VOGONS


Reply 100 of 339, by Bancho

User metadata
Rank Oldbie
Rank
Oldbie

Hi Tiido - This is such a great project and i'm sorry your not able to see it further currently. When you get back to normal i'd love to pick one up especially for the Wavetable header!!

A quick question, i currently have a 719/704 card, would your software be compatible with it? This is my card

msn0EJml.jpg

Last edited by Bancho on 2018-08-11, 23:44. Edited 1 time in total.

Reply 101 of 339, by Tiido

User metadata
Rank l33t
Rank
l33t

Thänk you for the kind words ~

It'll work as a generic YMF719 card to SETYMF due to reasons mentioned in my last post, the YMF704 is wired in a way that it is a MIDI expansion running off MPU401 of YMF719 while also replacing OPL3 of YMF719. It isn't an additional MIDI device and OPL3 to the ones already in YMF719. Some options will be hidden that are specific to my card but overall it works and controls all the things that can be controlled.

T-04YBSC, a new YMF71x based sound card & Official VOGONS thread about it
Newly made 4MB 60ns 30pin SIMMs ~
mida sa loed ? nagunii aru ei saa 😜

Reply 103 of 339, by Tiido

User metadata
Rank l33t
Rank
l33t

Hahaha, jaa, I'm TmEE on Sega-16. I actually have another sound mod cooked up but I haven't finalized it (been like that for years now 🤣).

I have also updated SETYMF, there was a cosmetic error. Version number was not shown correctly in the program itself, only on startup banner thing.

T-04YBSC, a new YMF71x based sound card & Official VOGONS thread about it
Newly made 4MB 60ns 30pin SIMMs ~
mida sa loed ? nagunii aru ei saa 😜

Reply 104 of 339, by 640K!enough

User metadata
Rank Oldbie
Rank
Oldbie
Tiido wrote:

It is fully OPL4 compatible as ar as sound generation stuff goes. CPU can be turned off and you can directly access the registers and make sound that way but the sample ROM is completely different beast without any similarity to YRW801 that most YMF278 things use.

That was a bit of an understatement, actually. Looking at the chart in the datasheet, not only is the sample ROM smaller, but they have stuffed more instruments in. The YRW801 includes only the standard drum bank, while this appears to have the usual nine of them. I strongly suspect that they must have made some significant compromises to get everything into half the space, much like they did for the standard InterWave ROM. If that's all you've ever heard, it may not seem to be lacking, but when you compare one with the other, the differences can be striking. That was the situation when switching between the 1 MiB InterWave ROM and it's 4 MiB counterpart. They are "based on the same sample data", but you start to really notice the shortened loops and truncated percussion once you've heard the larger version.

This still bothers me: if Fat-Seal certification is supposed to mean that the volume balance and instrument characteristics will be similar across GM devices, why is there such a dramatic difference between certified instrument banks? If we take the Audiotrix Pro (the very first Fat-certified card) demo files (not the 44-voice ones) and play them on your card, a Roland and the Fat-certified InterWave bank, the sound will still be quite different, in terms of balance, etc. What, then, was the point of the whole Fat certification exercise?

Reply 105 of 339, by Tiido

User metadata
Rank l33t
Rank
l33t

There's only one drumkit, less than 128 unique melodic instruments and not many multisamples (most have 3 samples, only piano has one per all the octaves). Some instruments are made by playing two sounds at the same time (which will negatively impact polyphony) and some just apply different vibrato and tremolo. It sounds more noisy due to 8bit samples but there's no obvious short loops or other things along those lines, most samples have same sized loops as things in the more fancier things, nothing actually sounds outright bad here.

Fat certification, as much as I have heard of examples, means nothing much 🤣. Marketing thing it seems. I have plans to make a MIDI synth in future and there will be headaches around balancing things...

T-04YBSC, a new YMF71x based sound card & Official VOGONS thread about it
Newly made 4MB 60ns 30pin SIMMs ~
mida sa loed ? nagunii aru ei saa 😜

Reply 106 of 339, by 640K!enough

User metadata
Rank Oldbie
Rank
Oldbie
Tiido wrote:

There's only one drumkit,

Really? I figured that, since they bothered to include the chart below in the datasheet, they actually had all of those in the ROM.

YMF721 Percussion Chart.jpg
Filename
YMF721 Percussion Chart.jpg
File size
267.54 KiB
Views
1726 views
File license
Fair use/fair dealing exception
Tiido wrote:

Fat certification, as much as I have heard of examples, means nothing much 🤣. Marketing thing it seems.

I don't think it's just marketting, though that's probably a big part of it. I have read about "Fat Labs sessions", where a bunch of changes were suggested, and each time, testing had to be repeated. Adjustment of instrument parameters can get to be rather tedious business. Still, I would have expected the volume balance between certified banks to be closer than what I have observed.

Reply 107 of 339, by Tiido

User metadata
Rank l33t
Rank
l33t

That table is definitely lies, all the drumkits sound same, no sort of variation sounds and also no extra samples for the alternate sounds.

I suppose the tests were all subjective and not rigorous recording, comparing and matching etc. so some level of variation depending on the assessing guys state will happen but who knows. The variations are rather big...

T-04YBSC, a new YMF71x based sound card & Official VOGONS thread about it
Newly made 4MB 60ns 30pin SIMMs ~
mida sa loed ? nagunii aru ei saa 😜

Reply 108 of 339, by Tiido

User metadata
Rank l33t
Rank
l33t

SETYMF has been updated. It was blocking MPU-401 and some other addresses used by YMF721 on my card on other YMF71x cards where such resources are free. Now all IO locations are choosable on random YMF71x cards.

There's also a problem with Creative's PnP programs, they want real checksums and don't accept "don't care" checksums that I'm using in my PnP EEPROM, thus violating PnP spec a bit. I will look into generating proper checksums sometime later.

There was an update earlier that allows to write custom PnP ID to the card on random YMF71x cards when one decides to overwrite the EEPROM. More info in the readme.

T-04YBSC, a new YMF71x based sound card & Official VOGONS thread about it
Newly made 4MB 60ns 30pin SIMMs ~
mida sa loed ? nagunii aru ei saa 😜

Reply 109 of 339, by stamasd

User metadata
Rank l33t
Rank
l33t

I've looked through my pile of random stuff, and I happen to have a donor card with a YMF704C-S chip on it... it's not a full sound card, but a weird little add-on module that I remember pulling out of a Micron-branded PII computer in the early 2000s; the computer was going to be recycled at work and I wasn't allowed to take it all, but could pull some parts from it. It's a small board about 4x3cm with 2 blocks of 3x2 connectors at each end through it plugged in the motherboard. I stuck it in a drawer and it stayed there for the past 15 years. I doubt I'll ever find a motherboard to plug it in.

image.jpg

I/O, I/O,
It's off to disk I go,
With a bit and a byte
And a read and a write,
I/O, I/O

Reply 110 of 339, by Thermalwrong

User metadata
Rank Oldbie
Rank
Oldbie
stamasd wrote:

I've looked through my pile of random stuff, and I happen to have a donor card with a YMF704C-S chip on it... it's not a full sound card, but a weird little add-on module that I remember pulling out of a Micron-branded PII computer in the early 2000s; the computer was going to be recycled at work and I wasn't allowed to take it all, but could pull some parts from it. It's a small board about 4x3cm with 2 blocks of 3x2 connectors at each end through it plugged in the motherboard. I stuck it in a drawer and it stayed there for the past 15 years. I doubt I'll ever find a motherboard to plug it in.

Here's where it fits on the Intel AL440LX motherboard, found in the Dell XPS D233/D266/D300/D333 computers and other OEM PCs 😁 The header isn't present but that's definitely the right part from what I remember of the AL440LX manual, it's afaik the last and maybe the first to have that modular wavetable connector, for pairing with the onboard YMF715/YMF718 chip. It should be specific to the motherboards that Intel designed, which is many of them from the early ATX era. It's not on anything later because the later boards all have the YMF7x4 instead, which practically integrates the wavetable.

2018-09-10 19_22_17-IMG_9709crop.jpg
Filename
2018-09-10 19_22_17-IMG_9709crop.jpg
File size
672.6 KiB
Views
1634 views
File license
Fair use/fair dealing exception
Last edited by Thermalwrong on 2018-09-10, 23:01. Edited 1 time in total.

Reply 111 of 339, by stamasd

User metadata
Rank l33t
Rank
l33t
Thermalwrong wrote:

Here's where it fits on the Intel AL440LX motherboard, found in the Dell XPS D233/D266/D300/D333 computers and other OEM PCs 😁 The header isn't present but that's definitely the right part from what I remember of the AL440LX manual, it's afaik the last and maybe the first to have that modular wavetable connector, for pairing with the onboard YMF715/YMF718 chip. It should be specific to the motherboards that Intel designed, which is many of them from the early ATX era. It's not on anything later because the later boards all have the YMF7x4 instead, which practically integrates the wavetable.

Interesting, thanks for the info. I have no idea what kind of motherboard that system had but I do know it was a PII-266, and the case was marked "Micron Millenia II"
I tried looking its specs up a few years ago but came out empty.

I/O, I/O,
It's off to disk I go,
With a bit and a byte
And a read and a write,
I/O, I/O

Reply 112 of 339, by 640K!enough

User metadata
Rank Oldbie
Rank
Oldbie
Tiido wrote:

There's also a problem with Creative's PnP programs, they want real checksums and don't accept "don't care" checksums that I'm using in my PnP EEPROM, thus violating PnP spec a bit. I will look into generating proper checksums sometime later.

Which checksum is causing the problem; the one after the card identifiers (byte 9, I think), or the last byte after the configuration data? If it's the latter, then it's a Creative bug; that one is explicitly allowed to be zero. That doesn't appear to be the case for the former, though, and many utilities will complain and/or fail in that situation.

Reply 113 of 339, by Tiido

User metadata
Rank l33t
Rank
l33t

The latter, which is why I said the tools break PnP spec. If the first was wrong then even Windows will not accept the card and I handle that checksum byte properly.

As far as that small module goes, I have seen it referenced to some DELL computers and laptops. The chip is given 33.8688MHz clock, serial MIDI data and i2s DAC link.

T-04YBSC, a new YMF71x based sound card & Official VOGONS thread about it
Newly made 4MB 60ns 30pin SIMMs ~
mida sa loed ? nagunii aru ei saa 😜

Reply 114 of 339, by 640K!enough

User metadata
Rank Oldbie
Rank
Oldbie
Tiido wrote:

The latter, which is why I said the tools break PnP spec. If the first was wrong then even Windows will not accept the card and I handle that checksum byte properly.

If you already have the first checksum value calculated correctly, the good news is that the last checksum byte is easier to calculate. You'll be able to do that one in a few lines of code, even if it's only to work around Creative's bug (or ignorance of the standard).

Reply 115 of 339, by techweenie

User metadata
Rank Newbie
Rank
Newbie
Tiido wrote:
The question is how cheaply. I wouldn't be able to create a sample but I do have a drawing made with all the measurements : http […]
Show full quote

The question is how cheaply. I wouldn't be able to create a sample but I do have a drawing made with all the measurements :
Bracket.png
(Those measuremests are takes from some random bracket, it would be nice to know what they actually are supposed to be and then add the custom bits on top of that).
There could be changes (such as going for RCAs instead), in the end it all comes down to cost and I'd like it to stay under 2€ range per bracket (shipping etc. included). Quantity would be 100 for starters.

I'm talking to the prototype shop today about the brackets. They're going to whip up a sample and I'll see if I can get an accurate price.

Reply 116 of 339, by Tiido

User metadata
Rank l33t
Rank
l33t

That sounds interesting ! I might actually have a solution from China, a friend found something from there but I have no clue yet what sort of quality they have etc. but the estimated price is between 1 and 2usd, depending on complexity involved. I also intend to change line out to RCAs on the next version. Once I know more that place could be a solution to the bracket issue to all the projects around here.

Speaking of next version, I am no longer poor but I still have some things to finish before I can put more effort into this.

T-04YBSC, a new YMF71x based sound card & Official VOGONS thread about it
Newly made 4MB 60ns 30pin SIMMs ~
mida sa loed ? nagunii aru ei saa 😜

Reply 118 of 339, by Tiido

User metadata
Rank l33t
Rank
l33t

Not anymore, last one went away a week ago. I expect to start making more next month but I don't think any selling happens until next year as I'll likely go to Norway for a month to see my GF in December and won't return until January.

T-04YBSC, a new YMF71x based sound card & Official VOGONS thread about it
Newly made 4MB 60ns 30pin SIMMs ~
mida sa loed ? nagunii aru ei saa 😜