Does it sound better than Cubase alone.

[valis]

It's been proved many times there is no difference in raising/lowering the master fader as opposed to single channels. And the floating vs integer debate has no weight either (assuming you're not clipping integer signal paths).

I believe my original case was a certain Behringer mixer that uses the same sharc chips that our scope cards do, being used simply for mixing a few channels of audio (not even more than a dozen.. Regardless of what theoretical perfect cases you might choose to use to support your end of the discussion, when proper respect to headroom isn't paid that mixer degrades the audio in ways that most mixers in Scope don't. It has little to do with clipping either, and I don't claim to have an expert ear for 'quantization error' and other such things, I simply know it sounded bad. Also, given an analog mixer by the same company and anything halfway decent made in the late 70's & early 80's, I know which mixer I would trust the master bus more on (assuming the older unit has been cared for). So in both the digital and analog realm I have experience (and I know I'm not alone here) with things that violate what you insist is utter truth.

Anyway I fail to see why this discussion keeps coming back to summing. It's like gamers who quote memory bandwidth figures and cpu stats while ignoring the fact that they're running a $15 motherboard.

Now I do understand your insistence on promoting the idea that musical content & the skill of the artist is far more important than splitting hairs over digital mixing systems. In fact if I had actually answered the original poster I probably would have told him that my opinion is that Scope's greatest strengths are in synthesis and there is Definately a noticeable quality difference here. And yet the math is still supposedly identical to what you can achieve in native software right?

Imo you're becoming pedantic in your responses and rather than being informative you come off as rather argumentative and steadfast without giving any real illumination to the subject. You might actually be right and we could be wrong for reasons that we've failed to notice, I can accept that. But let's be a bit more technical here and have all the info instead of vague references that act as broad sweeping dismissals. The real point of such discussions is that we're here to talk about Scope and that includes things that might be potentially splitting hairs.

[Liquid Len]

You're not the only one, Valis, that is using both the behringer 3216 mixer and scope mixers (as well as analog behringer & others). I can certainly notice the distortion problems with the 3216 especially when the channel inputs are running hot - you have to have cotton in your ears not to! Whether or not it's theoretically possible - my ears hear a difference between behringer's muddy sounding mixes, and scope's cleaner sounding ones. I've even considered going live using the scope card to replace the 'guts' of the 3216, but it would be more work than it was worth when all is said and done.

[darkrezin]

It's been proved many times there is no difference in raising/lowering the master fader as opposed to single channels.

Simply cannot agree. I'd even go as far as to say you're taking absolute shit here, everything my ears tell me, and everything I've learned about gain structure proves this to me.

You're obviously on some kind of mission to correct all us stupid subjective engineers out there. I wish you the best of luck in your futile quest

[katano]

so what's the tip for mixdown if the master level is clipping? reduce each channel or reduce the master level? how do you guys work out there?

[darkrezin]

Reducing each channel is what I would do (because this sounds best to me). However there may be other factors affecting this, like for example low end frequencies on tracks which don't need them.

[katano]

and why? let's say if none of the channel itself is clipping... ah, you just edited your post

[darkrezin]

It depends a lot upon the headroom of the mixer in my experience... as Valis says certain mixers have a higher headroom than others and won't clip as easily. I generally start a mix with all faders at -12dB and go from there.

[darkrezin]

As to *why* you should do this - if you reduce the master fader in:

1. an analog mixer - you reduce the signal-noise ratio

2. a digital mixer - you lose resolution (bits)

[garyb]

"It's been proved many times there is no difference in raising/lowering the master fader as opposed to single channels."

this right here is what seperates the trolls from the rest of us. poppycock, sir. poppycock.

[Immanuel]

Anyway I fail to see why this discussion keeps coming back to summing. It's like gamers who quote memory bandwidth figures and cpu stats while ignoring the fact that they're running a $15 motherboard.

Me too, I find it quite obvious, that the question in this thread is not intended to deal exclusively with summing - but rather with a 'mix' ... with all that it takes to do a mix.

[medway]

It's been proved many times there is no difference in raising/lowering the master fader as opposed to single channels. And the floating vs integer debate has no weight either (assuming you're not clipping integer signal paths).

I believe my original case was a certain Behringer mixer that uses the same sharc chips that our scope cards do, being used simply for mixing a few channels of audio (not even more than a dozen.. Regardless of what theoretical perfect cases you might choose to use to support your end of the discussion, when proper respect to headroom isn't paid that mixer degrades the audio in ways that most mixers in Scope don't. It has little to do with clipping either, and I don't claim to have an expert ear for 'quantization error' and other such things, I simply know it sounded bad. Also, given an analog mixer by the same company and anything halfway decent made in the late 70's & early 80's, I know which mixer I would trust the master bus more on (assuming the older unit has been cared for). So in both the digital and analog realm I have experience (and I know I'm not alone here) with things that violate what you insist is utter truth.

Anyway I fail to see why this discussion keeps coming back to summing. It's like gamers who quote memory bandwidth figures and cpu stats while ignoring the fact that they're running a $15 motherboard.

Now I do understand your insistence on promoting the idea that musical content & the skill of the artist is far more important than splitting hairs over digital mixing systems. In fact if I had actually answered the original poster I probably would have told him that my opinion is that Scope's greatest strengths are in synthesis and there is Definately a noticeable quality difference here. And yet the math is still supposedly identical to what you can achieve in native software right?

Imo you're becoming pedantic in your responses and rather than being informative you come off as rather argumentative and steadfast without giving any real illumination to the subject. You might actually be right and we could be wrong for reasons that we've failed to notice, I can accept that. But let's be a bit more technical here and have all the info instead of vague references that act as broad sweeping dismissals. The real point of such discussions is that we're here to talk about Scope and that includes things that might be potentially splitting hairs.

I agree we're beating a dead horse on some of these issues. Yet I do find discussing this kind of thing interesting though and I hope no one is taking offense here. I take my mixing seriously and find it important to get to the bottom of these types of details.

[medway]

It's been proved many times there is no difference in raising/lowering the master fader as opposed to single channels.

Simply cannot agree. I'd even go as far as to say you're taking absolute shit here, everything my ears tell me, and everything I've learned about gain structure proves this to me.

You're obviously on some kind of mission to correct all us stupid subjective engineers out there. I wish you the best of luck in your futile quest

What do you know about gain structure that leads you to believe this?

[medway]

It depends a lot upon the headroom of the mixer in my experience... as Valis says certain mixers have a higher headroom than others and won't clip as easily. I generally start a mix with all faders at -12dB and go from there.

Digital clips at 0. There's no gray area there.

Some of you guys seem to be confusing digital with analog.

[medway]

As to *why* you should do this - if you reduce the master fader in:

1. an analog mixer - you reduce the signal-noise ratio

2. a digital mixer - you lose resolution (bits)

Bits don't equal resolution. Bits define a potential to capture dynamic range. In a properly dithered system less bits only means more noise (which can be construed as a basis for quality but given the signal to noise ratio of most analog gear digital doesn't have to work too hard to match it). The whole analogy of "resolution" when talking about bits is a false one.

That being said -6db coming out of a master fader results in the same amount of bits regardless of if it was the master fader lowered or the track it self. It's the final outcome that matters. The master is just a multiplier.

Unfortunately this is one of those myths that will be very hard to shake as abotut 99.9% of engineers out there don't understand it. Largely due to manufacturers using it as a selling point.

[medway]

"It's been proved many times there is no difference in raising/lowering the master fader as opposed to single channels."

this right here is what seperates the trolls from the rest of us. poppycock, sir. poppycock.

It's very easy to test and prove. I'm sorry but you guys sound like you're still believing incorrect DAW info from about 5-6 years ago.

[Immanuel]

Bits don't equal resolution.

I suggest some basic reading and understanding of how bits work as building blocks. Here goes a simple explanation. Well, it is not an explanation, but ... 4 bit recording ...

8
8
8
8
84
84
842
8421

0000
0001
0010
0011
0100
0101
0110
0111
1000
1001
1010
1011
1100
1101
1110
1111

0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15


Now, removing one bit

8
8
8
8
84
84
842
842

000
001
010
011
100
101
110
111

0
2
4
6
8
10
12
14

Where did 1, 3, 5, 7, 9, 11, 13 and 15 go, when we removed a bit? Well it disapeared - simple as that ..... the resolution dropped.

[darkrezin]

As to *why* you should do this - if you reduce the master fader in:

1. an analog mixer - you reduce the signal-noise ratio

2. a digital mixer - you lose resolution (bits)

Bits don't equal resolution. Bits define a potential to capture dynamic range. In a properly dithered system less bits only means more noise (which can be construed as a basis for quality but given the signal to noise ratio of most analog gear digital doesn't have to work too hard to match it). The whole analogy of "resolution" when talking about bits is a false one.

That being said -6db coming out of a master fader results in the same amount of bits regardless of if it was the master fader lowered or the track it self. It's the final outcome that matters. The master is just a multiplier.

Unfortunately this is one of those myths that will be very hard to shake as abotut 99.9% of engineers out there don't understand it. Largely due to manufacturers using it as a selling point.

This was very badly expressed by me - thanks for picking up on it. In my experience, all digital mixers have a limited amount of headroom, but it is not the same across all of them. I have found that I get better results by not overdriving the mix bus and just keeping individual channel faders nice and low. In a digital mixer there's no need to have every channel touching 0dB, there's certainly no benefit from this. I find that if I keep channel faders low so I don't need to bring down the master fader or put some ugly ultra-limiter on it, the end result sounds more natural. Sadly I'm not a scientist when it comes to digital systems (I have no knowledge of DSP programming) so I can only go on what sounds better to me.

[valis]

I concur with what darkrezin says above. I'm limited on time atm but I'll try to dig out some of the references I recall that discuss gain structure (in analog and digital). Also I recall a nice discussion here on planetz on how floating point data can 'erode' through multiple steps. Memory tells me that it was Tom that posted it up in one of the discussions about why mixing in 32bit int might be better, but don't quote me on that 100%. I do recall it voiced something I learned in my programming days (15 years ago!) better than I can voice it now by memory.

Anyway I'm happy to see the discussion is a bit more back on track. I certainly meant no ill will towards medway, I just would like to get at the details if we can. And if we can't then just be clear where we're stating opinion or facts we can't back up.

[astroman]

... Memory tells me that it was Tom that posted it up in one of the discussions about why mixing in 32bit int might be better, but don't quote me on that 100%. ...

well, I posted something about numeric precision in general, as any digital system only has a finite resolution.
Anyone with even the most humble programming experience knows, that there are (usually) dozens of ways to implememnt a certain function.
It might explain differences in sound even though the same programming strategy (algorithm) was applied.

... for precision's sake any attempt IS bound to fail regardless of the initial number of bits used - it just happens later.

I got across this in the chapter Chaos wipes out every Computer from the book 'Chaos and Fractals' (Springer Publishing).
Quadratic equations with feedback aren't uncommon in audio processing afaik, so their 'experiments' even some real world appeal.

The first one was stunning, yet understandable to a degree, as the equation r+rp(1-p) was iterated on 2 scientific calculators with different numbers of digits (10 versus 12).
Quote:
... we noted that the tiny little deviation we noticed in the 10th decimal for the 6th iterate has migrated through all decimal places, i.e. after 40 iterations has been amplified by a factor of 10^10

the next 'experiment' was introduced with ...this is still not the end of the story. Things are even wilder as we've seen so far...
they run 2 versions of the equation above on ONE calculator i.e. p+rp(1-p) versus (1+r)p-rp^2
as you might have already guessed the results start to deviate at the 12 iteration...
it's a principle in this type of calculation and it will strike on a Casio pocket calculator as well as on multi-million $ supercomputer.
...

check the respective chapter in 'Chaos and Fractals' (Springer Publishing) in any bookstore, or buy the book... it's fairly interesting

cheers, Tom

[medway]

Bits don't equal resolution.

I suggest some basic reading and understanding of how bits work as building blocks. Here goes a simple explanation. Well, it is not an explanation, but ... 4 bit recording ...

8
8
8
8
84
84
842
8421

0000
0001
0010
0011
0100
0101
0110
0111
1000
1001
1010
1011
1100
1101
1110
1111

0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15


Now, removing one bit

8
8
8
8
84
84
842
842

000
001
010
011
100
101
110
111

0
2
4
6
8
10
12
14

Where did 1, 3, 5, 7, 9, 11, 13 and 15 go, when we removed a bit? Well it disapeared - simple as that ..... the resolution dropped.

Bits determine noise floor when properly dithered, nothing else. Again very easy to test. When you remove bits you don't lose resolution, you just lose low level signal into the increased noise floor. The key here is understanding how dither linearizes digital signals.