ChanServ changed the topic of ##yamahasynths to: Channel dedicated to questions and discussion of Yamaha FM Synthesizer internals and corresponding REing. Discussion of synthesis methods similar to the Yamaha line of chips, Sound Blasters + clones, PCM chips like RF5C68, and CD theory of operation are also on-topic. Channel logs: https://freenode.irclog.whitequark.org/~h~yamahasynths
<Lord_Nightmare>
Foone: The Impossible Mission voice, I believe but am not sure that that voice is of Dr. Forrest Mozer himself (and if it isn't, he encoded it at ESS, the company He and Fred Chan founded. They later became more well known for sound card ICs and nowadays DAC chips.)
<Lord_Nightmare>
and yes, Electronic Speech Systems and ESS (a'la audiodrive etc) are the same company.
<Lord_Nightmare>
Forrest mozer made a cottage industry of manually encoding speech segments for the TSI S14001a chip between 1975 and the early 80s (i.e. the arcade game berzerk, fidelity chess games, etc)
<Lord_Nightmare>
and after the licensing of that technology to National Semiconductor for the digitalker presmulably fell through around 1983ish, he and fred chan founded their own company to make a decoder for his patented 'mozer speech' stuff in software
<Lord_Nightmare>
that company was electronic speech systems
<Lord_Nightmare>
foone: you should ask plgdavid about this, he knows much more and has talked to Dr. Mozer
<Lord_Nightmare>
Or at least, what I said above is how I understand the story, which may very well be wrong
<cr1901_modern>
I've been trying to get him to come back to IRC since I the day this channel started lmao
_whitelogger has joined ##yamahasynths
_whitelogger has joined ##yamahasynths
<andlabs>
kode54: the mid-90s yamaha PCM synth chip that was used in their XG devices
<andlabs>
what vgmrips calls Multi-PCM is aoctually a GEW chip
l_oliveira has quit [Quit: ChatZilla 0.9.92-rdmsoft [XULRunner 35.0.1/20150122214805]]
<kode54>
ah
<kode54>
I suggest mentioning that to ValleyBell
<andlabs>
pretty sure ValleyBell already knows
<kode54>
aha
<Lord_Nightmare>
I think GEW8 was licensed by sega and renamed 'multipcm' and used on sega model1 and 2 (but not 2a which used the saturn scsp chip?) and system32 I think
<Lord_Nightmare>
while capable of using a secondary dsp for reverb, it didnt do this on sega hardware
<Sarayan>
I don't think I've ever seen a yamaha external effects chip on arcade boards
<Sarayan>
in general I've seen dsps here and there (the dasp57002 on gx and some zn, the ensoniq one...) and integrated stuff (swp30, aica)
<Sarayan>
arguably, qsound too
<superctr>
did Sega ever use the name "MultiPCM"
<superctr>
I always thought that was something that MAMEdev came up with
<superctr>
what makes the chip interesting is that it is silkscreened with Sega part numbers above a Yamaha internal one (FA1005 or something like that)
<superctr>
Sega using their part numbers on "generic" chips isn't too uncommon (It happened with the RF5C164 as well)
<superctr>
The Yamaha part number is very close the one marked on the YMF262 die, which is what made me propose that it was identical to the YMW258
<superctr>
when the TG100 ROM was dumped, I was able to confirm that the registers were compatible, and I believe someone in MAMEdev confimed the pinout as well
<superctr>
YMF262 is FA1006 lol
<Sarayan>
Yeah, I confirmed the pinout, it's exactly the same
<superctr>
I think Sega had a deal with Yamaha while the SCSP was still in development, where Sega bought left over stock YMW258 wafers and had them packaged with a custom part number etc
<superctr>
Probably the OPL4 was also under development at that time, since that would otherwise be a more logical choice (it has built-in timers. Sega used YM3438s on the Model1/2 sound baords to provide timing)
<Lord_Nightmare>
whitequark: you should ask kevtris abouyt 8051 math libs, iirc he dumped an 8051 or 8751 which contained a very optimized math library on it
<whitequark>
Lord_Nightmare: hmmm so my idea was to make provably optimal functions
<whitequark>
which is a bit different
<andlabs>
[07:08:42] <Sarayan>I don't think I've ever seen a yamaha external effects chip on arcade boards
<andlabs>
did yamaha ever sell external effects chips?
<andlabs>
or rather
<andlabs>
did anyone, for that matter?
<whitequark>
Lord_Nightmare: re chips: they're still in the box you sent me them in
<whitequark>
i hven't been able to do much with hardware in several months
<whitequark>
things will get better after uh
<whitequark>
who the fuck knows anymore?
<cr1901_modern>
"things will get better after uh who the fuck knows anymore?" That is definitely a recurring theme in general. :(
<superctr>
<andlabs> did yamaha ever sell external effects chips? <- they must have done, since you can find datasheets for some of them
<andlabs>
huh
<andlabs>
ok
<Lord_Nightmare>
whitequark: once the plague dies down, and assuming the world doesn't continue to head into anarchy, perhaps
<whitequark>
honestly, anarchy would be great
<whitequark>
it's totalitarianism that worries me
<whitequark>
e.g. hungary
<cr1901_modern>
Lord_Nightmare: Do not think of anarchy as chaos. That's all I'll say :P.
<whitequark>
can meet you at "continue straight to hell" :p
<cr1901_modern>
I've noticed I've been hyperfocusing lately to avoid dealing w/ existential dread. I acknowledge this is a problem. But I don't intend to change course b/c it's all I can do right now.
<whitequark>
so, i'm rederiving GCR encoding with my line code generator
<cr1901_modern>
And (32 choose (16 choose 2)) = 0... so fun fact, choosing is not associative
<cr1901_modern>
good to know
futarisIRCcloud has joined ##yamahasynths
<Sarayan>
cr1901_modern: Well, choosing 120 out of 32 is kinda difficult
<cr1901_modern>
Sarayan: Just apply Banach-Tarski to create additional objects on the fly
<cr1901_modern>
Sarayan: Btw, did you see my 68k question a few days ago?
Xyz_39808 has joined ##yamahasynths
<Stilett0>
<Sarayan> I don't think I've ever seen a yamaha external effects chip on arcade boards
<Stilett0>
I think someone had asked me to find a datasheet, something in the YSS series. Might have been someone from VOGONS circa 2002-2004 tho
<Lord_Nightmare>
Stilett0: didn't we find the datahseet or at least the chip type that creative labs used for the "ASP" on later sb16 and awe32 cards?
<Lord_Nightmare>
that would be nice to emulate, it was used by impulse tracker to do some fancy FIR(?) filtering if it was found
<cr1901_modern>
ej5: Unless the flash memory is storing the analog samples as, say, a charge on a capacitor that leaks veeeeeeery slowly, wouldn't the flash memory store the sample digitally :P.
<cr1901_modern>
Beautiful thread, btw
futarisIRCcloud has quit [Quit: Connection closed for inactivity]
<ej5>
thanks! it does store the analog voltage directly
<ej5>
but yes, it does slowly leak away over time.
<cr1901_modern>
._. ... how?!
<cr1901_modern>
and that's cool!
<andlabs>
well I guess I bought into the flux engine ecosystem for now
<andlabs>
mostly because it's cheaper and more immediate
<andlabs>
uhhh
<andlabs>
so I guess I'm asking this question for the last time now
<andlabs>
which 5.25" floppy drive should I go for
<andlabs>
TEAC FD-55GFR?
<andlabs>
I'll definitely look into getting glasgow later
<cr1901_modern>
FD-55GFR is the quintessential 5.25"
<cr1901_modern>
FD-55GFR is a high density floppy drive. Use with caution when writing double density disks.
<whitequark>
that DDF-52D is brutal, so much DIP
<andlabs>
I do need to look into doing bit perfect dumps of contact disks and maybe also laserdiscs soon and for that I'm hoping glasgow can work wonders
<andlabs>
of course first I need to plan my budget to actually afford to buy th sealed laserddiscs I'm after lol
<andlabs>
when they come up, I mean
<whitequark>
laserdiscs are analog though aren't they?
<whitequark>
so you'll need an AFE for Glasgow
<cr1901_modern>
whitequark: I would recommend DFE222A10A because I have a similar drive, and I've no complaints with it for the past several years. FD-55GFR would be a good secondary to test high density disks (and to compare disks written from the DFE)
<cr1901_modern>
My DFE is specifically used for experiments, btw. I.e. not deployed in a system
<whitequark>
can they both read and write high density disks?
<whitequark>
i don't know much about 5.25"
<cr1901_modern>
I don't know. My DFC222B02A _does not_ read/write high density disks.
<whitequark>
hmmm
<cr1901_modern>
(Oh, and it can prob be used as a sink strength test for Glasgow ;)... since I think the floppy drivers will source 64mA per pin)
<cr1901_modern>
(for the write signals)
<andlabs>
they are analog but they are also quantized and I'm not quite sure if that quantization brings the range of values in scale for digital
<andlabs>
there was that channel on freenode doing laserdisc stuff
<cr1901_modern>
#domesday86
<whitequark>
andlabs: oh, right
<andlabs>
I also don't know how Laserdiscs are read
<andlabs>
CDs sotre 0s and 1s as the transition between pits and lands
<whitequark>
glasgow can sample at about 200 Mbps if you *really* push it
<andlabs>
are Laserdsics using the actual lengths of pits and lands? or something else
<cr1901_modern>
I wonder if it's like a Diamond Disc, where groove depth == analog value
<whitequark>
i remember how laserdiscs work now, i think
<cr1901_modern>
(not the same as Victor records, where groove displacement from a neutral point == analog value)
<whitequark>
the signal is on-off but it's like sigma delta modulated, in modern terms
<andlabs>
ohh
<whitequark>
and if you filter it you get a video signal
<andlabs>
However, while the encoding is of a binary nature, the information is encoded as analog pulse-width modulation with a 50% duty cycle, where the information is contained in the lengths and spacing of the pits. In true digital media the pits, or their edges, directly represent 1s and 0s of a binary digital information stream.[20]
<andlabs>
wikipedia explained it the whole time it was just buried
<whitequark>
so, NTSC is what, 4 MHz bandwidth?
<cr1901_modern>
4.5 or 6 if you include sound
<whitequark>
if you're happy with 48 MHz sample rate that is trivial
<whitequark>
on glagow
<whitequark>
96 with minimal effort
<whitequark>
192 with some effort
<andlabs>
so yes, the "quantization" is just nyquist sampling
<andlabs>
so yeah, I still need a good 5.25 drive that can READ both 40 and 80 TPS disks
<andlabs>
not write, I have no plan on writing anything
<andlabs>
and an enclosure that can fit the drives, the psoc, and the picopsu
<andlabs>
or rather, two enclosures
<balrog>
whitequark: any chance you might make an Apple Disk II interface for the glasgow?
<balrog>
using PC floppy drives for Apple II disks doesn't exactly work well
<whitequark>
hm
<whitequark>
balrog: (a) why don't PC floppy drives work? different AFE needed?
<whitequark>
(b) are you prepared to pay the cost of an Apple Disk II ?
<whitequark>
or i mean i could probably work remotely too
<whitequark>
so if you gave me ssh that'd also work
<whitequark>
cr1901_modern: re: 64 mA drive strength
<whitequark>
it'll only sink that if i have a super strong pullup on glasgow side, right?
<whitequark>
i.e. that's the output transistor strength\
<whitequark>
since the pullups are on the motherboard/glasgow side of interface, this is immaterial for it
<cr1901_modern>
The floppy provides the pullups for the write signals
<cr1901_modern>
so it could end up sinking 64mA in the worst case on the Glasgow side
<whitequark>
that is a problem then, glasgow is rated for sourcing only 32 mA i think
<cr1901_modern>
(and yes, they are strong pullups. If your pullups on the Glasgow side are weak then I don't think they really factor in to how much current the Glasgow sinks)
<whitequark>
let me check
<cr1901_modern>
32mA _might_ work. In fact, I would guess it probably would work (seem to vaguely recall doing it in a pinch)
<cr1901_modern>
But if you have problems, that is the first place to look.
<whitequark>
yes, glasgow is rated for 32 mA
<whitequark>
64 mA was just absolutely not something i could fit in there