quattro usb-audio endianess conversion layer

quattro usb-audio endianess conversion layer

[Patrick Shirkey]

Has anyone had ideas on how to implement a layer that works to 
automatically convert the quattros big endian input into little endian 
to enable ease of use with most Linux audio apps?

It turns out the reason for the noise in the input channels all this 
time is due to the quattro only reading big endian at input. Conversely 
it only accepts little endian for output :(

This has got to be one of the most annoying workarounds in the history 
of ALSA audio drivers.


-- 
Patrick Shirkey - Boost Hardware Ltd.
Http://www.boosthardware.com
Http://www.djcj.org/LAU/guide/ - The Linux Audio Users guide
Http://www.djcj.org/gigs/ - Gigs guide Korea
Http://www.nana7.net - Bar Nana - Itaewon, Seoul, Sth Corea
========================================


Apparently upon the beginning of the barrage, the donkey broke 
discipline and panicked, toppling the cart. At that point, the rockets 
disconnected from the timer, leaving them strewn around the street. 
Tethered to the now toppled cart, the donkey was unable to escape before 
the arrival of U.S. troops.

United Press International
Rockets on donkeys hit major Baghdad sites

By P. MITCHELL PROTHERO
Published 11/21/2003 11:13 AM



-------------------------------------------------------
This SF.Net email is sponsored by: GNOME Foundation
Hackers Unite!  GUADEC: The world's #1 Open Source Desktop Event.
GNOME Users and Developers European Conference, 28-30th June in Norway
http://2004/guadec.org

Re: quattro usb-audio endianess conversion layer

[Clemens Ladisch]

Patrick Shirkey wrote:
> It turns out the reason for the noise in the input channels all this
> time is due to the quattro only reading big endian at input.

Does this happen all of the time?  There is a bug report which says
that it works sometimes.

> Has anyone had ideas on how to implement a layer that works to
> automatically convert the quattros big endian input into little endian
> to enable ease of use with most Linux audio apps?

With the patch below, it should work with plughw:X.  Does it?


HTH
Clemens

-- 
Index: alsa-kernel/usb/usbaudio.c
===================================================================
RCS file: /cvsroot/alsa/alsa-kernel/usb/usbaudio.c,v
retrieving revision 1.100
diff -u -r1.100 usbaudio.c
--- alsa-kernel/usb/usbaudio.c	5 May 2004 08:50:45 -0000	1.100
+++ alsa-kernel/usb/usbaudio.c	10 Jun 2004 17:00:40 -0000
@@ -2182,6 +2182,24 @@


 /*
+ * check if the device uses big-endian samples
+ */
+static int is_big_endian_format(struct usb_device *dev, struct audioformat *fp)
+{
+	/* M-Audio */
+	if (dev->descriptor.idVendor == 0x0763) {
+		/* Quattro: captured data only */
+		if (dev->descriptor.idProduct == 0x2001 &&
+		    fp->endpoint & USB_DIR_IN)
+			return 1;
+		/* Audiophile USB */
+		if (dev->descriptor.idProduct == 0x2003)
+			return 1;
+	}
+	return 0;
+}
+
+/*
  * parse the audio format type I descriptor
  * and returns the corresponding pcm format
  *
@@ -2217,17 +2235,13 @@
 			pcm_format = SNDRV_PCM_FORMAT_S8;
 			break;
 		case 2:
-			/* M-Audio audiophile USB workaround */
-			if (dev->descriptor.idVendor == 0x0763 &&
-			    dev->descriptor.idProduct == 0x2003)
+			if (is_big_endian_format(dev, fp))
 				pcm_format = SNDRV_PCM_FORMAT_S16_BE; /* grrr, big endian!! */
 			else
 				pcm_format = SNDRV_PCM_FORMAT_S16_LE;
 			break;
 		case 3:
-			/* M-Audio audiophile USB workaround */
-			if (dev->descriptor.idVendor == 0x0763 &&
-			    dev->descriptor.idProduct == 0x2003)
+			if (is_big_endian_format(dev, fp))
 				pcm_format = SNDRV_PCM_FORMAT_S24_3BE; /* grrr, big endian!! */
 			else
 				pcm_format = SNDRV_PCM_FORMAT_S24_3LE;
Index: alsa-kernel/usb/usbquirks.h
===================================================================
RCS file: /cvsroot/alsa/alsa-kernel/usb/usbquirks.h,v
retrieving revision 1.33
diff -u -r1.33 usbquirks.h
--- alsa-kernel/usb/usbquirks.h	12 May 2004 06:29:22 -0000	1.33
+++ alsa-kernel/usb/usbquirks.h	10 Jun 2004 17:00:40 -0000
@@ -830,11 +830,42 @@
 	.driver_info = (unsigned long) & (const snd_usb_audio_quirk_t) {
 		.vendor_name = "M-Audio",
 		.product_name = "Quattro",
-		.ifnum = 9,
-		.type = QUIRK_MIDI_MIDIMAN,
-		.data = & (const snd_usb_midi_endpoint_info_t) {
-			.out_cables = 0x0001,
-			.in_cables  = 0x0001
+		.ifnum = QUIRK_ANY_INTERFACE,
+		.type = QUIRK_COMPOSITE,
+		.data = & (const snd_usb_audio_quirk_t[]) {
+			/*
+			 * Interfaces 0-2 are "Windows-compatible", 16-bit only,
+			 * and share endpoints with the other interfaces.
+			 * Ignore them.  The other interfaces can do 24 bits,
+			 * but captured samples are big-endian (see usbaudio.c).
+			 */
+			{
+				.ifnum = 4,
+				.type = QUIRK_AUDIO_STANDARD_INTERFACE
+			},
+			{
+				.ifnum = 5,
+				.type = QUIRK_AUDIO_STANDARD_INTERFACE
+			},
+			{
+				.ifnum = 7,
+				.type = QUIRK_AUDIO_STANDARD_INTERFACE
+			},
+			{
+				.ifnum = 8,
+				.type = QUIRK_AUDIO_STANDARD_INTERFACE
+			},
+			{
+				.ifnum = 9,
+				.type = QUIRK_MIDI_MIDIMAN,
+				.data = & (const snd_usb_midi_endpoint_info_t) {
+					.out_cables = 0x0001,
+					.in_cables  = 0x0001
+				}
+			},
+			{
+				.ifnum = -1
+			}
 		}
 	}
 },




-------------------------------------------------------
This SF.Net email is sponsored by the new InstallShield X.
>From Windows to Linux, servers to mobile, InstallShield X is the
one installation-authoring solution that does it all. Learn more and
evaluate today! http://www.installshield.com/Dev2Dev/0504

Re: quattro usb-audio endianess conversion layer

[Patrick Shirkey]

Clemens Ladisch wrote:
> Patrick Shirkey wrote:
> 
>>It turns out the reason for the noise in the input channels all this
>>time is due to the quattro only reading big endian at input.
> 
> 
> Does this happen all of the time?  There is a bug report which says
> that it works sometimes.
> 
> 
>>Has anyone had ideas on how to implement a layer that works to
>>automatically convert the quattros big endian input into little endian
>>to enable ease of use with most Linux audio apps?
> 
> 
> With the patch below, it should work with plughw:X.  Does it?
> 

Thanks Clemens.


I finally was able to test the patch and my first impression is that it 
has made a significant difference and should be committed immediately.

I recorded with
	
	arecord -f cd -d 10 -D plughw:2,0 test.wav

played back with

	aplay -f cd -d hw:2,1 test.wav

The sound is much clearer.


My next question is: Does anyone have a good idea how to apply this to a 
jack setup without sacrificing latency?


-- 
Patrick Shirkey - Boost Hardware Ltd.
Http://www.boosthardware.com
Http://www.djcj.org/LAU/guide/ - The Linux Audio Users guide
Http://www.djcj.org/gigs/ - Gigs guide Korea
Http://www.nana7.net - Bar Nana - Itaewon, Seoul, Sth Corea
========================================


Apparently upon the beginning of the barrage, the donkey broke 
discipline and panicked, toppling the cart. At that point, the rockets 
disconnected from the timer, leaving them strewn around the street. 
Tethered to the now toppled cart, the donkey was unable to escape before 
the arrival of U.S. troops.

United Press International
Rockets on donkeys hit major Baghdad sites

By P. MITCHELL PROTHERO
Published 11/21/2003 11:13 AM



-------------------------------------------------------
This SF.Net email is sponsored by The 2004 JavaOne(SM) Conference
Learn from the experts at JavaOne(SM), Sun's Worldwide Java Developer
Conference, June 28 - July 1 at the Moscone Center in San Francisco, CA
REGISTER AND SAVE! http://java.sun.com/javaone/sf Priority Code NWMGYKND

Re: quattro usb-audio endianess conversion layer

[Clemens Ladisch]

Patrick Shirkey wrote:
> > Patrick Shirkey wrote:
> >
> >>It turns out the reason for the noise in the input channels all this
> >>time is due to the quattro only reading big endian at input.
>
> I finally was able to test the patch and my first impression is that it
> has made a significant difference

Good.

> My next question is: Does anyone have a good idea how to apply this to a
> jack setup without sacrificing latency?

By adding non-native endianness support to Jack.  (Which requires
supporting different sample formats for playback vs. recording.)


Regards,
Clemens




-------------------------------------------------------
This SF.Net email is sponsored by the new InstallShield X.
>From Windows to Linux, servers to mobile, InstallShield X is the
one installation-authoring solution that does it all. Learn more and
evaluate today! http://www.installshield.com/Dev2Dev/0504