XMOS-based Asynchronous USB to I2S interface

highres spdif in kernel streaming?

Hi, Lorien, Thanks for your reply!

Did not get the E-mail.(until now...)
I work on W7 Ult. 64 bits.
Player is XXHighEnd , latest version.
I managed getting it work up to 24/192! In Wasapi however.
Trying in KS does not work higher than 16/44. Which is a pity because I miss a lot of settings in the program to let it sound optimal.
Maybe I2S-coax works in a different way. I will need some more time to work that out. I'll let you know,
All the best,
Ed

Maybe somebody out there expierenced the same as I do??
Asio or Wasapi no problems.
On KS only 16/44.
 
now high res when using KS

Hi, Lorien, Thanks for your reply!

Did not get the E-mail.(until now...)
I work on W7 Ult. 64 bits.
Player is XXHighEnd , latest version.
I managed getting it work up to 24/192! In Wasapi however.
Trying in KS does not work higher than 16/44. Which is a pity because I miss a lot of settings in the program to let it sound optimal.
Maybe I2S-coax works in a different way. I will need some more time to work that out. I'll let you know,
All the best,
Ed

Hi,
I should add to the above story that I CAN choose up to 192Khz, though it is not possible to choose higher bitrates than 16.(In KS)
Ed
 
HI Lorien

I received the waveIO today. Thanks for the excellent work. Both my Mac and Win 7 recognized it without problem.

I connect it to Ian's FIFO via I2S, I can see the "lock" LED on FIFO but the BIII does't lock. Test the signal path by Ian SPDIF and everything work fine. Any advice to diagnose the problem?
 
Anybody can give us the number of sample/buffer/latency of the last XMOS hardware with the best ASIO drivers available..

I want to buy this - http://www.minidsp.com/images/documents/Product%20Brief%20-%20USBstreamer.pdf

******* tired of companies not talking about asio latency of their drivers like if nobody is interested...

Also what about jitter of the xmos chip... (since it need to be the master clock)

I also need to know the driver company.. like centrace or others..
 
Last edited:
Anybody can give us the number of sample/buffer/latency of the last XMOS hardware with the best ASIO drivers available..

I want to buy this - http://www.minidsp.com/images/documents/Product Brief - USBstreamer.pdf

******* tired of companies not talking about asio latency of their drivers like if nobody is interested...

Also what about jitter of the xmos chip... (since it need to be the master clock)

I also need to know the driver company.. like centrace or others..

XMOS interfaces do not need a driver on OSX or linux platforms, as the XMOS is USB class 2 audio compliant. Third part drivers are only required on Windows OS, since Microsoft does not feel the need to compy with the USB class 2 audio standard.

Not sure what you mean about the jitter of the XMOS chip? Jitter at the output of the Wave IO as far as your DAC is concerned, will be the jitter of the masterclock oscillator as the MC is sent directly out over I2S. Of course there will be slight additional jitter if you use the isolated outputs (~20 pS or thereabouts), but the straight outputs should be just the inherent jitter of the MC, as the XMOS chip does not alter the clock. I would be more concerned about keeping the I2S lines to the DAC as short as possible, and terminated as well as possible. I suspect the mini bnc cables and jacks are good in this regard.
 
Member
Joined 2004
Paid Member
Its a well thought out design and supports 8 channels, which is unusual. It does not currently support external clocks but that may be a simple firmware tweak to enable.

Jitter must be evaluated as a system. Starting with low jitter is critical, but it can be "screwed up" easily in implementation. The clocks will pass through the XMOS chip in this design (although even that can be altered with firmware I hope) and a buffer on a noisy ground is all it takes to degrade SOTA jitter levels 20+ dB. Depending on the particular DAC the I2S lines may have little influence on the jitter as long as they meet the setup timing requirements.
 
I didn't try it yet.
I can't track your card due to the fact that there seems to be some problems with IT post database. Do you have any news about the delivery?

HI Lorien

I received the waveIO today. Thanks for the excellent work. Both my Mac and Win 7 recognized it without problem.

I connect it to Ian's FIFO via I2S, I can see the "lock" LED on FIFO but the BIII does't lock. Test the signal path by Ian SPDIF and everything work fine. Any advice to diagnose the problem?
Well, I usually try to keep things simple so, for debug only, I would advice you to try to wire WaveIO I2S output port directly to BIII I2S inputs. In addition, if you are using the isolated I2S port, make sure that it's has 3.3V power applied between V+ and "Isol. GND" pins.

Cheers,
L

EDIT:
modules arrived!
Thank you for letting me know! I'll mark them as delivered!
 
Last edited:
I can't track your card due to the fact that there seems to be some problems with IT post database. Do you have any news about the delivery?

I got it last monday. The pcb is beautiful :)
I'm waiting to build the psu, as I wont even hook it up usb-powered. ;)

I also use xxhighend player, so KS at 176.4k is very important to me as well, and I hope the 24 bit issue will be resolved soon. Til then i'll use only hardware attenuation.
 
You didn't awnser me about latency of the asio driver and how much sample buffer it can be set without failing..

DON'T YOU UNDERSTAND.. wish people would try to understand what all people producing music do... when they sell stuff to them..

it's not only audiophiles anymore... comprende

low latency asio driver running at 64-128sample is a must
also playback with asio should take the least amount of CPU..

it actually need to compete with RME and LYNX soundcard for the asio drivers, stability and CPU usage..

Also about jitter your saying there will be jitter induce by the master clock but your XMOS board have the master clock and force you to use this one.. So i guess it depends what people used as master clock on their xmos board.. it's kinda limiting to force the use of the onboard master clock..
 
hmmm...

You didn't awnser me about latency of the asio driver and how much sample buffer it can be set without failing..

DON'T YOU UNDERSTAND.. wish people would try to understand what all people producing music do... when they sell stuff to them..

it's not only audiophiles anymore... comprende

low latency asio driver running at 64-128sample is a must
also playback with asio should take the least amount of CPU..

it actually need to compete with RME and LYNX soundcard for the asio drivers, stability and CPU usage..

Also about jitter your saying there will be jitter induce by the master clock but your XMOS board have the master clock and force you to use this one.. So i guess it depends what people used as master clock on their xmos board.. it's kinda limiting to force the use of the onboard master clock..

It sounds like you are looking for a pro audio soundcard, perhaps Lorien created the Wave IO for DIY audiophiles, not for recording engineers? It just seems like this is not the product for you, nothing wrong with that, it just was not created to suit your use.
 
Member
Joined 2004
Paid Member
USB is not ideal for low latency applications. The USB architecture imposes a minimum latency that is much higher than PCI can achieve. I believe its 1 mS. It may be possible to get it to 250 uS. This seems to be an issue with PCIe as well to a much lesser degree and may just be intrinsic to serial interfaces. No easy or even hard work-arounds. The high speed serial interfaces are unidirectional in their communications with a certain amount of time to process and react required for a full duplex communication. All of the new interfaces (USB3, Thunderbolt, PCIe, etc.) have this same architectural limitation.
 
@audiolic I apologies if I didn't answer to your question earlier! There are few reasons for not doing this: in your earlier duplicate post (well, almost) you are referring to a product that it's outside of my knowledge, moreover, you're upset on companies that are not willing to share those infos that you are asking about. Well, please note that I'm NOT a company, just a simple guy that loves what is doing and wants to share it with rest of those who wants to accept it.
About recordings, I want to remind you that WaveIO card wasn't built for it just for playback only! (it doesn't have I2S DATA inputs) and, speaking of latency.. if I have to reclock a thousand time all the I2S signals just to met what I believe I have to achieve then I'll do it without care too much on latency.

it's not only audiophiles anymore... comprende
it is for me, according to my believes :)
In the end, I suggest to address your request to those who are producing the card in question or their driver's manufacturer hoping that they will answer you...

Kind regards,
Lucian
 
Last edited:
Why should you use a 3,3V PSU? Preferably, give WaveIO a good, separate 5V low-noise PSU. You could take the 5V from the Placid, but then consider to place larger heathsinks on it. I've build a separate PSU with Belleson superregulators. It works flawlessly. If you follow Lucian's advice you can't go wrong:cool:
kind regards and success with your build
Henri