On 02/13/2010 12:01 PM, Rick Kunath wrote: > On Saturday 13 February 2010 02:38:24 pm Leigh L. Klotz, Jr WA5ZNU wrote: > >> Dave's point about sample rate conversion is a set of "favors" being >> done for users by the API or mid-tier software, not the USB Audio driver >> itself, which is essentially the same across all USB cards regardless of >> what chipset is used or how much it cost. Choosing OSS or PortAudio >> lets you decide which mid-tier software is doing you which favors. Most >> of the favors there are done for music recording, and the performance >> and operations decisions aren't necessarily optimal for ham use. >> > My question Leigh is whether or not a USB sound device can handle multiple > sample rates natively just as a PCI card can that can support these sample > rates natively? > Stelios or Dave would know for sure, but based on the time I've spent in the Linux driver code, the only way to do this would be with a device which produces multiple independent data streams from more than one ADC. Sampling at the native rate and then converting in one of the mid-tier APIs is the only way I am aware of, as the simple USB devices do not produce multiple streams of data. The EMU-0404 and EMU-1616 have multiple analog inputs, and produce multiple channels, but those are PCI devices and I don't know if you can configure them internally to split a single input into multiple ADC channels. > I'm assuming so based on what you wrote. > > I guess the question then is how to identify good USB hardware for ham use? > The CPU load is going to be pretty much the same for all USB devices. So you're looking at the distortion figures and dynamic range as the main differentiator. I use a Creative iMic (which is a branded version of a no-name; they change the chipset regularly) for fldigi, and an EMU-0202 for Quisk panadaptor. The sampling rate of the '0202 is up to 96Khz, but high sample rate isn't necessary for fldigi.
Here's N8LP's review of sound cards: http://www.telepostinc.com/soundcards.html It's biased toward Windows, but many of the hardware specs should be the same, as I suspect he understands how to avoid software resampling. The ARRL also had a two-part set of reviews in QST a year or two ago. > Another thread has some discussion about the mixer elements of these USB sound > devices, or in this case, the lack of. Wondering how that's handled for card > inputs and outputs? Also seemed to not be supported in oss-alsa emulation, > i.e. there was no /dev/dsp1 mixer created for the card. Though I don't know if > the USB device was aliased correctly or not. > > Rick Kunath, k9ao > This level of Linux sound mystifies me, though Stelios has tried to explain it to me several times. Leigh/WA5ZNU _______________________________________________ fldigi-alpha mailing list [email protected] https://lists.berlios.de/mailman/listinfo/fldigi-alpha
