Support Head-Fi.org by
starting all of your
Amazon.com shopping by
clicking here.
____________________________________________________________________
Today's Featured Head-Fi Blog: Jude's Blog
____________________________________________________________________
Please help
support Head-Fi by becoming a Contributing Member
CLICK
HERE -- Contributing Members, thank you
for your generous support! --
...I have tested the firmware that Benchmark is using and it does not trick the Windows driver into transparency for 16/44.1 data since that is just not possible. The firmware registers the USB device as only supporting 24 bit streams but kmixer is still slightly changing the bits due the way volume processing is implemented.
Thomas,
I would like to know more about your tests. Could you please explain the method you used to achieve your conclusion?
thanks thomaspf for the info!
I live in italy and i want to buy a dac1 usb in the next few days.
But i'd like to know if benchmark will release soon a new
version of the dac1 ( a dac 2).The price here is 1100 euro
and for me is a money effort.I've tried to ask to Elias
but got no reply about a new dac.I would not buy something
that becomes obsolete in the next 6-8 months due to the
release of a new product
thanks
cheers
Francesco
I replied to your PM, but I would also like to state this publicly.
It is our company policy that we do not discuss future products. I don't want to seem coy or aloof about this, but we learned this lesson the hard way.
Our default reply to this question is, "We currently have no announcements regarding new products." Again, please don't be put off by this. It is very important that we follow this protocol.
I replied to your PM, but I would also like to state this publicly.
It is our company policy that we do not discuss future products. I don't want to seem coy or aloof about this, but we learned this lesson the hard way.
Our default reply to this question is, "We currently have no announcements regarding new products." Again, please don't be put off by this. It is very important that we follow this protocol.
Thanks,
Elias
Ok Elias thanks for the reply.Mine it was just curiosity before purchasing
such a fantastic product as yours...no intention to mess with
your company policy...really
Maybe a better way to say this is that there is not a way for any USB DAC to fool the KMIXER into bit perfect accuracy.
Hey guys this is no big deal. Bypass the KMIXER as it is told everywhere or go to Vista for better results.
I do agree with others though and heck I would not have bought the DAC1 if they would not have said it was bit perfect out of the box in all os's. I think they should change the marketing on that one.
Thanks
Gordon
Gordon,
I think you are misunderstanding the premise. We are not claiming to 'fool' kmixer. Kmixer does not need to be fooled. (Thomas seems to contend otherwise, and I plan to work with him to determine where our differences originate). Our tests indicate (and Microsoft has said) that bit-transparency is completely possible through kmixer. This has nothing to do with marketing, just like listing a THD+N spec or other performance specs. Please don't confuse this.
On another note, Gordon's efforts with USB Audio at Wavelength Audio have apparently yielded important results: he has asynchronous USB support working in one of his DACs.
I am somewhat skeptical that any difference in sound quality can actually be heard between async and standard USB audio, but I haven't tested it myself so this is just speculation.
But it is an important milestone, because in theory this feature finally makes an external DAC completely independent of the computer it is attached to for the timing of the data for playback (assuming the computer can keep up with a perfect 44.1 playback speed).
The downside of this breakthrough is that it is only available on ridiculously expensive hardware (sorry Gordon). Maybe a company like Benchmark could license this technology and make it available at a more appropriate cost so a much greater number of people could one day appreciate it (assuming it actually sounds better)?
My understanding of ASYNC USB is that the USB-audio device dictates the sample rate, and the OS must sync to the USB-audio device's clock. Because, if so, this would put the computer into sample-rate conversion mode to accommodate the exact frequency of USB device's clock. As that correct, Gordon, or am I misunderstanding something?
Whereas SYNC USB would allow the computer to operate at the native sample-rate of the recording, and the USB device would simply accept the data at the sample-rate it is received. This mode (sync) will cause the digital audio to have more jitter, and this is a problem for DAC's that are affected by jitter. Luckily, the DAC1 USB is not affected by jitter, so it can run in this mode flawlessly. Also, sample-rate conversion is not invoked, so there is less DSP distortion.
...but perhaps I'm misunderstanding the ASYNC mode.
Now that I a bold statement I seriously doubt! Would you be in a positon to share the name of the person who told you that in a PM or is that a secret?
I'd just want to check whether that is an official Microsoft position ...
As I stated before in my tests the firmware in the DAC1 can't seem to impact how Windows works internally and therefore I can't seem to reproduce bit perfect playback other than using kernel streaming.
Cheers
Thomas
P.S.: Or just send me your phone number and we can chat a bit more about this
Thomas,
I'll have to go back through my old emails and find out. In the mean time, please check my last post to you, re: your testing methods. I am very interested in your findings.
I'm not sure what you're referring to, but I'm assuming it has to do with AudioMIDI Setup. No, there are no MIDI settings that are applicable. With regards to the DAC1 USB, all applicable settings in AudioMIDI setup are under the 'Audio' tab.
Elias, yes that's what I mean don't mess with any of the options in the AudioMidi setup Audio tab.
Can you please elaborate? I'm very curious about Leopard. What did you test for? What were your results, specifically?
G5 Dual OSX10.5.1=iTunes 7.5=>DAC 1 USB=>Prism dScope III
I have 16 and 24 bit digital created oscillator files in iTunes. I can output these and then test for drop outs, thd and other things. I also have 16 and 24 bit JTEST files I can send over and test.
I personally use Faber to test most of my stuff. It's oscillator and combined testing ability are really worth the $. On the PC I use Dr. Jordan Audio Test Suite the same way.
We use Audio Precision for all waveform generation and testing. I'll look into these, however.
The problem with the Ap is that you cannot do any native real USB Audio testing. They require you to play files in an application like iTunes, Media Player. The problem with that is that these can corrupt the data before it actually gets to the USB driver and therefore the data is not responsible.
In the Prism software it can kernel stream to any device on the computer and therefore can fully test natively any USB Audio product.
You sent a Benchmark DAC1 USB to Microsoft? They already have one...we gave them one during development so that we could simultaneously analyze the results as the firmware development progressed. However, before development began with the DAC1 USB, the engineers at Microsoft assured us that bit-transparency was, in fact, possible with XP at resolutions up to 192 kHz, 24-bit, through Kmixer. In fact, this was the basis for our quest to develop firmware that achieved bit-transparency at high-resolution through kmixer. If they had believed it was not possible, we would have never attempted it. But, alas, they said it was possible, and we have found it to be accurate. We also found similar problems with ASIO, as you mentioned.
Yes basically because of your claim of bit perfection. Hakon Strande who is in charge of the Windows Sound Group told me the following statement.
There is no way in XP to get bit perfect data through a USB PCM device without bypassing the KMIXER.
I asked if he wanted to check out the Benchmark and he said he didn't need too. So I sent it too Thomas who did the testing in the Audio Test lab and concluded that Hakon statement was true in XP.
I think you are misunderstanding the premise. We are not claiming to 'fool' kmixer. Kmixer does not need to be fooled. (Thomas seems to contend otherwise, and I plan to work with him to determine where our differences originate). Our tests indicate (and Microsoft has said) that bit-transparency is completely possible through kmixer. This has nothing to do with marketing, just like listing a THD+N spec or other performance specs. Please don't confuse this.
Not possible as per Hakon statement.
BUT in Vista if you had the 24 bit option checked and sent 16 bit padded data to 24 bits then Vista did do bit perfect data without any drivers, bypassing or other BS that we have to do in XP.
My Beta version of Vista had a weird behavioral quirk where it would sample-rate convert everything to the highest possible sample rate, even if the appropriate sample rate was selected. Microsoft engineers indicated that this was intentional (although I can't remember why...I'll have to go back and check my emails...) However, we haven't tested the official release of Vista yet, so I'll keep you posted as to what we find.
Why don't you have Vista? Vista and Lepoard are extensively better at handling audio than their predisessors. I would say it's a requirement to have these in your testing suite.
Thanks
Gordon
__________________
J. Gordon Rankin
Wavelength Audio, ltd.
My understanding of ASYNC USB is that the USB-audio device dictates the sample rate, and the OS must sync to the USB-audio device's clock. Because, if so, this would put the computer into sample-rate conversion mode to accommodate the exact frequency of USB device's clock. As that correct, Gordon, or am I misunderstanding something?
Elias,
No there is no sample rate conversion. Basically the feedback pipe dictates a speed control of the amout of data to send. The CPORT of the TAS1020 therefore does not change and can be rock solid. Or as I am doing with my stuff is piping a 0.5ps low jitter MASTER clock into MCLKin. Thereby bypassing the internally developed clocks (which are now stopped to keep the system quite) and derive the Bit Clock and Word Clock from them.
~~~~~~~~~~~
Originally Posted by Scrith View Post
On another note, Gordon's efforts with USB Audio at Wavelength Audio have apparently yielded important results: he has asynchronous USB support working in one of his DACs.
I am somewhat skeptical that any difference in sound quality can actually be heard between async and standard USB audio, but I haven't tested it myself so this is just speculation.
But it is an important milestone, because in theory this feature finally makes an external DAC completely independent of the computer it is attached to for the timing of the data for playback (assuming the computer can keep up with a perfect 44.1 playback speed).
The downside of this breakthrough is that it is only available on ridiculously expensive hardware (sorry Gordon). Maybe a company like Benchmark could license this technology and make it available at a more appropriate cost so a much greater number of people could one day appreciate it (assuming it actually sounds better)?
Scrith,
Actually you can cruise over to AudioAsylum there are some really great comments on ASYNC and how much better it sounds under the PC Audio section of their forums.
I can't really talk about what is going to happen with ASYNC but I can tell you I have four new ASYNC products that are going to be released at CES and some maybe better suited to your price point.
Thanks
Gordon
__________________
J. Gordon Rankin
Wavelength Audio, ltd.
This is a pretty good article about adjusting gain staging using the DAC1's output attenuator. It makes great sense to adjust the gains this way, especially with the DAC1 as the preamplifier in an active monitor setup (which is how I would / will be using the DAC1). It will also prevent me from blowing out my ears with a bump of the volume knob.
However, I remember reading in the manual for Mackie HR824s that turning down the input sensitivity knob on the monitors could affect the frequency response and noise floor, and that it should be left up... Two conflicting opinions, maybe?
