• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Topping D90 Balanced USB DAC Review

mocenigo

Major Contributor
Joined
Dec 8, 2018
Messages
1,301
Likes
1,084
DSD is not audio madness, there are real audio benefits to DSD, which apparently you have not investigated. Perhaps you might like to investigate such before making such inaccurate statements, this is OT here (accept for the fact that the D-90 allows for Direct DSD mode, which IME with the D-90 allows for its best possible sound quality).
There are downsides to DSD as well of course (no free lunch applies here), such as the inability to edit it in its native state.

There often appear to be a lot of blanket statements made at these forums which do not apply actual science-I would love to see an actual scientific discussion of DSD recording (A/D) and playback (D/A), without prejudice, on these forums.

BTW, also note that the best measuring D/A converter Amir has tested (Mola Mola Tambaqui) uses a single bit DSD converter in its final stage-there is good reason for this, and for the excellent measurements.

I said as a FORMAT. For storage or distribution, implied. As an intermediate stage for reproduction, it is a different matter – there are advantages to delta-sigma modulations even though the problem of settling time is in pure 1-bit DSD unsolvable (I have been literally torn to pieces by the wolves here – yes, I know that advanced modulation techniques can reduce large settling time to corners cases, I know that multi bit sigma-delta can reduce this to a set of measure almost zero, but mathematically the problem is still there with 1-bit DSD – the fact that they have designed DACs that made it into substantially a non-problem does not change that). In fact, at the moment, delta-sigma gives the best results in distortion and noise. And exceptionally good sound. But we do not really need to use DSD as a medium. Probably it is excellent for recording, but then the usefulness stops until the last step of reproduction – where it may in fact be replaced by "multi-bit sigma-delta" which is just a marketing name for "low resolution, ultra-high sampling rate PCM" – as in the Tambaqui, BTW. If you have to edit, 1 bit DSD is useless, unless you just mean splice and cut.

So it is a device to make good recordings and good reproduction chains. But as a format, it is a waste of disk space. In my opinion, of course.
 

barrows

Active Member
Joined
Jan 24, 2020
Messages
150
Likes
221
I said as a FORMAT. For storage or distribution, implied. As an intermediate stage for reproduction, it is a different matter – there are advantages to delta-sigma modulations even though the problem of settling time is in pure 1-bit DSD unsolvable (I have been literally torn to pieces by the wolves here – yes, I know that advanced modulation techniques can reduce large settling time to corners cases, I know that multi bit sigma-delta can reduce this to a set of measure almost zero, but mathematically the problem is still there with 1-bit DSD – the fact that they have designed DACs that made it into substantially a non-problem does not change that). In fact, at the moment, delta-sigma gives the best results in distortion and noise. And exceptionally good sound. But we do not really need to use DSD as a medium. Probably it is excellent for recording, but then the usefulness stops until the last step of reproduction – where it may in fact be replaced by "multi-bit sigma-delta" which is just a marketing name for "low resolution, ultra-high sampling rate PCM" – as in the Tambaqui, BTW. If you have to edit, 1 bit DSD is useless, unless you just mean splice and cut.

So it is a device to make good recordings and good reproduction chains. But as a format, it is a waste of disk space. In my opinion, of course.

OK, thanks for clarification, it appears that we are (mostly) in agreement. I would add that I would prefer DSD 128 as a format for distribution/archive of classic analog recordings. I also find Channel Classics DSD recordings (mixed through an analog console in real time and capured in high rate DSD) to be almost universally excellent. I am to sure what the fuss is about storage space though, as storage is now not something I consider a "problem" for private libraries.
I agree that sigma delta modulator design is critical to getting he best out of DSD, regardless of the original file format, hence the advantages I perceive with HQPlayer4.
 

mocenigo

Major Contributor
Joined
Dec 8, 2018
Messages
1,301
Likes
1,084
OK, thanks for clarification, it appears that we are (mostly) in agreement. I would add that I would prefer DSD 128 as a format for distribution/archive of classic analog recordings.[...] I am to sure what the fuss is about storage space though, as storage is now not something I consider a "problem" for private libraries.

It depends. I have a library of about 15000 (as in fifteen-thousand) ripped discs (mutual backup of my library, over 7000 records, and of the libraries of two close friends) – the vast majority being redbook, but about 500 are SACDs, 30 Blu-ray Discs. I have the SACD ISOs stored, but I converted then to 24/88.2 for my normal library, since I apply EQ, a bit of room correction... I also keep stuff in PCM because my DAC is the Soekris DAC1541, which is native PCM. I have almost 7 Tb of losslessly stored files (ALAC) and having them in DSD128 would make everything three times as large. This way, I carry everything with me in a single SSD (I have two further copies/backups).

But indeed I realise now you are speaking of analog recordings. There may be advantages in transferring them to DSD128 or higher, but I have only a very few of those, including a DSD256 transfer of music by Howard Hanson that, the music being great nonetheless, sound quality wise is not something to call home about.
 

mocenigo

Major Contributor
Joined
Dec 8, 2018
Messages
1,301
Likes
1,084
DSD is a solution in search of a problem...

Here I disagree. As a mathematical representation of signal to be fed to a very simple DAC, a DAC-less DAC almost, just a flip flop with an output filter, it is pure mathematical genius. Converting to DSD solves then many problems, even though it causes a few of its own (and, as I said above, mostly of academic nature). But as a distribution and "private" storage format, naaa.
 

barrows

Active Member
Joined
Jan 24, 2020
Messages
150
Likes
221
Here I disagree. As a mathematical representation of signal to be fed to a very simple DAC, a DAC-less DAC almost, just a flip flop with an output filter, it is pure mathematical genius. Converting to DSD solves then many problems, even though it causes a few of its own (and, as I said above, mostly of academic nature). But as a distribution and "private" storage format, naaa.

Indeed, this is IMO, the best reason for DSD at the conversion stage, it is fairly simple to convert DSD to analog hardware wise.

BTW, in no sense I am suggesting that one convert their files to DSD to storage! A good sized majority of my library is made of of Redbook rips, as I suspect is true for many people. I just prefer to playback DSD, so I convert in software on the fly to DSD 256 for conversion. I feel that these conversions can be better performed in a computer, than with the limited processing power available in most DACs (hardware). Nowadays I also purchase DSDv versions of things, when that is their native format.

Although there is the Tambaqui, which uses 3 AD SHARC processors, to do its internal conversions, and given what I have heard from that DAC, I suspect their algos for conversion are very, very good.
 
Last edited:

maverickronin

Major Contributor
Forum Donor
Joined
Jul 19, 2018
Messages
2,527
Likes
3,311
Location
Midwest, USA
Here I disagree. As a mathematical representation of signal to be fed to a very simple DAC, a DAC-less DAC almost, just a flip flop with an output filter, it is pure mathematical genius.

So basically you just think that it's cool.

Converting to DSD solves then many problems, even though it causes a few of its own (and, as I said above, mostly of academic nature). But as a distribution and "private" storage format, naaa.

You're going to have to turn it back into PCM at some point for DSP...
 

da Choge

Active Member
Forum Donor
Joined
Dec 31, 2018
Messages
233
Likes
220
Location
DFW
It appears that the MQA version is now available:
Hallelujah !!! And I don't really care about MQA and hardly believe in it (listened both with and without and ???). But I've really wanted this DAC for my bedroom setup, and I just made it my thing to hold out until the D90 came out with MQA (guess I'm a gadjet junkie :rolleyes:). Due to a world situation beyond our control, I thought the release of the D90-MQA would be significantly delayed, but maybe not so. Waiting for it to be offered on Amazon (got too many AMEX points for my own good).

BTW, I DON'T intellectually really care about MQA, but I couldn't bring myself to buy a D90 w/o it, if there was going to be one with it -- pretty sad :( -- Do I need therapy?
 
Last edited:

majingotan

Major Contributor
Forum Donor
Joined
Feb 13, 2018
Messages
1,542
Likes
1,826
Location
Laguna, Philippines
BTW, I DON'T intellectually really care about MQA, but I couldn't bring myself to buy a D90 w/o it, if there was going to be one with it -- pretty sad :( -- Do I need therapy?

Nah. I myself don't care about MQA benefits, but seeing 352.8 KHz (DXD sample rate) on my DAP (2x AK4499 DAC chip) with fancy MQA blue light is enough to put smile on my face even if there's no audible difference from the CD version. It's a psychoacoustic behavior that I acknowledge as part of this hobby

11214119.jpeg
 

majingotan

Major Contributor
Forum Donor
Joined
Feb 13, 2018
Messages
1,542
Likes
1,826
Location
Laguna, Philippines
@majingotan Is this the uhh, long awaited A&K dual 4499 DAP they teased for a while? Isn't this thing like a billion dollars?

Yep. This one was released last year May. I've had this DAP before the start of the Covid outbreak since it's my first time owning a AK DAC chip (I have 4 other DACs and DAC/amp with different DAC chip brands) and want to have that AK 4499 DAC chip in a portable format. Lol at the billion dollar price and you can actually buy 30 million units with that kind of money. Definitely good to have as a collector's item only and nothing more to write about it in terms of audio performance
 

majingotan

Major Contributor
Forum Donor
Joined
Feb 13, 2018
Messages
1,542
Likes
1,826
Location
Laguna, Philippines
Like your style AND your avatar!

Thanks! You should try that album in my avatar (comforting background music), and it's well mastered to sound great on many systems. It's not a bad thing to have full functionality on a DAC to me as I see it as a bonus (though you should not factor the additional cost of MQA Decoder in order to think that way lol. Otherwise, you wouldn't think of getting it if you are aiming for cost savings or absolute performance for the least amount of money)
 

Tks

Major Contributor
Joined
Apr 1, 2019
Messages
3,221
Likes
5,500
Yep. This one was released last year May. I've had this DAP before the start of the Covid outbreak since it's my first time owning a AK DAC chip (I have 4 other DACs and DAC/amp with different DAC chip brands) and want to have that AK 4499 DAC chip in a portable format. Lol at the billion dollar price and you can actually buy 30 million units with that kind of money. Definitely good to have as a collector's item only and nothing more to write about it in terms of audio performance

May? Wtf, I could've sworn it was only hinted at around May, and wasn't even out until at least the fall. You must have some great connections >_>

Either that, or I've lost my mind. Oh and has it really been that long? I remember salivating at seeing what makers will be able to do with 4499 chips (I'm still dying to see what Matrix Audio or RME do with it if they ever undertake such an effort).

Oh and DAPs to me, anything over a couple hundred feels like a billion dollars. Simply because I feel their performance has been pretty poor even with relatively lower power output than desktop products, and not being hooked up to any outlets (I would assume mains region frequency would be superior which it is, but still after 1kHz, many DAPs have fumbled pretty spectacularly given their size and featureset exclamations). They're probably getting much better now, but I still think they fail primarly due to their reliance on Android (battery draining garbage of an OS no one can optimize to save their lives in the DAP business), and the form-factors are more conducive to devices that would make more sense as portable desk products, rather than something like what iPods were (seems EVERY DAP maker is literally incompitent, totally unable to emulate the form and wheel execution the Classic iPods were capable of, which is pretty sad). My real qualm with that approach is, if I wanted Android on my DAP, I might as well get a flagship LG V60 currently or something... I get a phone and a great mobile performer. The only way I'd get some of these A&K DAPs is if I had a need for a highly mobile and all-in-one device that (should) have great performance and outdo desktop products (the chips it sports disqualify it from fumbling on performance, and so does the price).

I wonder how that device actually performs... Super curious
 

boXem

Major Contributor
Audio Company
Joined
Jun 19, 2019
Messages
2,022
Likes
4,926
Location
Europe
Although there is the Tambaqui, which uses 3 AD SHARC processors, to do its internal conversions, and given what I have heard from that DAC, I suspect their algos for conversion are very, very good.
Please stop bringing the Tambaqui as whatever proof of DSD goodness. I repeat: Tambaqui is a PWM DAC, not a PDM DAC. DSD is PDM. PWM is multibit, not DSD.
I suspect the 3 SHARC DSP are there because the PCM to PWM conversion algorithm used is extremely ressources intensive. The way I see it is one DSP for clocking and upsampling the PCM flow of the two channels, then one DSP per channel for PCM to PWM conversion.
 

lazybonesxp

Member
Joined
Mar 24, 2020
Messages
8
Likes
0
If this was directed at me, as mentioned, I did not use any of those filters in my listening to the D-90, as I fed it only DSD 256 in direct DSD mode. In direct DSD mode the oversampling/digital filtering is completely bypassed.

Hmm, I've just looked again at the datasheet for AK4499 and it looks like in DSD "volume bypass" mode DF still works (even with DSD256). Am I wrong?
 

Attachments

  • DSD-DF-1.jpg
    DSD-DF-1.jpg
    126.7 KB · Views: 123
  • DSD-DF-2.jpg
    DSD-DF-2.jpg
    77.8 KB · Views: 126

lazybonesxp

Member
Joined
Mar 24, 2020
Messages
8
Likes
0
Sooo as always. Anything to back that up? You are proudly exclaiming these things yet you having nothing but your experience/anecdote, we can't take that as a truth. Nothing in this review, or on the forum (audiosciencereview) shows that the pre-amp mode should be disabled for optimal sound.

You are just grasping at straws. Please don't proclaim what you think or believe, personally, as facts.

But only in DAC mode you can avoid DS-modulation step in case of DSD.
 

majingotan

Major Contributor
Forum Donor
Joined
Feb 13, 2018
Messages
1,542
Likes
1,826
Location
Laguna, Philippines
May? Wtf, I could've sworn it was only hinted at around May, and wasn't even out until at least the fall. You must have some great connections >_>

Either that, or I've lost my mind. Oh and has it really been that long? I remember salivating at seeing what makers will be able to do with 4499 chips (I'm still dying to see what Matrix Audio or RME do with it if they ever undertake such an effort).

Oh and DAPs to me, anything over a couple hundred feels like a billion dollars. Simply because I feel their performance has been pretty poor even with relatively lower power output than desktop products, and not being hooked up to any outlets (I would assume mains region frequency would be superior which it is, but still after 1kHz, many DAPs have fumbled pretty spectacularly given their size and featureset exclamations). They're probably getting much better now, but I still think they fail primarly due to their reliance on Android (battery draining garbage of an OS no one can optimize to save their lives in the DAP business), and the form-factors are more conducive to devices that would make more sense as portable desk products, rather than something like what iPods were (seems EVERY DAP maker is literally incompitent, totally unable to emulate the form and wheel execution the Classic iPods were capable of, which is pretty sad). My real qualm with that approach is, if I wanted Android on my DAP, I might as well get a flagship LG V60 currently or something... I get a phone and a great mobile performer. The only way I'd get some of these A&K DAPs is if I had a need for a highly mobile and all-in-one device that (should) have great performance and outdo desktop products (the chips it sports disqualify it from fumbling on performance, and so does the price).

I wonder how that device actually performs... Super curious

Not really. IIRC, they were offered firstly for public demo at CanJam SoCal 2019. Also, you might be right that the first available purchase was around September as A&K keep pushing them back. I first heard the AK SP2000 at the same time I demoed the Raal SR1a with their HS1a amp and Jotunheim R amp as well as the Dan Clark Voce + STAX SR009 / BHSE pairing back in late February/early March this year.

With my brief demo, the AK can certainly drive efficient headphones like Focal Utopia or Stellia on its unbalanced port without sounding rolled-off in bass frequencies as common with Realtek on-board audio. I didn't have a 2.5mm to 4pin XLR adapt that time to try some inefficient planar magnetics and high impedance headphones like the HD800. Then again, I don't like wearing headphones on public and would rather have IEMs which the AK has zero issues driving them on both unbalanced and balanced port.

Regarding DAPs, we'll probably get desktop level performance out of them soon as more manufacturers are aware of Amir's efforts of exposing their design flaws through thorough measurements. Unfortunately, almost all of them are turning to Android as consumers are pushing for that streaming functionality with their DAPs (and of course they only run on Android). Competition also drove R&D efforts to not even consider designing their own custom OS from ground up (this approach would allow click-wheel functionality like from old iPods) and attaching a skin UI while stripping other Android functionality is far easier to achieve thus less costly and less headache for troubleshooting for end users. IMO, it's not that Android is highly inefficient as you can see how long lasting those flagship phones go with their 5000 mAh battery. I blame it more on manufacturer's approach of turning off power savings on the flagship DAPs as they believe that power saving states will cause worse sound quality (not audio performance IMO). The AK doesn't even have a standby mode that other DAPs have (instead it has a user adjustable timer for turning off the unit completely when not in use). That's why I see the AK DAP as "luxury/collector's item" (just look at the build quality and chassis design) rather than a functional DAP (smartphones and DAPs with tons of dedicated buttons that have over 20 hrs of music playback)
 

Pluto

Addicted to Fun and Learning
Forum Donor
Joined
Sep 2, 2018
Messages
990
Likes
1,635
Location
Harrow, UK
DSD is not audio madness, there are real audio benefits to DSD
No, not madness, but the advantages of DSD, these days, are dubious. Let's look at some history. DSD was developed as a concept during the late 90s when conventional PCM converters typically ran at a maximum of 48kHz and an accuracy of 14 or 15 bits, if you were lucky. While there were converters that would deliver 24/96 format, their true performance was highly questionable. There were two significant realisations that happened, more or less, at the same time.

The first of these was when the record industry realised that its principal asset – its vast library of original recordings dating back 50 years or more – was quite literally falling apart. The older recordings were degrading rapidly (as entirely predicted by the chemical and magnetic characteristics of the tape on which they were created) and many of the newer recordings were meeting a somewhat premature end thanks to a mistake in the design of the tape formulation, causing the tape stock to self-destruct in spite of the highly optimised conditions in which the precious recordings were stored. The industry sought a "future proof" digital format on which to safeguard its asset. The mantra within the recording industry was that it had to operate at a level of quality an order of magnitude superior to whatever was being sold to the consumer, and in a world in which the Compact Disc had become the primary format for consumer delivery, that was not an easy ask. Hence the development, largely within the Sony laboratories, of the DSD format with its performance measurably (if not audibly) superior to the 16/44 standard of the time.

The other issue that caused record business executives to cry in their champagne related to the area that we now term "rights management". In its earliest days the CD had been an answer to a record executive's dream insofar as the things were inherently uncopiable, were it even possible; the 600MB capacity of the CD would completely overwhelm the storage capacity of even the largest hard disks available to the consumer. But the landscape was changing. The arrival of CD-R technology coupled with the ever increasing capacity of hard disks had broken down the inherent immunity of the medium to illicit copying and manufacture.

Sony suggested that its DSD technology could become the basis of the next consumer format but this time, there would be no mistakes. The system would be sold in a highly encrypted format, locked in by technologies that would remain highly proprietary to the licensees. And so the SACD was born, the record industry breathed a brief sigh of relief and returned to its indolent and extravagant existence. But while the record industry slept, the world moved on.

Over the next few years, while the record industry was fighting a losing battle to sell the world yet another new format, necessarily expensive, touting an improvement in sound quality to a market that loved the CD (yet 10 years previously had been largely happy with the sound quality of Compact Cassettes), the technology was moving on apace. Manufacturers (essentially the integrated circuit industry) were achieving better and ever better performance with analogue to digital conversion and vice versa. Sampling rates and bit depths were improving in orders of magnitude with the actual accuracy of conversion vastly exceeding the capability of the human ear.

I have been involved in many well-controlled industry tests over the years in which the capability and resilience of many formats and technologies have been examined and at no time has any evidence arisen that DSD-based technology has any inherent superiority to equivalent PCM. Comparing 16/44 PCM to DSD produces the results that are to be expected given the increase in data rate but once you increase the PCM format to the 24/96 region, DSD is left trailing in the dust.
 

BDWoody

Chief Cat Herder
Moderator
Forum Donor
Joined
Jan 9, 2019
Messages
7,109
Likes
23,721
Location
Mid-Atlantic, USA. (Maryland)
Perhaps you might like to investigate such before making such inaccurate statements...

https://web.archive.org/web/2007092...te/diplomarbeiten/dsdvspcm/aes_paper_6086.pdf

7. SUMMARY
These listening tests indicate that as a rule, no
significant differences could be heard between DSD and
high-resolution PCM (24-bit / 176.4 kHz) even with the
best equipment, under optimal listening conditions, and
with test subjects who had varied listening experience
and various ways of focusing on what they hear.
Consequently it could be proposed that neither of these
systems has a scientific basis for claiming audible
superiority over the other. This reality should put a halt
to the disputation being carried on by the various PR
departments concerned.
Only four of the 145 completed tests, or 2.76%, yielded
results within the range of “critical probability” (i.e. less
than 5% probability of guesswork). These four tests
were conducted with two-channel listening material
which was played back through headphones, while in
the 100 completed tests using surround recordings, not a
single test result achieved the critical probability level.
Though less readily formulated with mathematical
equations, the high level of frustration felt by many
subjects during their tests left quite a strong impression.
These people, for the most part, were well accustomed
to critical listening on a professional level, but they
found that they could not even begin to recognize any
sonic differences. A further frequent topic in personal
conversations right after the test was the appearance of
“would-be” differences—sonic illusions, so to speak.
That is an issue which certainly has special importance
to the Tonmeister; there needs to be a specific personal
clarity concerning its causes. Developing a thorough
understanding of this theme should profit both the
musical and the sonic/technical aspects of a
Tonmeister's work.

Has much changed? Seems that anything I've seen has either had problematic methodology or inconclusive results.
 
Top Bottom