Plugins Internal Clipping

[luctellier]

I'm using a lot the Waves SSL 4000 for my drums and I always been unsure about one thing, the internal plugin clipping versus the audio channel clipping. Since I never been sure I always lower the gain so the plugin never clip and my audio channel don't clip also. My question is, does it "really" matter if the plugin clip internally but the audio track don't? By audio track I mean my track in Cubase...

[killahertz]

it may have if waves designed the algorithm to operate optimally within a certain gain range.

[taturana]

i always lower the sound also..

doe anybody just let it rip for effect?

[Methlab]

I believe Bob Ohlsson does. He posted about this.

You can overdrive the SSL and URS plugs a little and they will react like their hardware counterparts. Not exactly, but i believe they are modeled to react the same way.

[Jesse Skeens]

Originally Posted by luctellier

I'm using a lot the Waves SSL 4000 for my drums and I always been unsure about one thing, the internal plugin clipping versus the audio channel clipping. Since I never been sure I always lower the gain so the plugin never clip and my audio channel don't clip also. My question is, does it "really" matter if the plugin clip internally but the audio track don't? By audio track I mean my track in Cubase...

Depends if the plug is outputing floating point compared to integer.

As far as the URS ones I serously doubt they model that, especially considering the light cpu load they have. In fact they were 24 integer early on and very easy to clip. Now I belive they raised them to 48bit to solve that.

With a plug outputing float and cubase you wont be clipping anywhere on the channel, just at the mater fader going to your converters.

If you want to test it just try a sine wave, its easy to hear clipping right away.

[luctellier]

Originally Posted by taturana

i always lower the sound also..

doe anybody just let it rip for effect?

Well. that's the main point of my question, with the SSL I noticed that a bit of clipping adds some attack on the kick and snare but hey, is it attack or "nice clipping"?! ...

[Absolute]

You didnt say what DAW you use

32 bit float does not clip even if the light comes on unless someone coded the plugin wrong. (and I wouldnt rule out that some are) Nuendo's plugins dont even have clip lights because they cant clip.

Im sure someone will 43 pens in his pocket is going to debate that..but its gonna be someone without Neundo.

[Empty Planet]

Think your answers might be in this thread, starting around here.

Prepare to be troubled.

The upshot, though, is that one can simply mix so that all channels peak at -6db or less -- until the mix buss -- and you'll avoid the intersample errors.


Cheers.

[Bob Yordan]

Good topic.

I always try to adapt the output level, so that the next plug in the chain
does not introduce unwanted distorition etc.

If a plug does not provide functions for adjusting out level it all becomes a bit more complicated.

I think I've read that audio channel can handle clipping better than plugs can internally. But I mostly try
to solve it with plugs instead.

[luctellier]

Originally Posted by Empty Planet

Think your answers might be in this thread, starting around here.

Prepare to be troubled.

The upshot, though, is that one can simply mix so that all channels peak at -6db or less -- until the mix buss -- and you'll avoid the intersample errors.


Cheers.

Thank you so much. I bought the Bob Katz's book "Mastering Audio" and already started to read this 300 pages book. So interesting! I guess they will always be the "hot level" and "low level" guys... I prefer to keep my signal as clean as possible.

[defjamm]

Originally Posted by Methlab

I believe Bob Ohlsson does. He posted about this.

You can overdrive the SSL and URS plugs a little and they will react like their hardware counterparts. Not exactly, but i believe they are modeled to react the same way.

this is not true for the urs eqs on native daws...if you clip them you clip your audio channel/daw, there is no modelling.

i think it was you in another thread some months ago who claimed that and there was a method posted how you could test it yourself and it's not true. i tried it myself.

i don't know about the waves stuff.

http://gearslutz.com/board/showthrea...highlight=neve

here is the method described.

[Methlab]

Well try the test on the SSL. If you clip it, it does not sound like shit, and the URS don't either really. I havent messed with the SSL channel in a month or so. I am still using the URS EQs.

The thing I don't get, is why model an API EQ and not factor in the clipping sound? Thats half the reason people like them in the first place.

[luctellier]

Until someone prove to me that internal clipping affect the sound in a good way, I will continue to mix and keep levels so that it doesn't clip anywhere!

[Methlab]

Its really up to you and how you hear things. I have overdriven plugs to make them sound a certain way, but I usually keep them out of the red and it sounds fine.

[luctellier]

Yeah I know.. sometimes it's funny 'cause it's all about SOUND. So many people talks about VU, Led indicator, Frequency Spectrum etc.. as long as it sounds good, where's the problem? The user at home that will listen to the song don't even care about what brand of EQ you used or what compressors settings were used, if it sounds good, it sounds good, that's it!

[Methlab]

yeah..for me, I bought the SSL Waves plugs and I had already bought the URS. Well, I thought for sure Id be using the Waves all the time, and I really just use the Bus compressor, and not even on the 2bus! I love what the URS plugs do more, even though popular opinion on here leans toward the Waves.

Personal opinion and experience.

Try driving those plugs and see if you like the sound. Only way to go.

[Mr B.]

It s all about intersample peaks!

[Jesse Skeens]

Originally Posted by Methlab

Well try the test on the SSL. If you clip it, it does not sound like shit, and the URS don't either really. I havent messed with the SSL channel in a month or so. I am still using the URS EQs.

The thing I don't get, is why model an API EQ and not factor in the clipping sound? Thats half the reason people like them in the first place.

Ya you'd wonder why but that's just the way it is with most plugs. Even the UAD stuff doesn't model clipping. They do their models based on a optimum level within the hardware unit.

Distortion done right would just take too much dsp it seems.

[Jesse Skeens]

Originally Posted by Methlab

Its really up to you and how you hear things. I have overdriven plugs to make them sound a certain way, but I usually keep them out of the red and it sounds fine.

Clipping can sound good depending on the source but pretty much all these plugs are clipping the same, flat tops, nothing special.

[heathen]

I think these emulations do try to emulate saturation when pushed hard, so for impact on drums or the like a touch of yellow and red can't be all bad (waves SSL), this is my conclusion though I always leave plenty of HR for gentle clean sounds.

[Methlab]

I have a very experienced engineer (nonslut) buddy who told me the same thing. He drives his plugs and got me to try it. It was cool on this one singer I was mixing, but not something I usually do.

If Bob O sees this, I am sure he ill chime in.

I don't mean to be OT, but do any of you actually get resluts driving the URS plugs?

I get a REALLY nasty nocking noise from drums if I push them in the red @ ALL.
(So I don't.)tutt

[max cooper]

Originally Posted by no ssl yet

I don't mean to be OT, but do any of you actually get resluts driving the URS plugs?

I get a REALLY nasty nocking noise from drums if I push them in the red @ ALL.
(So I don't.)tutt

Yep.

[Bob Olhsson]

What you don't want to do is digitally clip the inputs. It's the output of the waves ssl that emulates analog clipping.

I use the e channel a in non-analog mode unless something is too zingy to begin with.

[Tubthumper]

Originally Posted by max cooper

Yep.

Hi Max. Can't say I've attempted this with my URS plugs. Any hints for us URS plugins overdriving virgins???

[heathen]

Originally Posted by Bob Olhsson

What you don't want to do is digitally clip the inputs. It's the output of the waves ssl that emulates analog clipping.

I use the e channel a in non-analog mode unless something is too zingy to begin with.

Exactly my point, though I usually don't like to see red on the waves ssl channel outs but it can add a little extra bite. A mastering project I did the other day for someone was exactly like this, they mixed it in ptools with the waves ssl, I got them to drop all the levels back 3 db and the mixes no longer worked as they did, was very edgy stuff and it needed it, first time I've ever actually advised someone they should leave the levels like that in a digital system, I hate contradicting myself but I had too in this case, though I made him aware not to do this on all material or with any old plugin, just stuff that needs bite with the ssl bundle.

[luctellier]

Originally Posted by Bob Olhsson

What you don't want to do is digitally clip the inputs. It's the output of the waves ssl that emulates analog clipping.

I use the e channel a in non-analog mode unless something is too zingy to begin with.

THANKS, that's exactly what I wanted to know.

[Methlab]

Yeah this is where defjamm and I disagreed before. I was saying I overdrove my URS plugs and I liked the sound for a certain application. He said he did a test and they nulled. Max says he does it too. Bob overdrives his SSL's with good result.

It's a mystery I guess. Go by what sounds good. Don't clip your channels though!

[defjamm]

i provided a link where the method is described, there is no mystery. i'm talkin about the native versions of the urs eqs in a native daw (i tried it with the neve one, living sounds too when i remember it right)...there is no modelling involved.

if you overdrive them you overdrive your daw/sequencer/master bus and it's the same for most plug-ins.

like i said, i didn't try it with the waves ssl stuff. and i really don't care if it's a grammy-engineer who's doing it with the urs plugs...if it's native he's simply overdriving the master bus of his sequencer (be it sonar or cubase or samplitude or adobe audition or what ever).

and there is nothing wrong with doing that, it could work for some signals because it's some sort of distortion you'll get if you overdrive your native 32bit floating point daw.

@no ssl yet and max cooper: are you on tdm?

[defjamm]

Originally Posted by Methlab

Well try the test on the SSL. If you clip it, it does not sound like shit, and the URS don't either really. I havent messed with the SSL channel in a month or so. I am still using the URS EQs.

The thing I don't get, is why model an API EQ and not factor in the clipping sound? Thats half the reason people like them in the first place.

i'm no developer but i know a little bit. the modeling of urs is very basic, they use basic algos. you can duplicate most of their eqs if not all (didn't try all) with the old waves q10. you'll get the same curves and the same phase response.

the reason why people/engineers like them is because it's easy to get something usable fast because you don't have to many options to ruin your sound. but the modeling of the original units is not very advanced. and it seems that faithful distortion modeling is still something plug-in developers struggle with.