DCAudioDIY.com

DC Area Audio DIYer's Community
It is currently March 28th, 2024, 6:38 pm

All times are UTC - 5 hours [ DST ]




Post new topic Reply to topic  [ 11 posts ]  Go to page 1, 2  Next
Author Message
PostPosted: June 16th, 2021, 11:11 am 
Offline
Site Admin
User avatar

Joined: February 28th, 2013, 10:38 am
Posts: 1682
USB seems to be the go-to interface for PC based music playback, but is it the right choice? The current vogue seems to be to add expensive PCIe USB cards when building a serious bridge. When you're talking that kind of money, there are options (mainly intended for use in recording studios) for PCIe to AES/SPIDIF/Toslink. Any thoughts on the pros/cons of these two approaches? :confusion-confused:

Roscoe

_________________
I can explain it to you, but I can’t understand it for you.


Top
 Profile  
 
PostPosted: June 16th, 2021, 11:24 am 
Offline
User avatar

Joined: January 13th, 2016, 9:14 pm
Posts: 787
I still believe that the I2S Bus interface is the better interface for an external DAC than USB or S/PDIF-AES/EBU. As long as it is an impedance controlled connection, works well. It is the interface that USB, S/PDIF and AES/EBU all have to end with as that is what the DAC uses for input.

Interestingly, many Intel CPU's that are SoC's have direct I2S Bus interfaces available to the DAC as well as I2C for DAC control. The RPi2 on also have it.

Only thing the others give is typically longer interface cables.


Top
 Profile  
 
PostPosted: June 16th, 2021, 11:28 am 
Offline
Site Admin
User avatar

Joined: February 28th, 2013, 10:38 am
Posts: 1682
HAL wrote:
I still believe that the I2S Bus interface is the better interface for an external DAC than USB or S/PDIF-AES/EBU. As long as it is an impedance controlled connection, works well. It is the interface that USB, S/PDIF and AES/EBU all have to end with as that is what the DAC uses for input.

Interestingly, many Intel CPU's that are SoC's have direct I2S Bus interfaces available to the DAC as well as I2C for DAC control. The RPi2 on also have it.

Only thing the others give is typically longer interface cables.


Unfortunately, getting to the I2S on most things other than an rPi is rather difficult. Not particularly widespread on DACs either, but still more common than PC availability.

Roscoe

_________________
I can explain it to you, but I can’t understand it for you.


Top
 Profile  
 
PostPosted: June 16th, 2021, 11:42 am 
Offline
User avatar

Joined: January 13th, 2016, 9:14 pm
Posts: 787
The Aaeon UP and UpSquared SBC's both have I2S Bus interfaces on their 40 pin GPIO connector. They have Intel Atom CPU's with SoC's and lots of I/O as well. MS Visual Studio 2019 has an x64 audio system driver to compile for learning how to program it.

Runs Windows 10 or Linux with a lot more capability than a RPi4. Linux already has support for the I2S Bus for the UP board. The UP is very similar form factor to the RPi's.

Fun stuff to experiment with.


Top
 Profile  
 
PostPosted: June 16th, 2021, 12:08 pm 
Offline

Joined: July 17th, 2016, 6:24 am
Posts: 1099
Roscoe Primrose wrote:
USB seems to be the go-to interface for PC based music playback, but is it the right choice? The current vogue seems to be to add expensive PCIe USB cards when building a serious bridge. When you're talking that kind of money, there are options (mainly intended for use in recording studios) for PCIe to AES/SPIDIF/Toslink. Any thoughts on the pros/cons of these two approaches? :confusion-confused:

Roscoe


As far as I can remember, Toslink was not considered highly for hi-def transmissions. Even SPDIF did not have a good reputation in audio community.

Here is the comparison between USB and SPDIF. USB Wins.
https://www.audiosciencereview.com/foru ... tter.1943/


Top
 Profile  
 
PostPosted: June 16th, 2021, 12:49 pm 
Offline
Site Admin
User avatar

Joined: February 28th, 2013, 10:38 am
Posts: 1682
Cogito wrote:
Roscoe Primrose wrote:
USB seems to be the go-to interface for PC based music playback, but is it the right choice? The current vogue seems to be to add expensive PCIe USB cards when building a serious bridge. When you're talking that kind of money, there are options (mainly intended for use in recording studios) for PCIe to AES/SPIDIF/Toslink. Any thoughts on the pros/cons of these two approaches? :confusion-confused:

Roscoe


As far as I can remember, Toslink was not considered highly for hi-def transmissions. Even SPDIF did not have a good reputation in audio community.


The toslink bashing was a result of the early optical transceivers being junk. That problem was corrected 20+ years ago. https://6moons.com/audioreviews/toslink/toslink.html

Quote:
Here is the comparison between USB and SPDIF. USB Wins.
https://www.audiosciencereview.com/foru ... tter.1943/


Maybe. Depends on the DAC as the article shows. He certainly didn't test enough DACs to say USB is never worse.... And, if you look at the test protocol, in EVERY case SPDIF included USB since he didn't have a way to get to SPDIF w/o using USB, so you wouldn't expect SPDIF to be able to be better than the USB feeding it... He's also using a very old USB to SPDIF converter, so his conclusions are hardly convincing...

Roscoe

_________________
I can explain it to you, but I can’t understand it for you.


Top
 Profile  
 
PostPosted: June 16th, 2021, 1:04 pm 
Offline

Joined: July 17th, 2016, 6:24 am
Posts: 1099
In real-time streaming like Audio, quality of the source and destination clocks and the clock synchronization are critical elements.

The USB clocks that are available on the motherboard and/or standard USB PCI cards are of very low quality. They are not accurate and they tend to drift. In USB communications, there are multiple paths (wires) for communication, so the clocks are synchronized every millisecond even in asynchronous mode. Where as in SPDIF and Toslink, there is no facility to either asynchronous communications or asynchronous clock synchronization. As the clocks drift, the quality of the audio degrades. This is not a big issue for real-time video as the video stream only determines the state of a single pixel at a time. 4K video has 4000 pixels per line and 2160 lines, and 16mil colors for each pixel. So the small drift in clock is unlikely be noticeable in video.

Hope that helps.


Top
 Profile  
 
PostPosted: June 16th, 2021, 2:14 pm 
Offline
User avatar

Joined: February 28th, 2013, 3:31 pm
Posts: 1780
Last weekend, the Philadelphia Area Audio Group had a zoom meeting with Taiko Audio, the builders of the $45,000 Taiko Extreme music server. They spoke at great length about the importance of each piece of the product. One of the things that was interesting to me was they thought that USB resulted in better sound than SPDIF or AES/EBU.

Tom


Top
 Profile  
 
PostPosted: June 16th, 2021, 4:38 pm 
Offline

Joined: February 28th, 2013, 3:31 pm
Posts: 363
I believe the main motivation is consumer audio equipment. USB is capable of handling 32 bit PCM and double rate DSD. The others top out below that level and anybody selling a dac these days (except Berkeley) has to include usb to compete on features and make it easy for anyone playing music from a laptop or PC. That said, usb implementation has gotten much better in the last decade. It can sound excellent.

I used to avoid it at all costs, but think it has the best sound for the buck at this point. You have a get into pretty pricey network streamers and/or dacs to do better and allow high rate DSD playback without usb.

Also keep in mind the subjective. I don't think anyone would argue the major improvements from the JCAT cards or similar are things like increased soundstage height and/or width/depth and texture. Better instrument placement in the soundstage, etc. If your speakers and room aren't going to let that happen, no reason to invest. I was excited to try one of Charlie's early modified CAPS with an LPS in my system as it made a huge difference in his compared to the bridge we were both using at the time. Needless to say, my horn speakers weren't going to create that huge soundstage behind the loudspeakers and it essentially sounded just like my existing bridge.


Top
 Profile  
 
PostPosted: June 16th, 2021, 5:00 pm 
Offline
User avatar

Joined: January 13th, 2016, 9:14 pm
Posts: 787
The advance with USB Asynchronous interfaces is that the DAC requests the data and has the low phase noise clocks to run the DAC close to the DAC IC. The DAC requests the variable size buffers of audio data at an 8KHz rate of varying size buffers. The USB receiver like XMOS or VIA handles the clocks and data buffer send to the DAC vis I2S Bus internally. The USB Audio spec gives a very good explanation of the process that I posted awhile ago.

S/PDIF and AES/EBU have to encode the data and clock together and then recover the clock and data streams to feed the I2S Bus to the DAC IC. The PLL clock recovery process in the interface has higher clock jitter than a low phase noise clock from the specs I can find.

The lower the clock phase noise (jitter) feeding the DAC IC, the better it works. We have gone from picoseconds of jitter to femtoseconds of jitter in some DAC systems. Some DAC's go to Rhubidium or Cesium standard clocks for lower phase noise. It all depends on the audibility of the jitter and how the DAC IC reacts to it in the design.

Everytime I hear an S/PDIF or AES/EBU vs USB design with the same DAC, the USB interface sounds better.

YMMV


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 11 posts ]  Go to page 1, 2  Next

All times are UTC - 5 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 59 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron
Powered by phpBB® Forum Software © phpBB Group