On Fri, Nov 01, 2002 at 10:47:33AM -0500, Mark T. Valites wrote: > > I'd like to hear this confirmed from other sources, but I think the > > problem is that the audio output on the drive is not connected to the > > sound hardware on the motherboard. It appears that the cs4231 chipset > > supports this, but I have never seen a stock Sun connected this way, > > and I've had trouble even finding motherboard jacks for it (SS5, U1, > > U10). You probably could get XMMS to play -- I think there's an > > option to extract the digital audio data and dump it to the sound > > hardware (effectively, ripping it to play it.) I gave up and just > > ripped all my CD's to MP3's, which is more flexible anyway. > > Are we talking something isn't soldered together in there? Or that ultra > 5's don't ship with an audio cable, or something else entirely? Hmm. I'm at work, and just happened to have an open Ultra 10 nearby (which has the same motherboard as the U5, IIRC). It does have an audio cable from the CD drive to the motherboard. Which means at worst, all you'd need is the cable (no soldering required), and it may already be in place. (I know I couldn't find a motherboard jack on my SS5, and I don't think I could on the Ultra 1, either.) The problem is that I'm not sure how you would tell the cs4231 sound hardware to route the input from the CD to the output jacks. Sparc sound hardware appears significantly different from PC hardware. PC's have a real "mixer", in the sense that there are separate volume controls for each sound source (DSP, simple synth, CD, etc.) plus a master, for both play and record. And the output is delivered to all output jacks. The Sparc hardware (implied from the output of 'audioctl -a') has controls for the output (play.*) and input (record.*) of the DSP -- you get to pick the level, and any combination of the output and input ports / sources (but it's just on / off; for instance, there's no way to record the line_in signal at half the gain of the microphone signal, which you *can* do on a PC.) Hmmm. Sometimes just talking about it makes the answer clear: There's a monitor_gain setting. I'll bet that whatever you have selected for record.port is routed to whatever you have selected for play.port, with the level controlled by the monitor_gain setting. I'm not in front of a box that I can try this on right now -- I'll try later. If you beat me, try: audioctl -w monitor_gain=128 output_muted=0 play.gain=128 \ play.port=0x1 record.port=0x4 Then play a CD. Later, Rob.
Attachment:
pgp4REgWWvPYZ.pgp
Description: PGP signature