[Frank Stearns]

Mike Rivers writes:

On 7/12/2014 3:38 PM, Frank Stearns wrote:
Judging by some of the
specs of these supposedly 24 bit consoles, there really didn't seem to be 24 bits
inside, but something less (18? 20?), with the extra bits perhaps being used for the
results of larger multiplies. Is that the case? Or did they just bit-shift and let
the LSBs fall off?

How are you judging? By the S/N or dynamic range specifications? When

DNR specs (-109) and for a digital console, lousy cross talk (-80, which I assume
must be happening in the analog front ends and D-A outputs).

Just guessing, but it seems that given two not so great choices, the one with the
smaller but untruncated word might be better sonically than fixed truncation steps,
especially if your input level was low (for whatever operator oversight -- perhaps
the reason for paying careful attention to gain structure).

I'm not sure what you're getting at here, but you really shouldn't
operate a digital mixer any differently than you'd operate an analog

Agreed.

Multiple operators use the old Yamaha M7CL in a theater where I do the occasional
live sound gig.

When I got there last year, I thought the system sounded pretty bad. As I grew to
know the system as a whole, it seemed that one key reason was that the lead operator
ran preamp gains very low, with the master faders at zero (plus input faders at -10
to -20). With those low preamp trims, the equivalent "0 VU" was around a -50 through
the console. Inputs barely flicker the -60 channel meter LED.

The whole system has excess gain for what we normally do, on the order of 30 dB. In
part this is because of the long-standing stupid practice of running power amp gains
wide open. I'm chipping away at this, and later next fall might convince the
political powers to trim those amps more rationally. (In spite of this, the system
has a low noise floor; my issue is with distorsion from running so "low in the
mud" through the console.)

One victory was to get the output bus trims down -10 dB. That helps a little, but my
preference is to run the masters at a -15 to a -20 and increase the head amp gains
by 10 to 20 dB. Input metering is then more rational. I also run channel faders
around 0, and then overall it sounds wayyyy better. I am assuming the master faders
are dithered but that dither likely does not occur elsewhere.

The sonic difference is noticeable; slowly but surely this is sinking in on other
operators. But in making a convincing case, I've been assuming that internally the
console is something less than 24 bit, with summing maxed at 24 bit. So the question
is what is the architecture? Given the age and relatively low price point, it's no
Euphonics with an 80 bit summing bus (IIRC) -- so Yamaha is doing many tricks to
pack a good punch for the buck on for-its-day limited silicon -- but you have to be
very careful about things like gain.

I suspect the other operator is happy with low levels because he doesn't hear any
hum or buzz, but does not yet fully grasp how much better the thing sounds with a
more rational grain structure while going through all the channel strip DSP.

So, my goal is to better understand the architecture in order to create more
convincing arguments. With some folks in the PA world, ears aren't enough. If
gearheads, they want gearhead arguments. (As the only recording guy to use that
system, I sometimes get teased for "hearing too much". It's good natured, but does
summarize what I'm pushing against.)

Anyway.

Frank
Mobile Audio

--