PDA

View Full Version : Computer as cdplayer - USB/S/PDIF converters



Historicus
25-10-2009, 13:17
Here in Italy on an hi-fi fair we heared a comparison test between an expensive Spectral cdplayer and the sampe cd, ripped with a personal computer and played through an penstyle USB/S/Pdif connection.
The results are impressive. All the faults of the cdplayer seem solved.
That's why here many hi-fi stores don't sell any cplayer as mere mechanics anymore.
I still wait to test this penstyle device (actually it's sold out) and, if there is some interest, I could post here my impressions in future.
For this purpose I'll use an 24bits/192 kHZ DAC and some good vinyl rips at 24/96.
The most important thing is to bypass windows audio limitations (kernel mixer in directstound) that is now possible with drivers that interact in kernel streaming mode without downsampling the original files.
Do you have any experiences?
Here you can read something more about these USB/S/PIF devices:
Hhttp://www.m2tech.biz/products.html

http://www.head-fi.org/forums/f7/usb-spdif-converters-shoot-out-emu-0404-usb-vs-musiland-monitor-01-usd-vs-teralink-x-vs-m2tech-hiface-449885/

Marco
25-10-2009, 20:13
Ciao Piero :)


Here in Italy on an hi-fi fair we heared a comparison test between an expensive Spectral cdplayer and the sampe cd, ripped with a personal computer and played through an penstyle USB/S/Pdif connection.
The results are impressive. All the faults of the cdplayer seem solved.


I'm not surprised, as I've heard similar things myself! See my post #31 here for an explanation:

http://www.theartofsound.net/forum/showthread.php?p=75054#post75054

I would add that 'proper' CDPs from the heyday of the technology do not exhibit the faults that you mention (at least not to anything like the same degree), and so compare very well to the best computer streaming devices when both are being used with high quality DACs. It's to do with basic solid engineering, which is essential to get the best from CD.

The sad fact today is that you can no longer take it for granted that expensive CD players use top-notch disc transport mechanisms, nor are they always fundamentally well-engineered! :(

If you want to hear what CD is really capable of, and can't afford the likes of a Wadia or an Esoteric player, then buy a top Sony or Marantz CDP from the late 80s/early 90s and get a company such as Audiocom International to update some key components, and trust me you'll be stunned at the results ;)

I'd be very interested in reading your impressions of the "penstyle device" when you try it in due course.

Marco.

Historicus
25-10-2009, 21:18
Marco: I own the "panzer" Sony SCD 777ES as transport. The only thing I added is a good cable with trannies for galvanic isolation of the output of this player. Combining modern DACs to this player used as transport the results aren't so good as the original (sound is only louder but has less details, harmonics). The reason why I want to use a computer as source is because I want to play 24bits/96kHZ recordings directly without downsampling them to the cd standards (16/44.1).
Although my analogue chain with the original viniles, perfectly treated, sounds better then any cd. Unfortunately I use cds, as many does, only to warm up my gear. :-)

Marco
25-10-2009, 22:44
Marco: I own the "panzer" Sony SCD 777ES as transport. The only thing I added is a good cable with trannies for galvanic isolation of the output of this player. Combining modern DACs to this player used as transport the results aren't so good as the original (sound is only louder but has less details, harmonics).


I'm not surprised, Piero. The DAC in the 777ES is top-notch (not a classic TDA 1541, but very good nonetheless), and so most modern DACs would struggle to better it, musically, particularly with the way that Sony have implemented the design.

However, the 777ES CDP comes into its own when partnered with the Sony DAS-R1 DAC (which I use). When both are updated/modified by Audiocom they form a formidable partnership, and in terms of the 777ES, it processes musical data as accurately as the best computer streaming device, and with minimal error.

But it takes serious engineering in terms of transport quality and the use of the highest quality associated electronics to achieve this from a CD player or standalone CD transport. This is very rare to find these days and why so often computer audio devices outperform even the best CDPs, which is the point I was making :)

Marco.

Themis
26-10-2009, 19:09
I fairly understand the difficulties of a good transport and the advantages of a computer-based reliable transport.
Although, I consider that on pristine records these differences should be minimal, if any. And also, that for computer-based transports, the problem is shifted to the extraction process. Software solutions exist for error-free extraction (of non-pristine condition CDs), but, still, one has to use one of them. The transport itself cannot re-create lost (or erroneously interpolated) information.

On the other hand, when we separate transport and D/A conversion, we have to deal with the less-than-perfect S/PDIF or AES/EBU interfaces. Let alone optical converters. So, a separate dac is -by design- less accurate than a transport-integrated one. Unless we use an external master clock, but, then, how many hifi setups use such a thing ?
I also understand that modern dacs use several techniques (ASRC) trying to minimize jitter, and thus, they are more immune to certain types of jitter than they used to be.

But what I cannot understand, is why would I introduce an additional conversion (here an USB->S/PDIF) when I am able to buy a PC-card with S/PDIF output ?
Furthermore, how could possibly this additional conversion's downsides (added jitter and distortion) be more accurate than an average CD transport ?
Is there an explanation to this ?

EDIT: On the head-fi.org review, the reviewer states several times "big differences between usb cables". Needless to say, that the only possible explanation (if the reviewer is right) is added jitter and noise on the USB interface.

Historicus
27-10-2009, 12:39
A computer transport would make sense only to play 24bit/192 kHZ sources without converting them to the DVD-A format.
This USB/S/PDIF key has antijtter circuits but I understand it could be an unnecessary funnel.

Gentle Giant
27-10-2009, 14:14
I have s/pdif on my motherboard and a caiman. Whats the benefit with USB/S/PDIF converters?

John
27-10-2009, 14:53
Not everyone has S/PDIF outputs apart from that not sure

Themis
27-10-2009, 15:03
I have s/pdif on my motherboard and a caiman. Whats the benefit with USB/S/PDIF converters?
There is no benefit for you. Except if you consider that some additional jitter and/or distortion might produce a "better" sound. ;)

Historicus
28-10-2009, 06:21
I don't know if it could be truth but the companies that produce USB/S/PDIF converters sustain that a direct interface computer/DAC will cause jitter for the different clocks between the two devices. The converter instead that contans a "neutral" anti-jitter component would mitigate this phenomenon.
Anyway take these considerations with care. It sound to me like Aesop's tale of the fox and the grapes.

NRG
28-10-2009, 09:10
There is no benefit for you. Except if you consider that some additional jitter and/or distortion might produce a "better" sound. ;)

Well there might be a benefit in common mode noise rejection over distance as USB is a differential balanced interface, it's also impedance matched (to a degree, the tolerance is quite wide)....Single ended SPDIF with non matched phono connectors and variable impedance of the cable used, technically, it looks worse...it could cause the receiving device issues with noise and signal reflections.

Themis
28-10-2009, 09:43
Well there might be a benefit in common mode noise rejection over distance as USB is a differential balanced interface, it's also impedance matched (to a degree, the tolerance is quite wide)....Single ended SPDIF with non matched phono connectors and variable impedance of the cable used, technically, it looks worse...it could cause the receiving device issues with noise and signal reflections.
Most isochronous USB interfaces are adaptive, thus introduce more "problems" (high levels of jitter) than what the impedance matching is suppose to solve. Some -very rare- USB interfaces use asynchronous and, thus, are supposed to be better than the usual BB USB chips. See http://www.stereophile.com/digitalprocessors/ayre_acoustics_qb-9_usb_dac/ for some details.

I would understand the potential benefit if USB actually replaced S/PDIF (although implementations may vary, and real benefits are difficult to evaluate without referring to a particular setup) but... in our case, it comes on top of S/PDIF...
Moreover, most of these devices (M2Tech or others) clearly target their marketing on how much their S/PDIF interface is "superior" jitter-wise (whatever that may mean, and often without explaining that their "miracle" re-sampling is what all today's modern dacs use...) , often completely ignoring to add details about their USB implementation (à part from the usual blah-blah about "kernel streaming" and various "ASIO drivers"...).

At least, that's how I see it.