I highly doubt those are the only two protocols Thunderbolt will support. Maybe in the first computers it will only support this but I imagine Intel and Apple are thinking way ahead on this.
This is from the PDF on Intel's web site.
"For time-sensitive data, such as video and audio during creation
and playback, data transfers can be critical to the success of the
work. Thunderbolt technology was specifically designed with
video and audio applications in mind with inherently low latency
and highly accurate time synchronization capabilities.
For
...
I will begin posting some board images to get user feedback soon. The board will be bus powered. It will be stereo. Also the idea is to keep it Opus sized if possible, so I won''t be adding any unnecessary stuff.
...
I was wondering where you are at with the boards. Any chance on posting them?
Russ has mentioned that he's had to completely redo the boards to work with the L2 chip. The software for the L2 is also different to the L1, which is taking some time to develop. From what I understand the L1 software was pretty much a reference design already produced by Xmos to demonstrate the chip, whereas the L2 is not. The bulk of the code should be similar, but working it to work with the different chip could be a bit of a pain, especially perhaps if the second core is being used.
I thought the swap was mainly based on the number of inputs/outputs that the L1 had, rather then needing the extra processing power.
From what I've gathered this swap was necessary to work with 8 channels?
Russ has also mentioned that there will be isolation and that it won't use optos, I'm interested to see what this is!
I thought the swap was mainly based on the number of inputs/outputs that the L1 had, rather then needing the extra processing power.
From what I've gathered this swap was necessary to work with 8 channels?
Russ has also mentioned that there will be isolation and that it won't use optos, I'm interested to see what this is!
There's an L2 reference design as well as an L1 design. I'd be somewhat surprised if the L01A-LQ64 parts are IO bound, much less anything larger, though it does depend on how Russ has the board set up. See the schematics in the reference designs' hardware manuals.
Caps would be the simple, cost effective choice. LVDS magnetics or similar would work too.
The choice for L2 over L1 has to do with supporting more than 2 channels at 192khz or more. 
As for the isolation, it is not caps. And it will not be on the USB board itself. It is a separate module.
But the more I research the more I think I will stick with stereo for now. I am still very much interested in this coming to fruition.
As for the isolation, it is not caps.
And it will not be on the USB board itself. It is a separate module.But the more I research the more I think I will stick with stereo for now. I am still very much interested in this coming to fruition.
As for the isolation, it is not caps.
Is it bigger than a breadbox?
So is everyone else.
Works for me, as I expect I'd be following the L2 with an ADAU1445.As for the isolation, it is not caps.
Doesn't work for me as I have a requirement for six channels of ASIO input. 44.1 would be fine though---the 1445's not short of ASRCs. Any chance of offering say, stereo @ <up to some high rate>, 4 channels @ <up to some high rate> / 2, 8 channels @ <up to some high rate> / 4?
The separate module(s)..... w/ I2S MUX (+ differential transmission modules for I2S (+I2C) over HDMI) and magnetic isolationAs for the isolation, it is not caps.
I am assuming that you're going to be using the 1445 to create the crossover filters for an active loudspeaker setup. May I ask why you need 6 input channels? Surround sound?
You do realise that you only need to provide one stereo I2S stream to the 1445, then inside the DSP use the T connector, in sigmastudio, to split this off into as many stereo pairs as you want.
A combination of per channel delay equalization and capabilities and pricing of currently available hardware. Time reversed IIR is not feasible on streaming DSPs and I've been unable to identify DSCs which can support it without requiring a little too much pain in some aspect of the implementation for me to find it worthwhile. A VisualDSP++ license for the audio SHARCs is probably the most elegant solution but is out of my budget. And, even if it were affordable, it wouldn't be competitive with a fanless PC on an NRE basis. The moreso in my case as I've already sunk the cost for reverse time VSTs and setting up Firewire audio interfaces.
Since my reverse time processing is unlikely to move off the PC any time soon it would be nice to simplify the ASIO -> Firewire -> audio interface -> SPDIF -> I2S part of the playback chain. Unfortunately xmon's current driver cost structure requires a DIY demand aggregator like Twisted Pear or exaDevices to be competitive. I could write my own device driver but, as that's not a way I particularly care to spend my free time, it's likely one of the two companies---if not both---will beat me to it.
I too had hoped it would be multi (8 or more) channel to do 4 way crossover and equalisation on the source PC which will be under utilised and capable of very fine meshed equalisation, rather than adding the complexity of an AD144x between receiver and DACs.
I was hoping to keep my build all TPA for simplicity but the exaDevices product would meet my requirements. As I am still in the planning phase of my project, I can only hope TPA xmos USB mkII might be multichannel and released before I finish building the rest.
I was hoping to keep my build all TPA for simplicity but the exaDevices product would meet my requirements. As I am still in the planning phase of my project, I can only hope TPA xmos USB mkII might be multichannel and released before I finish building the rest.
Can't wait to see some pics!Lets just say I am planning an order of prototype PCBs in the next few days.
Can't wait to see some pics!
waiting is the one thing you should prepare for!
and you´re not alone
best
Leif
- Status
- Not open for further replies.