Lorien,
in a direct comparation with M2Tech and Open source your XMOS modules is the winner.
http://www.diyaudio.com/forums/digi...2s-m2tech-vs-xmos-vs-open-source-qnktc-2.html
in a direct comparation with M2Tech and Open source your XMOS modules is the winner.
http://www.diyaudio.com/forums/digi...2s-m2tech-vs-xmos-vs-open-source-qnktc-2.html
USB Cable
I actually use minimum length possible. In my case about 6 inches.
Prior to WaveIO I experimented with cables. I found shorter is better. Better cable is better. Best was home made. I simply took the ends off a good quality USB2 cable with gold plated ends and replaced their cable with two runs of CAT5 twisted pair.. the type that uses Teflon dielectric. This was far superior. I found the WaveIO to be far less cable sensitive than the pcm2706 converter it replaces. I still does well on this DIY cable. Ultimate setup would be WaveIO tight to MB and DAC board.
I actually use minimum length possible. In my case about 6 inches.
Prior to WaveIO I experimented with cables. I found shorter is better. Better cable is better. Best was home made. I simply took the ends off a good quality USB2 cable with gold plated ends and replaced their cable with two runs of CAT5 twisted pair.. the type that uses Teflon dielectric. This was far superior. I found the WaveIO to be far less cable sensitive than the pcm2706 converter it replaces. I still does well on this DIY cable. Ultimate setup would be WaveIO tight to MB and DAC board.
Did you happen to print or otherwise save the diagnostic equipment screen shots showing the superior results? I suppose you performed all of the necessary USB compliance tests on your home made cables with the appropriate equipment...
come on.
Standard USB is often not even twisted, so transporting the four lines each as twisted pair having the common earth [probably?] at the sending equipment [Mac or AirPort] using CAT6 cable (even tighter wound and higher bandwidth than Cat5), using the foil shielded version even, should perform better - disregarding compliance tests that can only be done in the proper labs . . .
At least this was my plan - and solder the receiving end of the cable directly to the WaveIO board is how I fancy it.
wlowes did not suggest he'd taken measurements but his subjective report is credible on technical grounds.
The nominal impedance of a USB cable is IIRC 90 ohms and of a LAN pair 100 ohms, tolerances both about 15 per cent. IOW, they are close and overlap towards the limits with short lengths apparently making impedances less critical in any case. (BTW, if USB data pairs are not twisted, they are out of spec.)
Supplying the "B" connector's Vcc externally (i.e. no link to the host's Vcc) definitely makes for better sound as can, where practicable, omitting screening (lowers capacitance) and connecting host and DAC grounds direct, not via the USB.
These are strictly DIY tricks and not commercially practicable but they work well and cost nothing. I'm grateful to wlowes for the idea, first aired on AA.
Frankly, I think the usb cable discussion isn't unique to xmos/waveIO, perhaps its a big enough topic there should be a thread of its own on the subject if there isn't one already?
News from xmos has a new addition to their lineup of reference designs - USB Audio 2.0 Multichannel Reference Design | XMOS
Lucian we've been begging for a while for xmos based mulitchannel usb-i2s, what can you do for us?
News from xmos has a new addition to their lineup of reference designs - USB Audio 2.0 Multichannel Reference Design | XMOS
Lucian we've been begging for a while for xmos based mulitchannel usb-i2s, what can you do for us?
Lorien's gift to us is beautiful. His foresight in providing ways for the DAC and WaveIO to share digital power resolves all but the smallest amount of jitter. When that has been done by providing external power one can only credibly argue that alterations to any part of the bit delivery path (that includes USB) repair something that was not 'bit perfect' at the start.
I was referring to the B connector, i.e. the connection to the device, not to the host. The data lines, which the spec labels D- and D+, are indeed a differential pair.
I agree that a dedicated DIY USB cable thread might be interesting.
News from xmos has a new addition to their lineup of reference designs - USB Audio 2.0 Multichannel Reference Design | XMOS
Lucian we've been begging for a while for xmos based mulitchannel usb-i2s, what can you do for us?
I'd love to see that! A really high quality interface for active digital multiway systems would open so many doors at reasonable cost. Exa's device looks good but is pretty pricey for me
Last edited:
ah yes, Wolfsin can recall . . .
Providing synchronous stereo to headphones is good but I am truly skeptical of its value in multichannel scenarios considering the velocity of sound in air and the lack of precision in channel delays required. Multiple digital buffers of varying depths would require software management of much greater complexity than exist in the Thesycon drivers. Picosecond precision
While this may be the future, and it may capture a following this time, I remember all too well the promises being made in the 90's and how many friends still complain about elusive 'lip synch'. Any how many others went thru multiple versions of hdmi and . . .
It may be that the MP3 crowd will awaken, listen with astonishment to HDaudio, and not recall the last fleece-job of multichannel but even the most risk loving investors would raise an eyebrow at a business plan predicated on that chain of events.
Providing synchronous stereo to headphones is good but I am truly skeptical of its value in multichannel scenarios considering the velocity of sound in air and the lack of precision in channel delays required. Multiple digital buffers of varying depths would require software management of much greater complexity than exist in the Thesycon drivers. Picosecond precision
While this may be the future, and it may capture a following this time, I remember all too well the promises being made in the 90's and how many friends still complain about elusive 'lip synch'. Any how many others went thru multiple versions of hdmi and . . .
It may be that the MP3 crowd will awaken, listen with astonishment to HDaudio, and not recall the last fleece-job of multichannel but even the most risk loving investors would raise an eyebrow at a business plan predicated on that chain of events.
I'm more interested in the possibility for computer based eq/crossover of multi-way active loudspeakers paired with a multichannel dac. All of it is predicated on foregoing any non-computer based audio source. None of this will be cheap.
Surround sound may be fun to try too. I will likely leave the surround part to a HT receiver and focus on the L/R channels in isolation.
I wouldn't like to try to justify many of the diy projects on this forum as a profitable product/business when marketed to a non-diy consumer!
Surround sound may be fun to try too. I will likely leave the surround part to a HT receiver and focus on the L/R channels in isolation.
I wouldn't like to try to justify many of the diy projects on this forum as a profitable product/business when marketed to a non-diy consumer!
thesycon? yeah its unfortunate thats the best windows has to offer, its soundly beaten for free in Linux or Mac. why it has taken 3 years to still not have a reliable UAC2 driver for win i have no idea.
digital processing WILL induce delay, it may get beyond perception one day, but anyone who didnt figure out they had to accurately delay the picture to match the audio delay in the 1st ior 2nd generation was a bit slow on the uptake.....
the xmos is designed for digi XO's and pro audio mixers/instruments, not home theater
digital processing WILL induce delay, it may get beyond perception one day, but anyone who didnt figure out they had to accurately delay the picture to match the audio delay in the 1st ior 2nd generation was a bit slow on the uptake.....
the xmos is designed for digi XO's and pro audio mixers/instruments, not home theater
Of course. That makes sense. For pro / recording multichannel USB may have a market and for Hochopepers case.
Wolfsin has developed for Windows, Mac, and Linux at the app level and at the kernel level for two of those. Updates are handled differently in each case. IMO there is nothing wrong with thesycon for stereo.