USB Sound Card devices???
Tom Diecidue
least 8 i/o) USB sound card devices? I currently have a Roland UA-100. It is
a USB sound device that works great with my laptop. However, it is limited
to 2 i/o's. I would like to continue to work on my laptop. It is just more
convenient.
Todd Sorg
multi I/O device for it. Especially 8 I/O.
Regards,
Todd Sorg
Digital Audio Labs
David Terry
http://home.sprynet.com/~jimroseberry/HTM/SNDCARDS.HTM
Look down at the PCMCIA Audio hardware section. Check out there main page,
they've got some cool stuff.
http://home.sprynet.com/~jimroseberry/default.htm
Tom Diecidue <diec...@paonline.com> wrote in message
news:7rr7ed$1...@hope.harvard.net...
Kevin Perry
> First some history. "Hydra" was the code name used for the open host
> controller designed by one of the companies originatinig the OHCI spec.
> The Hydra design is licensed for use by silicon vendors who require a USB
> host controller and find it a better for financial and resource reasons to
> license a design than design their own from scratch. This practice is
> very common in the silicon industry for widely used functions. The design
> is usually placed in whole into another IC such as large mulit-function
> south bridge chips used on motherboards or in simpler PCI-to-USB chips
> such as the OPTi device which is found on most retail add-in USB cards.
>
> The "Hydra bug" was discovered shortly after it was released to the
> licensees and the design was fixed by the company that originally designed
> it, who immediately released a version without the bug. Unfortunately,
> most of the licensees, for unknown reasons, did not update their design to
> the new version and now their are tens, if not hundreds of millions of
> defective systems in use around the world.
>
> Without getting too technical about it, the problem manifests itself when
> a low-speed IN packet is followed by a full-speed OUT packet. [Some
> explanation is probably in order here. USB has two signalling speeds: 1.5
> megabits per second (defined as low-speed) and 12 megabits per second
> (defined as full-speed). IN signalling is defined as device-to-host (also
> known as upstream), OUT signalling is defined as host-to-device (aka
> downstream). Most USB devices use full speed signalling. Low speed
> signalling is mostly used by Human Interface Devices (HID) such as
> keyboards, mice, and game controllers since their speed is limited by
> human input which is very slow, and the low-speed spec enables relatively
> significant cost savings for these very cost sensitive devices.] In a
> *Hydra* Open Host Controller when a low-speed IN packet is followed by a
> full-speed OUT packet the full speed OUT packet is corrupted; its data
> gets scrambled. Subsequent full speed out packets will be trans! mitted
> okay. There is an error correction mechanism in USB called cyclic
> redundancy check (CRC), unfortunately, the CRC is performed on the
> corrupted data, not on the data before it is corrupted, so the device is
> unaware that it has received corrupted data.
>
> The most common situation that exposes the Hydra bug by users is when
> they
> buy a USB mouse and move it while another USB device is functioning. For
> example, moving the mouse while speakers are playing music - the speakers
> will make a nasty noise, like pop, every time a corrupted packet is
> received. Since the mouse (low-speed IN) can send packets many times per
> second and there is a high packet rate to the speakers (full-speed OUT),
> the result is immediately perceptable and very annoying to humans. It is
> less so with other devices.
>
> Any device with full-speed OUT signalling is subject to the Hydra bug if a
> low-speed IN device is used at the same time. Here are a few devices with
> their signalling for *most* of their data transfers:
>
> Speakers: full-speed OUT
> Printers: full-speed OUT
> Scanners: full-speed IN
> Ethernet adapters: full-speed IN and OUT
> Mass storage: full-speed IN and OUT
> Cameras: full speed IN
>
> So, not only USB speakers are effected, but printers, mass storage (like
> Zip, superdisk, and hard drives) and networking devices will receive
> corrupted data. How a specific device behaves depends on the device.
> Speakers will make an annoying scratchy popping noise if the mouse is
> moved. Printers will either abort the print job (HP) or print a scrambled
> documnet (Epson). Scanners generally aren't effected - the data direction
> is upstream. The effect with networking and mass storage devices is
> insidious and ugly; the device and the user are unaware that the data is
> corrupted so the user can write bad data to their disk and not discover it
> until he/she reads it at a later time.
>
> The Hydra design, and the bug, is found in the OPTi device and in
> motherboard chipsets from ALi (Acer Labs Inc) and SiS (Silicon Integrated
> Systems). Don't expect to see bug free silicon from all of these vendors
> until the end of the year. Note that the only widely available chipsets
> for AMD K6-x systems are from ALi, SiS and VIA.
>
> There are other OHCI silicon suppliers for motherboard and PCI cards that
> do not have the Hydra design. These include Lucent, CMD, and the AMD
> chip
> set for Athlon. I've tested devices with these host controllers and they
> perform very well.
>
> In Darren's original post, he mentions that OHCI has problems with USB
> audio. I haven't seen that to be the case. What he is referring to is
> USB's "isochronous" mode, many speakers use this mode as it guarantees
> data will be sent every USB frame (one millisecond). Frame timing is
> important but I've don't have data on frame timing accuracy among
> different controllers. Any USB device should behave identically as
> perceived by the user regardless if it is connected to a OHCI or a UHCI
> system.
>
> Because there are many millions of Hydra parts in use, Microsoft took it
> upon themselves to develop a software workaround for the hardware design
> flaw. It is available in a Windows update, in Windows 98 Second Edition,
> and in Windows 2000 (NT5). Apple uses Open Host Controllers, and they've
> implemented a Hydra bug patch in their MacOS to cover the contingency that
> the device loaded on the board is one of the defective designs, so the
> user never sees the problem.
>
> Gary Pittmon
> Hewlett Packard
> _______________________________________________________________
>
> Jeffery Roberts posted Message 5022 in USB-IF Public Discussion Forum
>
> Dated : September 14, 1999 at 14:28:55
>
> Finally, an admission from the Big Cheese.
> I was right back in June, when I said the Win98SE made things worse with
> VIA chipset motherboards. Pardon me while I pat myself on the back and
> raspberries to all those who admonished me for my observations. Please
> note that the article URL only provides the information that the problem
> exists. The file itself is not there. Maybe this fix will work maybe not.
> Excuse my scepticism, I've seen lots of patches before. Well I'm off to
> dig out the actual patch. If and when I find it I'll let you know.
>
> Jeff
>
> Jeff
> : For those with the above, Microsoft has posted a knowledge base article:
> http://support.microsoft.com/support/kb/articles/Q240/0/75.ASP
>
> : Hopefully this helps with a lot of problems the VIA folks are having
--
Kevin Perry
Sonic Energy Authority
http://freespace.virgin.net/kevin.perry/
M.Pellerin
StudioPort AMX ... has two (2) TRS 1/4" Balanced Analog Inputs ... four (4) TRS
1/4" balanced Analog Outputs (in back) ... Digital S/PDIF I/O (in back) ... four
(4) MIDI In / four (4) MIDI Out (in back) for 64 MIDI channels ... combines all
features of DATportA+ and MIDIport 64 above plus more ... audio status and
clipping LED indicators, attenuation controls on analog inputs ... 19" rack
mountable with USB cables. OpCode's SMPTE to MTC conversion (all forms) and
unique Auto-MIDI-Thru (works with or without computer attached)... ).
Don't know what's going on with it.
Mario