24bit ASIO with SB Live?

LfmC · Jun 12, 2003, 01:20 AM

Is it possible to make the kx driver work in 24bit with SB live cards?
Will we have to wait for an update from kx or is it just impossible?
I'm sorry for the silly question but I really don't know much about hardware so I was wondering can it be done @ all.
And what about higher sample rates (96Hz)?

Max M. · Jun 12, 2003, 02:01 AM

>I was wondering can it be done @ all.

in short: No.

>And what about higher sample rates (96Hz)?

not on SBLive...

Daniel Drummond · Jun 12, 2003, 03:38 AM

Ok, but wouldn't it be possible to "fake" it? I mean, something like this: a filter (or something like that) divides a channel input into two separate channels and each one of this is sampled with 16bit quality (simultaneously). Than, this information is put together to make 24bit or 32bit recordings... Something like that...

Max M. · Jun 12, 2003, 03:54 AM

Yes, this is how it is done (in kx) with Audigy models now (in pre/alpha mode) - but that won't work with emu10k1 due to some glue in hardware when each sample gets masked by its neighbourhoods (e.g. kind of "resampling") - so such method will produce "pseudo 24bit" sound (which has nothing to do with original and even objectively worse then current 16bit one)... i think we will keep to investigate this problem but... also (and maybe this is more important then above reason) what a sense to play 24 bits to 16bit DAC (as well as record 24 bits from 16bit ADC)?

Daniel Drummond · Jun 12, 2003, 09:12 PM

Yes, makes no sense... But it's fun anyway! hehe Well, at least for us who are not into programming... Or maybe the real fun is in programming such weird stuff and see it work.. hehe

Mata Hari · Jun 12, 2003, 10:00 PM

The sense in it, might be recording software synths at 32 bits... As one of the big things I use kX for, is to be a virtual audio cable... Being able to do this in 24/32 bits, would get "scientificly" better music... (although, maybe not playable by the live, mastering from 24/32 bits samples, is always better then from 16 bits)

Only better increase ofcource would be a higher sampling rate, wich brings me to a thing I have just come up with...
the internal dsp is 48.000kHz, but the input and output can be sampled higher and lower...
is a signal splittable in frames, to have 2 sampleblocks be processed the same time? (so, having 16 bits 96 input, splits the signals in 2 blocks of 48000, then, when going out to recording, turn it back to 1 of 96)

the use of this all is pure and alone, for internal asio recording, I know this... but it IS a good way to produce the highest possible internal recordings of software synths possible (for programs, not capable of recording/exporting themselfs)

The use? probably close to 0.... but such a thing would go above and beyond the "accaptable" uses of the live... (32 bits, 96 khz, on a emu10k1... with the tram memory copy, it should be possible)

Daniel Drummond · Jun 13, 2003, 05:23 AM

Mata Hari, I guess that was what I was trying to say. Let's see what this crazy folks will do about it...

**Eugene Gavrilov** · Jun 15, 2003, 12:47 AM

unfortunately, there will be no support for >16\48 recording -- live's spdif inputs are truncated by design
however, it might be [theoretically] possible to implement 24\96 playback for all the 10kx cards provided the 'splitted' 24/96 signal is merged again -- but one will need an external device for that

that is, a special 'merger' connected to the digital outputs of the card...

however, I doubt this is interesting, simply because all the functionality of the 10kx chip will be disabled -- no realtime effects, no volume / pan control, no multichannel etc.etc.

so, probably, the only thing I promise to implement is 24/48 playback -- as soon as we discover the 'magic bit' (ac-3 bypass) for 10k1 -- (might be of limited use due to the limitations of the digital outputs of the card...)

/Eugene

Max M. · Jun 15, 2003, 02:34 AM

yep, although it would not be to hard to make pan/vol for "24/96 splits"...
BUT... an "external merger"?! guys, you must be crazy... :)

Jun 12, 2003, 01:20 AM	#1
LfmC DriverHeaven Newbie Join Date: Jun 2003 Posts: 4 Rep Power: 0	24bit ASIO with SB Live? Is it possible to make the kx driver work in 24bit with SB live cards? Will we have to wait for an update from kx or is it just impossible? I'm sorry for the silly question but I really don't know much about hardware so I was wondering can it be done @ all. And what about higher sample rates (96Hz)?

Jun 12, 2003, 02:01 AM	#2
Max M. h/h member-shmember Join Date: Dec 2002 Location: Evil Empire Posts: 2,639 Rep Power: 69	>I was wondering can it be done @ all. in short: No. >And what about higher sample rates (96Hz)? not on SBLive... __________________

Jun 12, 2003, 03:38 AM	#3
Daniel Drummond DriverHeaven Senior Member Join Date: Dec 2002 Location: Brazil Posts: 1,137 Rep Power: 0	Ok, but wouldn't it be possible to "fake" it? I mean, something like this: a filter (or something like that) divides a channel input into two separate channels and each one of this is sampled with 16bit quality (simultaneously). Than, this information is put together to make 24bit or 32bit recordings... Something like that... __________________ www.palcomp3.com.br/ed

Jun 12, 2003, 03:54 AM	#4
Max M. h/h member-shmember Join Date: Dec 2002 Location: Evil Empire Posts: 2,639 Rep Power: 69	Yes, this is how it is done (in kx) with Audigy models now (in pre/alpha mode) - but that won't work with emu10k1 due to some glue in hardware when each sample gets masked by its neighbourhoods (e.g. kind of "resampling") - so such method will produce "pseudo 24bit" sound (which has nothing to do with original and even objectively worse then current 16bit one)... i think we will keep to investigate this problem but... also (and maybe this is more important then above reason) what a sense to play 24 bits to 16bit DAC (as well as record 24 bits from 16bit ADC)? __________________

Jun 12, 2003, 09:12 PM	#5
Daniel Drummond DriverHeaven Senior Member Join Date: Dec 2002 Location: Brazil Posts: 1,137 Rep Power: 0	Yes, makes no sense... But it's fun anyway! hehe Well, at least for us who are not into programming... Or maybe the real fun is in programming such weird stuff and see it work.. hehe __________________ www.palcomp3.com.br/ed

Jun 12, 2003, 10:00 PM	#6
Mata Hari HardwareHeaven Lover Join Date: Feb 2003 Posts: 213 Rep Power: 0	The sense in it, might be recording software synths at 32 bits... As one of the big things I use kX for, is to be a virtual audio cable... Being able to do this in 24/32 bits, would get "scientificly" better music... (although, maybe not playable by the live, mastering from 24/32 bits samples, is always better then from 16 bits) Only better increase ofcource would be a higher sampling rate, wich brings me to a thing I have just come up with... the internal dsp is 48.000kHz, but the input and output can be sampled higher and lower... is a signal splittable in frames, to have 2 sampleblocks be processed the same time? (so, having 16 bits 96 input, splits the signals in 2 blocks of 48000, then, when going out to recording, turn it back to 1 of 96) the use of this all is pure and alone, for internal asio recording, I know this... but it IS a good way to produce the highest possible internal recordings of software synths possible (for programs, not capable of recording/exporting themselfs) The use? probably close to 0.... but such a thing would go above and beyond the "accaptable" uses of the live... (32 bits, 96 khz, on a emu10k1... with the tram memory copy, it should be possible)

Jun 13, 2003, 05:23 AM	#7
Daniel Drummond DriverHeaven Senior Member Join Date: Dec 2002 Location: Brazil Posts: 1,137 Rep Power: 0	Mata Hari, I guess that was what I was trying to say. Let's see what this crazy folks will do about it... __________________ www.palcomp3.com.br/ed

Jun 15, 2003, 12:47 AM	#8
Eugene Gavrilov kX Project Lead Programmer and Coordinator Join Date: Dec 2002 Posts: 3,119 Rep Power: 75	unfortunately, there will be no support for >16\48 recording -- live's spdif inputs are truncated by design however, it might be [theoretically] possible to implement 24\96 playback for all the 10kx cards provided the 'splitted' 24/96 signal is merged again -- but one will need an external device for that that is, a special 'merger' connected to the digital outputs of the card... however, I doubt this is interesting, simply because all the functionality of the 10kx chip will be disabled -- no realtime effects, no volume / pan control, no multichannel etc.etc. so, probably, the only thing I promise to implement is 24/48 playback -- as soon as we discover the 'magic bit' (ac-3 bypass) for 10k1 -- (might be of limited use due to the limitations of the digital outputs of the card...) /Eugene

Jun 15, 2003, 02:34 AM	#9
Max M. h/h member-shmember Join Date: Dec 2002 Location: Evil Empire Posts: 2,639 Rep Power: 69	yep, although it would not be to hard to make pan/vol for "24/96 splits"... BUT... an "external merger"?! guys, you must be crazy... :) __________________