Skip to contentSkip to conversation

4D-NXs (was Re: Sync Issues)

DISCUSSION Billy Biggs 6 participants 12 messages Oct 27, 1999 – Oct 28, 1999 Sound

lore
Share & tools
lore
#0 lore raw
Quoted 7 lines · Paul Barton-Davis
Quoted 3 lines

Drivers had better be forthcomming. Yes, I'm talking about cards like
those, such as the Event Layla and Darla, etc. Without them, it makes Linux pretty useless as a multitrack platform...

the manufacturers of such cards don't have much of an interest in Linux. i did a little work at AES to try to change that, but its going to be a long uphill battle.

Well, if there were a more powerful audio API, and applications that
took advantage of it, that would all change pretty quick...

Quoted 8 lines
Quoted 3 lines

Well, at least pretty annoying. Personally, I have two AudioPCI cards
and a dedicated MPU-401 card for MIDI. Each of the AudioPCI cards have two dsps, giving me four /dev/dsp devices.

if you had 2 4D-NX's, you'd have 64 openable (mono) channels for playback (*)... is that enough for you ? :))) total cost $78+postage, plus you'd have two excellent MIDI interfaces as well.
(*) but just 4 mono channels for recording, sigh.

4D-NX seems to be just the chipset. Which card are you referring to
here? And, how does it get so many seperate openable channels?

More importantly, would I be able to sync channels on the same sound
card?

--
Billy Biggs [email protected]
http://www.div8.net/billy [email protected]

#1 lore raw
Quoted 2 lines · Billy Biggs

4D-NX seems to be just the chipset. Which card are you referring to
here? And, how does it get so many seperate openable channels?

32 stereo channels, mixing is performed by hardware

Quoted 2 lines

More importantly, would I be able to sync channels on the same sound
card?

Yes, with the new PCM v2 ALSA API (currently in development - I welcome testers).

1) load the playback buffers
2) send SYNC GO command to the driver

You may create more syncronized groups, of course.

Jaroslav

-----
Jaroslav Kysela <[email protected]>
SuSE Linux http://www.suse.com
ALSA project http://www.alsa-project.org

#2 lore raw
In reply to Billy Biggs
Quoted 10 lines
Quoted 7 lines
Quoted 3 lines

Drivers had better be forthcomming. Yes, I'm talking about cards like
those, such as the Event Layla and Darla, etc. Without them, it makes Linux pretty useless as a multitrack platform...

the manufacturers of such cards don't have much of an interest in Linux. i did a little work at AES to try to change that, but its going to be a long uphill battle.

Well, if there were a more powerful audio API, and applications that
took advantage of it, that would all change pretty quick...

actually, no.

i've had many discussions with different companies. the problem has *never* been "Oh, Linux's audio API isn't worth us developing for". Its also never been "Well, what applications are there ?" (remember, most of these companies simply *assume* the existence of apps, or alternatively, write their own).

Instead, its a combination of:

* "some other people wrote a Linux driver for our board and it sucked"
* "we don't have enough programmers to do that"
* "we don't have any written documentation to give you guys - we wrote the driver by having the software group sit in with the hardware group"
* "we think our hardware's proprietary secrets will be revealed if there is a source code driver"
* "Linux ? Is that like Cakewalk ?"

Quoted 2 lines

4D-NX seems to be just the chipset. Which card are you referring to
here?

go to www.hoontech.com and look at the SoundWave NX.

--p

#3 lore raw
In reply to Billy Biggs
Quoted 1 line · Paul Barton-Davis

* "we don't have enough programmers to do that"

If the drivers are being written for them by volunteers, I dont see how this is relevant.

Quoted 3 lines

* "we don't have any written documentation to give you guys - we wrote the driver by having the software group sit in with the hardware group"

This should be a warning sign to anyone thinking of purchasing their hardware. If a company cant be bothered to internally document the hardware, what happens if key engineers leave the company? Oh dear, their project is *permanently screwed*, which means zero support for end users. This is no way to run a company.

Quoted 2 lines

* "we think our hardware's proprietary secrets will be revealed if there is a source code driver"

Uh, isnt this what patents are for? If someone reverse engineers their card, they are *completely screwed* unless they have patent protection.

-Dan

#4 lore raw
In reply to Billy Biggs
Quoted 10 lines · Paul Barton-Davis

Instead, its a combination of:

* "some other people wrote a Linux driver for our board and it sucked"
* "we don't have enough programmers to do that"
* "we don't have any written documentation to give you guys - we wrote the driver by having the software group sit in with the hardware group"
* "we think our hardware's proprietary secrets will be revealed if there is a source code driver"
* "Linux ? Is that like Cakewalk ?"

Another one:

* "Can you give an estimate of how many Linux users would be interested in buying our card, if there was a driver?"

That's a tough one... How many users do you get without real cards? And how many real cards and applications do you get without lots of users...?

Well, it's our own "fault"; this is a Free/Open Source world, and *we* have to get things moving if anything is to happen. Therefore, an API that makes it easier for us to reuse code, and for users to integrate existing applications into working solutions, will probably help a lot.

Would IBM, Corel, SyBase and co have cared about Linux if it wasn't already a big player in the server field? Probably not... They've been hurt by MS and other concurents too many times to jump into something like Linux, before they see that it actually works in real life.

I'm afraid we have to get Linux boxes into studios. Before that happens, performance superior by orders of magnitude to that of Windows and MacOS, is little but exciting news for the techies.

//David

·A·U·D·I·A·L·I·T·Y· P r o f e s s i o n a l L i n u x A u d i o
- - ------------------------------------------------------------- - -

    ·Rock Solid                                      David Olofson:
    ·Low Latency    www.angelfire.com/or/audiality   ·Audio Hacker
    ·Plug-Ins            [email protected]        ·Linux Advocate
    ·Open Source                                     ·Singer/Composer
#5 lore raw
In reply to Billy Biggs
Quoted 4 lines
Quoted 1 line

* "we don't have enough programmers to do that"

If the drivers are being written for them by volunteers, I dont see how this is relevant.

if they had a driver written for them by 4Front, and it had problems, their desire to have volunteers do another one is low. i won't name the company that had this problem, but they are a real soundcard company.

Quoted 9 lines
Quoted 3 lines

* "we don't have any written documentation to give you guys - we wrote the driver by having the software group sit in with the hardware group"

This should be a warning sign to anyone thinking of purchasing their hardware. If a company cant be bothered to internally document the hardware, what happens if key engineers leave the company? Oh dear, their project is *permanently screwed*, which means zero support for end users. This is no way to run a company.

Yamaha (the company that this quote comes from) seems to be doing quite well :)

Quoted 2 lines

Uh, isnt this what patents are for? If someone reverse engineers their card, they are *completely screwed* unless they have patent protection.

and the cost of a patent is ? look, there's very little truly clever stuff in these cards, and what is patented is often stupid (est told me last week about someone who has patented putting an LFO in hardware on a soundcard with a wavetable synth). most soundcard technologies don't last more than a few years, and investing the X-thousand dollars in seeking patent protection for a nifty hack that will be irrelevant in that time is pretty unjustifiable for most things. security for obscurity really does work when you only want it for a little while ;)

i think david's point is an important one that i didn't mention, and he's right that the only thing, in the end, that most of the companies are going to respond to is more than a couple of studios saying "we run linux - we want to buy your cards/devices and there are no drivers".

#6 lore raw
In reply to Billy Biggs
Quoted 6 lines · Paul Barton-Davis
Quoted 3 lines
Quoted 1 line

* "we don't have enough programmers to do that"

If the drivers are being written for them by volunteers, I dont see how this is relevant.

if they had a driver written for them by 4Front, and it had problems, their desire to have volunteers do another one is low. i won't name the company that had this problem, but they are a real soundcard company.

Still irrelevant to the *statement being made*, which is "*we* dont have enough programmers to do that". Why has the number of programmers *they* have, have anything to do with us volunteers writing a driver for them for free? *boggle*.

Quoted 10 lines
Quoted 8 lines
Quoted 3 lines

* "we don't have any written documentation to give you guys - we wrote the driver by having the software group sit in with the hardware group"

This should be a warning sign to anyone thinking of purchasing their hardware. If a company cant be bothered to internally document the hardware, what happens if key engineers leave the company? Oh dear, their project is *permanently screwed*, which means zero support for end users. This is no way to run a company.

Yamaha (the company that this quote comes from) seems to be doing quite well :)

No, Yamaha *does* have documentation on the native PCI stuff. They just wont release it.

Quoted 10 lines
Quoted 2 lines

Uh, isnt this what patents are for? If someone reverse engineers their card, they are *completely screwed* unless they have patent protection.

and the cost of a patent is ? look, there's very little truly clever stuff in these cards, and what is patented is often stupid (est told me last week about someone who has patented putting an LFO in hardware on a soundcard with a wavetable synth). most soundcard technologies don't last more than a few years, and investing the X-thousand dollars in seeking patent protection for a nifty hack that will be irrelevant in that time is pretty unjustifiable for most things. security for obscurity really does work when you only want it for a little while ;)

I guess what we need are some PCI snooper hardware so we can start putting the shits up (various unnamed companies). Start posting reverse engineered specs and stuff will surely get their attention ;)

Didnt RMS say that the FSF was going to start doing some reverse engineering with PCI snoopers?

Quoted 4 lines

i think david's point is an important one that i didn't mention, and he's right that the only thing, in the end, that most of the companies are going to respond to is more than a couple of studios saying "we run linux - we want to buy your cards/devices and there are no drivers".

Im more concerned about chip vendors than card manufacturers. If we convince the chip vendor to release docs then theres really nothing a card manufacturer can do to stop us.

We really only have three problematic chip vendors here - Yamaha, Aureal, and Creative Labs. Although Creative Labs may be coming around finally with the emu10k stuff.

The trend toward FPGA makes things interesting, of course.

-Dan

#7 lore raw
In reply to Billy Biggs
Quoted 4 lines

Still irrelevant to the *statement being made*, which is "*we* dont have enough programmers to do that". Why has the number of programmers *they* have, have anything to do with us volunteers writing a driver for them for free? *boggle*.

they don't *want* volunteers writing drivers for them. period. end of story.

Quoted 2 lines

No, Yamaha *does* have documentation on the native PCI stuff. They just wont release it.

i have been told by one senior level management guy at yamaha and one senior programming guy from their main japanese plant that there is *NO* documentation for the DSP Factory.

Quoted 3 lines

Im more concerned about chip vendors than card manufacturers. If we convince the chip vendor to release docs then theres really nothing a card manufacturer can do to stop us.

i have the "full specs" to the (ancient) YSS225 FX processor from yamaha. its installed on the Tropez+. the specs are useless because it has downloaded microcode to set it up, and they only document 5 microcode programs. Yamaha wrote a special one for Turtle Beach, and neither they nor Yamaha appear have any record of what it used the dsp memory locations for.

i've used dosemu to decode part of it, and reverse engineered the windows FX driver too, but the end result is a GUI with 512 buttons that inc/dec values in a DSP location to let me try to understand what it does. since many memory locations' function is tightly coupled with several others, its a pretty hopeless case. i did manage to find where the global stereo echo delay time was set, but that was all.

you could argue that this means that we don't have the specs, but Yamaha certainly doesn't see it that way.

--p

#8 lore raw
In reply to Billy Biggs
Quoted 5 lines · Paul Barton-Davis
Quoted 4 lines

Still irrelevant to the *statement being made*, which is "*we* dont have enough programmers to do that". Why has the number of programmers *they* have, have anything to do with us volunteers writing a driver for them for free? *boggle*.

they don't *want* volunteers writing drivers for them. period. end of story.

If thats what they wanted, why didnt they just say it. Stating it that way is purposely deceitful.

Quoted 5 lines
Quoted 2 lines

No, Yamaha *does* have documentation on the native PCI stuff. They just wont release it.

i have been told by one senior level management guy at yamaha and one senior programming guy from their main japanese plant that there is *NO* documentation for the DSP Factory.

I received a statement from one of their japanese engineers that they *do* have documentation for their native PCI audio block for the YMF724 and YMF744, but that it was only released to OEMs and hardware integrators.

So we have conflicting statements from the same company. Not too suprising really.

Quoted 17 lines
Quoted 3 lines

Im more concerned about chip vendors than card manufacturers. If we convince the chip vendor to release docs then theres really nothing a card manufacturer can do to stop us.

i have the "full specs" to the (ancient) YSS225 FX processor from yamaha. its installed on the Tropez+. the specs are useless because it has downloaded microcode to set it up, and they only document 5 microcode programs. Yamaha wrote a special one for Turtle Beach, and neither they nor Yamaha appear have any record of what it used the dsp memory locations for.
i've used dosemu to decode part of it, and reverse engineered the windows FX driver too, but the end result is a GUI with 512 buttons that inc/dec values in a DSP location to let me try to understand what it does. since many memory locations' function is tightly coupled with several others, its a pretty hopeless case. i did manage to find where the global stereo echo delay time was set, but that was all. you could argue that this means that we don't have the specs, but Yamaha certainly doesn't see it that way.

Well one would have to weigh the effort of reverse engineering it against the gains. Is there any point reverse engineering such ancient hardware? Compared to eg reverse engineering the emu10k or aureal chips.

FWIW the SAM9407 documentation was pretty inadequate but gerd rausch managed to come up with a pretty decent driver anyway.

-Dan

#9 lore raw
In reply to Billy Biggs

Paul Barton-Davis spake thusly<[email protected]>:
[x]

Quoted 4 lines

i think david's point is an important one that i didn't mention, and he's right that the only thing, in the end, that most of the companies are going to respond to is more than a couple of studios saying "we run linux - we want to buy your cards/devices and there are no drivers".

That's one thing altho they're more likely to be saying "we want to run linux because ..." - anyone running linux right now in a professional studio in other than an experimental way would be thought to be out of their minds given that travelling engineers/producers expect to see the familiar software "faces".

IMHO the greatest hope we have for the near term for software anyway is Fear! If Cubase thinks Logic is doing a port or visa versa, they'll be off and running. I suspect that some high-end sound card drivers would follow on this.

New killer apps in Linux audio software could drive this too but I think the process would take longer - the software has to get out there, be recognised for what it is, be talked about etc.

I think a survey is in order! - software and drivers... asking what stage they're at with linux. Something like ...

-eh? what's Linux?
-we're keeping a close eye on Linux developments
-we're doing a feasability study
-we're working on a port right now
-it's out next week!

Anyone who has particular people they'd like to see surveyed please email me... Likewise anyone with relevant email addresses that get to real people. I'll put the results up on Linux MusicStation including those who didn't answer.
tt
John

Signature · 1 line
http://linuxmusic.cjb.net
#10 lore raw
In reply to Billy Biggs
Quoted 8 lines
Quoted 5 lines
Quoted 4 lines

Still irrelevant to the *statement being made*, which is "*we* dont have enough programmers to do that". Why has the number of programmers *they* have, have anything to do with us volunteers writing a driver for them for free? *boggle*.

they don't *want* volunteers writing drivers for them. period. end of story.

If thats what they wanted, why didnt they just say it. Stating it that way is purposely deceitful.

the fault is more likely in my wording of it. i wasn't trying to quote directly. the company concerned has just been burned by having a 3rd party write drivers for it. i don't think they were deceitful with me, i just didn't present a very accurate picture of what they said.

Quoted 3 lines

Well one would have to weigh the effort of reverse engineering it against the gains. Is there any point reverse engineering such ancient hardware? Compared to eg reverse engineering the emu10k or aureal chips.

what do you think the expected lifetime of any of the current generation of chips is, given how long every previous equivalent has lasted ? if we're reverse engineering a chip more than 18mths after its first use in a PC soundcard, i'd say we might well be wasting our time, in the long haul.

Quoted 2 lines

FWIW the SAM9407 documentation was pretty inadequate but gerd rausch managed to come up with a pretty decent driver anyway.

well, mine works just fine. its just that the non-linear relationships between the memory locations on the DSP make it more or less useless to try to use the driver to alter the operation of the downloaded microprogram unless you're interested in random FX :) instead, you get the DSP in something close to passthru mode.

--p

#11 lore raw
In reply to Billy Biggs
Quoted 8 lines · Paul Barton-Davis
Quoted 3 lines

Well one would have to weigh the effort of reverse engineering it against the gains. Is there any point reverse engineering such ancient hardware? Compared to eg reverse engineering the emu10k or aureal chips.

what do you think the expected lifetime of any of the current generation of chips is, given how long every previous equivalent has lasted ? if we're reverse engineering a chip more than 18mths after its first use in a PC soundcard, i'd say we might well be wasting our time, in the long haul.

The Cirrus Logic chips have been around a long time. The ESS chips have been around a long time too. The Creative Labs stuff has been around for more than 10 years. If these chipsets would have had to be reverse engineered, it would definitely not have been a waste of time since the data would still be useful today.

Many chipsets go through incremental improvements (eg CS42xx, ESS1xxx) so reverse engineering one may prove to be useful later since future chips may be (mostly) backwards-compatible, with a few changes. I suspect this is the case with the YMF7xx chips and maybe even the AU8xxx ones.

Definitely the YMF7xx chips and the DS-1x architecture look like theyre going to be around for some time.

-Dan