>A more technical discussion of dac design/configuration is going in the 'ES9038Q2M Board' thread, in case you would like to read/participate in that.
Thanks a lot!
>It's not. Use voltmeter at the speaker terminal.
On music signal?
The measurements stand is located separately from the aural comparison.
>What about salesmen posting on internet forums? Have you spotted any on this thread?
I just have a fellow involved into the HiEnd trading business, I’m afraid a salesman has no reasons to spend time discussing technical aspects. The more vital to the fellow (for instance) questions are rather “how much” and “what’s the discount/tax/sales volume/etc”.
If you could look to the subject through a businessman eyes, you will be able to understand such point of view easily.
Once I discussed the subject with him, his answer was laconic – “I need to sell DACs, revealing good or bad parameters is senseless to me. Even I find a “worse” DAC I would need to sell it anyway” ;-)
Thanks a lot!
>It's not. Use voltmeter at the speaker terminal.
On music signal?
The measurements stand is located separately from the aural comparison.
>What about salesmen posting on internet forums? Have you spotted any on this thread?
I just have a fellow involved into the HiEnd trading business, I’m afraid a salesman has no reasons to spend time discussing technical aspects. The more vital to the fellow (for instance) questions are rather “how much” and “what’s the discount/tax/sales volume/etc”.
If you could look to the subject through a businessman eyes, you will be able to understand such point of view easily.
Once I discussed the subject with him, his answer was laconic – “I need to sell DACs, revealing good or bad parameters is senseless to me. Even I find a “worse” DAC I would need to sell it anyway” ;-)
>Have you checked the ringing (time domain distortion) due to FIR filter?
Yes, nothing special was found.
Yes, nothing special was found.
You need to play a tone when adjusting. Haven't you researched "level matched double blind test"?
Looks like you haven't spotted any audio salesman on this thread. Try a forum search on "FUD". You will see who spreads it.
>You need to play a tone when adjusting.
Maybe, I would just need to gather all the equipment together again and gather all the listeners. ;-)
Then, create/write a tone/wav to a CD. Just the latter is easy enough.
> you haven't spotted any audio salesman on this thread.
On paid basis? ;-)
What you would like to hear from them?
Are they aware of what THD is? ;-)
> You will see who spreads it.
Sorry, spreads what?
Maybe, I would just need to gather all the equipment together again and gather all the listeners. ;-)
Then, create/write a tone/wav to a CD. Just the latter is easy enough.
> you haven't spotted any audio salesman on this thread.
On paid basis? ;-)
What you would like to hear from them?
Are they aware of what THD is? ;-)
> You will see who spreads it.
Sorry, spreads what?
From Ben Duncan, High Performance Audio Power Amplifiers, Section 8.2, Page 316.
"It is well known that small differences in level can confuse inexperienced listeners, who hear a 'brighter' sound when one amplifier has a tad more gain or drive level. Levels much be very accurately matched. At high SPL's and low frequencies, differences of as little as 0.1 dB can be readily audible. But matching is conventionally carried out at 1 to 3khz, at least for a full range amplifier. In an active system, the geometric mean center of the frequency band would be optimal on this count, while the equal loudness contours demand that matching is performed at the more sensitive region of the band. For matching, precision equipment (AC millivoltmeter and high resolution, dB reading DMM or other analyser) able to resolve below 0.1dB and with commensurately high accuracy, must be used."
>small differences in level can confuse inexperienced listeners,
IME 0.1db is not aurally recognizable at all in “without looking” test.
Would be curious if you try as well.
For me the changes started from 0.5db minimum and perceived as pure volume change only (as it should be)
How to: a PC with music player, Equalizer APO, speakers/headphones
In AP0 add +0.1 db (or -0,1db) pre-amplifcition filter and ask somebody to switch it on/off without looking.
For me that “can” doesn’t mean “does”
IME 0.1db is not aurally recognizable at all in “without looking” test.
Would be curious if you try as well.
For me the changes started from 0.5db minimum and perceived as pure volume change only (as it should be)
How to: a PC with music player, Equalizer APO, speakers/headphones
In AP0 add +0.1 db (or -0,1db) pre-amplifcition filter and ask somebody to switch it on/off without looking.
For me that “can” doesn’t mean “does”
The reason for setting two devices to the same output level with .1dB is because it is low enough to make sure one won't sound louder than the other. Actually, at very low volume level differences it isn't exactly that one DUT sounds louder than the other. Perceptually it can seem more like one sounds slightly 'better' than the other. A little more punch, dynamics, tonal fullness, whatever. Just a little bit better.
Regarding experienced listeners verses inexperienced, it isn't the number of hours spent listening if there is no useful learning during that time. It is about learning what to listen for and how not to get fooled.
Regarding experienced listeners verses inexperienced, it isn't the number of hours spent listening if there is no useful learning during that time. It is about learning what to listen for and how not to get fooled.
Were the FFT data collected with the DACs connected to the actual loads used in listening?
Some loads inject current distortion into their sources, which is less detrimental if the source has a low output impedance. If you measure the DACs without the load they are supposed to drive you can miss such issues (same goes for current noise and ground-loop artifacts).
Some loads inject current distortion into their sources, which is less detrimental if the source has a low output impedance. If you measure the DACs without the load they are supposed to drive you can miss such issues (same goes for current noise and ground-loop artifacts).
Last edited:
Have you looked into the stereo separation numbers on the poor sounding dac? Kinda obvious, but in practice I've noticed that I need to adjust the interspacing between speakers when switching dacs (or amps) to create a similiar sound presentation.
>Were the FFT data collected with the DACs connected to the actual loads used in listening?
The measurements and audio comparison were made separately.
I understand the logic though.
Measurement equipment has a high input impedance,
Audio comparison was made with 47kOhm input amplifier.
All the DACs pairs were compared aurally with the same amplifier
and only one has unexplainable bad sound.
The DAC has (obviously not the best but) decent FFT.
However its sound was undoubtedly distorted aurally.
It sounds similar to very old PC embedded sound card.....
Tried with the amplifier and several headphones connected to headphone output of the DAC directly. The result was always the same
The question is what I have missed in the measurements to reveal such distortions reason
For now there are 2 reasonable hypotheses to verify
- RF signals on the output that intermodulate real sound but not detected with FFT
- Noise floor modulation in sigma delta DACs, Still have no idea for how to measure it, or see it anyhow
EDIT
- Recheck ringing (time domain distortion)
The measurements and audio comparison were made separately.
I understand the logic though.
Measurement equipment has a high input impedance,
Audio comparison was made with 47kOhm input amplifier.
All the DACs pairs were compared aurally with the same amplifier
and only one has unexplainable bad sound.
The DAC has (obviously not the best but) decent FFT.
However its sound was undoubtedly distorted aurally.
It sounds similar to very old PC embedded sound card.....
Tried with the amplifier and several headphones connected to headphone output of the DAC directly. The result was always the same
The question is what I have missed in the measurements to reveal such distortions reason
For now there are 2 reasonable hypotheses to verify
- RF signals on the output that intermodulate real sound but not detected with FFT
- Noise floor modulation in sigma delta DACs, Still have no idea for how to measure it, or see it anyhow
EDIT
- Recheck ringing (time domain distortion)
Last edited:
>Have you looked into the stereo separation numbers on the poor sounding dac?
Thank you I will try.
However the unexplanable difference was not tiny, it was very noticeable:
via the amplifier and via a set of headphones connected to the DAC's headphone output directly
Thank you I will try.
However the unexplanable difference was not tiny, it was very noticeable:
via the amplifier and via a set of headphones connected to the DAC's headphone output directly
I recommend to repeat your measurements at -60dB. Also a jitter test tone (J-test signal) could reveal the differences.
Last edited:
Harmonic distortion measurements have been bringing a nill useful result in audio for at least 70 years. Do we need persist? 🙄
oh, I do not claim that...certain members here do
(I replaced opams in the same dac and heard significant differences most of the time)
some claim all midranges sound same too 🙂
DAC blind test: NO audible difference whatsoever
What you claimed is, "according to some all dacs sound the same". Who made such statement? Some would mean more than one person. Can you quote at least two?