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
cr1901_modern1 has quit [Quit: Leaving.]
cr1901_modern has joined ##yamahasynths
futarisIRCcloud has quit [Quit: Connection closed for inactivity]
futarisIRCcloud has joined ##yamahasynths
andlabs has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
andlabs has joined ##yamahasynths
<whitequark>
cr1901_modern: what *is* in the floppy drive chip?
<whitequark>
i tried to take a DD floppy and tape the density window so it'be like SD and read it
<whitequark>
i expected the rotation speed to change
<whitequark>
instead it works at the same 300 rpm but the data pattern is... completely different
l_oliveira has quit [Quit: ChatZilla 0.9.92-rdmsoft [XULRunner 35.0.1/20150122214805]]
andlabs has quit [Ping timeout: 252 seconds]
andlabs has joined ##yamahasynths
<protosphere>
whitequark: I think SD floppies use FM rather than MFM? At least that’s the case with the 8” ones I have
<futarisIRCcloud>
All the PC formats were MFM based. Some earlier machines used GCR or FM.
<whitequark>
i meant take a HD floppy and tape the window so it'd be like DD
<whitequark>
but my brain is fried
<protosphere>
Ahh
<whitequark>
anyway i'm thinking about recording at a much higher density
<whitequark>
i'm limited by read channel bandwidth
<whitequark>
but only in frequency domain
<whitequark>
in phase domain i'm only limited by motor speed variation, write precompensation flexibility, and sample rate
<whitequark>
so of course my PLL needs to be able to train to the incoming sequence, but i think RLL is unnecessarily limiting
<whitequark>
MFM is basically 2-PM
<whitequark>
well, 3-PM
<whitequark>
i could make that any number i want so long as it doesn't exceed jitter imposed by the medium and mechanics
<TD-Linux>
8b/10b?
<whitequark>
not actually the best you can do, i think!
<whitequark>
think about this
<whitequark>
you could use 8-PM by requiring that each symbol spans between 8 UI and 16 UI
<whitequark>
and still be within the spec of a regular HD floppy
<whitequark>
now the problem is how the fuck do you train a PLL on this?
<whitequark>
and the answer is i think twofold
<whitequark>
first, you use a predictable fixed sequence that gently nudges the PLL into the right frequency, similar to gap/sync bytes
<whitequark>
second, you use a LFSR to ensure your data is normally distributed and the PLL wouldn't have a reason to slip into false lock
<whitequark>
whitening, essentially
<whitequark>
or i mean you could use AES, that actually works a bit better
<TD-Linux>
I feel like you could design your code to already have that property
<whitequark>
you mean like 64b/66b?
<whitequark>
that uses an LFSR
<TD-Linux>
yeah. or like CDMA
<whitequark>
azonenberg did suggest 64b66b
<whitequark>
but i think i can do better
<TD-Linux>
I think using erasure coding would also help a lot
<whitequark>
that's the next layer
<TD-Linux>
ya
<whitequark>
without erasure coding i'm not sure if you could go beyond like... 5-PM?
<TD-Linux>
I have no idea what the DD/HD line does on a floppy drive :/
<whitequark>
@brouhaha on twitter suggested it changes read channel bandpass filter
<whitequark>
which looks extremely like what i see on the histogram
<TD-Linux>
virtually no fdds change speed, with the exception of 3.5" trimode PC98 drives, and a lot of USB fdds
<TD-Linux>
no idea why the trimode drives didn't implement it with two different bit clocks rather than changing drive speed
<whitequark>
mhm
<Foone_>
yay trimodes
<Foone_>
I have one trimode and it seems to suck at reading any disks
<whitequark>
Foone_: what do you think about my phase modulation idea
<Foone_>
it's not really my area but it sounds like it could be a good idea. I was talking to the guy who did applesauce and apparently they've been able to read HD disks from a DD drive by doing something similar, I think?
<whitequark>
huh.
<whitequark>
really?
<Foone_>
I think so? it was only a brief talk at the VCF:PNW, so it's possible they're doing something different.
futarisIRCcloud has quit [Quit: Connection closed for inactivity]
ej5 has quit [Quit: Leaving]
andlabs has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
andlabs has joined ##yamahasynths
andlabs has quit [Ping timeout: 255 seconds]
andlabs has joined ##yamahasynths
andlabs has quit [Ping timeout: 276 seconds]
andlabs has joined ##yamahasynths
Sarayan has quit [Ping timeout: 258 seconds]
_whitelogger has joined ##yamahasynths
futarisIRCcloud has joined ##yamahasynths
andlabs has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]