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
Xyz_39808 has quit [Ping timeout: 245 seconds]
<andlabs>
The original HardSID (1999) was a card for the ISA bus (instantaneously anachronizing the item),
<andlabs>
wikipedia throwing some shade there
<cr1901_modern>
Considering it needs 12v tho, that was probably the easiest way to interface one to a PC
<cr1901_modern>
To do e.g. a USB MIDI version you'll need a boost converter- actually I'm not sure I would try to bus-power a SID
andlabs has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
andlabs has joined ##yamahasynths
andlabs has quit [Ping timeout: 268 seconds]
andlabs has joined ##yamahasynths
<TD-Linux>
I am unconvinced that buming flac chunk sizes greater than 4GB will do anything for compression of the resulting file
<TD-Linux>
*bumping
<TD-Linux>
flac uses LPC predictors which use << 4GB of history
<TD-Linux>
actually no blocks are limited to 64KB. so I don't have any idea what the 4GB limit is about
<andlabs>
again, why does it have to be flac that you use?
<andlabs>
I suggested alac but that's not the only other lossless encoding out there
<TD-Linux>
alac is flac but worse
<TD-Linux>
you could use anything
<TD-Linux>
though the real deal is that any lossless encoding is going to peak at about 50% compression
<whitequark>
wtf is an audio dvd
<TD-Linux>
an audio cd but with drm
<TD-Linux>
(it's actually totally different than audio cd coding, and wildly unpopular)
<Lord_Nightmare>
48khz, and handles 5.1 or 7.1 audio
<Lord_Nightmare>
i think 5.1
<Lord_Nightmare>
one of the weird al cds was released in simultaneous CD on one side and audio-dvd on the other
<Lord_Nightmare>
the latter mixed in 5.1
<Lord_Nightmare>
interestingly it was set up to play the audio dvd stuff in a karaoke mode from a dvd menu, so maybe it abused video dvd mode rather than audio? I don't know
<TD-Linux>
yes, video dvd audio is separate from audio dvd
<TD-Linux>
those combination cds/dvds are technically Totally Wrong™ because the plastic for the audio side is the wrong thickness
<andlabs>
the dark side of the moon immersion collection from 2012 comes with an audio dvd with the album encoded 5 different ways for some reasons
<andlabs>
it has a video part probably so that it can be played
<andlabs>
also
<andlabs>
WavPack is another lossless format
<andlabs>
I'm not sure if this is a bit-perfect preservation project but if it is then yes, the preservation should override the compression concerns
ej5 has quit [Read error: Connection reset by peer]
<Sarayan>
Hey wq, how is the die imaging setup going?
<whitequark>
i'm stuck in hk for time being, roommate is fetching the microscope
<whitequark>
so it's progressing.
<Sarayan>
oh, you're in hk? Cool ! (or maybe not, given what's going on there)
<whitequark>
the politics in hk is cool. i mean, prc sucks, hk people are good, etc
<whitequark>
but hk isn't good for my health sadly
<Sarayan>
I have access to my stuff now, when do you want me to try sending the votrax again?
<whitequark>
hmm I'm not sure how long I'll be in HK, let me think about it for a bit
<Sarayan>
no problem
<Sarayan>
I'll be gone for a little more than a week starting tomorrow, so if it's not on it will be at a minimum then :-)
<whitequark>
ack
<Sarayan>
I've added in the box a ym3438 and a zn security chip, the cp01
<Sarayan>
the latter is just for my curiosity, because we've seen it in another system where it works differently (the cat702, which the cp01 is one of)
l_oliveira has joined ##yamahasynths
Xyz_39808 has joined ##yamahasynths
Xyz_39808 has quit [Ping timeout: 245 seconds]
andlabs has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
andlabs has joined ##yamahasynths
andlabs has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
andlabs has joined ##yamahasynths
andlabs has quit [Ping timeout: 272 seconds]
Xyz_39808 has joined ##yamahasynths
andlabs has joined ##yamahasynths
Xyz39808 has joined ##yamahasynths
Xyz_39808 has quit [Ping timeout: 246 seconds]
andlabs has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
Xyz39808 has quit [Ping timeout: 245 seconds]
Xyz_39808 has joined ##yamahasynths
andlabs has joined ##yamahasynths
Stilett0-out is now known as Stilett0
<cr1901_modern>
Huh, nice we've broken 40 for the first time in months
<cr1901_modern>
l_oliveira: That Genesis w/ all the logic probes... why is it required to debug the Genesis side when the Sega CD side is failing?
<l_oliveira>
Well, the MD side boots everything
<l_oliveira>
the CD sub CPU has no ROM, it runs from RAM
<l_oliveira>
so at the bring up time the MD CPU loads a compressed chunk of data from it's own ROM and decompress into the sub CPU work area then resets it and have it run it's own BIOS
<cr1901_modern>
Is the MD CPU the one that's faulting when decompressing?
<l_oliveira>
the idea is learn why it is failing to boot
<l_oliveira>
it's not even starting correctly
<l_oliveira>
so the sub CPU never got to start properly. His idea originally was look what the sub CPU was (not) doing by connecting probes at the CD work RAM but I contacted him privately and told him to debug the main CPU first and after a back and forth privately he found out that the reads from the BIOS are corrupted. I suggested it could be a severed address trace.
<l_oliveira>
he then had the idea of connecting a everdrive with a BIOS loading from the cart slot (which made the BIOS work) but it would still crash
<l_oliveira>
I told him that likely it still fails because the BIOS is unable to access the ASIC registers, again due to the same fault which is preventing the original BIOS from reading back properly
<l_oliveira>
that's where we stopped, he said he will continue debugging during the week so keep watching
<cr1901_modern>
will keep watching
<cr1901_modern>
I'm very fascinated by the feasibility of hooking up an entire 68k bus to a logic analyzer and that there are old logic analyzers w/ 68k support built-in
<l_oliveira>
that's half way to ICE probes lol
<cr1901_modern>
Is that not what your friend was doing (logic analyzer)?
<l_oliveira>
have you seen pictures of the CPS2 devkit?
<cr1901_modern>
or was it an actual ICE?
<cr1901_modern>
with probes*
<l_oliveira>
well a ICE is different from that because the ICE is actually the CPU
<l_oliveira>
that is us just spying on what the CPU is doing
<cr1901_modern>
Never seen a CPS2 devkit. Like most of them that do anything more than "load your custom program", I expect it to be horrifying
<l_oliveira>
it has a socket for connecting a ICE probe
<cr1901_modern>
bottom left?
<cr1901_modern>
right*
<l_oliveira>
24bit dips
<l_oliveira>
mame had them implemented as a kludge
<l_oliveira>
XMEN COTA reads that and freezes even on retail boards
<l_oliveira>
real CPS2 reads 0xFFFF on open bus
<l_oliveira>
mame reads 0x0000
<cr1901_modern>
so... patch it :P?
<l_oliveira>
and that is why they need the kludge
<l_oliveira>
the game thinks someone did fling the freeze debug dip
<cr1901_modern>
fling the freeze?
<l_oliveira>
'fling' as flip
<l_oliveira>
they're the these old style flip switches
<cr1901_modern>
what does "freeze debug" do then? Does it stop the whole board to keep it in sync with the ICE?
<l_oliveira>
it's like the neogeo freeze, makes the game stop processing (does not stop the CPU, just make the game stop processing frames)
<andlabs>
[15:26:58] <+andlabs>lol Microsoft's ISOs for Visual Studio 2015 Pro and Enterprise do not have a valid Dotfuscator specification
<andlabs>
[15:29:24] <+andlabs>either that, or it's just the fact I'm trying to install over VS2013
<andlabs>
[15:29:37] <+andlabs>I'll need to get MSDN ISOs instead of just using random go.microsoft.com links though
<andlabs>
[15:29:42] <+andlabs>they ARE from microsoft, though...
Xyz_39808 has quit [Ping timeout: 272 seconds]
<KitsuWhooa>
I installed 2015 on a VM that had 2013 on it
<KitsuWhooa>
It would refuse to load Visual C++
<KitsuWhooa>
or was it 2012
<KitsuWhooa>
I gave up in the end and installed 2017 instead of 2015
<KitsuWhooa>
what a nightmare
<ValleyBell>
VS2010 and VS2017 work well here.
<ValleyBell>
At work I have 2015 and 2017 on the same system - works fine as well.
<andlabs>
oh this project was for 2017 instead of 2015 heh
<andlabs>
I should probably get that while it's still a thing
<andlabs>
but IIRC they don't provide offline installers anymore
<andlabs>
restoring nuget packages is taking forever
<andlabs>
yep no it depends on C++17 crap
<andlabs>
heh
<andlabs>
er
<andlabs>
C#17 or whatever it's called
<andlabs>
hmmm, the VS2017 web installer isn't working smoothly either so now I'm wondering if it's another IPv6 issue
andlabs|2 has joined ##yamahasynths
<andlabs|2>
yep - IPv6 issue
<andlabs|2>
good going Microsoft
<andlabs|2>
not sure if that fixes 2015 but I imagine it would (for an online install)
andlabs has quit [Ping timeout: 252 seconds]
andlabs|2 is now known as andlabs
<andlabs>
did that messag ecome through
<Lord_Nightmare>
andlabs: I just asked the domesday guys and tl/dr: they worked around the flac issues by using ffmpeg/libav directly somehow, and ffmpeg uses an ogg container around flac chunks and works around the issue by itself
<Lord_Nightmare>
so they are using flac
<andlabs>
weird o.o ok
<andlabs>
I hope they're using ffmpeg and not libav; libav has rightfully died
<emily>
(meanwhile in another channel) I hope they're using libav and not ffmpeg; ffmpeg has rightfully died
<Foone>
I had an interesting conversation with one of the domesday guys at VCF:W, apparently they support "basically all laserdisc players" and not just a few specific models like I'd thought
<Foone>
and also they think their method should work on CED players but haven't yet been able to test
<Foone>
so I'm hoping I can get mine in working order and then helping them out with directly ripping CEDs
<Foone>
(I have two players, both broken in different ways)
<andlabs>
I hope that was sarcasm
<andlabs>
by 2017 people started realizing that ffmpeg was miles ahead of libav in terms of bug fixes and supported features and everyone realized that the stupid politics nonsense was hurting open source, and when debian and ubuntu both flipped back to ffmpeg it was over
<andlabs>
oh! *snaps figner* unless you're gstreamer, in which case your ffmpeg branch is still called gst-libav for some dumb reason
<andlabs>
in fact I was told it may not work with ffmpeg??? I don't remember
<Ultrasauce>
gst-libav uses ffmpeg
<Ultrasauce>
i think it's just called that because libavcodec
<Ultrasauce>
which is the library provided by both the ffmpeg and libav projects
<Ultrasauce>
this is not confusing at all
<Lord_Nightmare>
andlabs: they're using libav because they're using ubuntu 18.04, but they said they'd try it with mainline ffmpeg at some point
<andlabs>
ok
<emily>
I mean, the "stupid politics nonsense" was pretty bipartisan in this case.
<emily>
sometimes free software maintainers are childish.
<cr1901_modern>
what happened to ffmpeg? Did they have a suckless moment?
<emily>
(the name "libav" was probably picked to be as confusing as possible, re libavcodec)
<cr1901_modern>
"moment"*
<emily>
cr1901_modern: a section of the devteam was unhappy and tried to execute a coup
<emily>
their control of the domain got revoked, so they forked instead
<emily>
now everybody hates each other forever or whatever
<cr1901_modern>
Is libav the fork?
<emily>
yeah
<Ultrasauce>
the worst part was some distros tried to support both
<TD-Linux>
but whenever I try to play anything, it drops the needle for about 10 seconds but then retracts. presumably something in the amplifier section makes it not detect the carrier, but I've REd a large chunk and have no idea what
<andlabs>
anyway all this is depressing
<andlabs>
so eh
<andlabs>
at least the users clearly won in the end
<andlabs>
get the version with all the bugfixes
<Foone>
TD-Linux: fun.
<andlabs>
comapred to glibc, which went nowhere until ulrich drepper quit his job
<cr1901_modern>
It's a pity his DRAM articles are still basically the best introduction to that whole fiasco
<cr1901_modern>
Not a pleasant person at all
<TD-Linux>
for more fun there's virtually no VHD documentation whatsoever and I've yet to find a schematic of a single player
<andlabs>
techmoan =P
<emily>
I remember debian being on eglibc for years
<emily>
tbh it would be nice if glibc had just died and musl grew to replace it or something
<andlabs>
that would be nice
<andlabs>
I can tell you from experience that is unrealistic
<emily>
shrug
<emily>
musl could be a lot more compatible with glibc if it really wanted to
<emily>
(it probably doesn't want to though)
<TD-Linux>
I can't see the main VHD carrier signal on my scope at all, but it could just because the needle is sliding all over
<TD-Linux>
strfry or bust
<andlabs>
I take it by VHD you don't mean VHD the format created by JVC at this point
<TD-Linux>
I do mean that one
<andlabs>
ok
<andlabs>
ask Techmoan for pictures of the machines he had to repair? =P
<TD-Linux>
doesn't really help. I have pictures of my machine :)
<TD-Linux>
he never really repaired them. he bought more until one worked
<cr1901_modern>
Wish I had that "disposeable" income lol
<andlabs>
he probably does too =P
andlabs has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]