FrankWW said:
At least the link to Jocko's 1970 Plymouth Cuda was fun : www.diyaudio.com/forums/showthread.php?postid=79429#post79429
x-pro said:
As noted above, it seems that indeed the target sr is 110kHz, so that would mean the src would always be active?
Jan Didden
00940 said:
That last link is very clear. So, if they indeed resample everything to 110kHz, and the sr concerter is always resampling, what's with the notion that this is Not Good?
Is there necessarily a difference in sound quality depending on what sr the source is, or would it be possible to have a src technology that results in the same high quality when resampled to 110kHz whatever the soure sr? I have the feeling that John S. hinted at that.
BTW Chuck Hansesn's measurements don't directly address these issues. He shows analog data like fr, IMD, THD and waveforms at 0dBFS square wave and -90dBFS sinewave.
Jan Didden
x-pro said:
I'm affraid those manuals were written by the marketing dpt.
In a huge thread titled "dac 1 now available with usb" on head-fi, the application engineer from Benchmark said that the asrc was set at 110khz.
See: http://www.head-fi.org/forums/f46/benchmark-dac1-now-available-usb-223006-post2885961
In many posts, the benchmark engineers go on about how 110khz is optimal for 96khz inputs. Why would they have done differently for the dac1 usb ?
00940 said:
The only reason would be to provide a bit-perfect data delivery from USB, as USB can be asynchronous and stream the data on demand from an internal DAC clock. However I have no exact information on that. Perhaps I should borrow the DAC1 USB for a while and do some measurements on it... .
Alex
x-pro said:
Alex,
What I understood from John Siau, is that the DAC-USB gets its USB data on demand, issued by the DAC. I didn't know this was possible, but apparently it is.
Jan Didden
janneman said:
Jan, I see why you have some confusion. The model I would use is a pure asychronous data feeding interface to a completely isolated DAC environment. There is no need to generate any clock or timing from the source, which to me can eliminate any transfer of jitter.
There are plenty of devices around that use an ASRC to take out the slop between even two sources with nominally the same clock rate i.e. the data is never bit perfect.
This is drifting away from the core of the blowtorch focus, at least in terms of understanding the impact of detail in analog electronics (even obsessive perfectionism in oscillators) and into the more typical hunting around the highly controversial issues in digital audio, like sample rate converters, asynchronous usb etc. These issues are being discussed ad-nausium elsewhere and without a lot of resolution. And by people with deep experience and very strong opinions.
We were exploring some much more immediately interesting stuff in precision oscillators (and had plenty of evidence that a very good crystal oscillator with attention to its power supply etc. could make a difference, under the right conditions and why).
I would be more interested in exploring issues around reducing noise coupling between boxes/stages either analog or digital. Like why milled blocks of aluminum are the current fashion (Blowtorch, Ayre) etc. And how to keep the incredibly universal RF from mucking up the sound, without the fix being worse than the problem.
Or, in a DAC, dealing with the digital-analog interface/interaction. With the current sota DAC's all running on 5V or 3.3v how do you get 24 bit accuracy (do you?) and keep the noise as low as it needs to be. Very much a blowtorch like challenge.
We were exploring some much more immediately interesting stuff in precision oscillators (and had plenty of evidence that a very good crystal oscillator with attention to its power supply etc. could make a difference, under the right conditions and why).
I would be more interested in exploring issues around reducing noise coupling between boxes/stages either analog or digital. Like why milled blocks of aluminum are the current fashion (Blowtorch, Ayre) etc. And how to keep the incredibly universal RF from mucking up the sound, without the fix being worse than the problem.
Or, in a DAC, dealing with the digital-analog interface/interaction. With the current sota DAC's all running on 5V or 3.3v how do you get 24 bit accuracy (do you?) and keep the noise as low as it needs to be. Very much a blowtorch like challenge.
Well said, Demian. It is more about new technology and measurements, rather than finding what makes 'good' better. Or the ideal difference between 'mid-fi' and 'hi-end'. For some contributing here, there is no difference, but in reality, there is.
For example, checking with Jam, who has actually seen the schematics of the Benchmark, and listened to the unit, the power supply isolation has a lot to be desired for
For example, checking with Jam, who has actually seen the schematics of the Benchmark, and listened to the unit, the power supply isolation has a lot to be desired for
Figuring out how to look at problems is a large part of fixing them. If attaching a scope probe to the circuit introduces more noise than the circuit has it becomes very hard to reduce the noise effectively. A lot of this is technique. And learning how to see what is going on. Some of the mumbo-jumbo of audio becomes clear when you can actually see the phenomena.
In one measurement of a low phase noise crystal oscillator, sensitivity to fluorescent lights was detected, on a battery powered supply. Looking at the supply circuit I see the use of an LED for biasing. I know from experience, that LED's are sensitive to external light, and that's why I would not use them for biasing. But this is a technique that isn't well known. It may not be the cause of the sensitivity but who knows?http://www.febo.com/pages/oscillators/wenzel_uln/supply.html
In one measurement of a low phase noise crystal oscillator, sensitivity to fluorescent lights was detected, on a battery powered supply. Looking at the supply circuit I see the use of an LED for biasing. I know from experience, that LED's are sensitive to external light, and that's why I would not use them for biasing. But this is a technique that isn't well known. It may not be the cause of the sensitivity but who knows?http://www.febo.com/pages/oscillators/wenzel_uln/supply.html
LEDs
Hi Demian,
Apart from sensitivity to light, it is said that LEDs are also noisy. Why and to what extent?
Noisier than a buried Zener diode for example?
Regards,
Edmond.
Hi Demian,
Apart from sensitivity to light, it is said that LEDs are also noisy. Why and to what extent?
Noisier than a buried Zener diode for example?
Regards,
Edmond.
And now I can demonstrate how little I know about solid state physics. I really can's say with confidence why LED's are noisy but this may be part of it. First all semiconductors are light sensitive to some degree. years ago someone was chopping the tops off of power transistors and driving them with LED's for a high end amp. Concept didn't survive, except as some specialty semiconductors.
Second, the physics of how an LED generates the light, with the electrons shifting from shell to shell (if I'm not confusing semis with lasers) would not be a linear process with clear steps.
Third, why would an LED manufacturer try to make an LED quiet?
A buried Zener is the absolute opposite extreme in technology.
Second, the physics of how an LED generates the light, with the electrons shifting from shell to shell (if I'm not confusing semis with lasers) would not be a linear process with clear steps.
Third, why would an LED manufacturer try to make an LED quiet?
A buried Zener is the absolute opposite extreme in technology.
scott wurcer said:
I see. I'm currently trying to measure the results of ASRC by a Behringer SRC2496. I have temp access to an AP 2722 Dual Domain analyzer. Seeing some funny things, not sure it isn't operator trouble though. I'm learning a lot!
But I'll get out of the way of the On Topic guys here 😉
Jan Didden
1audio said:
Indeed.
Now, I believe there is no question that a case of milled blocks of aluminum, like the one in Blowtorch, is the crème of the crème of SOTA audio.
I wish I could afford one for my DAC – but I cannot afford it.
So, financial considerations force me to move a step down, hopefully still remaining at SOTA level.
I believe next step down will be aluminum case made of separate 6 outer plates and some inner separation plates – or is there another option?
What would be the minimum thickness of aluminum plates for a case of medium budget project aimed at approaching SOTA audio gear?
DSD VS assorted PCM impulse response
Please note the poor impulse response of the bit rate and word count of the PCM systems being discussed, hardly high end stuff.....
Why is this thread getting off the analog topic , and worse, why is it focusing on an inferior technology....Any serious studio these days is already working ONLY in 192/24 and many are converting to double that, the DXD PCM system, and are still noting inferior sound to pure DSD. Also note that some recent DSD systems are now operating at 2X the standard DSD rate of 2822.4MHz, 5644.8HMz.
If you guys want to talk HIGH END ANALOG, as it related to some aspects of digital audio, I can promise you that there is great need for improvement in the analog stages of even the best DSD recording and playback equipment. The current crop of equipment is crippled with cheap op-amps, lots of feedback in many analog stages, and low performance caps in the signal path, everything that this thread is dedicated to the removal of...and that I would argue just might make a larger impact on the sound that a few PPM of PCM jitter. I'm not saying that asynchronous jitter is not important, DO NOT read that conclusion iinto this comment, only that there are other basic issues in the encode and decode digital audio chain that are in dire need of improvement too. Every aspect of the entire process IS important, but please, jitter and other related subjects are being well discussed on other threads. Lets try to keep it's inclusion here directly related to the spirit of this ultra-performance analog thread.
OK, I have gone and done it again, upset the apple-cart!
Please note the poor impulse response of the bit rate and word count of the PCM systems being discussed, hardly high end stuff.....
Why is this thread getting off the analog topic , and worse, why is it focusing on an inferior technology....Any serious studio these days is already working ONLY in 192/24 and many are converting to double that, the DXD PCM system, and are still noting inferior sound to pure DSD. Also note that some recent DSD systems are now operating at 2X the standard DSD rate of 2822.4MHz, 5644.8HMz.
If you guys want to talk HIGH END ANALOG, as it related to some aspects of digital audio, I can promise you that there is great need for improvement in the analog stages of even the best DSD recording and playback equipment. The current crop of equipment is crippled with cheap op-amps, lots of feedback in many analog stages, and low performance caps in the signal path, everything that this thread is dedicated to the removal of...and that I would argue just might make a larger impact on the sound that a few PPM of PCM jitter. I'm not saying that asynchronous jitter is not important, DO NOT read that conclusion iinto this comment, only that there are other basic issues in the encode and decode digital audio chain that are in dire need of improvement too. Every aspect of the entire process IS important, but please, jitter and other related subjects are being well discussed on other threads. Lets try to keep it's inclusion here directly related to the spirit of this ultra-performance analog thread.
OK, I have gone and done it again, upset the apple-cart!
As far as real information goes, Demain, your second submission on the osc vs power supply is the most informative input I have had for weeks.
When it comes to digital playback, I did some measurement of the high speed response of the typical power supply IC regulators used for setting the DC voltage of different IC chips in a CD playback scheme about 15 years ago. I found that most 3 terminal IC regulators were virtually transparent IN BOTH DIRECTIONS to digital garbage. What a potential problem!
When good engineers just look at specs and performance in specific chips, this is an area that seems to be often neglected. Now, I doubt that Scott Wurcer would make such an oversight, as he helped ME 25 years ago in doing first class cap bypassing when I was designing an RF amp for a major company. (Thanks again Scott, you are not forgotten in this.) However, is everyone here up to speed on what I am pointing out?
When it comes to digital playback, I did some measurement of the high speed response of the typical power supply IC regulators used for setting the DC voltage of different IC chips in a CD playback scheme about 15 years ago. I found that most 3 terminal IC regulators were virtually transparent IN BOTH DIRECTIONS to digital garbage. What a potential problem!
When good engineers just look at specs and performance in specific chips, this is an area that seems to be often neglected. Now, I doubt that Scott Wurcer would make such an oversight, as he helped ME 25 years ago in doing first class cap bypassing when I was designing an RF amp for a major company. (Thanks again Scott, you are not forgotten in this.) However, is everyone here up to speed on what I am pointing out?
john curl said:
I believe I am.
It's after midnight here.
Tomorrow I'll have a minor eye operation (cataract removal).
One of the next following days I'll post few questions regarding power supplies.
john curl said:
Yep. 15 years later, we have way much better three terminal analog regulators than the ancient 79xx/79xx (or anything else available then); some are specially designed for powering the digital/analog sections of A/D and D/A converters.
This is what I'm using, with excellent results, to power my PCM1798 D/A setup: http://focus.ti.com/docs/prod/folders/print/tps79133.html TI will send you samples for free if you ask.
even if they are superior, there is little to learn from black-boxes.
Me would appreciate a talk on discrete voltage regulators.
regards
Me would appreciate a talk on discrete voltage regulators.
regards
- Status
- Not open for further replies.