So far, all indications I have seen is that if SE is cleaned up enough of EMI/RFI and jitter problems, then it is closer to what sounds right to humans. That's why I said it goes against theory. I don't know exactly why, but I think the simple explanation some people offer: "some people like distortion" is not the right answer for all people who prefer SE. Humans are more complicated than such simple models.
EDIT: There is a back end story to this that I am not at liberty to talk about. I will just say to try it, give it some time, and see if you can start to hear what's good about it. Maybe it would help some if @nautibuoy can explain why he chose to listen that way. In any case, its very important to clean up the ground, power, jitter, and shielding problems. Maybe more than that, too. Its not all that easy to do well, but its worth the effort when played back on a good system.
Last edited:
That's very simple, for the breadboard/initial builds, short of dismantling one of my other DACs, I only had a means of using the SE output option available.
Just to be clear, I have not stated a preference for SE listening, only that when I did use the SE output option on the breadboard version I didn't experience the SQ issues you reported and was actually very impressed with the sound quality compared to my Valve DAC. I also noted that I wasn't using an Amanero on the input. My listening was done using a Noir Headphone amp and HiFiMan Sundara or BeyerDynamic DT-880 headphones, so a set-up well able to reveal issues and resolve subtleties.
With both my DSC2 and Valve DACs I have Lundahl transformers on the outputs to utilise the balanced outputs and they feed XLR connectors on the rear panel. To use with downstream SE equipment I use balanced-SE cables.
For my part, I don't have the knowledge or experience to perform meaningful technical measurements, or to hypothesise about the way our brains work, and only make use of one piece of, well used but reliable, test equipment.
We have to consider the understanding and experience of the majority in cases like this.
Technically speaking DAC puts out balanced outputs for a reason. There is HF noise also riding on signals so if only one leg (SE) is used then the noise flows through to the next stage. If both (BAL) are used then the differential action cancels out this HF noise that appears as common mode to input so a better quality output signal.
Technically speaking DAC puts out balanced outputs for a reason. There is HF noise also riding on signals so if only one leg (SE) is used then the noise flows through to the next stage. If both (BAL) are used then the differential action cancels out this HF noise that appears as common mode to input so a better quality output signal.
Most likely Noise modulation effects as a result of only using one leg of the DAC outputs in unbalanced mode - explained earlier, in line with theory. Use balanced mode to fix it
On the logic gate DAC I even mounted a 25 turn trimpot to equalize the gain in both path’s for a super exact tuning.
See image below showing both SE channels and their sum magnified by a factor 50 after having tuned the trimpot, resulting in ca 0.03% equal gain for both path's, just to be sure to get optimal suppression of CM components
On the RTZ Firdac I did the same measurement, but didn't have to go as far as installing trimpots because the results looked good enough for going on with further testing.
Hans
Attachments
Sorry Mark,
For a person usually being very careful for what you say, what you mention here holds no ground at all.
It's a terrible generalization throwing all SE and Balanced outputs on a big heap.
Because differential balanced signals suppress even harmonics, in some cases SE might sound more agreeable.
But there are many balanced outputs that are not differential at all, in those cases all even harmonics and CM components are still there for the full 100%.
And there are SE outputs made from both differential signals that have just as little even harmonics and CM components as the balanced output, see image below.
This is all just on the sending side, similar things can be said on the receiving side even with or without a transformer in between.
In this particular case, no one has to worry about even harmonics, because they are simply absent as my measurements have shown, but because of the aggressive digital environment, CM disruptions are manyfold and can only be suppressed in differential mode.
But even then, it requires a receiving side with a proper CMRR.
Hans
Attachments
Last edited:
@Markw4 ,
One important thing I forgot to mention in the above posting.
In those cases where you prefer SE over Balanced, it tells nothing about which side, sending or receiving, is causing the perceived difference, not even to mention the all important interlink.
When comparing things, you should change one thing at the time, but when changing from Balanced to SE where also having to change the interlink, you have changed three things, making it impossible to declare which of the three is responsible for the change in perceived sound.
I would be most interested to hear how you coped with this all to come to the conclusion that "SE sounds right to Human".
Hans
One important thing I forgot to mention in the above posting.
In those cases where you prefer SE over Balanced, it tells nothing about which side, sending or receiving, is causing the perceived difference, not even to mention the all important interlink.
When comparing things, you should change one thing at the time, but when changing from Balanced to SE where also having to change the interlink, you have changed three things, making it impossible to declare which of the three is responsible for the change in perceived sound.
I would be most interested to hear how you coped with this all to come to the conclusion that "SE sounds right to Human".
Hans
I have used HP-2 to listen in both modes.
I also suggested to listen, give it some time, and to see if you could hear why some people prefer SE. Instead of doing what I suggested what I am getting reads more like rationalizing about why listening is not worth the bother. I get the same type of reaction from other people when I suggest to try better clocking.
Curiously, Acko has gone to great lengths with clocks, yet the conventional wisdom and published research indicates humans can't hear absurdly low amounts of jitter. Strange he will try one 'impossible' thing but not another.
Also curious that Hans has a fancy Bel Canto dac which published research shows can't possibly sound any 'better' than a good Topping. Look at the measurements Topping can get. Their dacs are already better than the 'limit of hearing.' The experts at ASR can tell you all this with great certainty.
I also suggested to listen, give it some time, and to see if you could hear why some people prefer SE. Instead of doing what I suggested what I am getting reads more like rationalizing about why listening is not worth the bother. I get the same type of reaction from other people when I suggest to try better clocking.
Curiously, Acko has gone to great lengths with clocks, yet the conventional wisdom and published research indicates humans can't hear absurdly low amounts of jitter. Strange he will try one 'impossible' thing but not another.
Also curious that Hans has a fancy Bel Canto dac which published research shows can't possibly sound any 'better' than a good Topping. Look at the measurements Topping can get. Their dacs are already better than the 'limit of hearing.' The experts at ASR can tell you all this with great certainty.
Can you believe a normally rational person would say something like this:
"And believe me, blind listening tests are not needed to hear the difference, it is so obvious, at least in my case.
Although no satisfactory physical explanations exists and although many people are prejudiced that cables cannot make a difference, wouldn’t it be a challenge to give it a try and leave your comfortzone for once."
https://www.diyaudio.com/community/threads/zip-cord-for-speaker-test.371099/post-6686399
"And believe me, blind listening tests are not needed to hear the difference, it is so obvious, at least in my case.
Although no satisfactory physical explanations exists and although many people are prejudiced that cables cannot make a difference, wouldn’t it be a challenge to give it a try and leave your comfortzone for once."
https://www.diyaudio.com/community/threads/zip-cord-for-speaker-test.371099/post-6686399
You both are right (Hans and Mark): the cabling is usually different when comparing balanced with SE, and the circuit Hans gave splendid reason as to why, in many cases, the balanced doesn`t sound better, as the signal is generated from SE to begin with, needs extra parts to convert to balanced, uses different cabling and interconnects, then at the receiving side goes through the same type of EXTRA circuit to make it SE again. SE just omits a lot of electronics, unless someone wants to advocate more resistors and ic`s are a good thing..
Just look at the number of power amps that are not built bridged, and the 99,99% (100% ?) pre amps that do not have a 4 way potentiometer, which would be the only way to secure full symmetrical treatment of the signal from in- to output.
In short I`d say balanced circuitry involve more electronics, which can never be better for the signal than not having it, unless long wiring is needed, as well as admitting many things the eye sees (a measurement) isn`t always what the ears hear. Why that is or how significant, is a different story.
Build 2 identical dacs, one modified without significant measurable differences and I`m sure, on a good set, all will hear they`re different.
Just look at the number of power amps that are not built bridged, and the 99,99% (100% ?) pre amps that do not have a 4 way potentiometer, which would be the only way to secure full symmetrical treatment of the signal from in- to output.
In short I`d say balanced circuitry involve more electronics, which can never be better for the signal than not having it, unless long wiring is needed, as well as admitting many things the eye sees (a measurement) isn`t always what the ears hear. Why that is or how significant, is a different story.
Build 2 identical dacs, one modified without significant measurable differences and I`m sure, on a good set, all will hear they`re different.
Ok, I have added this SE vs BAL comparison to my research list
Other items:
DACs, discrete vs chips - this thread
Clocks, cheap vs OCXOs
PSUs, linear vs UltraCaps/batteries
FIFO vs sync reclocking
Ha, ha, this referred to a very specific case, comparing a 2.5meter LS Zip cable to the same cable equipped with LRC networks.
And even then I mentioned no blind testing needed "at least in my case".
Compare it to your separate 5Volt on the Amanero, that's very specific and it's your opinion.
How different is your fully unspecific remark about SE and Humans, completely incomparable to the above.
Our comment should serve to keep each other sharp and not to bully.
When I'm telling stupid things, I'm happy to be corrected.
Hans
Another thing possibly worth mentioning: When power supplies are on separate boards, it creates problems with IR losses, common mode noise, ground crosstalk between + and - rail ground returns, etc. In the end the modular approach is more complicated and more expensive to get right on a per unit manufacturing cost basis. Everything integrated into one board that has gone through as many design revisions as necessary to get it right will be less cost per unit for a given level of performance. However, it only makes sense to do if you can sell enough of them.
What I am going to do next with Marcel's dac makes little or no commercial sense other than to demonstrate what the dac can sound like under reasonably favorable conditions.
What I am going to do next with Marcel's dac makes little or no commercial sense other than to demonstrate what the dac can sound like under reasonably favorable conditions.
Actually, I got it from a guy who designs low-volume commercial dacs using Amanero boards. He talks about it directly with Dom, the guy who designed Amanero.
Moreover, I2SoverUSB boards are already designed for use with clean power on the clocks. Lot's of people have heard the difference on that board. Same problem though.
There are other things you don't know about either. Most of them I can't tell you.
Mark, why are you resorting to playground behaviour here? The people who used that on you have all been banned now. You should be above this, especially posting without context.