Jump to content

Why do some audiophiles spend thousands of dollars on a DAC? Are they searching for a "sound signature" they like, or just greater "accuracy"?


Recommended Posts

41 minutes ago, MLXXX said:

If exactly the same extent of attenuation occurred at each round trip through the audio interface,

 

That's what I am doubting.  On the first pass the ADC captures all it can, and misses some small amount.  On the subsequent passes it misses less each time.  You can't just calculate it simply unless you know the non-linear function for it's behaviour.

Link to post
Share on other sites
  • Replies 375
  • Created
  • Last Reply

Top Posters In This Topic

Top Posters In This Topic

Popular Posts

A consumer DAC is much more than just a chip.   If you take the PoV that anything in a hifi system degrades the sound ... people search for components that deliver the minimum degradation in

Yeah, honestly @MLXXX the whole premise of allowing YouTube viewers to judge sound quality of different DACs from a recorded video, played back on their device is massively flawed.  Others have alread

There's a 'Dorothy Dixer' if ever I heard one!     Andy  

2 minutes ago, frednork said:

As interesting as this may be, I fail to see how this relates to the topic under discussion. Am I missing something?

 

 

If a DAC (digital to analogue conversion device) were of very poor audio quality then even a single use could be expected to result in quite noticable degradation.  Similarly, If an ADC (audio to digital conversion device) were of very poor audio quality then a single use of it could be expected to  result in quite noticeable degradation.

 

File 2 involved two conversions:  DAC followed by ADC

File 3 involved 2 times "X" conversions:  X times (DAC followed by ADC)

File 4 involve 4 times "X" conversions:  2X times (DAC followed by ADC)

 

File 4 has been reported in this thread to be only barely noticeably impaired or not noticeably impaired if played in a standalone manner (as distinct from an immediate A B comparison manner) against file 1. Even played with immediate A B switching it has been reported as only slightly degraded.

 

It is clear at this point that the DAC and ADC devices in my mid-priced audio interface in a single use of them appear to impart very mild audible degradation (if any audible degradation) at the sample rate (44.1kHz) used.

 

kukynas performed a similar exercise with his more advanced audio interface ADC and two DACs and the resulting files from a singe re-recording showed only a very mild audible change (that is my assessment; I don't think there have been too many comments on his files).

 

So we are in the rather good place of having ADCs that appear to work well, and of DACs that appear to work well.

 

The goal was to demonstrate whether we could successfully replicate all the bits on a standard CD track without any loss whatsoever. Given the high linearity of the devices (extending beyond 16 bits ) and a very close to flat frequency response apart from frequencies approaching the Nyquist limit, that seemed feasible if the phase of the ADC sampling was locked to the phase of the DAC clocking [and adjusted for any latency effects]. 

 

If that goal had been achieved it would have demonstrated that the CD track could be reproduced perfectly right down to the 16th bit, raising the question what else might be pursued beyond complete accuracy.  That presumably would be a "sound signature".

  • Like 1
  • Thanks 1
Link to post
Share on other sites
2 hours ago, frednork said:

As interesting as this may be, I fail to see how this relates to the topic under discussion. Am I missing something?

I do not think that it is even interesting.  It is just a continuation of more confusion as to what the topic is ostensibly about. 

John

Edited by Assisi
word missing
  • Like 2
Link to post
Share on other sites
25 minutes ago, aussievintage said:

 

That's what I am doubting.  On the first pass the ADC captures all it can, and misses some small amount.  On the subsequent passes it misses less each time.  You can't just calculate it simply unless you know the non-linear function for it's behaviour.

I don't think the ADC ever "misses" the continuous 11.025kHz sine wave at the start of the file The DAC or the ADC (or both of them, each to some extent)  merely attenuates that sine wave to some extent. It is still there at each re-recording, just at a progressively more and more reduced amplitude (relative to the amplitude of the 1kHz sine wave at the end of the file).

 

Anyway we shall see, if someone does the calculations. (If no one else does, I will.)

Link to post
Share on other sites


1 hour ago, MLXXX said:

f a DAC (digital to analogue conversion device) were of very poor audio quality then even a single use could be expected to result in quite noticable degradation.  Similarly, If an ADC (audio to digital conversion device) were of very poor audio quality then a single use of it could be expected to  result in quite noticeable degradation.

Didnt think we were talking about comparing "very poor audio quality" dacs, In fact the opposite!!

 

1 hour ago, MLXXX said:

It is clear at this point that the DAC and ADC devices in my mid-priced audio interface in a single use of them appear to impart very mild audible degradation (if any audible degradation) at the sample rate (44.1kHz) used.

Not that I see how it applies but even in this you arent separating out the effect of the dac and adc

 

1 hour ago, MLXXX said:

The goal was to demonstrate whether we could successfully replicate all the bits on a standard CD track without any loss whatsoever. Given the high linearity of the devices (extending beyond 16 bits ) and a very close to flat frequency response apart from frequencies approaching the Nyquist limit, that seemed feasible if the phase of the ADC sampling was locked to the phase of the DAC clocking [and adjusted for any latency effects]. 

 

If that goal had been achieved it would have demonstrated that the CD track could be reproduced perfectly right down to the 16th bit, raising the question what else might be pursued beyond complete accuracy.  That presumably would be a "sound signature".

To be honest I have lost track of what has and hasnt been achieved as I still dont understand the point of the exercise besides confirming what is already known which is that degradation occurs when passed through an ADDA process. How that relates to a possible sound signature is not a linear extension of this for me.

 

However this is your thread and you should take it where you wish. I truly hope you achieve your aim. I will be interested to read it.

Edited by frednork
Link to post
Share on other sites
34 minutes ago, frednork said:

Not that I see how it applies but even in this you arent separating out the effect of the dac and adc

Yes it is true that the effects are not separated, but surely it is a more arduous test for audio to be put through the gauntlet of two domain conversions, than just to be subjected to one or the other conversion.

 

There is no suggestion of "compensating errors" as between the DAC effects and the ADC effects.

 

 

33 minutes ago, frednork said:

However this is your thread and you should take it where you wish. I truly hope you achieve your aim. I will be interested to read it.

Thanks.

Edited by MLXXX
Link to post
Share on other sites
On 21/11/2020 at 3:36 AM, frednork said:

Is this what you are trying to say ? that non faulty amplifiers cant have  this sort of difference.

 

No, but I'm saying if one amplifier "muddles the signal" then this will be measurable.

What do we have to measure amplifiers? Oscilloscopes, Distortion Analysers, Spectrum Analysers.

I'm not an expert, do you know of anything else?

 

But of course anything can be rebutted with a simple "I'm a firm believer the current measurements aren't up to it." if it can't be measured.

I don't know, but if that's not speculation then I don't know what is!

Once again, you're allowed to speculate, but lets just call it what it is.

 

Also, I'll give Bryston apparently in this case the benefit of the doubt that they know how to build a simple amplifier.

But put any amplifier past its limits into clipping and the signal falls apart (gets muddled).

Clipping is measurable by the way.

 

Electronics can have faults, maybe this is what happened.

My old man currently has an NAD amplifier with some sort of fault.

Turn the balance control Left or Right and it balances the opposite of what it should.

Swap the signal leads on just one end and then again turn the balance control Left to Right and it still balances the opposite of what it should, spooky!

Link to post
Share on other sites
5 minutes ago, Satanica said:

 

Also, I'll give Bryston apparently in this case the benefit of the doubt that they know how to build a simple amplifier.

But put any amplifier past its limits into clipping and the signal falls apart (gets muddled).

Clipping is measurable by the way.

 

 

Wow, S - you really like to clutch at every straw you can!  xD

 

The Bryston amp which I referred to had several 100 watts on tap (per channel).

 

If this was clipping into the big Dalis we were listening to ... then the 25w SE Class A amp which surprised us, in the amp comparison, would certainly have been clipping.  Yet this was the amp that projected the singer forwards of the spkr plane.

Andy

 

  • Like 1
Link to post
Share on other sites


On 21/11/2020 at 3:36 AM, frednork said:

The 2015 remaster was about fulfilling his original concept which inlcuded samples from 2001 A Space odyssey which were witheld by Kubrick while alive. Once he died the estate gave permission (and took the money)

 

Your system must go pretty low as many systems are rolling off on the low bits. 

 

Yes probably lower than most. Dirac says flat to 16Hz (I have two largish subwoofers). On the new remix\remaster I found Roger's bass just too hot through my system that it brings too much attention to itself. But there were aspects to the new mix\master that sounded even more spectacular than the original.

  • Like 1
Link to post
Share on other sites
14 minutes ago, andyr said:

 

Wow, S - you really like to clutch at every straw you can!  xD

 

The Bryston amp which I referred to had several 100 watts on tap (per channel).

 

If this was clipping into the big Dalis we were listening to ... then the 25w SE Class A amp which surprised us, in the amp comparison, would certainly have been clipping.  Yet this was the amp that projected the singer forwards of the spkr plane.

Andy

 

 

Hi Andy, well I was just giving example of what might happen not necessarily in this case, but in general.
And I'm allowed to speculate on here remember. 🤗

 

More importantly, how do you know which is the reference, the singer forward or in line with the speaker plane? 🧐

 

Edited by Satanica
Link to post
Share on other sites
1 hour ago, Satanica said:

 

More importantly, how do you know which is the reference, the singer forward or in line with the speaker plane? 🧐

 

 

You would appear to have missed DrSK's post on Thursday?  xD

 

Andy

 

  • Haha 1
Link to post
Share on other sites
2 hours ago, Satanica said:

No, but I'm saying if one amplifier "muddles the signal" then this will be measurable.

What do we have to measure amplifiers? Oscilloscopes, Distortion Analysers, Spectrum Analysers.

I'm not an expert, do you know of anything else?

Not an expert either but it would have to measure soundstage depth somehow. The rest are probably not that relevant.

Link to post
Share on other sites


2 hours ago, Satanica said:

No, but I'm saying if one amplifier "muddles the signal" then this will be measurable.

What do we have to measure amplifiers? Oscilloscopes, Distortion Analysers, Spectrum Analysers.

I'm not an expert, do you know of anything else?

 

12 minutes ago, frednork said:

Not an expert either but it would have to measure soundstage depth somehow. The rest are probably not that relevant.

 

Sorry to use your quotes as an example,  but as this thread evolves I  see more and more reason why the twain shall never meet.  Not you too, but the objective/subjective divide.    There really is no common language to explain one side to the other.

  • Like 2
Link to post
Share on other sites
On 13/11/2020 at 6:25 PM, andyr said:

OK ... let me put a scenario to you (involving 2 amps) and perhaps you would tell me what you would measure, in order to show why amp #2 did what it did.  :)  (The 5 people at this listening session all heard the same thing.)

As the question of what to measure for an amplifier that seems to alter the sound stage radically is still being discussed, I'll add in my 2 cents worth.

 

First off, width and depth are perceived by the ears in conjunction with the brain. You could program software to mimic that perception function though it wouldn't be straightforward and might only be feasible for tracking one instrument or vocalist at a time.  Information needed to perceive depth includes:

* compared with what could be expected if the source were close, the higher frequencies seem dull

* compared with what could be expected if the source were close, the sound seems soft

* compared with could be expected if the source were close, there is a great deal of audible reverberation of the source

 

All of the above attributes would be in play if you placed a single microphone a few metres in front of a  solo singer who was in front of a large massed choir, inside  a concert hall.  The soloist would be much louder than the choir; his or her recorded voice  would retain its high frequencies (not dulled by high frequency absorption that is progressively worse the greater the distance), and his or her recorded voice would have a high ratio of direct sound volume to reflected sound volume.

 

All an amplifier needs to do is to amplify in a neutral way (e.g. a flat frequency response, no strange phase shifts) to preserve the depth information in a recording.

 

So if two amplifiers are yielding dramatically different depth effects when driving speakers, for the same recording, at least one of the amplifiers must be "messing with" the sound (to use a word that has cropped up a few times already in this thread!).

 

Just the very basic usual battery of tests for measuring amplifier performance should be enough to reveal which amplifier is not delivering sound with a flat extended frequency response, low distortion, and a high damping factor. (I note that if the recording source is a standard CD there may not be too much point in testing with fast rise time square waves, as CD content is bandlimited to 22kHz, but it wouldn't do any harm to include such testing.)

 

Before typing this post I came across what I think is a very interesting webpage on how to create a recording mix that will provide perceptions of depth (and height) for people listening to it:   8 Tips for Creating Depth And Height In The Mix

Edited by MLXXX
Link to post
Share on other sites
14 hours ago, aussievintage said:

Not you too, but the objective/subjective divide. 

 

 I just had the thought that if you just think of the taste of a lemon for instance, your mouth starts to flow with saliva.

 

But one cannot measure the sweetness or tartness of the lemon you imagined.

 

If you even think of the word lemon.......

 

Sorry gents, back to the discussion at hand.

Link to post
Share on other sites


16 hours ago, frednork said:

Not an expert either but it would have to measure soundstage depth somehow. The rest are probably not that relevant.

 

The signal, its all about the signal and the rest are what it used to measure the signal.

Link to post
Share on other sites

 

19 hours ago, MLXXX said:

First off, width and depth are perceived by the ears in conjunction with the brain. You could program software to mimic that perception function though it wouldn't be straightforward and might only be feasible for tracking one instrument or vocalist at a time.  Information needed to perceive depth includes:

* compared with what could be expected if the source were close, the higher frequencies seem dull

* compared with what could be expected if the source were close, the sound seems soft

* compared with could be expected if the source were close, there is a great deal of audible reverberation of the source

 

All of the above attributes would be in play if you placed a single microphone a few metres in front of a  solo singer who was in front of a large massed choir, inside  a concert hall.  The soloist would be much louder than the choir; his or her recorded voice  would retain its high frequencies (not dulled by high frequency absorption that is progressively worse the greater the distance), and his or her recorded voice would have a high ratio of direct sound volume to reflected sound volume.

 

9 hours ago, Satanica said:

The signal, its all about the signal and the rest are what it used to measure the signal.

 

Yes, you are both allowed to speculate and that is exactly what the above is unless you can show practically how it does correlate with depth of image by either your work or someone elses.

 

I agree it is buried in the signal somewhere just dont agree there is a reliable measurement which correlates with it.

Edited by frednork
  • Like 2
Link to post
Share on other sites
2 hours ago, frednork said:

 just dont agree there is a reliable measurement which correlates with it.

In my reading of previous posts, no one has suggested a single measurement to explain depth.

 

As I say, all the amplifier needs to do is to reproduce the signal accurately. If it does that, any depth cues in the recording mix will be properly revealed.  [Subject of course to speakers or listening room acoustics not "messing" with the sound.]

  • Like 1
Link to post
Share on other sites
5 minutes ago, MLXXX said:

In my reading of previous posts, no one has suggested a single measurement to explain depth.

I agree, and just as happy for you to show any combination of measurements which do so and have been shown to correlate with depth perception.

 

9 minutes ago, MLXXX said:

As I say, all the amplifier needs to do is to reproduce the signal accurately.

Not  sure what this statement really means as no amplifier is truly accurate,  There are only different amounts and types of innaccuracy.  which type of innacuracy correlates with sound stage depth I cant say.

Link to post
Share on other sites
2 minutes ago, frednork said:

Not  sure what this statement really means as no amplifier is truly accurate,  There are only different amounts and types of innaccuracy.  which type of innacuracy correlates with sound stage depth I cant say.

This tangential discussion introduced by andyr was on the topic of  finding an explanation for a "wow" change to the apparent position of a female vocalist.

 

Such a dramatic change would not arise because of a minor inaccuracy in an amplifier.   (In fact it ought to be possible to substitute 10 different standard amplifiers from different manufacturers and hear either no, or only slight, changes in the apparent sound stage.)  Assuming the people involved in andyr's anecdote didn't accidentally bump one of the speakers when changing amplifiers, we are left with the conclusion that at least one of the amplifiers was non-standard in its inherent performance (or set to a non-standard setting).

 

I suggest we move on, as this thread is about DACs, not amplifiers and sound stage depth!

Link to post
Share on other sites
On 15/11/2020 at 4:29 AM, MLXXX said:

Late edit: Later measurements using an RMS measuring method have found the performance at 11025kHz to be almost spot on. (The original figures were thrown out by the use of a frequency that was exactly one quarter the sample rate.)  The updated results are:

 

RME ADI-2 DAC FS  response at 11025Hz:-  0.00565dB down (stereo) relative to the response at 1kHz

Topping E30  response at 11025Hz:-  0.04872dB down (stereo) relative to the response at 1kHz

 

The REME result is obviously very good indeed.

 

As shown above, I have had to correct my measurements for @kukynas's files. Although the amplitude of the 1kHz signal could be measured in various ways, the 11.025kHz signal required not only the highest sample value to be ascertained but the peak value of the reconstructed waveform.  There is a normally hidden analysis tool for Audacity (an RMS analysis) that gives the proper result.

Edited by MLXXX
Link to post
Share on other sites
On 23/11/2020 at 9:57 AM, aussievintage said:

That's what I am doubting.  On the first pass the ADC captures all it can, and misses some small amount.  On the subsequent passes it misses less each time.  You can't just calculate it simply unless you know the non-linear function for it's behaviour.

Hi, well the proof of the pudding is in the eating! I've just finished analysing the re-recordings. At 11.025kHz there was a loss of 0.55dB ± 0.03dB in each channel for each  round trip through the FCA 1616 interface, compared with the 1kHz level.   It was merely the same slight attenuation each time!

 

I'll provide detailed figures when I have more time to post. (Will also divulge how many round trips were involved for files 3 and 4.)

Edited by MLXXX
  • Like 1
Link to post
Share on other sites
3 hours ago, MLXXX said:

Hi, well the proof is in the pudding! I've just finished analysing the re-recordings. At 11.025kHz there was a loss of 0.55dB ± 0.03dB in each channel for each  round trip through the FCA 1616 interface, compared with the 1kHz level.   It was merely the same slight attenuation each time!

 

 

Fair enough.      Concede :)     Gee lookit that, measurements being useful :sorcerer:

Link to post
Share on other sites
On 14/11/2020 at 12:00 AM, MLXXX said:

Using my own equipment, a Behringer FCA1616 audio interface,  as both a DAC and an ADC, I was nowhere near achieving bit-for-bit accuracy. Although the "round trip" (DAC to analogue cable to ADC) frequency response was well maintained at low frequencies, it tapered off significantly from about 10kHz.  This was using a sample rate of 44.1kHz, the same rate as the music file. 

 

Via Google Drive I can share with the forum the re-recordings I've made. The files are:

 

  1. The original as downloaded,  but with opening and closing tones added, and at 24 bits 44.1kHz rather than the original 16 bits 44.1kHz format.
  2. The above after a single "round trip" of being played back with a DAC and simultaneously re-recorded with an ADC.
  3. A later generation copy, involving X round trips from the original downloaded file.  Forum members are welcome to guess how many round trips were involved.  Does X equal 2, 3, even more?
  4. An even later generation copy, involving exactly twice as many round trips.  

 

The files are labelled and it is not necessary to guess which is which. Here they are: The 4 test files     The only unknown forum members are being asked to guess is how many passes through the audio interface were involved for file 3 [being half as many as for file 4].   

The number of round trips involved in creating file 3 was  5 (a total of 10 conversions between digital and analogue domains).  File 4 involved 10 round trips (a total of 20 conversions between digital and analogue domains).

 

This table shows the progressive attenuation of the burst of 11.025kHz sine wave at the start of file 1, relative to the burst of 1kHz sine wave at the end of file 1:

               
  11.025kHz   RMS levels in dB ROUND TRIP CHANGE  
  COPY # LABEL LEFT RIGHT LEFT RIGHT  
  0 File 1 -9.03 -9.03 n/a n/a  
  1 File 2 -9.57 -9.58 -0.54 -0.55  
  2   -10.12 -10.15 -0.55 -0.57  
  3   -10.67 -10.7 -0.55 -0.55  
  4   -11.21 -11.26 -0.54 -0.56  
  5 File 3 -11.76 -11.81 -0.55 -0.55  
  6   -12.3 -12.37 -0.54 -0.56  
  7   -12.85 -12.93 -0.55 -0.56  
  8   -13.39 -13.48 -0.54 -0.55  
  9   -13.92 -14.03 -0.53 -0.55  
  10 File 4 -14.46 -14.6 -0.54 -0.57  
               

 

It would have been possible to "cheat" and determine the number of generations of re-recordings by comparing the attenuation  of the initial tone in file 2 (a single generation copy), with the attenuation of the initial tone in file 3 (the unknown generation copy), relative to the reference file, file 1.  This table shows how that could have been done:

               
  11.025kHz   CUMULATIVE CHANGE MULTIPLE OF FILE 2 CHANGE  
  COPY # LABEL LEFT RIGHT LEFT RIGHT  
  1 File 2 -0.54 -0.55 1.000 1.000  
  5 File 3 -2.73 -2.78 5.056 5.055  
  10 File 4 -5.43 -5.57 10.056 10.127  
               

 

 

Congratulations to Stereophilus for guessing (not cheating!) and getting pretty close to the correct answer of 5 round trips through the audio interface to arrive at the audible sound quality of file 3:

 

On 22/11/2020 at 10:18 PM, Stereophilus said:

I would guess 3 or 4, but it’s purely guess work.

 

 

 

 

Edited by MLXXX
  • Like 1
Link to post
Share on other sites

The picture below shows how the files looked on the DAW (digital audio workstation) display on the laptop I used for the re-recordings. The Behringer FCA1616 interface was connected via a USB cable to the laptop. The DAW software was Tracktion Waveform Pro 11. The analogue cable used for each of the two channels was 5 metres of stereo cable with 6.5mm plugs at each end. The capacitance of that cable would have contributed slightly to the attenuation at 11.025kHz.

 

The slanting text was added later and identifies the files with the same labels I have used in previous posts.

 

Astro-withFilesLabelled.thumb.png.57a2f2168d18085174a46c61d490df56.png

 

 

 

Link to post
Share on other sites

So, can we infer that some of the attenuation at 11kHz is due to the DAC... say 50%?  And if so, that this attenuation subsequently gets amplified in a typical hifi setup by maybe 10-15x downstream from the DAC before hitting the speakers?

Link to post
Share on other sites
9 hours ago, Stereophilus said:

So, can we infer that some of the attenuation at 11kHz is due to the DAC... say 50%?  And if so, that this attenuation subsequently gets amplified in a typical hifi setup by maybe 10-15x downstream from the DAC before hitting the speakers?

Yes whatever the extent of high frequency attenuation at the output of the DAC was, it would duly find its way to speakers, after being amplified.

 

Actually I'd assumed the interconnect cable was only 3m in length; 5m is getting a bit long! I may pick up shorter lengths of cable and retest.  (Another aspect is that I used a preamp input of the interface, not a line input, as the line input gain was very low.)

 

I'm happy though that the audible quality stood up well with some many trips through the interface for file 4, including going through 50 cumulative metres of interconnect cable!

 

The DAC would be rolling off a little at 11kHz I'd say with a 44kHz sample rate.  A complication is that an ADC operating at 44kHz  will also have roll-off because of its input bandlimiting filter to avoid aliasing. (I did not have an ADC available where I could disable such unnecessary input bandlimiting, unnecessary because the DAC was already delivering a bandlimited signal.) 

Edited by MLXXX
Link to post
Share on other sites

Hi gents,

 

Just want to point out that the “attenuation” doesn’t get “amplified”. The attenuation stays exactly the same from one end of the amplifier to the other, in dB.

 

Unless the amplifier is sub-par.

 

43 minutes ago, MLXXX said:

The DAC would be rolling off a little at 11kHz I'd say with a 44kHz sample rate.

 

This shouldn’t be happening with an over-sampling DAC. It is certainly not a necessary consequence of using 44.1k sampled data for storage.

 

Assuming your measurements are OK, and the unit is performing to spec, and the designers were competent, then this is deliberate tailoring, IMO.

 

cheers

Grant

Edited by Grant Slack
Link to post
Share on other sites
14 minutes ago, Grant Slack said:

Just want to point out that the “attenuation” doesn’t get “amplified”. The attenuation stays exactly the same from one end of the amplifier to the other, in dB.

Yes of course.  It is almost too obvious to say, but if the treble has been rolled off by a DAC, that roll-off will remain after the DAC signal is amplified, assuming of course a standard amplifier with a frequency response that is flat.

Link to post
Share on other sites
32 minutes ago, Grant Slack said:

This shouldn’t be happening with an over-sampling DAC. It is certainly not a necessary consequence of using 44.1k sampled data for storage.

 

Assuming your measurements are OK, and the unit is performing to spec, and the designers were competent, then this is deliberate tailoring, IMO.

 

23 minutes ago, Ittaku said:

Does the DAC concerned use a "slow roll-off filter"? By design they roll off at the top of the audible range.

 

The FCA1616 audio interface provides 8 balanced DAC outputs. I chose 2 of these on my unit at random for my tests. (I never use this interface for listening to music per se, only for making and quickly reviewing recordings of live music.) I've never looked into the DAC filter design, e.g. whether it uses a slow roll-off filter.  (I've seen largely favourable comments on the net about the DAC performance.)

 

I might try to get an answer by practical means, e.g. using an ADC operating with a 96kHz sample rate to see how the reconstruction filters of the FCA1616 DACs seem to apply for a sample rate of 44.1kHz.

 

It's interesting to review the results I found with kukynas's test recordings. No "slow roll-off" filtering apparent there! :-

           
  RME RMS level at 1000 Hz RMS level at 11025Hz dB down at 11025Hz relative to level at 1000Hz  
   
  Left -9.08159 -9.08713 -0.00554  
  Right -9.07955 -9.08327 -0.00372  
  Stereo -9.07955 -9.0852 -0.00565  
           
           
           
  TOPPING RMS level at 1000 Hz RMS level at 11025Hz dB down at 11025Hz relative to level at 1000Hz  
   
  Left -9.23363 -9.28215 -0.04852  
  Right -9.23142 -9.28033 -0.04891  
  Stereo -9.23252 -9.28124 -0.04872  
           

 

Only a 20th of a dB attenuation at 11.025kHz with the Topping DAC [incorporating the effect if any of the RME ADC].  Even less apparent attenuation with the RME DAC.

Link to post
Share on other sites
1 hour ago, Stereophilus said:

Interesting that this 20th of a dB is nonetheless audible, albeit in a minor way.

 

You cannot conclude this is the slight difference we are hearing.  In fact, I'll bet it isn't

Link to post
Share on other sites
40 minutes ago, aussievintage said:

 

You cannot conclude this is the slight difference we are hearing.  In fact, I'll bet it isn't

Just play the 11kHz test tones (not the music) back to back a few times. There is an audible difference.

Link to post
Share on other sites
1 hour ago, Stereophilus said:

Just play the 11kHz test tones (not the music) back to back a few times. There is an audible difference.

Kukynas's files as supplied by him have not been normalised to the same reference level at 1kHz. 

 

Below are measurements I've done using the RMS analysis tool available in Audacity.  (To load this tool, use the add plug-in option under the Analysis tab.)  The RMS measurements can vary in the last two decimal places depending on exactly how much of the waveform you select.

 

The Topping file is 0.153dB quieter at 1kHz, and 0.196dB quieter at 11.025kHz than the RME file.  0.196dB can be rounded to 0.2dB or a fifth of a decibel.

 

(For my personal listening tests I normalised kukynas's files in their Left and Right channels to match the 1kHz level in file 1. This can be achieved by using the split stereo file option in Audacity at the start of an Audacity  track, selecting the track content for the relevant channel, and then using the effect "amplify", set for the appropriate amount of amplification.  An easier and automatic method would be to use foobar 2000 with ABX plug-in and allow it to normalise the files, though foobar might not give exactly the right gain correction for these particular files. ) 

 

           
  RME RMS level at 1000 Hz RMS level at 11025Hz RMS level discrepancy at 11025Hz relative to 1000Hz, in dB  
   
  Left -9.08159 -9.08713 -0.00554  
  Right -9.07955 -9.08327 -0.00372  
  Stereo -9.07955 -9.0852 -0.00565  
           
           
           
  TOPPING RMS level at 1000 Hz RMS level at 11025Hz RMS level discrepancy at 11025Hz relative to 1000Hz, in dB  
   
  Left -9.23363 -9.28215 -0.04852  
  Right -9.23142 -9.28033 -0.04891  
  Stereo -9.23252 -9.28124 -0.04872  
           
           
  FILE 1 RMS level at 1000 Hz RMS level at 11025Hz RMS level discrepancy at 11025Hz relative to 1000Hz, in dB  
   
  Left -9.03091 -9.03106 -0.00015  
  Right -9.03091 -9.03106 -0.00015  
  Stereo -9.03091 -9.03106 -0.00015  
           
           
    MISMATCH RMS stereo level at 1k Hz RMS level discrepancy at 1000Hz compared with file 1, in dB  
     
    FILE 1 -9.03091 0  
    RME -9.07955 -0.04864  
    TOPPING -9.23252 -0.20161  
           
  • Like 1
Link to post
Share on other sites
11 hours ago, Stereophilus said:

Just play the 11kHz test tones (not the music) back to back a few times. There is an audible difference.

 

Then something else is changing.  You cannot hear 1/20th of a db

Link to post
Share on other sites
  • Recently Browsing   0 members

    No registered users viewing this page.




×
×
  • Create New...