How can you describe DAW's sound?

[Sigma]

Originally Posted by dbbubba

Vernier...

You are showing your age.
No one knows what this post means!

Funny joke though.
It's sort of the same as what I posted earlier.

Are you from San Francisco?

pardner..The Shins did a cover of it not long ago

[epp]

So to all you nay-sayers here, who try to 'make sense';

don't you know that 'just math' isn't 'just math', as there in reality is no such thing. Math is flawed, numbers doesn't necessarily add up perfectly, the fact that many number-chrunching softwares turn out different numbers given the same starting point shouldn't really be that much of a suprise.

Most people programming software don't know the difference between good audio and middle of the road. There are millions of ways of screwing up digital audio.... and you can hear the proof for that on the radio. If you null test radio channel one to radio channel two - you'll find that they both suck.

If all DAWs are created equal, then that should mean that they all are perfect, or close to perfect, am I right?

Then why is everything digital so far removed from analogue, in terms of sonic qualities and likability?

[nobo]

Originally Posted by epp

So to all you nay-sayers here, who try to 'make sense';

don't you know that 'just math' isn't 'just math', as there in reality is no such thing. Math is flawed, numbers doesn't necessarily add up perfectly, the fact that many number-chrunching softwares turn out different numbers given the same starting point shouldn't really be that much of a suprise.

Most people programming software don't know the difference between good audio and middle of the road. There are millions of ways of screwing up digital audio.... and you can hear the proof for that on the radio. If you null test radio channel one to radio channel two - you'll find that they both suck.

If all DAWs are created equal, then that should mean that they all are perfect, or close to perfect, am I right?

Then why is everything digital so far removed from analogue, in terms of sonic qualities and likability?

Woah - there's a whole load of mixed up issues in there.

Yes, there are millions of ways of screwing up digital audio - just listen to one of my mixes!

But:
- Perfect? That's a tricky word to use. And in audio, usually (but not always - see below) subjective.
- Digital vs analogue is a whole different debate. Analogue is far from "perfect" technically, but may be perfect sounding to a given set of ears in a given application.
- Removing the panning law issues (which differ in some DAWs), Lynn's awesome DAW summing test shows that most (N.B. not all!) DAWs will null against each other. (BTW, he also compares DAWs to analogue desks). Any DAWs that nulls against another DAW in such a summing test perfectly matches the other for those test conditions. If the files null, there is no audio difference in those files. So in that sense, most DAWs are perfect under those conditions.
- Where things get trickier - and where there curiously still seems to be a lack of consensus on this thread - are some outstanding issues, e.g.:
- bit depth affecting plug in resolution (and therefore sound) - a 64 bit mix engine is going to do a different ("better") job than a 24 or 32 bit one.
- handling of non-linearities - e.g. distortion/clipping (one of the reasons PARIS does sound different to other DAWs is that you can push it well into the red - clip lights going off all the time - and it doesn't give you that horrible digital mess but instead sounds more analogue - adding warmth and compression).

- As for maths not being perfect, that's a big and complicated issue. Chaos theory shows us that maths is very sensitive to initial conditions and that in a system a small starting inaccuracy can have massive repercussions down the line in. But there it's not maths itself that flawed, but - for example - roundings in the initial numbers causing problems. Generally, the sort of maths a DAW is doing in controlling levels and summing is simple and shouldn't be flawed. At least, until you start clipping audio (but then, PARIS apart, you definitely don't want to be clipping your DAW!). Maths as an abstract concept is pretty much perfect: it's a theory. Though admittedly, when you get to crazy PhD and beyond maths (I've a few friends who are doing this), it becomes a slightly different issue.


There's loads already on DAW summing - some of it already on this forum.

DAW Summing Comparison - A new testing angle

DAW software sonic differences

Awesome DAWSUM Samples - 3dB


Summing Accuracy Demonstrated - 3dB

And this one, though not for the faint of heart. It is 23 pages long.

Difference Signals - 3dB ______________

And, to quote Lynn on the awesome DAWSUM testing:

" Each platform was given identical audio files (11 stereo pairs) and generated a stereo audio file that was the sum of those. We compared those, both sonically and mathematically. Then people commented on the results. Many, but not all, were able to match identically on a value by value basis for the 69M character result. Some were not able to match.

I will add this. Some people listened to numerically identical files and heard different things. Resolving that issue is not something anyone has been able to do successfully yet, nor do I suspect it will ever be scientifically explained, primarily due to the intricacies and errancy of human perception and interpretation."

[Clem Snide]

Originally Posted by epp

So to all you nay-sayers here, who try to 'make sense';

don't you know that 'just math' isn't 'just math', as there in reality is no such thing. Math is flawed, numbers doesn't necessarily add up perfectly, the fact that many number-chrunching softwares turn out different numbers given the same starting point shouldn't really be that much of a suprise.

Most people programming software don't know the difference between good audio and middle of the road. There are millions of ways of screwing up digital audio.... and you can hear the proof for that on the radio. If you null test radio channel one to radio channel two - you'll find that they both suck.

If all DAWs are created equal, then that should mean that they all are perfect, or close to perfect, am I right?

Then why is everything digital so far removed from analogue, in terms of sonic qualities and likability?

I commend you on the sheer concentration of misapprehension in one post.

[vernier]

Clearly, science is taking over.
'

[theRog]

Hi there

if you can show me someone who can hear a difference between two DAW's (in a double blind test - that is) I'll start believing in this.

But until then they all sound the same as it is all digital...

AD/DA converters definitely sound different.
Different plugins on different DAW's sound different.

But don't go into float vs. fixed as it all gets too complex and again I doubt somebody would hear a difference!

That's it - choose your DAW based on features and of course your experience working the tool ;-)

Move the Mic an inch and the whole difference in DAW sound if there is one becomes quite irrelevant.


Cheers
Roger

[deaf bitch]

Originally Posted by theRog

Move the Mic an inch and the whole difference in DAW sound if there is one becomes quite irrelevant.

one more: start making sound by yourself - instead of expecting it by your gear.

[Mr.HOLMES]

Originally Posted by deaf bitch

one more: start making sound by yourself - instead of expecting it by your gear.

I am wondering that we have to explain that sound is not coming from the gear....sound is 90% performance of the musicians.... the gear can not write a SONG!!!

[cedar80]

Each DAW (must I say, each audio software) has its own sound. the differences are subtle but noticeable. there's no question.

Sometime marketing doesn't lie. When brands say "new audio engine - better sound", it's not bullshit.

Digital audio is not a matter of mathematics only. It's far more complicated. During bouncing, the software sums the tracks (and also re-calculates the word clock), by using proprietary codes. Every serious man knows that. And there's other factors...

But it's clear that these differences are NOT a valid reason to move to another DAW. And I think every "big" DAWs (as all mentionned in the thread), are "pro results" capable.

Skills, talents and determination are the only way.

[Corran]

Originally Posted by cedar80

Each DAW (must I say, each audio software) has its own sound. the differences are subtle but noticeable. there's no question.

I think there IS a question! Experiments say otherwise!

Not much in audio can be "quantified" in terms of math and experimental outcomes. I think this is one of the ones that can. If tests show they sum the same, then THEY DO.

[joeq]

Originally Posted by Corran

I think there IS a question! Experiments say otherwise!

Not much in audio can be "quantified" in terms of math and experimental outcomes. I think this is one of the ones that can. If tests show they sum the same, then THEY DO.

why do I have the feeling that cedar80 is trolling? Could it be that he really did not read any of the 10 pages of this thread before he posted? Or could it be that he knows how to annoy people by trying to take everyone back to Square One?

[Corran]

Well yeah that might be...

[cedar80]

Again, I'm sad to see that bi-valent logic rules the world... but I forgive you - you're not prepared.

I heard differences and I believe in my ears. Theses are subtle BUT noticeable.

My impressions are based on my own ears and more than ten years of working on PT (TDM), Cubase and sometimes Ableton Live . And they DO sound different, as true as hi-end converters sound close but don't exactly sound the same.

With amateur hardware with poor DA converters, eg Motu or Digi002 stuff, two different hi-hend AD converters will sound the same. For example, try to tell the difference between a Mytek and a Weiss by listening to it using cheap DA converters: it will be extremely hard if not impossible. If you replace the cheapo DA converter by an accurate one, eg Mytek, Lavry, Weiss or so, theses differences will become "much" more obvious. That's the same with DAWs.

I didn't listen to the examples posted on any forum because I don't give ANY **** about an hypothetic shootout made by a member, simply because on a forum, everybody is a potential liar. How to verify the authenticity of the examples???? And who has the time to make such a shootout? Who owns ALL the DAWs? there's no way to prove that. And screen caps don't prove anything.

And, how to prove which DA converters, which monitors are used by the members who affirm "there's no difference"? It's not a secret that the MAJORITY of gearslutz people have amateur/semi-pro hardware.
I know some of you have "good" gear. I know some of you, as me, are recording/mixing engineers.
But I know the rest has motu, digi, presonus, etc. This is a fact. Since many members are not pros and have a job, they don't "slutz" at home. The "slutz" at job, with crappy multimedia speakers (or earphones) and AC97 soundcard.

Not to mention quality of the human ear. We are talking about very small differences. Some ears won't notice it. Some others will.

There's also another factor: the placebo effect. Blindly believing in something can alter perceptions, that's why I'm always DOUBTING about what people say and of course what I say. Talking about hihend "neutral" stuff, I mean stuff where difference are very small (hihend converters, cables, DAWs...), if you are a 100% slut (I mean if you follow the leaders of the thread), and If you want them to sound the same, then they WILL sound the same.

My initial convictions about theses differences were confirmed by listening to the "Awesome DAWSUM Sampler" CD by 3DAudio. not cheap but worses the penny. I recommend it to everyone. By listening this CD you definitively CAN'T affirm they all sound the same. I really would like to share it but there's too much risk to have problems with law, so sorry.

I have much more confidence in the 3DAudio guys than in gearslutz members because until now they are the ones who tend to prove the authenticity of their shootout (don't like this word), by the existence of their company and their experience. I may be wrong.
And you know, gearslutz is clearly considered by more and more people as.... a place for real slutz. a place for guys who aren't capable of thinking by themselves, a place where opinions and manipulations have replaced facts and real proofs. I'm not talking about the founders or administrators. I'm talking about some members (who sadly are leaders). Fortunately I know everybody here isn't like that.

And, I forgot (I know it's not a proof, but I prefer to make you remind that): At the SAE (Paris - but I guess it should be identical in others SAE) I was also teached that main DAWs on the market sound all slightly different, with a little superiority of DSP-based host (TDM, Pyramix... but I'm really doubting). Can someone else could confirm he has leaned that?

Concerning VERY technical explanations (in french, sorry) about differences of sound between DAWs (in this case, ProTools TDM vs ProTools LE vs other natives DAW - read all the post written by the member named Labroue, a rec/mixing enginner mainly in latin music, former consultant, also specialized in digital audio, with an enormous scientific knowledge and skills, and a very opened spirit):

http://www.forum.audioaddict.fr/viewtopic.php?t=553

But, I can be wrong and I want to say: Don't believe me.
Maybe we don't agree on that subject ("difference of sound between DAWs").

But we all agree that the choice of the DAW doesn't matter, and it won't make your sound.

[joeq]

Originally Posted by cedar80

d.


There's also another factor: the placebo effect. Blindly believing in something can alter perceptions, that's why I'm always DOUBTING about what people say and of course what I say. Talking about hihend "neutral" stuff, I mean stuff where difference are very small (hihend converters, cables, DAWs...), if you are a 100% slut (I mean if you follow the leaders of the thread), and If you want them to sound the same, then they WILL sound the same.

and if you want them to sound different they WILL sound different. Even if they really are the same. The placebo effect cuts both ways.

Which is why scientific proof that a large number of DAWs sum identically is so compelling.

My initial convictions about theses differences were confirmed by listening to the "Awesome DAWSUM Sampler" CD by 3DAudio. not cheap but worses the penny. I recommend it to everyone. By listening this CD you definitively CAN'T affirm they all sound the same.

again, it PAYS to read the thread you are posting to. The Awesome DAWSUM experiment is the one that everyone has been talking about here for the last 10 pages.

While it found differences between some DAWS, null tests using the files from this CD confirm that many of the very programs that people had been claiming had 'subtle differences' actually created identical files when summing.


No one is saying they all sound the same (again I refer you to reading the entire thread) but people are saying that the ones that output numerically identical files must sound the same.

[Dean7]

Originally Posted by joeq

and if you want them to sound different they WILL sound different. Even if they really are the same. The placebo effect cuts both ways.

Which is why scientific proof that a large number of DAWs sum identically is so compelling.



again, it PAYS to read the thread you are posting to. The Awesome DAWSUM experiment is the one that everyone has been talking about here for the last 10 pages.

While it found differences between some DAWS, null tests using the files from this CD confirm that many of the very programs that people had been claiming had 'subtle differences' actually created identical files when summing.


No one is saying they all sound the same (again I refer you to reading the entire thread) but people are saying that the ones that output numerically identical files must sound the same.

I 100% agree. Good post.

[cedar80]

Joeq, you're right. I must confess I didn't read every single post. But I am sure I'm not the only one in that case.

To be honnest I started reading the 4 first pages. I was bored cause it looked quite sterile and took a speculative way. I jumped to page 9 then 10. Always the same thing: believers and non-believers with no valable arguments.

In our jobs (passions, hobbies etc), as you know, there's ONLY ONE JUDGE: the human ear. Of course, as I stated, with a mind as independent as possible. That's why I always turn off my screens when "listening in order to judge". I don't care about my eyes when I'm judging sound quality. I don't care about screenshots. Don't care about any phase cancellation which is SUPPOSED to prove something. Don't care about nothing but HOW IT SOUNDS.

How can someone affirm a null test cannot be wrong?

I knew a guy who is designing headphones with noise cancellation system for a very young company. As you know, theses systems uses the same principe as the null test (phase cancellation).

He affirmed that, inverting the phase of one, two very close (but different) files can be completely nulled. To him there IS a very tight tolerance. A tolerance he uses in his work. Believe me if you want. (Of course that tolerance becomes greater as the bit depth decreases)

and if you want them to sound different they WILL sound different. Even if they really are the same. The placebo effect cuts both ways.

I never meant the contrary. I just said that placebo thing falsifies judgement.
No need to say I'm not considering me as a 100% slut. And when I began recording/mixing I knew only one software and thought all DAWs had the same sound. I never wanted them to sound different.

[joeq]

Originally Posted by cedar80

How can someone affirm a null test cannot be wrong?

I knew a guy who is designing headphones with noise cancellation system for a very young company. As you know, theses systems uses the same principe as the null test (phase cancellation).

Phase cancellation for the purpose of noise abatement is a far cry from the 100% cancellation found in some of these tests. In a computer, digital gain can be applied to bring out the tiniest difference signal. If you gain it up and up and up and there is still nothing there, you have 100% cancellation. You can try this for yourself making a copy of any audio file you have handy. A copied file, being identical, will null 100%

Modify one of the files in the tiniest possible fashion: move one pan knob from 100 to 99 (something nobody could claim they could Hear) and you will get a difference signal- often quite a significant one. This means that the null test is also capable of proving two files are different - even to a greater sensitivity than the most "Golden" of human ears.

One does not have to look at this from the point of view of "hearing" or how "quiet" the difference signal is. The proper application of the null test is rather subtraction. An audio file can be thought of as a VERY large number. When you subtract one number from another number and you get zero, those two numbers MUST have been identical.

No other conclusion is mathematically possible.

Zero! Not .0000000000001, but zero. 44,100 times a second every sample came out with the same exact 24 bit number. There is no wiggle room for the mystical subtleties that supposedly flit like ghosts through our software's algorithms if the two programs in question produce nulling files.

If they don't null, and someone can repeatedly pick them out in a blind test, that's another story. I personally don't think very many people can, but that's just an opinion, not a mathematical fact.

He affirmed that, inverting the phase of one, two very close (but different) files can be completely nulled. To him there IS a very tight tolerance. A tolerance he uses in his work.

His tolerance, however "tight" in his world, only has to be good enough for noise abatement, not mathematical analysis of two files. If the files are different, they will not completely null. They may null to the point where someone may not hear any sound coming out with his unaided ear, but when gain is applied, a difference signal will be found down there somewhere.

Believe me if you want.

It's not that I don't WANT to believe you. It's that I CAN'T believe in Magic when the science is so unequivocal.


I understand there is a cash prize waiting for the person who can correctly and consistently differentiate two nulling files in a blind test. Find the person whose ears are so good, he can reliably hear a difference where the math says there is no difference, and I will eat my words.

[LPK]

don't you know that 'just math' isn't 'just math', as there in reality is no such thing. Math is flawed, numbers doesn't necessarily add up perfectly, the fact that many number-chrunching softwares turn out different numbers given the same starting point shouldn't really be that much of a suprise.

There's only one thing:
The resolution.
For most hosts it's 32bit, for Protools it's 48bit. Both is high enough to give a flawless output of 24bit.
1+1 is always 2. There's nothing, which can't be done wrong here. The phrase "just math isn't just math" is just untrue.

Most people programming software don't know the difference between good audio and middle of the road. There are millions of ways of screwing up digital audio.... and you can hear the proof for that on the radio. If you null test radio channel one to radio channel two - you'll find that they both suck.

?¿?

Each DAW (must I say, each audio software) has its own sound. the differences are subtle but noticeable. there's no question.

Funny ...
I did some phase tests with some sequencers and the differences were always somewhere at ca. -150db (so barely on the last bit of an 24bit file).
And those probably came from a wrong setting somewhere ... had no ambition to waste any more time for that stuff to check out.

I heard differences and I believe in my ears. Theses are subtle BUT noticeable.

It's sometimes (especially for such things) not always the best thing to only trust your ears. Because your brain can play a trick using the placebo-effect. I think nobody is immune to this effect.



Thing is:
There is absolutely no reason -NO REASON- for a digital sequencer to have any sound at all. ABSOLUTELY NO REASON! It's digital .. it's exact, it's absolutely neutral.
But some people think, they noticed something like that ... but why don't they just do a phase-test to really make it sure ? If they'd do, they'll see, there is nothing like the "Protools-Sound" or the "Logic-Sound" or the "Cubase-Sound" or whatever.

Such stuff is just esoteric ...
(I even saw people discussion about, that the CPU in a pc had an influence to the sound in audio-production.)

How can someone affirm a null test cannot be wrong?

There is nothing to affirm. It's a fact, which is right per definition.
A null test shows you exactly the difference of two files. The moment you are questioning that, you are questioning the meaning of the operator "minus".

[Dean7]

Originally Posted by cedar80

In our jobs (passions, hobbies etc), as you know, there's ONLY ONE JUDGE: the human ear.

When it comes to music, you couldn't be more right. As long as it sounds good, who cares? As long as people like it, who cares?

When it comes to audio, you couldn't be more wrong. I'm not saying the human ear is completely useless, but it's not a scientific measuring tool.

-Dean

[Electronique]

Sorry If im repeating, i didnt read through all the posts.. But

Hasnt it got to do with Dithering?

Im not really up with what it is.. Noise added to fill in the gaps??
I know Sonar has several different dither settings.. NOw wouldnt the various DAW dither algorithms differ the sound... a little?

[dzn]

okay, both files null after rendering and switching the phase on one. so they must be the same. but maybe (don't flame me!! *g*) the sound in playback mode before bouncing is slightly different.

why?

to give the listener the tiny illusion it sounds "better", "richer", "warmer", more "analog"...
some people claim that their cubase-mixes sound different after mixdown.

keep in mind, i'm not saying sequencer really do sound noticeable different.

[Dean7]

Originally Posted by Electronique

Sorry If im repeating, i didnt read through all the posts.. But

Hasnt it got to do with Dithering?

Im not really up with what it is.. Noise added to fill in the gaps??
I know Sonar has several different dither settings.. NOw wouldnt the various DAW dither algorithms differ the sound... a little?

Well, first off you should read Massey's post on dithering:

smassey.com

Anyway, you're right though and dithering could be one thing that might make the tests not null (even though the difference would be so minute that the human ear could probably have no hope of picking it up). But dithering usually comes in to play when you change downsample or change bit depth (I think some people think otherwise though, based on Massey's post).

Originally Posted by http://en.wikipedia.org/wiki/Dither

Dither most often surfaces in the fields of digital audio and video, where it is applied to rate conversions and (usually optionally) to bit-depth transitions; it is utilized in many different fields where digital processing and analysis is used — especially waveform analysis. These uses include systems using digital signal processing, such as digital audio, digital video, digital photography, seismology, RADAR, weather forecasting systems and many more.

The premise is that quantization and re-quantization of digital data yields error. If that error is repeating and correlated to the signal, the error that results is repeating, cyclical, and mathematically determinable. In some fields, especially where the receptor is sensitive to such artifacts, cyclical errors yield undesirable artifacts. In these fields dither results in less determinable artifacts. The field of audio is a primary example of this — the human ear functions much like a Fourier transform, wherein it hears individual frequencies. The ear is therefore very sensitive to distortion, or additional frequency content that "colors" the sound differently. The ear is far less sensitive to random noise at all frequencies.

So, basically re-quantization might introduce some slight distortion, and the thinking behind dithering is: white noise is better than distortion. Anyway, I bet if you down-sample tracks using different software they might not null out (depending on dithering and whatnot). But I also bet the difference between the audio files would be pretty small, and I'm not sure any human ear would be able to pick it up.

[Dean7]

Originally Posted by dzn

okay, both files null after rendering and switching the phase on one. so they must be the same. but maybe (don't flame me!! *g*) the sound in playback mode before bouncing is slightly different.

why?

to give the listener the tiny illusion it sounds "better", "richer", "warmer", more "analog"...
some people claim that their cubase-mixes sound different after mixdown.

keep in mind, i'm not saying sequencer really do sound noticeable different.

Well, it's possible that one of these DAWs runs the audio (during playback) through some sort of "analogifier" plug-in, but I find that highly unlikely because if people found out they'd be furious. And what good is a DAW if the mixes sound different bounced than they do during playback?

-Dean