Head-Fi.org › Forums › Equipment Forums › Computer Audio › Help me understand USB to S/PDIF converters?
New Posts  All Forums:Forum Nav:

Help me understand USB to S/PDIF converters?

post #1 of 15
Thread Starter 
Hi,

I was recently having a chat with an experienced Headfi poster regarding my system which I am in he process of upgrading (new amp) and they were advising me that I might be limiting my systems performance because of my DAC's onboard USB-S/PDIF interface.

They were recommending that I purchase a separate USB-S/PDIF converter to place between my iMac and DAC like these:

http://www.audiophilleo.com/
http://www.berkeleyaudiodesign.com/products.html
http://www.empiricalaudio.com/products/off-ramp-converter

My current DAC, a Grace Design m903 has up to 24/192 USB class 2 audio and asynchronous transfer mode licensed from Gordon Rankin (Wavelength Audio capability http://www.usbdacs.com/Concept/Concept.html). It is my understanding that this was considered very good technology in terms of jitter management/elimination.

Current Configuration:

iMac running iTunes or Decibel playing material between 16/44 to 24/96
Grace Design m903 connected via USB
DNA Stratus 2A3 connected via unbalanced cables
Sennheiser HD800s

One of these units would go between the iMac and the DAC with USB into the converter with S/PDIF cable connected to the DAC. Can someone kindly explain to me why these units will improve my sound over my units current inboard capability?

Confused,

Bob
Edited by iim7V7IM7 - 10/9/12 at 3:10pm
post #2 of 15
Quote:
Originally Posted by iim7V7IM7 View Post

Hi,
I was recently having a chat with an experienced Headfi poster regarding my system which I am in he process of upgrading (new amp) and they were advising me that I might be limiting my systems performance because of my DAC's onboard USB-S/PDIF interface.
They were recommending that I purchase a separate USB-S/PDIF converter to place between my iMac and DAC
[..]
Can someone kindly explain to me why these units will improve my sound over my units current inboard capability?
Confused,
Bob

 

They won't. Don't waste your money - there will be no difference. Not for the better anyway, unless you spend more on the interface than your DAC is worth. All you would achieve is a placebo effect similar to the benefits of a high performance power cable. I'm probably gonna get flamed for this statement :)

 

This is digital technology and numbers are numbers. If someone is telling you that 2+2 is smaller than 2+2+0 you know what to do with a statement like that. Yet in this field we see more witchcraft and superstitions than in many fantasy movies. Very often people confuse symptoms with reasons and draw wrong conclusions. The myth of superiority of S/PDIF over USB is one of them.

 

I have explained differences between these interfaces several times in this forum, so please search the archives if you want to better understand why.

post #3 of 15

There's a place for USB to S/PDIF, like when I want to play music  which resides in my laptop on my home theatre system. My pre/pro only has IEC 958 coaxial and Toslink optical inputs to its DAC so I need a converter. Right now, I use a Creative Sound Blaster X-FI HD to convert. I am happy with the results, with most everything sounding as good as the same material heard direct on CD or SACD.


Edited by sterling1 - 10/10/12 at 5:30am
post #4 of 15
Thread Starter 
I only use the USB input and not the S/PDIF or TOSLINK. If my DAC already takes asynchronous control over the USB buss data transfer from my iMac and the method is Gordon Rankin's what benefit would an external converter do? I am just confused as to why an external unit is being recommended when I already have jitter control?
post #5 of 15
Quote:
Originally Posted by sterling1 View Post

There's a place for USB to S/PDIF, like when I want to play music  which resides on my laptop on my home theatre system. My pre/pro only has IEC 958 coaxial and Toslink optical inputs to its DAC so I need a converter. Right now, I use a Creative Sound Blaster X-FI HD to convert. I am happy with the results, with most everything sounding as good as the same material heard direct on CD or SACD.

 

You are absolutely right: converters are useful when you can't connect your source with the receiver any other way, but stating that they will improve the sound quality is raising expectations a bit too high. This may happen if the converter has more accurate clock than the DAC, but in most situations it's the other way round.

post #6 of 15
Quote:
Originally Posted by PleasantSounds View Post

 

You are absolutely right: converters are useful when you can't connect your source with the receiver any other way, but stating that they will improve the sound quality is raising expectations a bit too high. This may happen if the converter has more accurate clock than the DAC, but in most situations it's the other way round.

 

You  know which hog ate the cabbage; but, there are some young folks here, at least I hope they're young, no excuse if they're not, who have, it appears, mindsets cast out of gobbledygook. They have succumbed to audio equipment propaganda which has appealed to their emotions rather than to their developing intellects. I wish it were not so. They should reserve their emotion for the music and their intellect for the equipment, not the other way around. But, that's today's world I suppose.

post #7 of 15

In my view there is no performance advantage to using an external USB-SPDF converter.  I think it could only detract from the performance of the built in USB system.  When using the m903 USB input (or any asynchronous mode USB DAC for that matter) the m903 is the clock master and the computer adjusts the data rate to suit.  When using a SPDF/TOSLINK/AES3 input, the sampling clock is recovered from the incoming data stream and the m903 follows the incoming clock rate.  This means that the m903 must synchronize to the frequency of the source device.  This is done with a PLL (phase locked loop).  The m903 employes an extremely high performance phase locked loop (sLock) to do this and it's jitter rejection and intrinsic jitter rival that of a fixed crystal oscillator.  However, there is no PLL like NO PLL and adding a conversion back to SPDF can only add jitter to the system.

I hope this makes sense!

Cheers,

Michael

post #8 of 15
Thread Starter 
Quote:
Originally Posted by mbgrace View Post

In my view there is no performance advantage to using an external USB-SPDF converter.  I think it could only detract from the performance of the built in USB system.  When using the m903 USB input (or any asynchronous mode USB DAC for that matter) the m903 is the clock master and the computer adjusts the data rate to suit.  When using a SPDF/TOSLINK/AES3 input, the sampling clock is recovered from the incoming data stream and the m903 follows the incoming clock rate.  This means that the m903 must synchronize to the frequency of the source device.  This is done with a PLL (phase locked loop).  The m903 employes an extremely high performance phase locked loop (sLock) to do this and it's jitter rejection and intrinsic jitter rival that of a fixed crystal oscillator.  However, there is no PLL like NO PLL and adding a conversion back to SPDF can only add jitter to the system.
I hope this makes sense!
Cheers,
Michael

Thank you. Your answer is crystal clear and jitter free...:-)
post #9 of 15

All I can share is my experience, as I have tried a couple of mid-fi converters as well as a cheap one with three different DACs, and I experienced three different results:

 

  • With the Bel Canto DAC2, a converter is required as it only has coax and toslink inputs. But the DAC2 has a rock-solid master clock, and as its asynchronous sample rate converter upsamples it takes care of jitter. According to one reviewer, it "employs the latest SPDIF receiver technology and a dual PLL reference clock recovery scheme using both analog and FIFO-based digital phase-lock circuitry for the lowest possible jitter performance. A local crystal oscillator drives the DAC directly." There was no benefit to using my Firestone Bravo 24/96 Tenor TE0722-based converter with separate Supplier PSU, as it didn't sound any better than an inexpensive eBay 16/48 Burr-Brown PCM2902E-based USB-powered converter.
  • With the Matrix Mini-i, the adaptive 16/48 USB input is definitely a weak link, and sounds totally inferior to the SPDIF and AES/EBU inputs. I heard a drastic improvement with the Bravo/Supplier. There was no benefit to using the inexpensive 16/48 converter however, that pretty much just duplicated the Mini-i's built-in USB receiver. I recently upgraded to the XMOS-based V-Link192 so I could take advantage of the 192k sample rate, as well as the "balanced digital" AES/EBU output. I heard further improvement, but it wasn't as drastic a difference as comparing the Bravo to the cheaper converter or the Mini-i's built-in USB input.
  • With the Schiit Gungnir, the difference wasn't that clear-cut. It has no AES/EBU input (my only real gripe with it), and its optional USB receiver is excellent: "Technically, Adapticlock is the industry’s most advanced jitter-reducing reclocking system. It automatically switches between VCXO and VCO reclocking, depending on the quality of your source. High-quality sources run on the VCXOs, for best jitter performance. Lower-quality sources that deviate from the range of the VCXOs are routed automatically to the VCOs, and an LED on the front panel comes on. We call this the “buy better gear” light... Sources that are outside of the industry-standard specified range of clock frequencies... still get... clock regeneration from the VCOs, but not the better regeneration performance of the VCXOs." I did hear slight improvement with the V192 but the Gungnir wasn't burned in yet. Now it's got about 150-200 hours on it, but through the RCA coaxial input. I am now proceeding to burn in its USB input and will compare again soon. It was so close that I can't rule out placebo.

 

So for me it depends on a couple of variables: how good is the DAC's built-in USB receiver (if it has one), and how good is the converter. My recommendation as always is to compare for yourself, in your own system, with your own ears. I would guess that the 903's USB input is exceptionally good, and that it would probably take an uber-expensive converter to improve on it. And while it may sound different with one of these converters, different isn't always better.

 

This is a great idea for a thread, I hope there is much more participation to come!


Edited by grokit - 10/10/12 at 9:54pm
post #10 of 15
Thread Starter 
Thanks for sharing your experiences.

What I don't understand is that say for example hat I placed a USB to S/PDIF converter between my iMac and my m903 (USB in/SPDIF out) the m903 the units phase lock loop is going to attempt adjust the incoming signal from the converter and will still run off the units internal clock. How can this be superior to the unit taking control over the transfer via its own Internal clock via USB in terms of jitter control? If the DAC did not have these two techniques, I could understand where a capable conversion unit could help. But when it does, I just see adding jitter or SPDIF signal degradation via cable. Unless I am missing something.
post #11 of 15

The m903 is doing much more than the S/PDIF protocol requires. In a standard S/PDIF connection the timing would be controlled by the sending device, while for USB connections it always happens on the DAC side. Older DACs with basic USB support didn't do a very good job timing the signal, while many higher end CD players with optical output had fairly decent clocks and controlled the timing quite well. Some users compared sound of a CD played from a CD player with the same material ripped loslessly on a PC and drawn their conclusions.

That led to a generalization that S/PDIF connection gives superior results to USB.

post #12 of 15
Thread Starter 
Thanks very much to all who contributed to this post. After careful discussion with others, I have come to the following conclusion:

Since the m903's crystal-based audio clock masters how the computer's USB bus is synced with the DAC and Grace Design chose asynchronous code to be written by Gordon Rankin at Wavelength Audio who is a pioneer in asynchronous USB audio, it is very close to being jitter free. In addition, the USB port is completely ground isolated from audio ground to ensure no noisy computer grounds.

Given this, I cannot conceive how any external any device, no matter how well it converts USB to S/PDIF can be as jitter free as when DAC clock itself directs the data transfer. Particularly since all S/PDIF signals incoming to the m903 will be altered by the DAC's own phase loop lock which will attempt to run off the local clock at the nearest target frequency. This cannot be as good as the direct asynchronous method jitter wise nor noise wise because it adds a length of S/PDIF cable.
post #13 of 15
Quote:
Originally Posted by iim7V7IM7 View Post

Thanks very much to all who contributed to this post. After careful discussion with others, I have come to the following conclusion:

Since the m903's crystal-based audio clock masters how the computer's USB bus is synced with the DAC and Grace Design chose asynchronous code to be written by Gordon Rankin at Wavelength Audio who is a pioneer in asynchronous USB audio, it is very close to being jitter free. In addition, the USB port is completely ground isolated from audio ground to ensure no noisy computer grounds.

Given this, I cannot conceive how any external any device, no matter how well it converts USB to S/PDIF can be as jitter free as when DAC clock itself directs the data transfer. Particularly since all S/PDIF signals incoming to the m903 will be altered by the DAC's own phase loop lock which will attempt to run off the local clock at the nearest target frequency. This cannot be as good as the direct asynchronous method jitter wise nor noise wise because it adds a length of S/PDIF cable.

 

iim7V7IM7,

 

curious how you connect your Grace m903 to your computer?  I presume USB?  what USB cable are you using (i know they're all the same, right : )?


Edited by aphinity - 4/20/13 at 12:18am
post #14 of 15
Quote:
Originally Posted by PleasantSounds View Post

 

You are absolutely right: converters are useful when you can't connect your source with the receiver any other way, but stating that they will improve the sound quality is raising expectations a bit too high. This may happen if the converter has more accurate clock than the DAC, but in most situations it's the other way round.

I agree and think it would most likely create a bottleneck if anything with the transfer speed differences in both cable types.

post #15 of 15

Sorry to insist, but I wonder if I can improve the sound quality on my following system. I am using a PC with a Creative X-Fi Platinum Fatal1ty Champion Series (pffff, could have used a shorter name) sound card that has front panel and can output S/PDIF in both coaxial and optical, from the front header and directly from the back of the card. I am using the digital output to feed the signal to a Denon AVR-1509 receiver, using it's DAC to convert the digital signal and finish the rest of the job to the speakers. Problem is from what I understood, since the clock generator (used to create digital square signal from data on the disc) is somehow under control of the OS (I use Windows 7 x64), there will be lots of IRQs that will somehow alter the accuracy of the square signal and thus creating jitter. I am directly outputing the signal bypassing the signal to the receiver, without card and Windows' mixer control (much), so little issues here. But the problems arise in the jitter.

 

Now, I read that outputting the signal directly via an USB-S/PDIF adapter to the DAC will eliminate the problem, since it uses it's own clock generator, outside of Windows' control. Although I wonder if this is the case, since Windows controls USB ports as well. But my question is: is it worth spending a lot of money on expensive USB-S/PDIF converters, or any cheap one will do the job just as well, of course apart from the supported formats ? My DAC supports 24 bit 96 KHz only, so I wonder if any cheap USB-S/PDIF converter will eliminate the jitter just as efficient as an expensive one.

 

I am looking forward to your answers. Thanks.

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Computer Audio
Head-Fi.org › Forums › Equipment Forums › Computer Audio › Help me understand USB to S/PDIF converters?