Computer Audio Asylum

Music servers and other computer based digital audio technologies.

Return to Computer Audio Asylum

Message Sort: Post Order or Asylum Reverse Threaded

dac frequency response..

112.209.25.246
Posted on July 18, 2012 at 08:25:12
Les C
Audiophile

Posts: 195
Joined: November 27, 2003
 just one of my random questions:

when a dac says frequency response and it says it's just 20-20,000 - is it really limited to 20khz even if it can work with sampling rate of up to 192khz ?

 

Hide full thread outline!
    ...
RE: dac frequency response.., posted on July 18, 2012 at 09:24:11
AbeCollins
Audiophile

Posts: 19504
Location: USA
Joined: June 22, 2001
Contributor
  Since: February 2, 2002
 That's the analog output audio bandwidth, not the digital sampling rate.

 

RE: dac frequency response.., posted on July 18, 2012 at 15:15:27
bwb
Audiophile

Posts: 1286
Joined: July 25, 2011
Contributor
  Since: January 4, 2013
 .

I think that's what he meant... Is the analog limited to 20-20K?

The answer is no, the theoretical limit is 1/2 the sample rate but depending on the DAC they may decide to roll off the output before that.

I think they spec 20-20K plus/minus some dB because that is what people are used to seeing.

.

 

RE: dac frequency response.., posted on July 18, 2012 at 16:26:16
AbeCollins
Audiophile

Posts: 19504
Location: USA
Joined: June 22, 2001
Contributor
  Since: February 2, 2002
 For practical purposes the analog output is limited to 20Hz - 20KHz which is the most common spec that is cited for the human audible range. Whether it's rolled off at the limits is up to the manufacturer, and of course his hearing.


 

Question for you, posted on July 19, 2012 at 00:11:47
Frihed89
Audiophile

Posts: 13064
Location: Copenhagen
Joined: March 21, 2005
 Do DACs, as a general rule, have a flatter frequency response than amplifiers with less roll-off and with less variation between DACs than amps?

I don't know the answer; it is only my vague impression from measurements I have seen in Stereophile that this is the case.
Observe, before you think. Think before you open your yap. Act on the basis of experience.

 

"for practical purposes" ...... what does that mean?, posted on July 19, 2012 at 06:28:36
bwb
Audiophile

Posts: 1286
Joined: July 25, 2011
Contributor
  Since: January 4, 2013
 .


They are either limited or they are not. What does "practical" have to do with it?

DACs aren't limited to 20-20K any more than amps or preamps are.

.

 

RE: dac frequency response.., posted on July 19, 2012 at 08:28:09
Archimago
Audiophile

Posts: 572
Joined: January 18, 2002
 "Do DACs, as a general rule, have a flatter frequency response than amplifiers with less roll-off and with less variation between DACs than amps?

I don't know the answer; it is only my vague impression from measurements I have seen in Stereophile that this is the case."

IMO this is correct. Sound differences between DAC's are quite small for standard designs aimed at flat frequency response. I've tested good inexpensive vs. "orthodox" top end units and the difference is miniscule (ie. not worth bothering about). Differences I've heard are usually due to factors like NOS or tube analog outputs... Whether one thinks this is a good thing is subjective.

-------
Archimago's Musings: A 'more objective' audiophile blog.

 

What DACs did you test?, posted on July 19, 2012 at 09:20:52
Sordidman
Audiophile

Posts: 11242
Location: San Francisco
Joined: May 14, 2001
 Can you list some, all?



"Asylums with doors open wide,
Where people had paid to see inside,
For entertainment they watch his body twist
Behind his eyes he says, 'I still exist.'"

 

Time for more resolving system, perhaps?, posted on July 19, 2012 at 13:28:17
carcass93
Audiophile

Posts: 5417
Location: NJ
Joined: September 20, 2006
 This is so way off base - it's not even funny.

 

RE: "for practical purposes" ...... what does that mean?, posted on July 19, 2012 at 13:33:41
AbeCollins
Audiophile

Posts: 19504
Location: USA
Joined: June 22, 2001
Contributor
  Since: February 2, 2002
 Look at the OP. He says the DAC is spec'd at 20-20,000 so if this is the case, then he is talking about the analog output from this particular DAC.

And "for practical purposes" how many of us actually hear beyond 20 - 20,000Hz. And "for practical purposes" I answered the OP's question.

I think you're over complicating what was originally a very simple question requiring a very simple answer, not a dissertation on Nyquist sampling theorem.



 

actually, you completely ignored his question, posted on July 19, 2012 at 14:37:13
bwb
Audiophile

Posts: 1286
Joined: July 25, 2011
Contributor
  Since: January 4, 2013
 ,

Of course he was talking about the analog output... that's why I was surprised your answer pointed out that 20-20K was not the sample rate... he never hinted that it was.

I didn't complicate things, I answered his question... His question definitely pertains to Nyquist and the relationship to the 20-20K... why else would he ask.

"when a dac says frequency response and it says it's just 20-20,000 - is it really limited to 20khz even if it can work with sampling rate of up to 192khz"

so your response definitely did not answer that question..



He didn't ask if it was practical or not, only if it really was limited..

.


.

 

RE: actually, you completely ignored his question, posted on July 19, 2012 at 15:02:15
AbeCollins
Audiophile

Posts: 19504
Location: USA
Joined: June 22, 2001
Contributor
  Since: February 2, 2002
 If you take his question at face value, he is already stating up front that the DAC's frequency response is 20Hz - 20 KHz, so unless you have proof to the contrary, then this particular DACs response is limited to 20Hz - 20KHz regardless of the digital sampling rate.

Do we know the manufacturer of this DAC? Not at this point, but if the manufacturer says 20 - 20,000Hz then "for practical purposes" that IS the DAC's frequency response.

 

really?, posted on July 19, 2012 at 15:23:34
bwb
Audiophile

Posts: 1286
Joined: July 25, 2011
Contributor
  Since: January 4, 2013
 .


When a manufacture states his amp, or preamp, or whatever has a response of 20-20K that doesn't mean there are brick wall filters to limit the response. It doesn't mean that those are the limits. To state otherwise is just silly. I'm surprised you would even go there.

His question was clearly about analog out as it relates to Nyquist.

nice chatting with you

.

 

Huh??, posted on July 19, 2012 at 15:41:12
Dynobot
Audiophile

Posts: 2700
Location: USA
Joined: December 8, 2011
 >>>Sound differences between DAC's are quite small for standard designs aimed at flat frequency response.

Wow between you and someone else the whole front-end of audio makes no difference in sound.

[Sources sound the same]
[Dacs sound the same]

Can I get someone who thinks preamps, amps or speakers all sound the same. Lets go for a full-house!

Better yet, how about all music sounds the same?


Dynobots Audio
Music is the Bridge between Heaven and Earth - 音楽は天国と地球のかけ橋

 

RE: dac frequency response.., posted on July 19, 2012 at 17:54:22
Tony Lauck
Audiophile

Posts: 10387
Location: Vermont
Joined: November 12, 2007
Contributor
  Since: February 24, 2009
 A DAC can't go much above 20,000 at 44.1 kHz. If the DAC goes higher it would be only at higher sample rates. Some products quote frequency response at various sample rates. You will also see this in some reviews.

Frequency response is hardly the be all and end all of DAC performance. In fact there isn't even a standard method of measuring some of the really critical aspects of DAC performance. You have to listen to see how a DAC sounds.

Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

RE: actually, you completely ignored his question, posted on July 19, 2012 at 18:02:52
Tony Lauck
Audiophile

Posts: 10387
Location: Vermont
Joined: November 12, 2007
Contributor
  Since: February 24, 2009
 "if the manufacturer says 20 - 20,000Hz then "for practical purposes" that IS the DAC's frequency response."

Marketing literature has to be read with extreme caution. It is often complete BS and is seldom written by engineers.


Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

RE: What DACs did you test?, posted on July 19, 2012 at 18:36:44
Archimago
Audiophile

Posts: 572
Joined: January 18, 2002
 I won't list all because I suggest you guys try it for yourself... I'm just an anonymous Internet poster just like the rest of you.

Make sure to volume equalize the outputs.

One of the expensive DAC's I tested is the Weiss DAC202.

-------
Archimago's Musings: A 'more objective' audiophile blog.

 

RE: Huh??, posted on July 19, 2012 at 19:07:13
Archimago
Audiophile

Posts: 572
Joined: January 18, 2002
 No, speakers don't sound the same. :-)

-------
Archimago's Musings: A 'more objective' audiophile blog.

 

RE: Time for more resolving system, perhaps?, posted on July 19, 2012 at 19:12:24
Archimago
Audiophile

Posts: 572
Joined: January 18, 2002
 Lets just say the test equipment should be *very* adequate if such a thing is defined by $$$, Stereophile "Class A", and The Absolute Sound "Best of The Year" recommendations.

I do not expect to change any minds who have been made up. However, for Frihed89 who pondered about the sound difference, my experience is consistent with what he/she wrote and what the graphs suggest.

-------
Archimago's Musings: A 'more objective' audiophile blog.

 

RE: actually, you completely ignored his question, posted on July 19, 2012 at 20:38:40
AbeCollins
Audiophile

Posts: 19504
Location: USA
Joined: June 22, 2001
Contributor
  Since: February 2, 2002
 "Marketing literature has to be read with extreme caution. It is often complete BS and is seldom written by engineers."

Of course, but lacking additional information to the contrary one can't pick some hypothetical number out of thin air either.

 

Stereophile Class A, posted on July 20, 2012 at 03:13:04
fmak
Audiophile

Posts: 9053
Joined: June 1, 2002
 A particular preamp that got an A sounded to me like Class B, and that was in comaprison with truly Class A units.

It's a function of the entire system and preference.

 

RE: Stereophile Class A, posted on July 20, 2012 at 06:18:21
Dynobot
Audiophile

Posts: 2700
Location: USA
Joined: December 8, 2011
 Just curious, was the price consistent with Class A products?

How about the looks or the companies reputation.

For example a Jeff Rowland preamp costing $20K will be more likely to show up as Class A no matter what the sound vs. a $1k Chinese preamp. Example shows extremes but you get the idea.

Being made in China = never a glowing review no matter what the sound
Being made in Japan = a favorable benefit of the doubt for good sound
Being made in the USA = Rose colored glasses on, benefit of the doubt applied and a glowing review if it sounds good and a avg review even if it sounds bad.

Where components are made not only flavors the sound but also flavors the review. I've read so many times "Classic British sound....", no doubt it probably does but those words alone denote bias.


Dynobots Audio
Music is the Bridge between Heaven and Earth - 音楽は天国と地球のかけ橋

 

RE: Stereophile Class A, posted on July 20, 2012 at 06:59:39
fmak
Audiophile

Posts: 9053
Joined: June 1, 2002
 It was not cheap, made in USA by a well respected designer/manufacturer. It was a minimal stage Class A device but the sound was governed by the coupling capcitor - nice and mellow but lacked dynamics.

 

He did NOT write anything on the difference in sound of the DACs. As for graphs - ..., posted on July 20, 2012 at 08:19:35
carcass93
Audiophile

Posts: 5417
Location: NJ
Joined: September 20, 2006
 ... you're probably one of the very few people on Earth who listen with their eyes.

Apparently, in your opinion all the equipment that measures the same in THD % and frequency response sounds the same?

I have strong feeling that you're in a wrong forum...

 

Nothing's impossible. Speakers can easily be made sounding the same - ..., posted on July 20, 2012 at 08:42:18
carcass93
Audiophile

Posts: 5417
Location: NJ
Joined: September 20, 2006
 ... if you connect them to source, DAC, pre and amp that sound the same.

Of course, you'll have to listen with ears which couldn't distinguish any differences between anything at all, and just to be sure, program the brain to filter out any differences - if, God forbid, they get past those first several barriers. Some posters here seem to be pretty good at that.

Music sounding all the same will be the last, and most important, frontier.

 

RE: dac frequency response.., posted on July 20, 2012 at 09:01:09
fmak
Audiophile

Posts: 9053
Joined: June 1, 2002
 With 192k, the digital filter roll off point is around 96k, but the dac itself may have an addtional analog filter that rolls off at around 21k, to cater for CD replay.

One dac I built has anaolg filters that roll off near half the sampling frequency (44.1 and 96k) and this sounds very good. I believe that not many do, hence the 20k spec.

Both the analog and digital filters affect sound quality, easpecially in PCM in which transient and square wave responses are 'skewed'.

 

RE: He did NOT write anything on the difference in sound of the DACs. As for graphs - ..., posted on July 20, 2012 at 09:49:04
Archimago
Audiophile

Posts: 572
Joined: January 18, 2002
 "... you're probably one of the very few people on Earth who listen with their eyes."

Curious why you presume this. I'm making a point that IMO, the data fits my experience. ABX / blinded / unblinded - I've done these as best I can to understand and experience the qualitative difference. I've taken the time to control the variables like cables, same music pressing, same/similar playback transports, obviously same amp & speaker system.

Furthermore, I've done this with a number of audiophile friends and when controlled in this way, (after doing this about 5 times on different systems over the years) NOBODY has been able to reliably differentiate good quality DAC's on opposite spectrum of the price range (we're comparing decent orthodox DAC designs from $300 to $6000+).

I don't see why to the audiophile crowd, the cost of something should be so important. It's like saying an expensive LV purse can hold things "better" than a decent $50 knock-off. The only difference is in the pride of ownership (and I guess some people care about this to the point of blurring the boundaries with sound quality in audio world).



-------
Archimago's Musings: A 'more objective' audiophile blog.

 

RE: dac frequency response.., posted on July 20, 2012 at 09:58:41
Tony Lauck
Audiophile

Posts: 10387
Location: Vermont
Joined: November 12, 2007
Contributor
  Since: February 24, 2009
 Elaborating a bit on what you said...

A common way to build a multi-rate DAC is to resample to a multiple of the input sample rate (e.g. 384 kHz or 352.8 kHz or higher) using a digital filter that depends on the input sample rate. The result is then converted to analog and filtered with a fixed analog filter calibrated according to the high master clock frequency. The variable digital filter removes images according to the sample rate, and the fixed analog filter removes any images from the conversion to analog at the master clock rate. The net result is different frequency response at each sample rate. With a NOS DAC it will be necessary to use a different analog filter for each sampling rate to achieve the best frequency response at each sample rate.

Any real-world low-pass filter will have a transition band. Below the transition band the signal is passed "essentially unchanged" and above this transition band the signal is "completely" rejected, where the filter design includes the size of the transition band and the definition of "essentially" and "completely". There are mathematical relationships between the width of this transition band and the filter properties, in particular if the transition band is narrow then the filter will ring for a long time. If you want to avoid excessive ringing then it will be necessary to widen the transition band and this implies rolling off some audio frequencies below 1/2 the sampling rate or passing some spurious frequencies above 1/2 of the sampling rate. In audible terms there is a three way conflict between good transient response, good frequency response, and rejection of spurious frequencies. This tradeoff will depend on the recording, the playback chain and the listener.

Companies that make high-end converters use bespoke filters that take into account psychoacoustic tradeoffs. For example, at the higher sampling rates the spurious frequencies are well above the audible range and it becomes possible to use filters with a wide transition band that have very little ringing without rolling off audio that is likely to be on the recording. For low sampling rates, e.g. 44.1 kHz, where the margin between audible frequencies and spurious frequencies is much less, narrower transition band filters are used. You can see this in the specs and white papers of companies such as DCS and DAD.

One can get filtering performance similar to that of the bespoke filters used in high end DACs by upsampling in software in the PC using a good filter and sending the result to a DAC operating at a higher sample rate. Stacking the filters may add additional roundoff errors and/or dither noise, but with a 24 bit or 32 bit DAC this can be at very low levels, so results may be nearly as good as with the high end DAC, or possibly better if one implements filters more suited to one's system. On my system, the Mytek Stereo192-DSD DAC sounds best playing all PCM sample rates by upsampling to 5.6 MHz (DSD128) in the computer. There is not a big difference over using the stock filters in the SABRE DAC, but there does seem to be an audible difference. I have been using the "poly sinc short mp" filter and the "SDM5" modulator of HQPlayer for this conversion.



Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

No, none of us are anonymous. , posted on July 23, 2012 at 11:15:03
Sordidman
Audiophile

Posts: 11242
Location: San Francisco
Joined: May 14, 2001
 The MOST IMPORTANT thing that we bring here to these forums is an historical perspective on our experiences in context.

If you can't relate your experiences, then you have less to offer, and less knowledge.



"Asylums with doors open wide,
Where people had paid to see inside,
For entertainment they watch his body twist
Behind his eyes he says, 'I still exist.'"

 

LOL, posted on July 23, 2012 at 11:20:11
Sordidman
Audiophile

Posts: 11242
Location: San Francisco
Joined: May 14, 2001
 "" NOBODY has been able to reliably differentiate good quality DAC's"'

Now that's super funny stuff. Clueless and ignorant that it is....

"Asylums with doors open wide,
Where people had paid to see inside,
For entertainment they watch his body twist
Behind his eyes he says, 'I still exist.'"

 

RE: dac frequency response.., posted on July 24, 2012 at 15:45:05
Les C
Audiophile

Posts: 195
Joined: November 27, 2003
 thanks guys. very informative and i learned a lot.. the usual case.

reason i asked was a lot of dacs have quoted frequency response of 20-20k but go up to 192khz or higher. makes me wonder what the benefits are of accepting certain sample rates but not having an analog output capable of reproducing the higher frequencies.

i have j river's parametric set to remove greater than 17khz and anything below 25hz. it has certain effect.

 

Page processed in 0.027 seconds.

[ Top Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2013, Audio Asylum, all rights reserved ] [ Take5 ]
[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]