Analog summing emulation idea

[Kingston]

when was the last time you guys had a wank?

I started with a warning, didn't I?

Here comes total transpotter wankery, excuse me everybody else.

[friteuse]

Ok, here's some point of view from a more musician than engineer...
First of all, I appreciate all kind of discussion about subjects like this, but please, keep it on a discussion level, don't become too personal...

I always found newer mixes of mine a bit "flat" and unpleasant, since I changed to an all digital environement. First, I didn't want to hear that, because the digital world offers so much great possiblities and is so easy to use.

But then, I did some tests with a few summing methods: Logics built in summing, ProTools summing and analog summing (on a Dangerous Music Mix Bus). The first look resulted in a much more transparent and pleasing mix with the analog summing. Even our trainee, that was very inexperienced with music productions, noticed an obvious difference doing the blind test.

I wondered, why. Maybe the dither from the DAs? So I tried to sort out DA/AD artefacts by directly resampling the bus outputs and summing them within Logic/ProTools. Different, but not that much better. Maybe because of resampling all busses instead of only one sum when analog summing. Analog summing was still better.

What's the point? The people from Dangerous Music told me, that analog summing sounds much better because of it's high precission and accuracy. Eh?? I don't have the data sheets anymore, but I remember, the tolerance of the mix bus was much smaller than 1%.

As a conclusion, I tried to optimize the "precission parameters" of DAW, and these are: sample resolution and sample rate. I found, with 96 kHz and Protool's 48 bit fixed point mix engine, adequate mixing results could be achieved, and made analog mixing accuracy for my needs useless.

This is only for the analog precission, not for the analog sound itself, which is something completly different. When it comes to distortion, the "feeling", in my humble opinion nothing beats the real thing... I get this feeling by recording with high quality mic preamps (though colouring), using real guitar amps and some hardware gear (less than I wished )

Ok, this is not a scientific test, it was done just for myself to find out, what worked better for me. Just wanted to let you know that

[Aleksey Vaneev]

camsr, phase absence you've noticed is ONLY applicable to DC and Nyquist frequency. Other frequencies have the phase information. But if we are talking about resolution it is as good as filter you use - there are no other limitations to it.

[Aleksey Vaneev]

friteuse I suggest you to capture your D/A - A/D chain's impulse response and apply it (through convolution) to the audio mix you've got from Logic or Protools. Paired digital/analog - analog/digital conversion may add a lot of coloration and so 'analog summing' may sound better. Not to say both D/A - A/D implement distortion plus the analog box itself may create some harmonic distortion as well. It is an unfair test. If you applied some non-linear VST plug-in to Logic or Protools output and then compared it to analog summing output the comparison would be a bit more fair.

[defjamm]

If it's so well known, wtf are the plugins that emulate this? No need to be nasty dude. I have followed the ITB vs Summing debates at forums where real record producers hang out. No reflection on the KVR community, but the people actually making records tends to be hardware slutz, rather than plugin slutz.

The people actually making hit records are going in for analog summing in a big way, if not full analog.

bob katz speaks the truth
http://gearslutz.com/board/showthread.php3?t=63827

you mean the 'pros' who claim digital changes the sound in a bad way and adds something, and then i do a phase test and everything nulls out and they still claim their bullshit? these guys don't hear, they believe. maybe they are pros, but a lot of them hear with their eyes.

didn't you ever wonder, why there are so few soundexamples of different things matched as close as possible?

[Left Headphone]

This debate will go on forever until somebody produces some test results. I think both sides are afraid of the results...

Just do what ever sounds good to you...

[deaf dunderkwac]

I won't provide test results, but the analog mixer 'adds' stuff, like phase errors, noise and other little distortions that the 'digital' version is incapable of.
Any noise or distortion added to the digital mixer is under *my* control, whereas there is no control over what an analog mixer adds (and depending on the mixer and the external wiring *that* can be alot)

[nuffink]

One of the better comments I ever read on kvr was "If all of this classic gear is so good why do all those old records sound so shit?" Hand's up who said it. It should be engraved above the door.

[Zoomby]

Hi,
I have no experience with analog gear, but I think since the effects analog stuff is doing with a sound is nothing magical but only some kind of eq, noise or distortion, the only thing needed is an algorithm. Then it will sound the same in the digital world. So it's only a question of DSP knowledge and CPU power. What do you think?

[friteuse]

friteuse I suggest you to capture your D/A - A/D chain's impulse response and apply it (through convolution) to the audio mix you've got from Logic or Protools. Paired digital/analog - analog/digital conversion may add a lot of coloration and so 'analog summing' may sound better. Not to say both D/A - A/D implement distortion plus the analog box itself may create some harmonic distortion as well. It is an unfair test. If you applied some non-linear VST plug-in to Logic or Protools output and then compared it to analog summing output the comparison would be a bit more fair.

Hi Aleksey,

thanks for your reply! You just said what I wanted to point out (maybe I used wrong words, it's so difficult in english )

So, what I wanted to say is that I noticed the difference but I was quiet unsure if it wasn't just the DA/AD thing. My result is, if I work in 96 kHz with high bit depth, digital world sounds very good to me and without a big difference to the analog summing. And this only under the aspect that analog summing gear doesn't add a noticable distortion (what is the aim of the Mix Bus).

Or did I get you wrong?

Btw, I further did a test with digital bus summing, and applied a dither plugin in each bus. This came nearer to the analog thing when working with 44.1 kHz, but wasn't as good as 96 kHz without dither.

With "distortion" I thought more the amps, compressors, tube boxes and so on, say the things that add wanted and very noticable distortion. Here, I really prefer the analog gear...

[deaf dunderkwac]

I have years o experience with analog mixers as a mixer and a 'chief'. So I'll repeat again what happens at the mx buss.
1. noise. Both broadband and hum-induced. This is one form of distortion. Some analog mixers perform better than others in this reguard, but they all add some.
2. Distortion. Besides all the distortions aded in the rest of th analog path, there is additional distortion of the signal as it's modulated by noise, the other signals (pure mixing is impossible to achieve in the analog world) and buss leakage (signals on another summing buss can and do leak into all the other busses, and vis-versa)
3. Phase distortion. It's analog. Try this experiment. Apply a signal to an input channel, reverse the phase of said signal and apply it to another input channel. Combine the input channels at exactly the same level and listen to the -not-pretty- results (there will be artifacts, because of phase errors introduced in the analog world.
repeat said experiment in a digital mixer. *hear* the difference. If your ears like the sound of the analog mixer better, perhaps you need to smear your signals before the mix process. I might suggest grouping simular signals to a 'verb, or add some additional group-delay (very short delay) or even a tiny bit of flanging to get the effect of what happens in an analog mix bus.
I use those types of effects when I'm mixing in the digital world, so this isn't something I'm not familiar with. I'm just suggesting that these introduced 'errors' are under your control in a digital mixer, but you are at the mercy of whatever analog mixer you use.
This isn't a rant, just trying to help you understand what is happening in a analog mixer that is 'missing' from the digital one.

[snareSpanker]

One of the better comments I ever read on kvr was "If all of this classic gear is so good why do all those old records sound so shit?" Hand's up who said it. It should be engraved above the door.

It was me! (I wish)

[Kebmaster]

Clearly some people dont understand what I am talking about or are clearly ignorant and inexperienced.

[nuffink]

Clearly some people dont understand what I am talking about or are clearly ignorant and inexperienced.

Clearly not. Who the f**k are you again?

[nuffink]

Got it. You're the knob who wrote...

Remember, another heavily overlooked factor is cables and leads. More expensive cable tends to add less colour than cheap cable. Cheaper cable tends to sound more muddy and bass heavy.

... and was, rightly, treated like the nutter on the bus. Even by the other nutters.