Live Audio Quality

[Meef Chaloin]

I have to agree with some of the statements in this thread. Its not so much the quality of the audio (although i have noticed some difference in my mixes, i presumed it was my own skills that needed brushing up), but the thing about the master overloading is certainly a valid point. It DOES overload far too easily. I don't really understand all the explanations about bit floats & all that but there is definitely something unusual about the headroom. In Cubase, Logic and Reason I hardly ever had any clipping on the master unless something was very loud.

It seems to me that this is a very important issue for a 'live' program - if you are doing things on the fly and improvised then you need to know that its not going to clip, more so than in a mixing environment in my opinion.

And in the defence of the original posters, its not exactly easy to find a specfic thread on this forum, the search is crap.

[wildcon]

I really hope the metering etc is sorted out in Live 6.

I have had songs distort completely (DJ songs - so no mastering issues etc)
when the metering has shown no clipping or red lining at all !

Very annoying.

[leisuremuffin]

here's an old thread on this topic:

http://www.ableton.com/forum/viewtopic. ... sc&start=0

and here's one of the messages in it:

This topic seems to be very esoteric.

First of all, if you believe that a compressor with an orange display sounds warmer then one with a grey skin and this helps you doing
good music i will be the last person to tell you that you are wrong.
If it works for you it is right.

But in this thread we are talking about a very brutal mater: DIGITAL AUDIO.

if you take a file which has been recorded with 44.1 kHz 24 bit
into any system on this planet which plays it back sample by sample
you will get the exact same file at the output. This is true for Cubase, Logic.... and for Live since version 2.0 as long as it`s sample rate is set to 44.1 kHz and the file is either unwarped and pitch is set to 0.00 or if it is warped and pitch is set to 0.00 and original tempo at every warpmarker is exactly the same as song tempo. No wapring, no interpolation between samples, just straight sample by sample readout.
I have been checking this many times. If you can show me single reproducable example where this is not the case i am very curious to see it.

Mixing: MIXINIG in a digital system means multiplying a signal with a value. It does not change the sound. ADDING two ore more channels is
a mathematical addition of two or more signals. It does not change the sound. There are some very theoretical differences if you go into the extremes ( i am talking of boosting a signal by 100 dB ) because then
single precission floating point math may be a bit unaccurate, but
Live uses the same math as Cubase Logic... so there will be no difference.

Dithering: dithering is a techique which shifts the spectral energy of quantisation errors once a file is put out into a digital analog converter.
It is never recommended to add dither as long as you stay in the digital domain and are not creating a finished product which will not go thru any mastering process including changing the level even by a very smal amount. Dithering should be applied at the mastering studio as the very last step by someone who knows what to do.

Unless you are doing music with _very_ quited parts you will not hear
the difference anyway. We are talking aobut the range of -80 dB and below....

Effects : of course different plug ins use differrent kinds of algorithms for creating compression, tube saturation, reverb or eq. Feel free to use
what ever VST and Audio Units offer you to find the one which sounds best for you. In Logic Cubase ... and in Live - no difference.

These are the facts. I have a history in professional mastering and vinly cutting and i am willing to discuss this topic in depth if needed. I can also point to literature.

I know how much psychological factors determine our perception
and as i said at the beginnig, i am perfectly fine with what ever one
believes as long as it leads to good results, but i am getting upset by this
discussion about "sound quality of an audio engine". This is DIGITAL, not
analog. My old ENSONIQ ASR-10 sounds great beacause the digital to ANALOG converters do funny things. If i am making a typo here it will be _my_ fault and it will be at the forum because i wrote it and not because my computer has a slightly different way of thinking then yours.

Robert

btw, you can use pretty much any search engine you want to search the forums. Just make the search engine look only within this domain.





.lm.

[knotkranky]

Yes its a Live thing and the pilot. Time stretching is the bottom line. The next thing is gain structure. Time stretching will always degrade your audio, especially on the fly stretching, even in complex mode. Offline time strtching helps alot in the biggy DAWs. Lives sucky meters make gain structure an ear only ugly job and It's also very very easy to muck up your audio quality in Live. A real test would involve not using any time stretch parameters in live (no warp) and good gain habits. You will find Live to be on par. You can certainly do professional work with Live, but Live is not a professional program. It's a load of fun to make music on but I wouldn't make audiofile arguments from it. I hope someday Live can be the only program i need, but they seem to have different priorities. Too bad for me, but good for many others. What can you do....

[Chris J]

How come all faders have to be half way through in Live if you don't want to distort on the main output where as the same tracks in SX would have the loudest tracks at zero db, and all other tracks with the whole fader run (from 0 to inf) to play with.
In live you end up with such a tiny margin to move your faders to mix.

Also the statement from Robert about simply adding numbers to mix tracks is contradictory to several articles I've read saying that each make has its own algorithm, therefore its own sound when adding tracks together.
So, is Robert in the know more than anybody else, and are specialized mags talking crap (very plausible)

[rbmonosylabik]

In Cubase, Logic and Reason I hardly ever had any clipping on the master unless something was very loud.

Reason has the worst software audio metering EVER, hands down.

[leisuremuffin]

mags are always talking crap.

remember that computer music article on osx optimization that came out a few years ago? "tracking the big cats"

The guy who wrote it claimed you would get a signifigant boost in cpu usage by removing fonts you didn't use from the computer, along with a whole lot of other bullshit. all complete nonsense.



engineers are a funny bunch, alot of them will tell you that they can hear the difference between different high end pre-amps and stuff like that, but when you put them to the blind fold test, not many really can. I've tried it, and i can tell the difference between the pre's on a mackie and someting actually good, but not between, oh, say neve and API. The guys who sy that different daws sound different are the same kind of people who would advocate buying some several thousand dollar analog summing device rather than use any of the daws summing busses. I think thats a bunch of horseshit. Does digital summing sound differnt than analog summing? Yeah, somewhat. Does it sound better? I don't think so just on the merits of it being analog summing. In this case, i think that the person doing the mixing will make a much bigger impact than the device they choose to mix with. And in the case of difference between DAWs, i'm going to have to say that i think it's highly unlikely that there is any difference, and if there is, nobody on earth could tell the difference blindfolded.



.lm.

[sweetjesus]

ive held back most of the time on issues relating to Live's audio engine and its perceieved quality in the past.

i like to consider myself one of those guys who is very precise in terms of production quality, not to say i have golden ears, but i can definately tell the difference between many fine tonal discrepencies and whatnot.

Recently i've been working on low tempo, minimal vocal stuff with Live (think norah jones, portishead etc) and ive been noticing some very very bizaare stuff in terms of the audio meters not being indicative of the actual clipping that occurs.

I've gone by words from people like Robert in the past as he has said that there is no way things should clip especially if channels are not going into the red, and this made sense considering Live is meant to have 32 bit internal engine. .. however it is clearly not the case.

I do find that it's very much required to mix with your ears in Live as opposed to using any visual metering reference, because somewhere along the line things start to clip and pump.

The compressors clip quite easily, so does the filter and the master meter does not have the proper headroom which I have come to expect of many DAW's.

I don't hate Live for this, but to be honest I haven't got my head around the matter yet. A track I have in progress makes me think that I will not just be able to export it, but I have to do a proper mixdown and start afresh with the gain structure and the ineterbalance of the elements and the amount of air they push, so to this effect it bugs me a little because it removes some of the immediacy of Live.

The crazy thing is, sometimes I've had tracks go into red (including the master) and this actually has helped push the headroom and air back into the audio playback so it's a very dark horse and one which is not as simple as doing bit for bit tests on input and output of a sin-wave (~).

I can assure you guys though that the audio 'quality' is not an issue in Live as I've had no issues about frequency response, but only clipping and distorting problems but for some the two are one and the same.

[leisuremuffin]

I do find that it's very much required to mix with your ears in Live as opposed to using any visual metering reference

This goes for any program or desk in the world.

IMO meters are for setting input levels or for fun blinky lights only.

Can't say that i've noticed anything odd about headroom on lives master bus, but then again, i haven't used any other soft in almost a year at this point. Actually, much longer than that, because every time i've used protools in the past few yearz i was mixing with a desk, not in the box.


One thing i'll agree with is that i've heard distortion from many of the ableton plugs without their little meters going into the red, but that's why i use my ears, not meters.



.lm.

[bassntreble]

all i know is what my ears tell me and I find it more labor intensive to get the sound i'm looking for when i'm mixing 16+ tracks in live and i prefer Protools for this.

Having said that, without live I wouldn't be able to manipulate audio as easily or quickly come up with arrangements or dj/play live sets for that matter. In that case the audio quality is 'good enough' for me.

Live does what it does better than anything else out there. As a compositional/improvisational tool it takes the cake hands down (or up in the air, depending on your bent )

Either way, personally I use the tools that work best for me in the situations they work best in...if that makes any sense...

[radeon]

IMO meters are for setting input levels or for fun blinky lights only.



.lm.

Do you make serious observing about meters you can not be the serious person of audio to thinking this.
but i say there is nothininks wrong with live audio sound it is the situation of always overload plugins vst and live plugins and impulse and simpler make sounds to loud and I dont like this situation becasue I make adjust to volume on plugins always which waste time and live is only software i ever use to make this problems. It is not serious I can say that but it is not good features for sure
And robert say right thing for audio quality but he does no explainings for why sounds are to loud so maybe you can explain this phantom

[radeon]

I can assure you guys though that the audio 'quality' is not an issue in Live as I've had no issues about frequency response, but only clipping and distorting problems but for some the two are one and the same.

I 100% agree with you and want to tell the peopels who keep to saying we talk no good that we are not to talk about live audio quality but the overloadings and bad meters.
I wish those dogs listen to us and stop making lies what we say and confused the story we tell

LIVE AUDIO IS GOOD IT IS BAD METER AND EASY OVERLOAD OF PLUGINS

[sweetjesus]

IMO meters are for setting input levels or for fun blinky lights only.



.lm.

Do you make serious observing about meters you can not be the serious person of audio to thinking this.

this is a bad habit from my past profession as a special fx guy for tv and film.. often you would have to meter things such as legal color limits by using tools such as vectorscopes and waveform monitors to make sure that you are within legal contrast thresholds and also you have legal phasing so you don't blow up a tv network in the process of showing an animation.

but it sounds like you haven't been around enough serious engineers to know that many of them rely on similar tools and visual reference devices such as spectrum analyzers and whatnot to get the job done.

i am not an engineer, i am a producer, i should not have to worry about being accurate with my mixdown if i pay someone else to do it for me, but i should be able to export a track that sounds hot enough to listen to for a reference as long as the meters don't indicate they clip.. . why else do meters on every program and mixer out there go red... because the coders want to give radeon an epileptic fit or because it serves a purpose... think about it man.

[leisuremuffin]

but it sounds like you haven't been around enough serious engineers to know that many of them rely on similar tools and visual reference devices such as spectrum analyzers and whatnot to get the job done.

Wow, that's nice of you. We were talking about the track meters, not a specrum analyzer or phase correlation scope, which live doesn't have anyway. If you know of any engineer who uses the track meters on the board to do something with mixing, please let me know. Good luck with that. Because track meters are used for setting input levels and that's about it.

I don't know, i'm kind pissed off at you right now for implying that i don't know what i'm talking about. As far as that other person goes, untill he stops using that fake bad english i won't address him.



.lm.

[djsynchro]

The compressors clip quite easily, so does the filter and the master meter does not have the proper headroom which I have come to expect of many DAW's.

With all due respect, dude...
What ARE you talking about? Maybe you make nasty clicking noises by driving the compressor hard , or clicky sounds with the filter/envelope, but that is not clipping. There IS no clipping in the internal signal path, you do NOT have to have your channels near unity gain like on an analgue mixer, you can do whatever you want.

Take a normal signal, put 6 uilities at full amplification on them, put 6 more at full attenuation... same signal is back bit-for-bit. Do the reverse, 6 attenuating, then 6 amplifying... your signal will totally disappear, then magically reappear.

If you mix normally to the masterbus (not more than 20db over) and every now and then the meter shoots in the read all you have to then is pull the masterfader back until it just touches 0db... render to 32 or 24bit... done.