Hmm, didn't see that, one downside is that it's only 16 channels with USB, also the USB feature isn't working yet,
There is a different between USB 2.0 or USB 3.0 are a Firewire "killer" and whether it is a competitor. USB only has to crowd Firewire out of spots of the market to grow and stunt the growth of Firewire. I'm not going to be surprised if 16 channels may prove to a boundary but we have gone from "USB is not viable be used for pro audio work" to "USB isn't viable be used for more than 16 channel pro audio work".
. Anyway unfortunately almost every other high end USB 2.0 audio interface has fallen flat on its a**, as an example the MOTU 828MkII USB model just didn't sell, and had inherent problems that a technician from MOTU claimed was due to the USB spec, whereas the Firewire model didn't suffer from these issues.
That was be USB 2.0 spec. Not the USB 3.0 one. The SuperSpeed channel is separate (different set of wires ) , new, and not backwards compatible at the wire level. So whatever happened with with USB 2.0 is somewhat moot. USB 2.0 had some features that conflicted with isochronous transport (e.g., computer polling all devices for data and limited bandwidth. ). That has changed. Additionally, they bumped up the data rate for ischronous transfer from 128Mb/s to 384Mb/s . (
http://www.usb.org/developers/presentations/SuperSpeed_USB_DevCon_Isochronous_Froelich.pdf ) If USB 2.0 was having problems keeping up with the data flow of 8 deeply sampled channels then this should help.
Likewise with the significant increase in speed requires a significant increase in more accurate data transmission. (otherwise when required to do resends on corrupted packets your real data throughput goes substantially down. )
I would be shocked if the data delivery performance of USB 3.0 was the same as USB 2.0. There is extremely little to indicate that it will be the same. Even more so if users take steps to separate their real time data traffic solely on the SuperSpeed channel and keep the mundane device traffic confined entirely on the legacy USB 2.0 channel. They both should be able to run in parallel with minimal interference.
This thread is suggestive that this could be a "fixable" problem with the 828 .
http://www.motunation.com/forum/viewtopic.php?f=14&t=24657&start=60
Some folks get stability with the device and many folks don't. Given the combo of USB implementations, drivers , and data rates it is probably was just a bit too aggressive of a design. There is nothing here to suggest that given an improved USB foundation that still would not work. Getting it off the USB 2.0 channel would likely be an improvement all by itself. (one of tweaks folks were trying to removing adjusting tweaks that other USB devices inject to try to manage flow control. If have a real time network ... surprising when shared with random stream stuff that get hiccups and burps? Not really. )
Again until hardware manufacturers start offering USB3 devices the point is mute,
The notion that no one has USB 3.0 products is strange in the light that back in May Renesas announced:
"Having shipped over three million units of its" USB 3.0 controllers.
http://www.renesas.com/press/news/news20100520.html
Sure it is a small number relative to the PC market. However, it isn't that small if compare to the niche high end pro audio/video markets being discussed here.
It seems like a no brainier that any current USB 2.0 high fidelity product could be flipped into a USB 3.0 with a modest amount of engineering and likely get increased real time data transfer stability.
Right now is a bit early to drop a revised box. However, this time next year won't be.
but it's hard not to be cynical, as USB 2 promised alot and failed to deliver in the audio field.
The USB 2.0 standard was finalized in 2000. The USB 3.0 standard was finalized in 2008. So your cynicism presumes than in 8 years a team of smart folks could not sit down and fix the problems being reported by those trying to use USB 2.0 in a audio context while being given the freedom that they didn't have to remain wire backward compatible with USB 2.0. It is not like Intel and others didn't have folks getting problem reports from vendors.
I'd be more cycnical if USB 3.0 had come 2-4 years after USB 2.0. But USB 2.0 is old as dirt relatively on a technology lifespan scale. There has been more than enough time to figure out how to fix some of the problems and more closely match Firewire performance in this area.
The aspect that I'd expect problems one is whether the early USB 3.0 implementations get the isochronous features implemented well. Likewise, the driver and firmware (in remote devices ) support for the incremental changes they made to the isochronous feature set and API are correctly implemeinted in the early versions.
As far as being over hyped.... The all do it. Even Firewire....
" Firewire will supplant SCSI, IDE, RS-232C, parallel I/O ports and every other manner of digital interface between a motherboard and a peripheral device."
http://www.ausairpower.net/OSR-0201.html
Some thing with the current "Light Peak will replace everything" writeups that have appeared over last year or so .
As for audio over ethernet, the last device I recall that used any such thing was by Waves, and it failed miserably, again in part due to massive latency (and a lack of support unfortunately),
Most likely probaly too early with too old a version of Ethernet. It has been tried by several
http://en.wikipedia.org/wiki/Audio_over_Ethernet
Again this will most likely not work if don't have a dedicated, private, "audio traffic only" Ethernet network. Folks who say it will work if on same network that folks are web surfing on, send data to printers, and generally shipping lots of TCP/IP packets ... are ignoring lost of potential problems. Likewise, avoiding Cat-6 (or fiber if can get affordable enough transceivers ) and 1000BaseT is a mistake. Those should be minimal starting points if going to be doing 192kHz sampling and multiple channels.
Not cost effective just yet, because restricted to the 10Gbps Ethernet, even using proprietary layer 2 protocols is going away.
http://www.theregister.co.uk/2010/04/19/rdma_over_ethernet/
http://en.wikipedia.org/wiki/Remote_direct_memory_access
Ethernet is competing with Infiniband which is driving changes to the standard to enable solutions for low latency data transfer. Firewire was never a big enough competitor to trigger a move, but Infiniband is at the higher end. Firewire 400 is certainly going to be cheaper for at least a couple of years though.
Well that is good. The long XP and Vista drought though didn't help FW800 momentum. I guess there are more industrial controllers running windows with FW cameras than I though.
Won't be surprising to see Windows 7 ship with USB 3.0 drivers long before Mac OS X does also.
I can't see it usurping firewire and PCIe for professional audio, for as forward thinking the audio market can be, it still holds on to legacy stuff that works, MIDI being a prime example.
I can see it doing quite well in the small studio and/or live recording context. ( a quartet with 4 mics and 4 instruments recording an hour session at a venue to a laptop with internal SSD drive and a 8 channel box hooked via USB 3.0 ) . The fancy 8+ channel large studio much less so. Likewise complicated large stage shows. The size in the venue doesn't make it "pro" or not in my view.
However, I stand by my original point. Once, "legacy equipment" becomes the primary rational .... there is no "legs" any more. The momentum is gone. Just a question of hold long can hold out. It can be a long time.
I'm still waiting for Light Peak anyway, that really looks like the shiznitz
Lightpeak likely has some isochronous aspects because need to be able to not starve the various encapsulated protocols flowing across. IMHO, the hype factor on Lightpeak though is way higher than that on on USB 3.0 though. I have huge doubts it is the panacea folks are hoping it will be. it can be daisy chained but is that a flexible tree (like Firewire) or a star-tiered (like USB). Given is it Intel and borrowed aspects from fiber channel version of USB 3.0 it is very unclear which one it is. if it is the latter it is not going to do Firewire as that is a major protocol conflict.
The danger of Lightpeak is that Jobs will use it as an excuse to "Steve" firewire (kill it) across the Mac line up because it has fallen into his "It is too old and is loosing momentum" classification.
