AW: [Digital_Hell] DX-7 Noise

[jhaible]

BJ wrote:

> Well, thats what i did belive you was saying, and still i dont agree
> into this because the FM  part is dead silent.Actually the TGG77/99
> is wery different from the TX802, different FM ASIC, different DACs
> different FM implementation of the FM synthesis.Dont forget that
> the TX802 is in fact one generation older then the DX7II and the DX7IIFD
> machines, you can definitively hear the difference between those DX FM
> generation the DX7FD are slightely less noisier then the TX802.
> 
> There are differences between the first generation of Dx7 to
> 10 bit DAC, 12 bit DAC, 14 bit ,16 bit DAC's was used between the
> models.
> (she's a model and are looking good, deooo, de, deo ,deoo , deoo de
> deooooow!)
> 
> Newertheless, i think you shall go out and buy your self a TG77
> its dead cheap these days, here i often se them at 3500skr=777DM
> some times even cheaper.And then you buy the RP from us so you can
> build your self a custom Remote Programmer(RP for short) for the TG77 
> using its 128 pot AD channels and 32 buttoms.(when its get finished)

Michael Bacich wrote:

>Oh yeah, I forgot to mention this - There were completely new tone
generation
>chipsets for each of the three series of Yamaha six-operator FM
instruments
>(series 1: DX7, TX7, TX816; series 2: DX7II, DX7s, TX802; series 3: sy77,
>SY99, TG77).

Ok, I am ready to believe every word you say, as my own "experience" with
the SY / TG synths is next to zero. What I heard when I considered
upgrading
to a SY77 may as well just have been badly programmed sounds. I was 
disappointed, and that stopped me from doing any further invertigations.
(Which might have been an error.)

Having sad that, now to a more precise question.

I'm aware that the *output* resolution (i.e. DAC) was increased from
earlier
to later models. (Btw, 10 bit for DX-7 Mk 1, does that mean 8-bit DAC chip 
+ 2 extra bits, or 10-bit DAC + 2 extra bits = 12 bits ?)
That's why I programmed as few dynamics as possible into the level of the
*carrier* for many sounds. That, and carefully adjusting the trimmer for
the
"extra bits" helped a lot to make sounds cleaner. Later models (even the
TX-802) would surely be more tolerant about carrier level dynamics.

But even then, there was the noise from *modulation* the carrier. As this
is just phase modulation, it would't affect the output amplitude (of the
carrier), and thus wouldn't change the amount of quantisation by the DAC.
But the mere phase modulation process introduced some extra noise,
so I came to believe that this was due to a limed resolution of the 
*modulation process*. Maybe adress length of the sine tables, or whatever.
never dug deeper into that.
So for me the big question is: Did they increase this *internal* resolution
from the DX to the SY / TG models ??

I really don't know, and I may be on the wrong track here, too, but now I'd
like to know.

JH.