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
<andlabs> ok, I'll instead assume that the software does not read the jumpers and interpolates the tunings for a 3.5mhz clock
<andlabs> see my comment
<andlabs> when he plays the space harrier vgm at 4Mhz it sounds very wrong
<andlabs> actually I'm not sure if it's even right at 3.5Mhz compared to vgmrips.net's player but then again either could be wrong
<andlabs> afk
<cr1901_modern> At least they actually route the IRQ line to the PC
_whitelogger has joined ##yamahasynths
Ultrasauce has quit [*.net *.split]
Ultrasauce has joined ##yamahasynths
nukeykt has joined ##yamahasynths
andlabs has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
ej5 has quit [Read error: Connection reset by peer]
andlabs has joined ##yamahasynths
andlabs has quit [Remote host closed the connection]
andlabs has joined ##yamahasynths
andlabs has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
andlabs has joined ##yamahasynths
<nukeykt> Spent few days on recreating ym2413 patches using extensive hardware tests
<kode54> andlabs: chip clock rates change the tuning
<kode54> you have to adjust the frequency values to account for the clock rate, or it will sound wrong
<kode54> the VGM players don't adjust any pitch rates from the logs, and assume the file was logged and has the correct clock rate in its header
<kode54> of course it won't play right if the chip is emulated at the wrong clock, or if you feed the log to a real chip with a different clock
<Xyz_39809> why not just use plogue's patch set?
<Lord_Nightmare> nukeykt: did that involve the debug mode and watching the operator and envelope regs for those two channels that they're visible on?
<Xyz_39809> I don't understand the six patches beneath the 15 default ones. There's only 5 ROM drums
<Lord_Nightmare> they're 3 drum patches but interleaved weirdly iirc
<nukeykt> yeah, i've used both 0xF register and D0/D1 data to check patches, i have two YM2413 running on parrallel, on first i've set original preset on second recreated and then compared data coming from D0/D1 pins
<andlabs> kode54: that would be correct if the inversion was the other way around
<andlabs> if it sounded wrong at 3.5mhz and right at 4mhz
<andlabs> but that's not what this video shows
<nukeykt> 0xF register layout is same as VRC7
<andlabs> and also the YM2151 is special and uses MIDI-like frequency schemes
<kode54> depends on the machine being logged, and whether the log was written correctly
<andlabs> so hmmm
<andlabs> might want to throw that up there I guess?
<andlabs> I'm guessing the machine was not logged correctly
<kode54> the arcade machine spec may not know the correct frequency
<kode54> the log has whatever the game's driver wrote to the chip
<andlabs> this is the x68000 version
<kode54> hmm
<andlabs> the arcade version uses a YM2203 (OPN-type chip)
<Xyz_39809> so is it like op13 op14 op15 op16 op17 op18? and then they're interleaved the way rhythm modes uses those operators? (don't have my notes with me)
<andlabs> afk again
<kode54> maybe the OPM emulators are wrong
<kode54> I wasn't aware any of the FM chips had "MIDI-like" tunings
<kode54> I thought they just used clock counts for pitches, and those varied with the clock of the chip
<kode54> or with a divisor of that clock
<Lord_Nightmare> i'm comparing these patches to the plogue ones
<Lord_Nightmare> so far the only differences are attack, decay and release of the modulator waveform, for the first 4 patches
<Lord_Nightmare> everything else is the same
<Lord_Nightmare> patch 5 plogue did not have the ks rate bit set for the modulator
<andlabs> yeah I don't know what's going on either
<andlabs> that's why I'm bringing it up
<andlabs> good night
<ValleyBell> kode54: I can confirm that the OPM has registers set as "note / octave" and "note fraction".
<ValleyBell> This is unlik all other OPx chips.
<cr1901_modern> ValleyBell: Using "note / octave" and "note fraction" for a single chip doesn't make much sense to me. I'm guessing OPS (whose reg set is not publicly known yet) also used it, since 2151 is a descendant of the multi-chip FM synths.
<cr1901_modern> Then Yamaha changed their mind and tried other approaches
<cr1901_modern> nukeykt: How'd you get two YM2413 in sync, including their global "cycle" counter and the register pipeline?
<nukeykt> IC resets cycle counter
<cr1901_modern> oh interesting... I didn't think it did based on oscilloscope traces Nemesis captured showing two runs of the same test ROM and same reg values starting on two different offsets of the decay table
<cr1901_modern> But I could be misremembering... I don't even know if Genesis circuitry routes /IC
<cr1901_modern> nukeykt: Does the decay table ever attenuate by more than 1 in a single sample, or is maximum attentuation rate acheived by "attenuate the amplitude by 1 every sample"?
<nukeykt> I think it does 4 and 8 steps on envelope rate 63
nukeykt has quit [Remote host closed the connection]
cr1901_modern has quit [Quit: Leaving.]
cr1901_modern has joined ##yamahasynths
<cr1901_modern> colorburst is the official "correct" frequency for 2151 according to the datasheet, but 4MHz is also somewhat common. x68k uses 4MHz (probably at 40% duty cycle on the early models), and I think Atari System I does so as well
TD-Linux has quit [*.net *.split]
<Lord_Nightmare> ym2164 on sfg-05 i think also uses 3.579545mhz
<Lord_Nightmare> but since the timer b runs at half the rate (well, it runs at the same rate but has one extra bit so it effectively takes twice as long?)
TD-Linux has joined ##yamahasynths
<Lord_Nightmare> it can be used for more accurate midi tempo
ej5 has joined ##yamahasynths