>> > I'll need some time to check what is needed for the Fast Track Pro. >> >> In fact it's not that tricky: just install a usb sniffer under windows, >> then play with the control panel and check the SELECT_INTERFACE >> messages for Interface/AlternateSetting. With the audiophile USB, it >> only changes: >> * if you change the bit depth >> * if you change the sampling rate >> * if you enable/disable the SPDIF input >> > > I'll do some investigations on that this weekend. Also note that when the SPDIF input is enabled, the device might try to synchronize to the external digital signal. This leads to additionnal synch errors when no signal is read on this digital input. >> > I think there is a big chance that many things are handled the same >> > way like Audiophile. At least the number of Channels and also the >> > mapping of digital an analog interfaces are the same. I tried to >> > connect directly to the 24-Bit mode after startup of the USB device >> > and I had the same "wrong" signal on both channels. This makes me >> > positve to believe that the byte alignement (like you described) gets >> > mixed up after mode switching. >> >> Yes that sounds exactly the same... >> >> > The distortion I head on both channels could have something to do >> > with endianess. Instead of a sine wave I had a 3rd Harmonic at level >> > of about -55dB (in the range of one byte) and the noise level was >> > about -80dB (the range of the second byte). >> > > I can say now a bit more about the actual signal I get. It has not only to do > with little or big endian. > > One problem is that it is dependant on something todo with initialisation. So > the signal I get is not allways the same. My problem was also dependent on initialization, and I didn't get extactly the same result each time: see https://bugtrack.alsa-project.org/alsa-bug/view.php?id=1801 > Sometimes even the working 16-Bit > signal is messed up. We talked about that the byte ordering might be > mixed up. I hadn't any problem with 16bits signals but I remember having seen some quirks for my device in the code in order to select usb interfaces by order of preference depending of their synchronization (adpative/synchronous, asynch). However, these quirks are no longer needed if the device is correctly initialized with device_setup because adaptive/synchronous interfaces are correctly choosen depending on the SPDIF input status. If the Interfaces/AltSets mappings are exactly the same for your device and the audiophile usb, why not try to recompile the usbaudio.c driver with: * the big-endian patch (adding your usb_id to the previously mentionned function) * add your usb_id (already used for the audiophile usb) to the following code: + + /* audiophile usb: skip altsets incompatible with device_setup + */ + if (chip->usb_id == USB_ID(0x0763, 0x2003) && + audiophile_skip_setting_quirk(chip, iface_no, altno)) + continue; > Basically I got to differt sorts of signals. I recorded a sine and > looked at the actual samples. > Lets assume that if we have little endian the right samples would be > organised > like that: > L1-L2-L3 R1-R2-R3 > (L1 is left least significant byte and R3 is right most sig. byte) > Can someone comfirm that this would be right? > > wat I get look like that: > L3-L3-L3 R3-R3-R3 in one case and > L3-L2-L3 R2-R3-R2 in the other case Could this be linked to the fact that the Interface/AltSet used is in fact synched to the inexistant signl on SPDIF input ? > Sometimes I'm not shure who (which application) I can believe or not. > Does anybody > now a way to check what realy comes from the USB device? Under linux if you eable debugfs and usbmon at the kernel compilation time you can read the usb bus data: see http://www.mjmwired.net/kernel/Documentation/usb/usbmon.txt Regards, Thibault ------------------------------------------------------------------------- Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job easier Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 _______________________________________________ Alsa-devel mailing list Alsa-devel@xxxxxxxxxxxxxxxxxxxxx https://lists.sourceforge.net/lists/listinfo/alsa-devel
