speak)

Actually, I'm not trying to aim this at any poster in particular ... I initiallly intended to just answer the post ... and so: I only add that those of you who might be developing the idea that AD or digital gainstage clipping is better than say, L2 limiting.. well, you really ought to try comparing 2dB of clipping with 2dB of limiting and see what you think as you get accustomed to the tradeoffs.

-dave

[TanTan]

Originally Posted by dave-G

Actually, I'm not trying to aim this at any poster in particular ... I initiallly intended to just answer the post ... and so: I only add that those of you who might be developing the idea that AD or digital gainstage clipping is better than say, L2 limiting.. well, you really ought to try comparing 2dB of clipping with 2dB of limiting and see what you think as you get accustomed to the tradeoffs.

-dave

L2 and A\d cliping are two totaly different sounding limiters , i don't think you can compare them

[neilwight]

absolutely somewhat to both the last two posts...
driving an AD like a lavry or prism works like a hard limiter with little impact on attack and slowly increasing distortion up to a point (way above lesser ADs).

using L2 (IMHO the best of the limiters) softens the sound by dulling attacks even on fastest setting and adds distortion once it actually starts to limit.

they are IMHO directly comparable in so much as their use is for the same result, increased volume though thats about all they share.

ive found, as have many that driving my lavry works much better than using L2. it still degrades the sound but with a less noticeable impact generally and much more controllability. distortion and preserved attacks generally work better than distortion and dulled attacks. this has been discussed many times on brads forum. there was a blind test done way before chris ever mentioned it and clipping won easily.
outwith the mastering world though i have noticed that this question seems to crop up almost daily now across forums.

all of this is trying to juggle the quest for loudness with an attempt to retain some quality, theres always a trade off. i for one have never change dmy position on that, ive said it a million times before and i'll say it again, you dont get gains without losing somewhere else in quests for level.
it shouldnt be the case, that labels and bands are sarcrifing quality to push things louder and louder, flooding the marketplace with often harsh products but the clients demand it and we are a service industry.

every ME i talk with knows the trade offs, what they can get from their gear etc. we have to deal with it every day. perhaps the downside of a more general mastering forum is that there are going to be people taking this info and applying it outwith this realm or without really considering it.

just to follow on from what dave g was saying:
if i was using apogee or the likes, it would need to be L2 all the way. they couldnt take clipping of the AD without imparting terrible cracking and it would be a no brainer which to use.
if i didnt have good outboard there would be no good gain and headroom to get the signals hot, again something like L2 would most likely sound better.
as such for the vast majority, clipping your AD will neither be a viable or best quality option.

the advocation of keeping away from clipping in general is a very good one. if we didnt then we'd be faced with audio that was cracking and falling apart at the seams much more than it is at the moment.
surely at the mixing a recording stages the aim should still be to keep it as good as it can be?
IMHO its not contradictory to say, i find clipping my lavry AD works better for volume in most cases than limiters to get the volume clients request but i hate the sound of most releases of recent times with their undynamic, smashed to pieces sound.

i do follow what you are saying though dave ref the loudness debate given the repeated response and interest in this level titbit.
its worth remembering though that the music business is just that, a business. for most its not art in a vacuum so the less cynical could argue that all they are responding to is consumer demand/choice

[dave-G]

Originally Posted by TanTan

L2 and A\d cliping are two totaly different sounding limiters , i don't think you can compare them

Of course you can .. You're right that they're different, but actually, I think those differences must be compared if you're professionally in a position of answering client requests for louder masters.

Originally Posted by neilwight

this has been discussed many times on brads forum. there was a blind test done way before chris ever mentioned it and clipping won easily..

If you're talking about the series of L2-TC-Clipping comparisons from last year (or so), I listened to all of those too, and did not come to the conclusion that the clipping won easily on every example.

But I think we agree for the most part on the other principles, and I absolutely think there's a place for all of it. I'm just commenting on the moderately scary feeling of enthusiasm I see in many of the non-ME's posts around here for the blanket-idea of clipping for loudness, as the other side of the coin is usually so derided around here.

Caveat limitor.

-dave

[neilwight]

dave, i also share your scary feeling.
ive seen this so many places already its unreal. like ripples in a pond. i for one can see it becoming the next multiband.
everyone using MBCs because thats what all MEs use (though someone forgot to tell us)

maybe we should all start denying it wholesale. oops, the cat ran out long ago

[TanTan]

Originally Posted by dave-G

Of course you can .. You're right that they're different, but actually, I think those differences must be compared if you're professionally in a position of answering client requests for louder masters.

-dave

sure , that's why i often use them both . i'm clipping a DCS 904 and adding 1-3 DB of GR from L2 \ L3 , as for clients wanting everything so damn loud my job is making it loud for them , i'm not asking them if they want it in a Waves or a DCS way , when i go to the garage to fix my car i want it fixed and i couldn't care less whether the mechanic is using a hammer a screwdriver or a vibrator as long as it works , and it's not up to me to decide , clipping the a\d does not work for every mix , and i find the L2\L3 usable for 99% of my mastering jobs when used gently, but when it works the DCS clipping sounds much louder and changing the mix balance and stereo image less then an L3 , and for me they are two totally different types of limiting characters

[Kiwiburger]

I've been convinced of the benefits of clipping/saturation vs limiting for some time.

An interesting way to look at is this: when you accelate a speaker cone, it has inertia. Just because you top the highest peaks, doesn't mean it will necessarily stop right there - it will tend to keep going and effectively restore the missing peak. Since the rest of the waveform is unharmed, the speaker can catch up with the waveform and carry on normally.

Limiting usually brings the waveform down for a little longer than necessary, effectively punching holes into your sound that follows the peak.

Somebody should analyse this analog convert clipping function and recreate it digitally.

[AnalogBob]

[QUOTE=neilwight]

Originally Posted by Kiwiburger

So from what I can see, they've discovered digital tape emulation ...
QUOTE]

ref clipping DAs from analog bob,
the one issue with your theory is that you will need to capture this clipped DA signal back into the DAW via an AD or you will not see any gains in level being physically passed on to the material. as such you arent leaving an AD out the equation anyways. theres also issues surrounding getting a hot signal. when clipping the AD you have the huge amounts of excellent gain and headroom on the analog outboard to drive the signal. this isnt available with your clipping DA theory so you would see very little practical gain from this stage when compared to doing it the other way.

with good converters the software would clip before the DA, though this is often exploited to get some extra gain (only around a db or so) and not use a limiter. limiting can often soften the sound too much by dulling attacks etc and while clipping DAW adds distortion, so does limiting anyways so atleast impact is kept to a minimum.

with poorer DAs these will also crack at anything slightly above 0dbfs, making it very hard to judge what is happening from where and being able to make sound judgements on what you are doing.

while it would obviously be better if none of this were requested and audio went a few DBs quieter, gaining tremendously in other ways, it doesnt change the fact that this isnt what most labels/clients want.

beyond a point, anything to gain volume is a trade off with quality and all an ME can do is attempt to control and limit impact.

Thanks for the explanations all, makes total sense to me, my questions and theories have been well covered.. I can sleep better