E1DA DACs Discussion Thread (PDV2, 9038S, 9038D, 9038X, HiFiToy, HPToy)
Aug 6, 2020 at 10:44 PM Post #886 of 1,483
Yes, I also had that unchecked as when using the G3 the M6 is a source rather than a DAC.

Oh no, you're referring to the "USB DAC" setting in Settings->Audio. What I'm talking about is the "USB output mode" in the settings sidebar of of the Music App (not device settings). Only when I set this to "OFF" could I get the G3 to work with the M6.
 
Aug 7, 2020 at 6:06 AM Post #887 of 1,483
Oh no, you're referring to the "USB DAC" setting in Settings->Audio. What I'm talking about is the "USB output mode" in the settings sidebar of of the Music App (not device settings). Only when I set this to "OFF" could I get the G3 to work with the M6.
OK, sorry for the misunderstanding. At any rate, I'm glad it's working for you now.
 
Aug 8, 2020 at 4:00 PM Post #888 of 1,483
I just posted my review of the E1DA 9038S Gen 3 to my blog: https://medium.com/@darmanskirata1136/e1da-9038s-gen-3-review-b94e0216f268

DSC07277.jpg


My readers are likely familiar with my long-stated belief that if a playback device is coloring the sound, something is wrong with the source device. My initial listening experience using the 9038S challenged that conviction. I could not shake the sense of improved clarity and detail retrieval offered by the 9038S over the sources I originally used to review the Smabat ST-10S. I decided to evaluate these impressions under more controlled circumstances.

I connected one earbud to The Element and the other to the 9038S and volume matched the two sources using my Dayton iMM-6 microphone to 1/10th of a dB. To do this, I aligned the center of the earbud face with the center of the tube coupler bore and secured the earbud to the coupler using micropore tape around the earbud circumference. I then played a Redbook copy of “Archetype” by Fear Factory and switched back and forth between the two sources.

To my surprise and in the face of my long-held conviction that all well-designed sources should sound identical, the earbud playing from the 9038S sounded tighter, more detailed, and more spacious, with better definition of and separation between notes, while the earbud playing from The Element sounded slightly smoother and warmer in tonality.

I repeated this test using the same track with the Moondrop S8. This time I took care to ensure that both The Element and the 9038S were set to 24 bit /44.1kHz in shared mode in addition to volume matching, which I had not done in my initial comparison using the ST-10S. My differing perceptions of the two sources remained audible.

I will note that this is a sighted comparison and not a double-blind test. However, I will also note my predisposition towards dismissing claims of audible differences between sources coming into this comparison, an attitude I have held since early in my audio journey. Additionally, I am mostly confident that I would not have been able to distinguish between the two sources examined here without the benefit of instantaneous switching.

There are two alternative explanations for the conclusion I have reached. One is that channel imbalance can explain my perception. To assess this possibility, I measured both channels with both sources. Based on my measurements, I am skeptical of this alternative explanation given the minute differences in matching between the left and right channels of the S8. There are minute variations in my measurements between the two sources in the upper-treble region but I am inclined to believe these are a result of tiny differences in insertion depth and bore angle during the measurement process as opposed to frequency response differences inherent to the hardware. The other alternative explanation is that I hear better in my left ear than my right ear. While my hearing range is indeed around 200 Hz higher in my left ear than in my right ear, I can perceive sounds above 19 kHz in both ears. Again, I am skeptical that this explains the audible differences between the two sources.

There is a lot of snake oil in the audio enthusiast marketplace, and it should be incumbent upon individuals who claim that different pieces of hardware sound audibly distinct to make sincere attempts to disprove alternative explanations to their claims, as I have done here. This is especially true for reviewers and others in positions of influence. I certainly benefited from my training in the physical sciences as I set up the experiments I detailed in this review, but these experiments are far from exotic or complex and could be easily replicated by others.
 
Last edited:
Aug 9, 2020 at 3:19 AM Post #889 of 1,483
Добрый день, я хочу купить 9038s smabat m2pro. Я в основном слушаю дэт-метал, это правильный выбор? или это лучше st10 s?
 
Aug 9, 2020 at 9:30 AM Post #891 of 1,483
Не могли бы вы написать на английском языке? Спасибо

Nice experiment. A good way to learn. But given the fact that you can hear 19kHz I conclude that you are still young and still firmly set in 'school' state of mind where independent thinking is not promoted (tests are scored on how closely you follow the curriculum). This is all very good for becoming a productive member of society (ie the slave force). But there is so much more to the world than what is taught in school. The way you tested is akin to all those 'scientific' reviews and 'influencers', specs and marketing talk. Measurements are fine if you measure the things that matter and know how to interpret the data. Inherently the data is hardly ever wrong (besides measuring errors or margins greater than 10%).

So here is my question to try to push your mind in a different direction: why do you think there is an audible difference when the frequency graph is all but identical?

And my answer: because we have been taught to focus on amplitude (there is an historic cause for that) but that hardly matters anymore nowadays (except for speakers, microphones etc). Digital sources and amplifiers are ruler flat. And that is perfect. Harmonic distortion etc is virtually absent. So... what's the reason you still hear a difference? It must be other than that.

If you look at a frequency graph you have 2 axis: the x and the y-axis; frequency and amplitude. But there is a 3rd axis that matters. And for that you don't need perfect hearing (only training to perceive and put it into words). That is timing.

Now imagine a waterfall graph in 3D. How high is the graph just before and after you measure? The total amount of energy is the volume under the graph. So if rise and decay are slow you have more energy, but it measures less (it is depending on the aperture of your time-window). This is why more and more people move to R2R dacs because its temporal behaviour is different (and more natural: no pre-ringing).

The 9038s sounds so clear is because there is less electronics between dac- and output-stage. So there is less smearing in the time domain. This is what you hear as better recreation of venue and size and timing of instruments. Lots of steps in the signal path can smear out the signal (lose information) but never closer to the original (can't create information). That is why I agree with Ivan on his approach to keep the signal path as short as possible.
Minimalism and K. I. S. S. (keep it simple, stupid).
 
Aug 9, 2020 at 9:52 AM Post #892 of 1,483
Не могли бы вы написать на английском языке? Спасибо

Nice experiment. A good way to learn. But given the fact that you can hear 19kHz I conclude that you are still young and still firmly set in 'school' state of mind where independent thinking is not promoted (tests are scored on how closely you follow the curriculum). This is all very good for becoming a productive member of society (ie the slave force). But there is so much more to the world than what is taught in school. The way you tested is akin to all those 'scientific' reviews and 'influencers', specs and marketing talk. Measurements are fine if you measure the things that matter and know how to interpret the data. Inherently the data is hardly ever wrong (besides measuring errors or margins greater than 10%).

So here is my question to try to push your mind in a different direction: why do you think there is an audible difference when the frequency graph is all but identical?

And my answer: because we have been taught to focus on amplitude (there is an historic cause for that) but that hardly matters anymore nowadays (except for speakers, microphones etc). Digital sources and amplifiers are ruler flat. And that is perfect. Harmonic distortion etc is virtually absent. So... what's the reason you still hear a difference? It must be other than that.

If you look at a frequency graph you have 2 axis: the x and the y-axis; frequency and amplitude. But there is a 3rd axis that matters. And for that you don't need perfect hearing (only training to perceive and put it into words). That is timing.

Now imagine a waterfall graph in 3D. How high is the graph just before and after you measure? The total amount of energy is the volume under the graph. So if rise and decay are slow you have more energy, but it measures less (it is depending on the aperture of your time-window). This is why more and more people move to R2R dacs because its temporal behaviour is different (and more natural: no pre-ringing).

The 9038s sounds so clear is because there is less electronics between dac- and output-stage. So there is less smearing in the time domain. This is what you hear as better recreation of venue and size and timing of instruments. Lots of steps in the signal path can smear out the signal (lose information) but never closer to the original (can't create information). That is why I agree with Ivan on his approach to keep the signal path as short as possible.
Minimalism and K. I. S. S. (keep it simple, stupid).
The explanation may be that The Element uses a minimum phase digital filter, while the 9038S uses a linear phase filter by default. I flashed the minimum phase slow roll-off filter firmware to the E1DA and recompared the two, and I no longer heard a clear difference.
 
Aug 10, 2020 at 2:35 AM Post #893 of 1,483
Hey guys. Which of the models you would recommend for E-mu Teak, theyre 25ohm. I would only use them at desktop.
Innitially i was looking for desktop combos like SMSL M500 or Sabaj D5.
How much do i lose out in SQ if i went with either 9038s or PVD2?
 
Aug 14, 2020 at 4:51 PM Post #894 of 1,483
Не могли бы вы написать на английском языке? Спасибо

Nice experiment. A good way to learn. But given the fact that you can hear 19kHz I conclude that you are still young and still firmly set in 'school' state of mind where independent thinking is not promoted ....
19kHz - Are dogs allowed as members in the forum? :gs1000smile:
 
Aug 14, 2020 at 7:56 PM Post #895 of 1,483
The explanation may be that The Element uses a minimum phase digital filter, while the 9038S uses a linear phase filter by default. I flashed the minimum phase slow roll-off filter firmware to the E1DA and recompared the two, and I no longer heard a clear difference.
so the difference is phase... and phase is? Timing. But that's just 1 aspect of what goes on in the time domain.
 
Aug 16, 2020 at 2:05 PM Post #896 of 1,483
Hi everyone

as of now which of all the DACs mentioned here is the most powerful? Looking for something to replace my great (but heavy) Fiio M15 on the go.

Thanks
 

Users who are viewing this thread

Back
Top