Topping D90SE Review (Balanced DAC)

[Deleted member 31709]

I`m pretty convinced that most of commenters won`t buy D90e because of the price. D90 vs D90e for someone who is not dedicated geek and already owns D90 , make no sense. For premium users it is still to cheap, but those do not care much about measurement. One thing Zeos misjudges in my opinion is statement that DAC is just DAC.

Have heard similar statements before. I feel in some way general perception of the Conversion in DAC sways people in wrong direction. People perceive the conversion process is infallible and perfect. When really a DAC generates an analog output.
Even if the analog stages of two DACs were absolutely identical in every aspect, how they implement (or not implement in case of NOS) the reconstruction filter will make a world of difference.
Perhaps next time you should link them to a good read on Shannon's sampling theorem and why it is impossible to achieve a perfect brick-wall analog LPF

 

[JohnYang1997]

Have heard similar statements before. I feel in some way general perception of the Conversion in DAC sways people in wrong direction. People perceive the conversion process is infallible and perfect. When really a DAC generates an analog output.
Even if the analog stages of two DACs were absolutely identical in every aspect, how they implement (or not implement in case of NOS) the reconstruction filter will make a world of difference.
Perhaps next time you should link them to a good read on Shannon's sampling theorem and why it is impossible to achieve a perfect brick-wall analog LPF

I like the way you are so confident yet so wrong.

 

[HDavidson]

Seeing your confidence , mine grows with each passing day.
If you have anything to add worthwhile to a discussion - do so in a constructive way. Stop the hypocrisy.

John designs technically good machines for a reasonable price, Amir tests them and gives the chifres. And you?

 

[abdo123]

If someone finds something objectionable in a statement, discuss that in a constructive manner. That is what Science is about.

Do you know what Science is also about? Evidence and reproducible results

So can you please go and substantiate your previous statements please.

In Particular this one:

Even if the analog stages of two DACs were absolutely identical in every aspect, how they implement (or not implement in case of NOS) the reconstruction filter will make a world of difference.

Or perhaps why this is necessary

it is impossible to achieve a perfect brick-wall analog LPF

 

[THW]

maybe I’m just dumb but I seem to recall that filters for nearly any decent DAC are good enough, maybe not “ideal brick-wall at 22kHz” per se but if being slightly off the ideal does not affect the audible range, why does it matter?

 

[JohnYang1997]

Are you asking me to provide evidence for a theorem?

Let me ask you, do you understand Shannon's sampling theorem and the premises for perfect reconstruction of original analog waveform?
I would humbly ask you to revisit https://www.sciencedirect.com/topics/computer-science/sampling-theorem
particularly the sections concerning the 27.4

"Unfortunately, the first condition is virtually unrealizable, since it is nearly impossible to devise a low-pass filter with a perfect cutoff. Recall from Chapter 3 that a low-pass filter with a perfect cutoff will have infinite extent in the time domain, so any attempt at achieving the same effect by time domain operations must be doomed to failure. However, acceptable approximations can be achieved by allowing a “guard-band” between the desired and actual cutoff frequencies. This means that the sampling rate must therefore be higher than the Nyquist rate. (In telecommunications, satisfactory operation can generally be achieved at sampling rates around 20% above the Nyquist rate—see Brown and Glazier, 1974.)"

Now what was so incorrect in my prior statement? Are you saying the steeper roll off can't be measured in an analyzer - well we definitely can.
So what's the fuss about?

I didn't want to bother but.

Back to the current noise of I/V circuit, the output impedance of 9038pro is low enough that there will be extra current noise for op1612 except the thermal noise of the resistance itself. This is already explained in this thread before. At this level of low output impedance it's the voltage noise that matters not current noise. The TI document was not wrong because the context was with output impedance in the KiloOhm range.

The the analogue low pass has nothing to do with reconstruction filter.
The reconstruction filter or basically the low pass filter after upsampling does not need to be perfect to reproduce perfect signal from 20hz to 20khz. The 20-22.05khz requirement can be relaxed.
Measurements showed there's no rise in distortion in high frequency with 3 of the filters, so signal is "perfectly" reproduced.

So basically you just read some books or application notes without actually understanding the underlying mechanics also are too confident that you correct someone without doing experiments.

 

[Billy Budapest]

components such as OP amps and power supply upgrades change the sound. like tubes but less noticeable. thousands of people agree, who cares if one or 2 parameters measure the same? or are "inaudible" if its inaudible than why do me and everyone who upgrades equipment from begining of time hear it? sometimes you can do what makes you happy in audio and not care if somone says i need to do a scientific study in order to have the opinion or share the opinion that some things sound different than others with similar measurements. who cares. its an opinion . i think THX amps sound a bit flat. compared to burson or schiit. some people love THX i would never tell either person its in their head even if the measurements are "inaudible" i believe they hear it.

See my earlier post. It is your brain that changes, and the equipment is not responsible for an “increase in detail,” “a veil was lifted,” “I heard things I’ve never heard before” reactions. The differences you hear are real, even if they are just in your mind. Because you are paying attention to the music and focusing on whether or not there are audible differences between the gear, you are actually hearing differences rather than your brain just glossing over the music.

Amir and others on this forum have pointed out this effect numerous times. It’s real, and makes perfect sense. This blog page was cited earlier but it identifies the various psychoacoustic mechanisms at play:

https://nwavguy.blogspot.com/2012/04/what-we-hear.html

Also, watch and listen here—“what we hear might not always be the truth”:

 

[abdo123]

So what's the fuss about?

I'm not arguing about whether these facts are true or not, I asked about their relevance.

you claimed a world of difference between two functional DAC filters, you got any evidence for that?

you mentioned that it is impossible to have a perfect brick-wall filter, why is this is actually important or relevant to you? have you experienced any issue in your playback that you think a perfect brick-wall filter will fix?

 

[bboris77]

On the subject of blind testing that has been mentioned in this thread repeatedly - I would like to propose an alternative test: I would like to know if @amirm, as a trained listener, can successfully pick out the D90SE against any DAC with its SINAD above 96dB.

If he can, the objectivist approach has an obvious problem. If he cannot, which is likely, then it will put things in perspective for a lot of people. It will indicate that this pursuit of ever more transparent measuring devices above a certain SINAD has had no effect at all on the listening experience. At least for human beings.
I’m not diminishing the engineering achievements here - they are real and measurable. However, the impact of these improvements on the end user are likely not objectively observable. The benefit to the end user would ironically fall in the psychological category - the knowledge that they possess the most transparent and best engineered DAC. This could likely improve their subjective listening experience due to the whole placebo psychological effect.

 

[abdo123]

If he can, the objectivist approach has an obvious problem.

To the contrary, I feel like if he could tell them apart (and prefer the DAC with the higher SINAD) then the way DACs are being ranked at this forum (based on SINAD) is perfectly good.

 

[Labjr]

If I was buying a flagship product I'd certainly want it to be compatible with every existing format. I wouldn't even consider the inexpensive Atom DAC unless they make it DSD compatible.

I think there are many who don't want MQA included in any product because they don't want MQA entrenched any further for fear it will become dominant and regular hi-res formats will cease to exist. I believe that was the goal of MQA. However, I don't think that's likely to happen at this point.

 

[abdo123]

You got any evidence for an amp with 0.01 % THD+N vs 0.0001 THD+N will make any difference? Do you guys selectively decide what's audible or not to human ears to your convenience?

there have been several studies on the audibility of Harmonic distortion actually.

 

[JohnYang1997]

You are wrong on so many fronts -
the whole point of reconstruction filter being digital is because we can NOT design steep filters in analog. Hence the need for up-sampling followed by digital LPF.
Perfect reproduction also implies reconstructing transient information between two samples. If you think hundred coefficients of ESS/ AKM are good enough to remotely achieve that and anything beyond that is placebo, let's not waste each other's time in that case.

Revisit your fundamentals boy!

You are the one that is wrong in every single way...

 

[Deleted member 31709]

there have been several studies on the audibility of Harmonic distortion actually.

This graph does not answer my question - do you have a case study that shows relevance of 3 or more leading 0s in a source equipment on the listener?

 

[bboris77]

To the contrary, I feel like if he could tell them apart (and prefer the DAC with the higher SINAD) then the way DACs are being ranked at this forum (based on SINAD) is perfectly good.

I was not being facetious. It would be a genuinely interesting blind test. Subjectivists have been challenged many times to perform these with no success. At least Amir has the knowledge and equipment to perform this type of test.

 

[abdo123]

This graph does not answer my question - do you have a case study that shows relevance of 3 or more leading 0s in a source equipment on the listener?

I did not make such claims. I'm actually a strong advocate on this forum that these things don't matter past a certain point.

But please, this is not about me, can you please go back to 'the world of difference' reconstruction filters make?

 

[gpauk]

Amir - can I add to the few requests for a bluetooth test, please?

 

[THW]

So basically you just read some books or application notes without actually understanding the underlying mechanics also are too confident that you correct someone without doing experiments.

this reminds me of this:
https://en.m.wikipedia.org/wiki/Dunning–Kruger_effect

 

[Mulder]

Please. Can you perhaps release this quarrel now so that we can return to discuss Topping's new DAC instead.

 

[AdamG]

General Warning! This is a product review thread. Stay on topic. Personal insults removed and poster given a thread reply ban. More will be handed out if some can not behave and stay on topic. Please and thank you.

Oh yeah forgot to mention. Don’t feed the Trolls. They end up living in your closet and wearing your favorite T-shirts.