g_man
g_man's picture
Offline
Last seen: Never ago
Joined: Mar 8 2006 - 3:09am
new Benchmark DAC1 with USB input
Elk
Elk's picture
Offline
Last seen: 3 years 7 months ago
Joined: Dec 26 2006 - 6:32am

I, too, am very curious.

I also wonder how they get "native USB" to transfer 24/96 files when USB is limited to 16/48.

EliasGwinn
EliasGwinn's picture
Offline
Last seen: Never ago
Joined: Feb 28 2007 - 8:09am

Hello,

I'm Elias Gwinn, an engineer at Benchmark Media. I wanted to answer the questions you have posted here.

The USB interface on the DAC1 USB is truly capable with 96/24 streaming audio. In fact, when using the DAC1 USB to playback audio, it achieves the same performance as the other digital inputs. Therefore, we do not discourage using the USB interface whatsoever.

I hope this helps!

Elias Gwinn
Electrical Engineer
Benchmark Media Systems, Inc
www.BenchmarkMedia.com
1-315-437-6300

Elk
Elk's picture
Offline
Last seen: 3 years 7 months ago
Joined: Dec 26 2006 - 6:32am

Elias, thanks for posting. It is great to hear from someone that really knows how a product works.

Am I correct that USB is natively limited to 16/48? How do you stream greater resolutions using the native drivers? I am not trying to get proprietary information, I just thought this wasn't possible using native drivers.

How are you dealing with jitter over USB? From what I have seen jitter is much greater over USB than the other interfaces. I know that the DAC1 stores the data in a buffer and reclocks the data. Is there anything special done for USB data? Why is word clock jitter rejection of USB transmitted data harder to accomplish?

What is the difference in the outputs between the USB version and the "regular" version? That is, what is the high current output? Is this simply a lower impedance output?

EliasGwinn
EliasGwinn's picture
Offline
Last seen: Never ago
Joined: Feb 28 2007 - 8:09am


Quote:
Elias, thanks for posting. It is great to hear from someone that really knows how a product works.

Thank you for that welcoming gesture. It's always a pleasure talking about these things. I hope to answer your questions as best as possible.


Quote:
Am I correct that USB is natively limited to 16/48? How do you stream greater resolutions using the native drivers? I am not trying to get proprietary information, I just thought this wasn't possible using native drivers.

Native USB is not limited to 48/16. In fact it is only limited by the bandwidth of the USB protocol. That is, USB 1.1 or 2.0. USB 1.1 is limited to 96/24. This is true for native and non-native.

The reason native USB is commonly believed to be limited to 48/16 is because the typical off-the-shelf USB solution for audio devices could not stream more than 48/16. The solution used in the DAC1 USB was engineered from scratch to utilize the 96/24 capabilities of USB 1.1, and it is the first to do so, as far as we have found.


Quote:
How are you dealing with jitter over USB? From what I have seen jitter is much greater over USB than the other interfaces. I know that the DAC1 stores the data in a buffer and reclocks the data. Is there anything special done for USB data? Why is word clock jitter rejection of USB transmitted data harder to accomplish?

Yes, jitter is a very serious matter with USB. In fact, one interesting piece of information we found (though we did not confirm) is that computer manufacturers will ADD JITTER to their clocks for the purpose of spreading the energy across a wider bandwidth to pass emissions testing. Apparently the spike at the fundamental clock frequency was causing interference of some sort.

The DAC1 USB is well suited to handle this jitter utilizing the same clocking technology that made the DAC1 so successful...UltraLock. With this clocking system, the DAC1 will perform exactly the same no matter how much jitter is on the clock...quite literally. There aren't too many converters that can make that claim, but we encourage you to compare.


Quote:
What is the difference in the outputs between the USB version and the "regular" version? That is, what is the high current output? Is this simply a lower impedance output?

Yes and no...the DAC1 USB has low impedance output, but its much more then that. Even low-impedance outputs will suffer from THD when it has to drive low impedance loads, high capacitive loads, and/or high inductive loads. The DAC1 USB is designed to drive these loads with no loss in THD performance. We refer to them as "High-Current" outputs because these loads require a lot of drive current, which will usually cause the THD to suffer. The DAC1 USB is designed to handle such loads without affecting the performance.

Hope this helps...I can clarify any of this if you like.

-Elias

Elk
Elk's picture
Offline
Last seen: 3 years 7 months ago
Joined: Dec 26 2006 - 6:32am

Elias, this is very interesting information - especially that relating to the limitations of USB and jitter. You provided information that is totally new to me.

The Benchmark DAC1 is a great unit with which I (and many others) have been very impressed.

I also have not heard of anyone getting better than 16/48 over USB. Benchmark has indeed accomplished something others have not so far using native drivers. Congrats!

Do you have any thoughts as to why other manufacturers have difficulty with rejecting wordclock jitter over USB? I am not interesting in naming names or slamming another's product. I am curious as to why USB appears to be more difficult than dealing with jitter over S/PDIF and other digital transmission protocols.

Thank you again.

Audio_newb
Audio_newb's picture
Offline
Last seen: 11 years 1 month ago
Joined: Mar 1 2007 - 9:12pm

Hello Elk, nice to see familiar faces from around the forums of audiophilia. Not to answer I question I really have no technical knowledge of (and please correct me if I'm wrong, Elias) but from my talks with Steve Nugent (of Empirical Audio) and Gordon Rankin (of Wavelegnth Audio) I gather that the biggest issues with USB jitter so far have stemmed from the fact that the only chips and receivers available have been consumer grade with focus on computer applications. Thus what could potentially be a transfer method with less jitter than SPDIF, as USB can transfer data in ways that don't intrinsically introduce more jitter, was initially inferior because of many years of audiophile designed SPDIF receiver chips.

Elias, I realize you can't reveal all of your secrets, but while we have an actual engineer here I'd love to ask you two things. Firstly, since Steve Nugent and a few others have started to evangelize the benefits of direct USB to I2S signal path systems, there has been much talk of the benefits of such an approach. I was wondering if you had a take on this, and whether it even mattered since the design of the benchmark is to reclock everything anyway. Secondly, while it was good to hear that you don't discourage people from using USB I was wondering if you had even so much as anecdotal comparisons between using USB and using the other digital inputs on the new DAC1. In any case thanks for taking the time to answer all our questions...but don't spend too much time here. You need to be engineering awesome products for audiophiles always looking for the next new upgrade.

EliasGwinn
EliasGwinn's picture
Offline
Last seen: Never ago
Joined: Feb 28 2007 - 8:09am


Quote:
In any case thanks for taking the time to answer all our questions...but don't spend too much time here. You need to be engineering awesome products for audiophiles always looking for the next new upgrade.

Thats really funny in a way that hits close to home. I love talking about this stuff...as does John Siau, our head of engineering and the pioneer behind Benchmark's digital audio products, as well as the new 4-channel mic pre. But, it is important to not spend too much time 'talking shop' vs. working in the shop!!

I'll try to answer as many of your questions as possible. I've answered a lot of these same questions on other forums, so I'll just copy and paste if you don't mind.

Yes, jitter is a very serious matter with computer audio. In fact, one interesting piece of information we found (though we did not confirm) is that computer manufacturers will ADD JITTER to their clocks for the purpose of spreading the energy across a wider bandwidth to pass emissions testing. Apparently the spike at the fundamental clock frequency was causing interference of some sort.

EliasGwinn
EliasGwinn's picture
Offline
Last seen: Never ago
Joined: Feb 28 2007 - 8:09am

We did measure the audio performance (Freq response, THD, IMD, etc) of the USB input, and it was completely similar to the all other digital inputs of the DAC1 up to 96/24. We also did listening tests. I am continuously conducting this test (as we speak ), as are several others here at Benchmark. Testing with a CD transport feeding the Coax input, and the computer feeding the USB with the same music, no one has been able to differentiate the two inputs.

EliasGwinn
EliasGwinn's picture
Offline
Last seen: Never ago
Joined: Feb 28 2007 - 8:09am

With regards to the statement about USB chips being built for the consumer market, I can't really say for sure. What I do know is... Jitter is not a concern with the DAC1 USB. The UltraLock clocking system in the Benchmark converters can handle more jitter then we can give it!! With an AP signal generator, we can induce various amounts of jitter onto the signal for testing purposes, and we can't add enough jitter to affect the performance of the DAC1. The DAC1 (original and USB) will have the exact same performance with lots of jitter as with no jitter...quite literally!! This is a claim not many converter companies will make, but we encourage you to compare.

gumby0610
gumby0610's picture
Offline
Last seen: Never ago
Joined: Mar 5 2007 - 7:57am

I was wondering whether Benchmark is the first to offer 96/24 USB streaming or whether Ron Sutherland's 12dAX7 Preamp streaming audio already has a DAC cable of 96/24, but not natively. I know the 12dAX7 Preamp did not get a favaroble review by Stereophile.

In order to reclock the digital bitstream, is the bitstream being read into a buffer first before is reclocked?

Can anyone recommend a CD transport (an afforbable one) that works well with the DAC1? I was thinking of building one, maybe some kits or sites related to DIY CD transports.

Elk
Elk's picture
Offline
Last seen: 3 years 7 months ago
Joined: Dec 26 2006 - 6:32am

Any CD player with a digital out will work wonderfully with a Benchmark DAC.

jackfish
jackfish's picture
Offline
Last seen: 10 years 5 months ago
Joined: Dec 19 2005 - 2:42pm

g_man, sorry if I'm changing the topic too much. But I have an Oppo DV-970HD and in the documentation it says output for SACD and DVD-A is through the 5.1 analog outputs. But I was wondering since I have a stereo setup, if I set the down mix to stereo, if the output could be put through the digital outputs to a separate DAC or will the Oppo only decode the 5.1 and output it through the 5.1 analog outputs and when down mixed only output to the mixed (stereo) analog outputs? It looks like if I set the SPDIF Output to PCM it will force a stereo down mix for digital output, but I don't know if I am reading that right. I can't seem to find specific reference to this issue.

struts
struts's picture
Offline
Last seen: 2 years 10 months ago
Joined: Feb 1 2007 - 12:02pm

gumby0610,

John Marks recommends cheap and/or second-hand DVD players with digital outs as excellent budget transports. I have no reason to doubt him.

Of course I just have to point out that transports are so twentieth century. If you rip your CDs in a lossless codec with error correction on, a second-hand Mac Mini should drive a DAC-1 USB with better audio quality than a transport (again see Gordon Rankin's compelling arguments in Art Dudley's review of the Wavelength Brick, Listening #33) and save you lifting your audiophile rear to change CDs!

DBZ
DBZ's picture
Offline
Last seen: Never ago
Joined: Mar 19 2007 - 3:54pm

Elias:

I really appreciate your participation in the forum. Wish more manufacturers would do that.

Here's my question: How does one get a 96/24 feed into the Benchmark without making the original recording yourself? If I go out and buy a DVD-Audio disk, is there any way to run the digital output directly to the Benchmark? In the past, I've been unable to do that.

Thanks again.

EliasGwinn
EliasGwinn's picture
Offline
Last seen: Never ago
Joined: Feb 28 2007 - 8:09am


Quote:
Elias:

I really appreciate your participation in the forum. Wish more manufacturers would do that.

Here's my question: How does one get a 96/24 feed into the Benchmark without making the original recording yourself? If I go out and buy a DVD-Audio disk, is there any way to run the digital output directly to the Benchmark? In the past, I've been unable to do that.

Thanks again.

Its a very good question, and its a pulse we've been trying to feel for sometime. The problem is, many DVD-A players don't actually transmit 24/96 via the digital output. Even those that say they will play up to 24/192....they will play the hi-res audio, but they down-sample (sometimes to 44.1!!) before the digital output!! We've been testing DVD-A players (mainly <$200) and we can't find one that plays bit-transparently!!

As for software that will play DVD-A and output 96/24, I don't know of any, but that is because I've never looked. I'm pretty sure its out there, but I don't know where.

Thanks,
Elias

DBZ
DBZ's picture
Offline
Last seen: Never ago
Joined: Mar 19 2007 - 3:54pm

Well, thanks for the response, even if it wasn't what I wanted to hear.

Maybe someone else will tell me they've figured out a way to put a DVD-A disk into their computer and get it to output 96/24.

Elk
Elk's picture
Offline
Last seen: 3 years 7 months ago
Joined: Dec 26 2006 - 6:32am

I have a Panasonic DVD-S77 that will output 24/96 when playing DVD-A on its coaxial S/PDIF.

At first I thought it would output only 48kHz. However I learned that if the video is turned off (audio circuits active only) or if the HDMI output is turned off it will then pass 24/96.

The 24/96 does indeed sound better.

The specs indicate that it decode higher resolutions as well but I have yet to burn a DVD-A with higher resolution files on it. (I never record over 24/96.)

EliasGwinn
EliasGwinn's picture
Offline
Last seen: Never ago
Joined: Feb 28 2007 - 8:09am


Quote:
I have a Panasonic DVD-S77 that will output 24/96 when playing DVD-A on its coaxial S/PDIF.

At first I thought it would output only 48kHz. However I learned that if the video is turned off (audio circuits active only) or if the HDMI output is turned off it will then pass 24/96.

The 24/96 does indeed sound better.

The specs indicate that it decode higher resolutions as well but I have yet to burn a DVD-A with higher resolution files on it. (I never record over 24/96.)

Unfortunately the DVD-S77 is no longer available. The Panasonic DVD-S47 and the Pioneer DV-578A-S are also capable of 96/24, but they are no longer available either!!

CECE
CECE's picture
Offline
Last seen: 5 years 3 months ago
Joined: Sep 17 2005 - 8:16am

Cool, now let's see the digital cable makers with all their nonsense about how their digital cable "sounds" better cus' it cures a problem that doesn't exist. I have the DAC 1 works great sounds fine. And other DAC too. Glad to hear the mfg says it sounds the same from USB or coax in. I'm sure the golden ear crowd ( are they still golden or mostly greyed out?) will claim to hear differences, that really don't exist! I will continue to use my generic cabling, from Parts Express or MCM. Benchmark makes stuff based on reality electrical designs, what a concept!!! Zander, take note ($45K misiwired CD only player). Benchmark and a few other makers that I own, have proved, top quality sound is available at mortal prices. Since teh products are based on real electrical design, not magicaly nonsense, designed in teh ad depts. Keep up the good work BenchMark

Audio_newb
Audio_newb's picture
Offline
Last seen: 11 years 1 month ago
Joined: Mar 1 2007 - 9:12pm

DBZ, take this with a grain of salt since I have never tried it myself, but I believe that both Powerdvd 6 (not sure about later versions) and the software suite that comes with creative audigy cards are capable of playing DVD-audio disks. Thus in theory you should be able to run 24/96 via usb on a pc from DVD-audio disks. If you happen to try this, let us know how it goes.

I think the lack of support over digital is mostly a copy protection thing, which is why I'll be excited to see audio gear starting to ship with HDMI (I hope) which can pass everything under the sun including multichannel DSD. Looking forward to the Benchmark USB/ HDMI edition, Elias.

EliasGwinn
EliasGwinn's picture
Offline
Last seen: Never ago
Joined: Feb 28 2007 - 8:09am

Unfortunately, I've heard that even HDMI has its issues. I'm not completely up to date on the technology so I can't elaborate, but I've heard some things about it that wasn't too promising. But, even with a perfect interface, SRC will only be avoided if the copyright folks allow it to happen.

Thanks,
Elias

Elk
Elk's picture
Offline
Last seen: 3 years 7 months ago
Joined: Dec 26 2006 - 6:32am

I just learned from a friend that the Oppo Digital 981 can be set to output 96 kHz and 192kHz. $230

Other Oppos may as well, but I don't know.

g_man
g_man's picture
Offline
Last seen: Never ago
Joined: Mar 8 2006 - 3:09am

Actually, I believe the Oppo 981 will only output DVD-A and SACD at full bandwidth via HDMI.

Elk
Elk's picture
Offline
Last seen: 3 years 7 months ago
Joined: Dec 26 2006 - 6:32am

I appreciate the suspicion; most DVD players will not output anything other than 16/48 on their digital outputs.

However, as the DAC he is using identifies the resolution of the incoming signal I'm pretty comfortable the Oppo is operating as described. But I don't know anything further. I don't have an Oppo to personally play with.

My Panasonic SD77 will output 24/96 as will a few other players. (Sometime I will record at 24/192, burn a DVD-A and see if it will output this as well - it has 192 capable DACs so the unit may not panic.)

CECE
CECE's picture
Offline
Last seen: 5 years 3 months ago
Joined: Sep 17 2005 - 8:16am

You don't have an Oppo to play with?????!!!! Sounds like a birth defect.

g_man
g_man's picture
Offline
Last seen: Never ago
Joined: Mar 8 2006 - 3:09am


Quote:
However, as the DAC he is using identifies the resolution of the incoming signal I'm pretty comfortable the Oppo is operating as described. But I don't know anything further. I don't have an Oppo to personally play with.

My Panasonic SD77 will output 24/96 as will a few other players. (Sometime I will record at 24/192, burn a DVD-A and see if it will output this as well - it has 192 capable DACs so the unit may not panic.)

OK - I was just going on the user manual and the issue about copy right control. I have an Oppo 981, but never use it for music. I would *almost* like my DAC1 to display resolution, but then that would be getting away from the simplicity of the unit, not to mention the great price!

Elk
Elk's picture
Offline
Last seen: 3 years 7 months ago
Joined: Dec 26 2006 - 6:32am

Good one, DUP.

Log in or register to post comments
-->
  • X