[tosehee]

So.. I thought about this source/transport for awhile, but can't seem to understand the logic behind this..

The reason why we want the USB is to provide the timing issue and avoid the SPDIF in the chain. However, this very device does nothing more than providing another SPDIF to the DAC since it can't provide I2S feed directly to DAC.

Why is this any better than having a direction connection from SC's optical/coax? This one comes with better timing device in the converter?

 

[gattari]

Quote:

Originally Posted by coolfungadget /img/forum/go_quote.gif 外置USB声å¡(USB Sound Devices) - MUSILAND ä¹ä¹‹é‚¦ ä¸*国 å®¢æˆ·è®ºå› - Powered by Discuz! You need Google Translate to read it. But I guess I will post a notice here if they release anything new.

How to register? is a mission impossible

for me

 

[coolfungadget]

Quote:

Originally Posted by gattari /img/forum/go_quote.gif How to register? is a mission impossible for me

pm me.

 

[mmerrill99]

Quote:

Originally Posted by tosehee /img/forum/go_quote.gif So.. I thought about this source/transport for awhile, but can't seem to understand the logic behind this.. The reason why we want the USB is to provide the timing issue and avoid the SPDIF in the chain. However, this very device does nothing more than providing another SPDIF to the DAC since it can't provide I2S feed directly to DAC. Why is this any better than having a direction connection from SC's optical/coax? This one comes with better timing device in the converter?

I agree that using SPDIF adds back some jitter,; that's why I went with the US version to get at the I2S signals. Having said that, this is a new USB interface & another option to using PCM27XX, TAS1020, TAS30?? chips. Whether this is shown, in the long term, to be better than products with these offerings, remains to be seen. Early indications are good but implementation is everything!

 

[tosehee]

There is a jitter from PC to USB Device. Then, there is another jitter from USB device to whatever DAC you are attaching to.

In short, I think this will add more jitters unless the timing chip on the USB device has superior reclocking and etc. Even then, I don't know if this is superior to optical straight to DAC.

Just my opinion without any formal testing and etc. Musiland is on my wishlist. When the mac driver comes out, I am definitely going to try out myself.

 

[slim.a]

Quote:

Originally Posted by jtam /img/forum/go_quote.gif According to Musiland, the master clock is generated by some algorithm in the chip instead of a simple PLL to achieve low jitter. As for their claim of +-1 ppm, it means the FPGA chip only creates +-1 ppm jitter when processing the clock from the 24MHz crystal. The final jitter actually depends on the crystal...

Hi guys,

Do you think that by replacing the 24MHz crystal with an ultra low jitter clock, we could get the overall jitter down to 1ppm ? If this is true, a modded Musiland should be the best usb to spdif around ? Am I wrong ?

 

[mmerrill99]

ppm is not a measure of jitter but of precision - this figure, when quoted for clocks, means that over a long time period (months/years?) the frequency will not have changed by more than 1 part per million. This is not relevant to audio or important for good sound. I don't know what the Musiland figure actually means - I think they use it because the Chinese clock manufacturers use ppm as a measure of clock quality.

I looked into Xilinx FPGA & Jitter and found that worse case figure quoted for jitter added by DCM (Digital Clock Management) in FPGA is 100ps. This is a pessimistic figure & real world is probably a lot less than this. Now Musiland may have a clever way of minimising this. We won't know unless they publish real jitter figures or somebody tests it.

So this jitter is added to the clock jitter or in this case crystal jitter to get the total jitter.

I had the same thought as you & figured a better clock might bring the jitter to shrinkingly low values. But I paused, because the clock is first fed through the Cypress USB controller chip (EZ-USB) and then onto the FPGA. Jitter may be added at this stage, I don't know & if so it might swamp the amount of clock jitter.

Having said all that, some early signs are that this is a low jitter unit - some measure <100ps - listening results indicate a low jitter sound.

So maybe it is worthwhile to change the crystal for a low jitter clock but the crystal itself could be low jitter, we don't know!

 

[slim.a]

Jkeny, tahnks for looking at the jitter of the Xilinx FPGA.
Even if the jitter is around 100ps, it is a very good number for usb to spdif device.
Looking at stereophile measurement of usb dacs, most of them (except the benchmark dac1) have high jitter from their usb input, usually higher than 1ns. They also measured the jitter from the digital out of the EMU 0404 usb at 8ns.

 

[mmerrill99]

I agree slim.a but let's wait until we have some agreement on jitter figures and the test bed etc. This is a notorious mine field. I guess sending a unit to Stereophile would give a comparative jitter result.

I'm hoping that in stock form it proves to be a very low jitter unit & I would look at the PS as the first area for improvement. Replacing USB power with external clean PS can only improve everything, including jitter - this will be my first mod. Possibly , separate power to crystal would be another consideration. I2S from FPGA, next & then maybe low jitter clock.

 

[DoYouRight]

I hope these stay cheap for those not yet onboard

 

[mmerrill99]

Yeh, DYR, I predict that these will be like the T-amp Tripath wave - not the ultimate in sound quality (and they'll have lot's of detractors) but getting most of the way there for very little money!

Shh, maybe we should go quiet now & buy up all we can

 

[slim.a]

Quote:

Originally Posted by jkeny /img/forum/go_quote.gif Yeh, DYR, I predict that these will be like the T-amp Tripath wave - not the ultimate in sound quality (and they'll have lot's of detractors) but getting most of the way there for very little money! Shh, maybe we should go quiet now & buy up all we can

I am going to order a second unit just to be safe

 

[K3cT]

I'm intrigued.

Will putting this between the Audio-gd Compass and my laptop to make use of Compass' SPDIF input optimize overall audio performance?

 

[slim.a]

Quote:

Originally Posted by K3cT /img/forum/go_quote.gif I'm intrigued. Will putting this between the Audio-gd Compass and my laptop to make use of Compass' SPDIF input optimize overall audio performance?

Yes it will probably improve upon the USB input of the Compass.
I have an Audio-gd DAC-100 which is basically the DAC section of the Compass and I had very good results putting the musiland between the computer and the spdif input of the dac. You just have to use it with a decent quality spdif cable.

 

[mmerrill99]

OK, I need to correct my jitter figures above and add some info:
- the maximum allowable jitter on the input clock is +/-300ps (so this defines the worst acceptable crystal jitter)
- the actual worse case added jitter is 1% of the input clock period (CLKIN) + 150ps
- so if the input clock is 100MHz, CLKIN is 10ns; 1% of this 0.1ns or 100ps added to 150ps for an absolute worse case of 250ps
- This is then added to the input clock jitter to give the total jitter!
- But as I said above this is with all gates firing and at a much higher speed than the 48MHz that this FPGA is operating at.

Just so as I didn't mislead anybody with optimistic figures!