OS2 World Community Forum
OS/2, eCS & ArcaOS - Technical => Hardware => Topic started by: Igor on October 15, 2023, 12:35:26 am
-
Hi!
No sound. I'm using the driver from Lars Erdmann:10.248#@OS/2 USB Audio 2.0 Device Driver
Any ideas on howto to fix it?
<<< Device Description >>>
Type : 01
USB Rev : 200
Class : Unknown class EF
Subclass : Unknown Subclass EF
Protocol : Unknown Protocol EF
Max. packetsize : 40
Vendor ID : 06CB
Product ID : 1595
Device Release# : 0000
Strings:
Manufacturer Name : Synaptics
Product Name : CX31993 384Khz HIFI AUDIO
Serial number : Not implemented
Number of Configurations : 1
Configuration 0 :
Lenght : 422
Name : Headset
Value : 1
Attributes : 0xA0
Power : 100 mA
Interfaces : 4
No Interface 0x04, Endpoint 0x05 or Classspec 0x21-0x025 Description is 0x0B
Interface 0
Alt Setting : 0
Name : <NONE>
Class : Audio
SubClass : AudioControl
Protocol : Error Protocol is set to 01
Endpoints : 0
Audiocontrol CS Interface Descriptor
Length : 9
Type : 0x24
Subtype : 0x01
Header :
Audio Device Class Release Number : 0200
Max size of AudioControl interface descriptor. : 22360
Number of AudioStreaming and MIDIStreaming interfaces : 0
Audiocontrol CS Interface Descriptor
Unexpected SubType
Length : 8
Type : 36
Subtype : 10
Dumped Data :
09 03 07 00 00 11
Audiocontrol CS Interface Descriptor
Input Terminal :
Length : 17
Type : 0x24
Subtype : 0x02
Unique Terminal ID : 1
Terminal Type : 0x0101
Assoc. Terminal ID : 0
Number of Log. Channels : 9
Spatial channel location : 0x0302
Name of Channel : Not Implemented
Name of Terminal : Not Implemented
Audiocontrol CS Interface Descriptor
Feature Unit :
Length : 18
Type : 36
Subtype : 6
Dumped Data :
02 01 03 00 00 00
0C 00 00 00 0C 00 00 00
04 0C
Unique Unit ID : 2
Source ID : 1
Size of Control Item : 3
ControlMap for Master : 0x000000
ControlMap for Channel 1 : 0x00000C
ControlMap for Channel 2 : 0x000000
Name of Feature Unit : Not Implemented
Audiocontrol CS Interface Descriptor
Output Terminal :
Length : 12
Type : 0x24
Subtype : 0x03
Unique Terminal ID : 3
Terminal Type : 0x0402
Assoc. Terminal ID : 4
Source ID : 2
Name of Terminal : Has Index 9(GetString Error -225)
Audiocontrol CS Interface Descriptor
Unexpected SubType
Length : 8
Type : 36
Subtype : 10
Dumped Data :
0A 03 07 00 00 0C
Audiocontrol CS Interface Descriptor
Output Terminal :
Length : 12
Type : 0x24
Subtype : 0x03
Unique Terminal ID : 6
Terminal Type : 0x0101
Assoc. Terminal ID : 0
Source ID : 5
Name of Terminal : Has Index 10(GetString Error -225)
Audiocontrol CS Interface Descriptor
Feature Unit :
Length : 14
Type : 36
Subtype : 6
Dumped Data :
05 04 03 00 00 00
0C 00 00 00 06 11
Unique Unit ID : 5
Source ID : 4
Size of Control Item : 3
ControlMap for Master : 0x000000
ControlMap for Channel 1 : 0x00000C
Name of Feature Unit : Headset
Audiocontrol CS Interface Descriptor
Input Terminal :
Length : 17
Type : 0x24
Subtype : 0x02
Unique Terminal ID : 4
Terminal Type : 0x0402
Assoc. Terminal ID : 3
Number of Log. Channels : 10
Spatial channel location : 0x0101
Name of Channel : Not Implemented
Name of Terminal : Not Implemented
Interface 1
Alt Setting : 0
Name : <NONE>
Class : Audio
SubClass : Streaming Audio
Protocol : Error Protocol is set to 01
Endpoints : 0
Interface 1
Alt Setting : 1
Name : <NONE>
Class : Audio
SubClass : Streaming Audio
Protocol : Error Protocol is set to 01
Endpoints : 1
Audiostreaming CS Interface Descriptor
Format Type:
Length : 16
Type : 0x24
Subtype : 0x01
Physical Channels : 0
Subframe size : 1
Bits used : 1
Frequency Range
- From 131072Hz
- To 3Hz
Audiostreaming CS Interface Descriptor
Format Type:
Length : 6
Type : 0x24
Subtype : 0x02
Physical Channels : 2
Subframe size : 16
Bits used : 7
Supported Frequencies
- 12586241Hz
- 524544Hz
- 293Hz
- 131584Hz
- 264448Hz
- 66049Hz
Endpoint 0:
Address : 01
Attributes : 0D
Packetsize : 00C0
Interval : 1
Endpoint Descriptor NOT Understood:
Length : 8
Type : 37
Subtype : 1
Dumped Data :
00 00 02 02 00 09
Interface 1
Alt Setting : 2
Name : <NONE>
Class : Audio
SubClass : Streaming Audio
Protocol : Error Protocol is set to 01
Endpoints : 1
Audiostreaming CS Interface Descriptor
Format Type:
Length : 16
Type : 0x24
Subtype : 0x01
Physical Channels : 0
Subframe size : 1
Bits used : 1
Frequency Range
- From 131072Hz
- To 3Hz
Audiostreaming CS Interface Descriptor
Format Type:
Length : 6
Type : 0x24
Subtype : 0x02
Physical Channels : 3
Subframe size : 24
Bits used : 7
Supported Frequencies
- 2100481Hz
- 524545Hz
- 293Hz
- 131584Hz
- 264448Hz
- 66305Hz
Endpoint 0:
Address : 01
Attributes : 0D
Packetsize : 0120
Interval : 1
Endpoint Descriptor NOT Understood:
Length : 8
Type : 37
Subtype : 1
Dumped Data :
00 00 02 02 00 09
Interface 1
Alt Setting : 3
Name : <NONE>
Class : Audio
SubClass : Streaming Audio
Protocol : Error Protocol is set to 01
Endpoints : 1
Audiostreaming CS Interface Descriptor
Format Type:
Length : 16
Type : 0x24
Subtype : 0x01
Physical Channels : 0
Subframe size : 1
Bits used : 1
Frequency Range
- From 131072Hz
- To 3Hz
Audiostreaming CS Interface Descriptor
Format Type:
Length : 6
Type : 0x24
Subtype : 0x02
Physical Channels : 4
Subframe size : 32
Bits used : 7
Supported Frequencies
- 8391937Hz
- 524545Hz
- 293Hz
- 131584Hz
- 264448Hz
- 2Hz
Endpoint 0:
Address : 01
Attributes : 0D
Packetsize : 0180
Interval : 1
Endpoint Descriptor NOT Understood:
Length : 8
Type : 37
Subtype : 1
Dumped Data :
00 00 02 02 00 09
Interface 2
Alt Setting : 0
Name : <NONE>
Class : Audio
SubClass : Streaming Audio
Protocol : Error Protocol is set to 01
Endpoints : 0
Interface 2
Alt Setting : 1
Name : <NONE>
Class : Audio
SubClass : Streaming Audio
Protocol : Error Protocol is set to 01
Endpoints : 1
Audiostreaming CS Interface Descriptor
Format Type:
Length : 16
Type : 0x24
Subtype : 0x01
Physical Channels : 0
Subframe size : 1
Bits used : 1
Frequency Range
- From 65536Hz
- To 1Hz
Audiostreaming CS Interface Descriptor
Format Type:
Length : 6
Type : 0x24
Subtype : 0x02
Physical Channels : 2
Subframe size : 16
Bits used : 7
Supported Frequencies
- 1576321Hz
- 524544Hz
- 293Hz
- 131584Hz
- 264448Hz
- 66050Hz
Endpoint 0:
Address : 81
Attributes : 0D
Packetsize : 0018
Interval : 1
Endpoint Descriptor NOT Understood:
Length : 8
Type : 37
Subtype : 1
Dumped Data :
00 00 02 02 00 09
Interface 2
Alt Setting : 2
Name : <NONE>
Class : Audio
SubClass : Streaming Audio
Protocol : Error Protocol is set to 01
Endpoints : 1
Audiostreaming CS Interface Descriptor
Format Type:
Length : 16
Type : 0x24
Subtype : 0x01
Physical Channels : 5
Subframe size : 1
Bits used : 1
Frequency Range
- From 65536Hz
- To 1Hz
Audiostreaming CS Interface Descriptor
Format Type:
Length : 6
Type : 0x24
Subtype : 0x02
Physical Channels : 3
Subframe size : 24
Bits used : 7
Supported Frequencies
- 2362753Hz
- 524544Hz
- 293Hz
- 131584Hz
- 722944Hz
- 196867Hz
Endpoint 0:
Address : 81
Attributes : 0D
Packetsize : 0024
Interval : 1
Endpoint Descriptor NOT Understood:
Length : 8
Type : 37
Subtype : 1
Dumped Data :
00 00 02 02 00 08
No Interface 0x04, Endpoint 0x05 or Classspec 0x21-0x025 Description is 0x0B
Interface 3
Alt Setting : 0
Name : <NONE>
Class : Human Interface Device (HID)
SubClass : No Subclass
Protocol : None
Endpoints : 1
HID Description:
Length : 09
HID Rel. Number : 111
Country Code : 0000
Descriptors : 1
Type : 22
Length : 0
Type : 07
Length : 0
Endpoint 0:
Address : 83
Attributes : 03
Packetsize : 0040
Interval : 7
-
can you please use latest lsusb.exe to dump this info ? What you have dumped is not complete and barely interpretable.
latest lsusb:
https://www.dropbox.com/s/ef45b19aqrc6c95/lsusb-20130526.zip?dl=0
-
can you please use latest lsusb.exe to dump this info ? What you have dumped is not complete and barely interpretable.
latest lsusb:
https://www.dropbox.com/s/ef45b19aqrc6c95/lsusb-20130526.zip?dl=0
Ok. Please see attache.
-
A few questions, if I may:
1. Does your UAC 2.0 device work correctly on USB 2.0 ports with EHCI controllers?
2. Have you tried using 10.244 drivers for this device?
3. Have you managed to use other UAC 2.0 devices on USB 3.0 (XHCI) ports?
-
I managed to play the sound from this device (with some problems).
First of all, I re -installed the USBAudio and USBWAV set from Lars on top of the installed AOS 5.1 system.
Most likely, the installer prescribed everything correctly this time.
Problems that have arisen:
1. The USBAudio.Sys driver does not load with a core 14.204. With the OS4 core, this driver is loading, but does not work.
At the same time, the USBAUD2 loads in both cases and play the sound.
2. With disk activity, malfunctions arise in the sound - it becomes intermittent, sometimes there are artifacts in the sound and in the calm
condition.
3. The device does not work in the USB3 port, the system does not see it.
All ports are checked - they work USB Flash.
-
Hi Igor.
I created this Wiki page based in your HW report: https://www.os2world.com/wiki/index.php/Synaptics_Conexant_CX31993
1) For what I know USBAUDIO.SYS is for some specific USB Audio devices (USB 1.x ?) and USBAUD2.SYS for some other kind of USB Audio devices (USB 2) with other chipsets. I'm guessing that your "Synaptics Conexant CX31993" works with "USBAUD2.SYS", so we can forget about "USBAUDIO.SYS".
2) What are your specific audio tests here? Are you running an MP3 with some specific player? and what disk activity makes the audio intermittent? Are you running MIDI? Audio from the WPS?
Does your machine has an Audio device that works with UNIAUD so you can test the difference between the two devices?
3) I don't know if Lar's "USBAUD2.SYS" works with AN USB 3.0 drivers. Maybe Lars can answer that. I assuming your machine has USB 2 and USB 3 ports that are independent because of your comments.
Regards
-
At the moment, as far as I know, there are no devices left that can work simultaneously in two UAC classes 1 and 2, each of the classes respectively needs its own drivers.
In your case, you can forget about USBAUDIO.SYS, as Martin correctly said.
For UAC2 devices to work correctly, it is enough to install only the USBAudio package via minstall \* , selecting only USBAudio 2.0.
Crackling and distortion in sound are not related to the operation of the HDD or SSD. They are observed even on a OS/2 5.1 system that runs ENTIRELY from RAM Drive and boots over the network. The distortions, in my opinion, originate from the incorrect operation of USBXHCD.SYS in drivers 11.14. On different USB controllers the type of distortion will be different and in some cases it will even seem almost unnoticeable.
But, most importantly, if you continue to torture your psyche and switch tracks with different formats, listening to crackling and wheezing, then most likely as a result you will see a screen of death with Trap 000e in the USBXHCD module.
-
I do not have a 14.204 kernel and I do not know where to get it from. But I can say that on the 14.203 kernel, I have no problems with neither USBAUDIO.SYS nor USBAUD2.SYS. I have USB audio devices of both types but of course, I only have one of each type.
There are problems with any audio driver if the WPS comes into play. If I play a sound via UNIAUD and access a USB drive, UNIAUD will immediately stop playing. For USB the situation is better in that it will not stop, but a hiccup can still be heard.
WPS is such a lousy piece of SW that it is beyond hope that it will ever become reliable. I have had many,many,many desktop freezes where the only cure was to hit the reset button and often enough, to completely turn off the PC.
USBXHCD.SYS has known problems with isochronous transfers. There is nothing that I can do about it. And it should be plain obvious that a trap in USBXHCD.SYS needs to be fixed in USBXHCD.SYS and nowhere else, regardless of if USB client drivers (like USBAUDIO.SYS / USBAUD2.SYS) are badly written or not. You should not accept that as an excuse.
Ask AN for proper USB isochronous support, in fact, ask them for proper USB audio client drivers too, just in case they answer you that "many USB client drivers are broken and do not follow specs". If they have no interest and it still is important for you, then move to another OS. It is that easy.
-
I now gave it a quick test with 14.204 kernel. USBAUDIO.SYS works just fine. Must be something else on your system that does not work.
-
What if I can’t find another operating system that produces sound of the same quality as OS/2?
If even specially prepared builds of Linux https://www.audio-linux.com/ and Windows https://www.highend-audiopc.com/audiophile-optimizer are not capable of outputting sound of the same quality as OS/2?
And I'm talking about sound quality, and not about the characteristics of the files being played.
Now there is only one problem in OS/2 - these are XHCI controllers to which it is not possible to connect UAC2.0 devices, so that the sound would be played without problems.
Recently, the system has learned to play any audio formats, in any resolution, on USBAudio devices, including over a network, like a DLNA client. It is also loaded into RAM over the network from an 18MB image.
So what, let's drop everything and run to persuade Microsoft to rewrite Windows so that it generates less jitter at the output of the audio stream?
Sorry but no.
-
What if I can’t find another operating system that produces sound of the same quality as OS/2?
If even specially prepared builds of Linux https://www.audio-linux.com/ and Windows https://www.highend-audiopc.com/audiophile-optimizer are not capable of outputting sound of the same quality as OS/2?
And I'm talking about sound quality, and not about the characteristics of the files being played.
Now there is only one problem in OS/2 - these are XHCI controllers to which it is not possible to connect UAC2.0 devices, so that the sound would be played without problems.
Recently, the system has learned to play any audio formats, in any resolution, on USBAudio devices, including over a network, like a DLNA client. It is also loaded into RAM over the network from an 18MB image.
So what, let's drop everything and run to persuade Microsoft to rewrite Windows so that it generates less jitter at the output of the audio stream?
Sorry but no.
Then you are sadly out of luck and would have to pay AN (maybe per crowd funding ?) to improve the USB 3.0 HC driver (plus writing the "correct" USB audio client drivers).
-
Hi Igor.
I created this Wiki page based in your HW report: https://www.os2world.com/wiki/index.php/Synaptics_Conexant_CX31993
1) For what I know USBAUDIO.SYS is for some specific USB Audio devices (USB 1.x ?) and USBAUD2.SYS for some other kind of USB Audio devices (USB 2) with other chipsets. I'm guessing that your "Synaptics Conexant CX31993" works with "USBAUD2.SYS", so we can forget about "USBAUDIO.SYS".
2) What are your specific audio tests here? Are you running an MP3 with some specific player? and what disk activity makes the audio intermittent? Are you running MIDI? Audio from the WPS?
Does your machine has an Audio device that works with UNIAUD so you can test the difference between the two devices?
3) I don't know if Lar's "USBAUD2.SYS" works with AN USB 3.0 drivers. Maybe Lars can answer that. I assuming your machine has USB 2 and USB 3 ports that are independent because of your comments.
1. Yes, right. This device is supported by the USBAUD2.Sys driver.
However, I also have other USB 2.0 devices supported by usbaudio.sys.
I already realized that I can’t use them together. At the same time, they are connected in two different ports, but even the USBRes scanner does not see them.
Separately - sees. I wonder why?
2. I use version 1.43 for the PM123 test (build 2022-08-02). As far as I understand, the player uses the DART for playing sound.
I play FLAC and mp3 files.
I also have an onboard device Uniaud, which works very well in all tests.
3. I have separate independent USB 2.0 and USB 3 controllers on my motherboard.
-
I do not have a 14.204 kernel and I do not know where to get it from. But I can say that on the 14.203 kernel, I have no problems with neither USBAUDIO.SYS nor USBAUD2.SYS. I have USB audio devices of both types but of course, I only have one of each type.
There are problems with any audio driver if the WPS comes into play. If I play a sound via UNIAUD and access a USB drive, UNIAUD will immediately stop playing. For USB the situation is better in that it will not stop, but a hiccup can still be heard.
WPS is such a lousy piece of SW that it is beyond hope that it will ever become reliable. I have had many,many,many desktop freezes where the only cure was to hit the reset button and often enough, to completely turn off the PC.
USBXHCD.SYS has known problems with isochronous transfers. There is nothing that I can do about it. And it should be plain obvious that a trap in USBXHCD.SYS needs to be fixed in USBXHCD.SYS and nowhere else, regardless of if USB client drivers (like USBAUDIO.SYS / USBAUD2.SYS) are badly written or not. You should not accept that as an excuse.
Ask AN for proper USB isochronous support, in fact, ask them for proper USB audio client drivers too, just in case they answer you that "many USB client drivers are broken and do not follow specs". If they have no interest and it still is important for you, then move to another OS. It is that easy.
Sorry, Lars. I inadvertently misled you regarding the driver usbaudio.sys and core 14.204.
Everything works correctly. The error arose due to the fact that the installer added another line to Config.sys with the driver's loading and the situation with double loading of different versions of the driver arose. As a result, an error occurred when loading the system.
There was no error with the OS4 core, because there I edit the Config file with my hands.
One way or another, but now everything starts and plays in both configurations.
However, there are problems with the stability of sound and Progress bar in PM123. I am surprised to observe when playing the sound of a jerk
Progress Bar back. Perhaps this is due to the incorrect work of isosynchronous transmission.
The sound has not croaks yet, but I spent few tests without the load of the system.
-
1) I have now installed PM123_1-43.exe. Don`t know where I got it from and if it is the latest version (looks like I got it from Hobbes)
2) I plugged in a USB audio class 1.0 device (I DO NOT mean a USB 1.x device), I switched the default audio device to the USB audio class 1.0 driver
3) I played an audio cd: that worked right away, no oddities, volume control works
4) I then loaded an .MP3 file: plays just fine, no hiccups, no hops backwards, whatever, volume control works, either with the global volume control or with the enabled PM123 equalizer pre-amp volume control
5) selected an ADPCM encoded .WAV file: plays just fine (but only if you have USBWAV installed because that transforms from ADPCM to PCM via the MMOS2 supplied filter functionality), global volume control works, PM123 equalizer pre-amp control also works
6) I then selected a PCM encoded .WAV file: plays just fine, volume control works either way
When I play a sound, no matter what sound material, things get problematic if you use the WPS to open a USB drive or if you use NDFS to access an FTP folder via the WPS while the replay takes place. As I mentioned already, the WPS is such a lousy piece of SW (at least in conjunction with SMP, I remember that the problems with the WPS were only minor for as long as we all possessed single-core CPUs) that it will massively block various threads in the system, seemingly also the threads that DART and/or the MMOS2 streaming manager create to manage the sound replay (feed audio data to the device driver, issue user callbacks to receive events to feed more data to the MMOS2 subsystem).
So, if you want to listen to music on an OS/2 system, do not touch and use the system while listening.
I you could provide me with a sound file that leads to what you are describing, feel free to attach them here and I will give it a try.
Under OS/2, I do not use USB 3.x (that is, I do have USB 3.x hardware but I have no USB driver loaded to support it). All USB audio devices, regardless if class 1 (USBAUDIO.SYS) or class 2 (USBAUD2.SYS) only use USB 1.x or USB 2.x. I think the ONLY USB HW that profits from USB 3.x is Flash memory sticks/HDDs and maybe USB network devices (the latter not being supported by OS/2 anyway, except for a few).
Therefore, as a test, I suggest that you comment out USBXHCD.SYS in your config.sys UNLESS you have USB 3.x only ports. In the latter case you are out of luck and would need to buy a USB 2.x plug-in PCI card and plug your USB audio device into that.
-
I have noticed that when using the Xcenter to use the drive folder, the drive menus slow to a craw and become unstable. Does this and the sound problem go back to the SIQ (the Single Input Que) that was mentioned years ago, but not lately?
-
As far as I remember the SIQ problem was fixed years ago, so that should not be the problem. But there are enough deficiencies in the WPS that lead to the effect you describe. And I have the suspicion that it relates to the introduction of SMP. What you can do as a test is to reduce to one core by using the /MAXCPU=1 switch of ACPI.PSD and see if that improves things. If it improves things, then my suspicion would prove to be correct.
-
Lars, thanks for the tip. I tried to use /MAXCPU=1, but the clicks were still audible in the sound, and after a while an error appeared for the XHCI driver, exactly the same as what I gave earlier. I tested different devices, they all output clicks and crackles over the sound and some crashed with an XHCI error, some became unavailable in the DART plugin after some time
-
Hallo Mikhail,
I was referring to David's post and that related to a general problem regarding any sound driver because there is a general problem with streaming audio data when the WPS comes into play. In your case, the USBXHCD.SYS driver is causing the problem or its interaction with the audio client drivers (USBAUDIO.SYS and/or USBAUD2.SYS).
-
Hi Lars
I added the switch and rebooted.
On the Xcenter, I selected the drives folder.
Then I selected drive C which is my USB Fat32 drive
The folders and files come up and then the system hangs.
The good thing is that TOP allows a reboot.
The ArcaOS drives with JFS do not hang when selected.
-
Hi David,
so, then we are already two people that experience this problem with FAT32 formatted drives. By the way, it does not hang indefinitely, you just need to wait around 30 seconds or so.
I know AN does not want to hear it but it's a problem with the FAT32 IFS that they are delivering. I do not have this problem with the latest FAT32 from Netlabs.
-
Hi Lars
I added the switch and rebooted.
On the Xcenter, I selected the drives folder.
Then I selected drive C which is my USB Fat32 drive
The folders and files come up and then the system hangs.
The good thing is that TOP allows a reboot.
The ArcaOS drives with JFS do not hang when selected.
Are you using 5.0.8 of the FAT32 driver? I am and cannot reproduce the issue but I may not have enough files on the volume or maybe I need directories with files or maybe without (nested, or I don't know).
-
I am using the fat 5.05 downloaded from the ArcaOS web site.
-
There is a switch for supporting EA's on the fat32 driver (fat32.ifs \eas). You might want to try with and without it to see if one or other helps. I always leave it off and haven't seen these hangs you describe...
Regards,
-
There is a switch for supporting EA's on the fat32 driver (fat32.ifs \eas). You might want to try with and without it to see if one or other helps. I always leave it off and haven't seen these hangs you describe...
Regards,
The /eas made no difference. The selection of the fat32 usb drive using the xcenter drive folder still hangs at the fat32 drive. This is with both Fat32 versions 5.04 and 5.05. Don't have version 5.08 to test. However, when selecting the fat32 drive from the drives folder on the desktop and it works ok.