[Dave_M]

Hi EliasGwinn, sorry if I have been repeating what others had said but I have not read through the whole thread.

There is nothing wrong with using cheap off-the-shelf ICs to get the job done! After all, the cheaper the DAC is to produce, the more competitive you can be with pricing. And besides, an all in one integrated solution can often work better than a custom one with many ICs and discrete components.

But I'm not sure I accept your argument of a native, bit-transparent USB audio solution. Since a "native solution" still needs drivers, it's just that those drivers were written by someone at Microsoft and come with the operating system. If a USB 2.0 one needs custom drivers, then it can be done just as well as "native solution", but just takes more time and money.

 

[EliasGwinn]

Quote:

Originally Posted by Dave_M /img/forum/go_quote.gif Hi EliasGwinn, sorry if I have been repeating what others had said but I have not read through the whole thread. There is nothing wrong with using cheap off-the-shelf ICs to get the job done! After all, the cheaper the DAC is to produce, the more competitive you can be with pricing. And besides, an all in one integrated solution can often work better than a custom one with many ICs and discrete components. But I'm not sure I accept your argument of a native, bit-transparent USB audio solution. Since a "native solution" still needs drivers, it's just that those drivers were written by someone at Microsoft and come with the operating system. If a USB 2.0 one needs custom drivers, then it can be done just as well as "native solution", but just takes more time and money.

You are correct. My use of the term 'native device' is defined as a device which uses native drivers. Theoretically, a custom driver may be designed as good as the native drivers. However, we have not seen any custom drivers currently available that are satisfactory. Meanwhile, the native drivers are very stable, very capable, and, most importantly, bit-transparent. This makes sence, as Microsoft's and/or Apple's programming team has a much broader and deeper familiarity of the system and environment in which they operate.

With that said, although all chips (both 1.1 and 2.0) are off-the-shelf chips, the firmware is not necessarily. There are some chips which have firmware included off-the-shelf; these are the chips used by most native USB audio devices. These off-the-shelf solutions are currently limited to 44/16 and 48/16 streams. Our solution uses a readily available USB chip (TI's TAS1020B) with custom-built firmware which enables it to stream bit-transparently at resolutions up to and including 96/24 using the native drivers. This firmware is what separates our USB solution from any others currently available. We are not aware of any USB audio devices which can stream 96/24 using native drivers.

Thanks,
Elias

 

[Jetlag]

Elias, does the DAC 1 USB upconvert the USB signal to 192KHz?

 

[schaqfu]

Quote:

Originally Posted by EliasGwinn /img/forum/go_quote.gif This is partly accurate. Let me clarify... The price of USB 2.0 chips was not a factor. However, there are no USB 2.0 chips available that support native USB audio. So, one of the trade-off's of using USB 2.0 is the necessity of a custom driver. As I've elaborated before, we tested several other (3rd party) USB audio devices that use custom drivers, and the results were not positive. In fact, throughout our testing, the only devices that achieved bit-transparency were native devices (devices which do not need custom drivers). Also, custom drivers often cause conflicts with other drivers. Even if the driver is designed well, another driver may not 'get along' with it. Using the native drivers (which are inherently installed with the operating system) gives you the guarantee that, if a 3rd party driver ever worked on these computers, they will work with the DAC1 USB running. Another point you made is accurate, and partly affected our decision. That is, it would have taken longer to develop a USB 2.0 device. This wasn't an issue about saving money as much as getting the product to market as soon as possible. If the other factors weren't already weighing against a USB 2.0 solution, we may have decided it was worth waiting to release this product until a USB 2.0 solution was developed. However, the decision was primarily based on the design goal of a native, bit-transparent USB audio solution.

This -- and Elias' following post as well -- is a totally legitimate explanation and rationalization of why they stuck with USB 1.1. This puts it into stark terms of why it was a market decision and not simply a lazy or errant one. I can't tell you how much I appreciate seeing this kind of response. It is very satisfying. I wish all manufacturers would do this. And, in fact, Elias, I highly recommend that you include something to this effect (perhaps more condensed) on the FAQ portion of your site. Right now the explanation posted sounds like a cop-out, seeming to say you chose USB 1.1 simply because it's widely available. The real answer is, this is the state of the art with respect to USB chips, while remaining within the (very well-received) design requirement of plug and play bit transparency.

 

[Dave_M]

Quote:

Originally Posted by schaqfu /img/forum/go_quote.gif This -- and Elias' following post as well -- is a totally legitimate explanation and rationalization of why they stuck with USB 1.1. This puts it into stark terms of why it was a market decision and not simply a lazy or errant one. I can't tell you how much I appreciate seeing this kind of response. It is very satisfying. I wish all manufacturers would do this. And, in fact, Elias, I highly recommend that you include something to this effect (perhaps more condensed) on the FAQ portion of your site. Right now the explanation posted sounds like a cop-out, seeming to say you chose USB 1.1 simply because it's widely available. The real answer is, this is the state of the art with respect to USB chips, while remaining within the (very well-received) design requirement of plug and play bit transparency.

I agree. Elias, you have retained all of your credibility when answering these questions. I think people can see through a lot of the BS you get from so many companies nowadays. Especially in this case anyone buying a DAC1 is not going to be your average schmuck.

It usually the computer world that leads the way in technology and the hi-fi world that clings to old tech like valves. In this case it is benchmark who are waiting to the computer world to catch up. One day I'm sure there will be a DAC1 with 24/192 over USB 2.0, but until then I think they made the right decision to support 24/96 native. As having it work without custom drivers is a major advantage.




Jetlag, I think all signals are re-sampled to 110 KHz. See post #265.

 

[EliasGwinn]

Quote:

Originally Posted by Jetlag /img/forum/go_quote.gif Elias, does the DAC 1 USB upconvert the USB signal to 192KHz?

Jetlag,

The short answer is... the USB signal is converted to 110 kHz, as Dave mentioned.

The long answer is... (deep breath)

The USB signal is first converted to I2S, maintaining its original sample-rate. I2S is a fundamental form of digital audio. When an AES/EBU or S/PDIF signal is streamed to the DAC1 via XLR/coax/optical, it is also first converted to I2S at its original sample-rate. The front panel switch chooses which of these I2S signals are sent to the next stage: the sample-rate converter (SRC) chip. This converts the sample-rate of the I2S signal to 110 kHz, regardless of the original sample rate.

The reason it is converted to 110 kHz is because this is the sample-rate at which the D-to-A converter chip is most efficient. The trade-off of SRC far out-weighs the more significant distortion of the filter in the D-to-A chip operating at sample-rates other then its most efficient rate.

Thanks,
Elias

 

[gregeas]

Regarding 24/192 input, I don't see a practical reason why you would need it, unless you work in a mastering studio.

For about five years I have done my best to enjoy hi-res audio (DVD-A and SACD), and just two months ago bought Arcam's new top-of-the-line univeral player. (I owned a variety of other universal players before that.) The problem is that after all these years there are still less than 25 hi-res disks that I care to own. And I don't see anything worthwhile on the horizon. (I'm not into classical.) Of course, there are some gems in my collection, and I hate to give them up, but how many times can I listen to Morph the Cat and Sea Change? Plus, the CDs of these albums sound damn good.

I did explore extracting the hi-res audio from DVD-A disks. This can be done but sounds very involved. So for now I'll live with 16/44.1 on my DAC1 and Transporter.

Does anyone else have a way to access hi-res audio files?

 

[Riboge]

Quote:

Originally Posted by EliasGwinn /img/forum/go_quote.gif ... With that said, although all chips (both 1.1 and 2.0) are off-the-shelf chips, the firmware is not necessarily. There are some chips which have firmware included off-the-shelf; these are the chips used by most native USB audio devices. These off-the-shelf solutions are currently limited to 44/16 and 48/16 streams. Our solution uses a readily available USB chip (TI's TAS1020B) with custom-built firmware which enables it to stream bit-transparently at resolutions up to and including 96/24 using the native drivers. This firmware is what separates our USB solution from any others currently available. We are not aware of any USB audio devices which can stream 96/24 using native drivers. Thanks, Elias

E-mu and M-Audio both offer USB sound cards they claim use USB2 and can take in and put out 24/96. How do these reconcile with what you have said above? Do they not use native drivers?

Also, my Stello 220 mkII has USB 1.1 input but is a 24/192 DAC. If I play recordings from the computer via USB that are 24/96, is that what reaches the DAC, and is it correct that the DAC can then upsample that to 192? How is this streaming or not streaming 24/96 via USB using a native driver?

It is remarkable how hard it is to get consistent and reliable information like this which seems basic to understanding a given DAC and transfer mode.

 

[EliasGwinn]

Quote:

Originally Posted by Riboge /img/forum/go_quote.gif E-mu and M-Audio both offer USB sound cards they claim use USB2 and can take in and put out 24/96. How do these reconcile with what you have said above? Do they not use native drivers? Also, my Stello 220 mkII has USB 1.1 input but is a 24/192 DAC. If I play recordings from the computer via USB that are 24/96, is that what reaches the DAC, and is it correct that the DAC can then upsample that to 192? How is this streaming or not streaming 24/96 via USB using a native driver? It is remarkable how hard it is to get consistent and reliable information like this which seems basic to understanding a given DAC and transfer mode.

The E-mu and M-Audio both use custom drivers. They do not use native drivers.

As for the Stello 220 mkII, I went to their website, but I couldn't find any spec's relating directly to the sample-rate capabilities of the USB input. You'll have to ask them about its USB capablities. The fact that it is a 24/192 DAC does not indicate anything about the USB sample-rate capabilities. For instance, the Benchmark DAC1 USB is also a 24/192 DAC, however the USB is limited to 96/24. If you play 24/192 audio on your computer with a USB audio device which cannot stream 24/192 via USB, the operating system will convert the sample-rate to one which the device is capable of streaming.

As for consistent and reliable information, I couldn't agree with you more. I just went to the website for the Stello, and I couldn't find any worthwhile information. I hope our site is a little more straight-forward then that.

I also hope that if anyone feels that our site is not straight-forward or easy to navigate, that he/she will let me know so we can do something about it.

Thanks,
Elias

 

[Riboge]

Quote:

Originally Posted by EliasGwinn /img/forum/go_quote.gif The E-mu and M-Audio both use custom drivers. They do not use native drivers.

So does your comment about your estimation of custom drivers you have tried apply to these, ie, not good enough?

 

[Graham Maynard]

Hi Elias,

Did Benchmark ever consider fitting an Alps motorised pot;- so that remote level control could be offered ?
Could one be fitted by a competent technician ?

Cheers ....... Graham.

 

[EliasGwinn]

Quote:

Originally Posted by Riboge /img/forum/go_quote.gif So does your comment about your estimation of custom drivers you have tried apply to these, ie, not good enough?

Well, the question of "good enough" is relative to the application, and what you expect from the interface. If you expect bit-transparency, then they are not good enough. The units I tested had low-distortion, but not bit-transparent. Also, at 24/96, they 'choke' a lot. Specifically, the M-Audio Audiophile USB...when trying to record or playback at 24/96, it frequently has 'hiccups'. 24/96 is the extreme of its capabilities, so it can't quite handle it apparently.

Thanks,
Elias

 

[EliasGwinn]

Quote:

Originally Posted by Graham Maynard /img/forum/go_quote.gif Hi Elias, Did Benchmark ever consider fitting an Alps motorised pot;- so that remote level control could be offered ? Could one be fitted by a competent technician ? Cheers ....... Graham.

We have an Alps pot here which we are testing, but we have not drawn any conclusions on its performance yet. The form factor of the assembly is much too large for the DAC1, however, so it would not physically fit the DAC1.

Thanks,
Elias

 

[Graham Maynard]

Thanks Elias,

It occured to me after posting that if necessary a DIYer could make up their own small "wall-wart" powered housing beside a racked DAC-1 if they needed a remotely controllable ALPS+buffer facility.

Cheers ........ Graham.

 

[choariwap]

Quote:

Originally Posted by EliasGwinn /img/forum/go_quote.gif Well, the question of "good enough" is relative to the application, and what you expect from the interface. If you expect bit-transparency, then they are not good enough. The units I tested had low-distortion, but not bit-transparent. Also, at 24/96, they 'choke' a lot. Specifically, the M-Audio Audiophile USB...when trying to record or playback at 24/96, it frequently has 'hiccups'. 24/96 is the extreme of its capabilities, so it can't quite handle it apparently. Thanks, Elias

so maudio's drivers dont do bit perfect? even in ASIO? wow...

i'm really itching to get a dac1 usb, but the closest i can get one is in hong kong. curses! soonest i can go is june-ish... sigh