Upsampling to anything other than your DAC's internal conversion rate

[Miska]

The DAC may not understand where the 192 came from, but 'we' know the file has undergone sample rate conversion.

 

It has gone through it anyway...

 

There maybe more processes out in the wild but I only know of Izotope that's built into Sound Forge and Audirvana +. The Izotope SRC can be change the steepness, max filter length, cutoff scaling, alias suppression and Pre-Ring. I don't have much of a clue how to set them (the point of another thread), let alone make conclusions how they work. On the occasion I do SRC to create a CD from a 96/24 file, there's nothing much lost, so by luck SRC works.

 

Sure, you can find comparisons for some of those here:

SRC Comparisons

 

The ones listed under title "Signalyst 2.9.1" still apply to current versions of HQPlayer.

 

I don't offer iZotope kind of adjustments, because those are not enough to describe all the filter properties, just a minor subset of it. And still they offer a way to shoot yourself in a foot.

 

By inference, you're telling me that upsampling SRC is transparent? If you apply mathematics, no matter how 64bit precision you make it, there will be a difference.

 

No, I'm telling that there's no way you can avoid upsampling SRC with any of the modern DACs. Either you do it in software before sending it to DAC (and DAC just thinks it's a hires recording). Or you let DAC do it. The DSP engines inside DAC chips are nowhere close in accuracy and quality to what you can do with software. In modern parts those are typically 32-bit and in older parts 24-bit. And quality of the filters is also severely limited due to the issues I said.

 

Another item is quality of the delta-sigma modulator in the DAC chip.

 

Overall, most DAC chips only run upsampling digital filters to 352.8/384k rate and then go up from there by copying samples or by using linear interpolation (AKA the dumb old mathematical mean). While for example HQPlayer can upsample with digital filter right to 24.576 MHz rate and doesn't ever use cheap methods like copying samples or linear interpolation to increase rate.

 

Some examples...

 

Here's output of 1 kHz -60 dBFS sine with 44.1/24 input data, you can see the early increasing modulator noise floor due to 3rd order modulator and some idle tones too:

 

And there's output of the 1 kHz -60 dBFS sine upsampled to DSD256, you can see that noise floor is now flat and the only extra peaks left are the low level noise peaks from the XMOS USB receiver (USB packet noise):

 

...of course you could still subjectively prefer the first one...

 

Or from another angle, 0 - 22.05 kHz sweep input at 44.1/24, standard digital filter selected, you can see images around multiples of 352.8 kHz:

 

Same source data, now upsampled to DSD512, no images at all because it has gone through proper digital filters to 24.576 MHz rate:

 

...of course you could still subjectively prefer the first one...

 

 

Sure it takes much more processing power to upsample 256x or 512x compared to 8x done inside the DAC chip. Especially because the precision is also immensely higher.

 

Others have posted that they prefer not to upsample and I would agree after listening to various sample rates for HQPlayer, including DSD256 for 18 months. I don't believe it is problem that the computer is working harder and is noisier as a result, it's the FLAC (uncompressing) vs WAV argument, no. Maybe Izotope is better.

 

I cannot know why you prefer something and cannot argue about it. I can just argue about objective things.

 

Maybe I drag out the TASCAM DA-3000 recorder and make some comparisons. Now, which filter to use?

 

I'm not sure what you are talking about. Your filter selection should be based on the material you listen and your personal preferences.

[Jud]

The DAC may not understand where the 192 came from, but 'we' know the file has undergone sample rate conversion. It may be all numbers in the mathematical domain, but we are not multiplying any easy equation, like 2 x 4 =8. The algorithms for SRC are surely as complex for upsampling process as they are for the downsampling process.

 

There maybe more processes out in the wild but I only know of Izotope that's built into Sound Forge and Audirvana +. The Izotope SRC can be change the steepness, max filter length, cutoff scaling, alias suppression and Pre-Ring. I don't have much of a clue how to set them (the point of another thread), let alone make conclusions how they work. On the occasion I do SRC to create a CD from a 96/24 file, there's nothing much lost, so by luck SRC works.

 

Actually it's more difficult to try to downsample with audible transparency than it is to upsample. Your intuition that "there's nothing much lost" when you create a CD from a 96/24 file is not correct. Once again this has to do with unavoidable mathematical tradeoffs between time domain and frequency domain performance that become more problematic the closer one gets to the audio band. Upsampling moves you away from the audio band (that's why it's been standard in the digital audio industry since before separate DACs even existed), downsampling gets you closer to it and thus is more difficult to do without audible problems.

 

By inference, you're telling me that upsampling SRC is transparent? If you apply mathematics, no matter how 64bit precision you make it, there will be a difference.

 

Nope, though some people may find it to be. Just saying that unless you choose from among a tiny handful of true NOS DACs, your choice is to upsample in software or in your DAC. Upsampling with competently done software is usually measurably better than doing so in the DAC by reason of greater computing resources when using the former. And if you do choose a true NOS DAC and run Redbook through it, there is a significant, measurable increase in harmonic and intermodulation distortion versus using upsampling. (Again, that's why upsampling has been standard in the digital audio industry for decades.)

 

Others have posted that they prefer not to upsample and I would agree after listening to various sample rates for HQPlayer, including DSD256 for 18 months. I don't believe it is problem that the computer is working harder and is noisier as a result, it's the FLAC (uncompressing) vs WAV argument, no. Maybe Izotope is better. Maybe I drag out the TASCAM DA-3000 recorder and make some comparisons. Now, which filter to use?

 

Certainly no one can argue with personal preference.

[Jud]

Jud: as you can see my feeble lawyer/VC mind struggles with the precision of these engineering concepts

 

Yeah, I often feel like I'm attacking the problem with a blunt instrument too.

 

The part I found interesting in playing around with the A+ parameters in iZotope were the necessary tradeoffs and learning much more about my own personal sensitivities to a) ringing, b) decay, c) transient attack, d) frequency flatness and others. Quite often things that sounded great on day one became tiresome after a few days or weeks. The longer I listened, the more less was more (I think you experienced some of the same in your progressive leaning toward less-invasive flatter filter slopes). It was that learning with iZotope that then allowed me much more quickly to choose which filters I liked in HQPlayer (even if I didn't know what fine-tuning parameter tradeoffs Jussi had made).

 

I also have to assume that in choosing some of these filters we are "making up" for limitations elsewhere in our system -- for example a bookshelf sized set of speakers might seem much larger with a bit more reverb and slower decay. With my giant Magnepans, exactly the opposite was true -- I tended toward leaner, more nuanced presentations.

 

I would rather think of one part of the system "working with" other parts. An example I've used before is that my main system speakers are set up (including the crossover filters) to be time and phase correct, which may be a reason I prefer linear phase to minimum phase filters there. It seems to me subjectively that this results in better imaging (and in fact I've done a bit of blind testing where I've preferred linear phase).

 

One area where Miska in particular has tried to make up for deficiencies elsewhere in the chain is by providing among the HQPlayer choices "apodizing" filters, to substitute the better transient performance of the HQPlayer filters for worse transient performance that is often characteristic of the filters used in the creation of CDs. On the other hand, there are people who don't particularly like apodizing filters, which leads to what I say below.

 

All that being said, I do think there is value in us collectively trying to come up with a list of starting points for different DACs. Put differently, I'd like to make sure I'm going through filter choices that make the greatest possible impact because I have chosen a DSD512 or 24/384 starting point wherever that is possible and thus knowing where to start then pushes the learning over to the other filter choices rather than choice of upsampling.

 

As noted before, I think choosing software upsampling rates is really as easy as upsampling to the max input rates for most DACs, with the exception of certain DACs (like the Chords) that accept DSD input but should be fed the highest possible PCM rates.

 

Beyond that, I doubt we'll find a lot of consensus, because people not only have different tastes, we actually hear differently from each other. As you said, a lot of choice in DACs comes down to filtering preference, and there's hardly unanimity in choice of DACs. But for me it's nice to try to find a better understanding of what the different filters and their choices of parameters are doing, so at least I have some idea of *why* I may prefer one filter to another.

[sdolezalek]

For example rock recorded in studio puts demands on the transient response, so a minimum phase filter that doesn't have pre-ringing is usually good fit, since there's usually not much real acoustics involved anyway. While classical music recorded in real acoustics puts demands to the sound field/space, so a linear phase filter is usually good fit, since there are no strong transients or at least very few.

 

Given that most music library software (like Roon) includes "Genre" tagging, would it be possible to create a genre-specific filter assignment rather than having to stop HQPlayer, change the filter and restart? I understand it would have to create a gap in playing as the filters are changed, but the ability to change filters "on the fly" is particularly relevant for those of us where the processing computer is in a completely different part of the house than our music system.

[ted_b]

Given that most music library software (like Roon) includes "Genre" tagging, would it be possible to create a genre-specific filter assignment rather than having to stop HQPlayer, change the filter and restart? I understand it would have to create a gap in playing as the filters are changed, but the ability to change filters "on the fly" is particularly relevant for those of us where the processing computer is in a completely different part of the house than our music system.

 

The Alchemy Desktop (HQPlayer 3rd party front end) has a genre-based (among other rules) filter map that does exactly this, but since my trial expired I haven't been able to try it. I know that the latest version has many updates, which likely allow for no HQP restart, but not sure.

http://www.computeraudiophile.com/f11-software/alchemy-desktop-touch-friendly-control-interface-hqplayer-windows-29567/

[orgel]

Like others, I'm really enjoying this thread. I think some of the info is starting to seep through the mostly impermeable barrier provided by my skull.

 

I know this is tangential, but I wanted to add a note to something @Miska said…

 

Note! I don't know details about the latest series ESS has put out! exaSound is first I've seen to use a chip from that series.

 

Based on the datasheet it should work up to DSD512 and there are some implementations. But it doesn't seem to be reliable.

 

The only requirement they have is that MCLK must be at least 3x the DSD BCLK and MCLK must be <= 100 MHz. I personally think it should work if one would use exactly the 3x clock. This is a bit hard of course with normal audio clocks. But if someone wants to experiment, someone could come up with some frequency like 90 MHz for MCLK and then I could add 30 MHz output to HQPlayer. That would of course require a bit of customization to the USB firmware too.

 

It appears that Resonessence is using the the same ESS ES9028PRO chip in its Invicta Pro, Invicta Mirus Pro, and Veritas DACs as exaSound is using in its e32. I don't think any of these does better than DSD256 — at least, none are advertised as doing DSD512.

 

--David

[徐中銳]

Another item is quality of the delta-sigma modulator in the DAC chip.

 

Actually, nearly all finished DAC products for sale are superfluous to your preference of upconverting to DSD256 ? In the sense that you'd rather have no further processing, just a low-pass filter ?

If so, what good are finished DAC products ? What audience, for what purposes are they marketed towards ?

[Miska]

An example I've used before is that my main system speakers are set up (including the crossover filters) to be time and phase correct, which may be a reason I prefer linear phase to minimum phase filters there. It seems to me subjectively that this results in better imaging (and in fact I've done a bit of blind testing where I've preferred linear phase).

 

As I've been trying to say before, this is one of the main decision driving factors among type of the content. I listen mostly through headphones, and mostly through Sennheiser HD800. Since it is single driver one-way solution like most headphones, it is time/phase coherent. But since I listen a lot things like prog rock (mostly, but really everything from classical to electronic club music), there's not much "imaging" to talk about in that genre (IMO), but I'm particularly sensitive how the drum/cymbal highs sound like and how their leading edge sounds like. Overall, I'm most bothered by problems at the top end of the audio band, like resonances from metal dome tweeters. Any problems on that area, like hissy tweeters make me feel uncomfortable. Problems in that area can be emphasized by certain leaky digital filters. As a Finnish person I guess I'm supposed to like Finnish speakers like Genelec studio monitors, but I really cannot listen those. Damn, almost all commercial Finnish speakers use metal dome tweeters. I love ribbon tweeters though, fast and clean and wide. Silk domes are fine for me too, although less precise and wide. That's why I now use both Elac speakers with their JET tweeter (ribbon in AMT configuration) and Dynaudio with their silk dome.

 

But that's me and I think I know how to switch my listening focus to other things and I know for certain that other people are sensitive to other things and listen different types of music. So there are various different options with emphasis on different things. I go to classical concerts too, especially because I like the new Music Centre at Helsinki.

 

My favorite album is Pink Floyd's Meddle. The highest resolution versions of it I have are all RedBook (couple of versions). But damn I have tried hard to make best out of it, and I believe I've got pretty far, IMO...

 

choices "apodizing" filters, to substitute the better transient performance of the HQPlayer filters for worse transient performance that is often characteristic of the filters used in the creation of CDs. On the other hand, there are people who don't particularly like apodizing filters, which leads to what I say below.

 

There are still both number of linear- and minimum-phase versions of the apodizing filters to choose from. But there are alternatives of course too, like poly-sinc-hb and closed-form that are non-apodizing and from practical perspective don't change anything in good or bad.

[Miska]

Actually' date=' nearly all [i']finished DAC products[/i] for sale are superfluous to your preference of upconverting to DSD256 ? In the sense that you'd rather have no further processing, just a low-pass filter ?

If so, what good are finished DAC products ? What audience, for what purposes are they marketed towards ?

 

I'm not sure I understand your question. They try to provide single-box solution manufactured to certain price point making certain amount of profit.

 

From my perspective, they have features that are not necessarily needed. But if one doesn't have, or doesn't want to have, external processing entity, they can be self-contained too. But most DACs that are DSD capable are based on COTS DAC-chips. For example the AKM's AK4490. It's 2.5€/piece, if it happens to have some PCM side to do rudimentary digital filters and modulation, it doesn't matter too much in the big picture (and doesn't do harm) as long as the DSD conversion is good quality. If someone want's to use the on-chip DSP engine needed with PCM-inputs, I don't really care.

 

But I'd rather have DAC manufacturers spending the money to make the actual digital-to-analog conversion as good as possible and leave the DSP processing to others. For example the T+A DAC8 DSD is good example of such for the DSD-side. It would be even better if they'd leave out the (to me unnecessary) PCM side out altogether and use the money to improve the DSD side.

 

The only thing that I feel comes close to what I want is dCS with their upsampler + DAC device combo. But the price is quite high and I feel I get quite a bit better performance for a lot cheaper now. So I wouldn't buy one even if I'd win in a lottery.

 

 

P.S. I have for example TEAC NT-503, it has lot of features, but the only thing I use from it is DSD256 input over USB. Everything else is unnecessary extra, but I don't mind since it doesn't interfere with the stuff I want to do. (the DSD upsampling they have built-in is horrible quality and nobody should use it)

[One and a half]

Some examples...

 

Here's output of 1 kHz -60 dBFS sine with 44.1/24 input data, you can see the early increasing modulator noise floor due to 3rd order modulator and some idle tones too:

[ATTACH=CONFIG]30009[/ATTACH]

 

And there's output of the 1 kHz -60 dBFS sine upsampled to DSD256, you can see that noise floor is now flat and the only extra peaks left are the low level noise peaks from the XMOS USB receiver (USB packet noise):

[ATTACH=CONFIG]30010[/ATTACH]

 

...of course you could still subjectively prefer the first one...

 

Or from another angle, 0 - 22.05 kHz sweep input at 44.1/24, standard digital filter selected, you can see images around multiples of 352.8 kHz:

[ATTACH=CONFIG]30011[/ATTACH]

 

Same source data, now upsampled to DSD512, no images at all because it has gone through proper digital filters to 24.576 MHz rate:

[ATTACH=CONFIG]30012[/ATTACH]

 

...of course you could still subjectively prefer the first one...

 

 

 

Thank you for the waveforms, it all looks good for sine waves and sweeps. I would agree that 20kHz cutoff is a real problem with any filter having trouble, since they do impact on the audio band no matter steepness etc.. It therefore makes sense to increase the sampling frequency of especially redbook to 176.4 and use a filter where it won't affect the audio band as much. Understand the similar approach with DSD64 and increase the speed and applying even smaller filters.

 

The selection of the filter is still a crap shoot with some really deep understanding of which filter works at its optimum with:

 

a) the music content, eg complex orchestral, jazz trio, or compressed rock

 

b) the type of DAC chip used.

 

You're all over it, I'm not, the more I look at filter selection, gives me the heebee jeebees, where's that crucifix. It begs the question then to introduce profiles for different types of music, gosh even down to the album (?), so the filter has an optimum, or better chance of working. Is this what needs to happen? No, this is not listening to music, this is wholesale tinkering, no thank you. No wonder people still prefer a turntable or tape. Even order harmonic distortion is a good thing then, music has gobs of harmonics, so makes sense to add it in, rather than being clinically cold.... any plans to add it as a "filter"? A+ does, I believe it's Mode 2.

 

Computer audio does make selection of music very easy, rather than shuffling through discs, but to get the optimum out of it, is a lot of work hardware wise and software wise to get it sound less digital. Makes me wonder and why bother with it all.

[Miska]

The selection of the filter is still a crap shoot with some really deep understanding of which filter works at its optimum with:

 

a) the music content, eg complex orchestral, jazz trio, or compressed rock

 

For that, I have the recommendations. But in any case you can just try out or make an educated guess based on the recommendations.

 

b) the type of DAC chip used.

 

That's about modulators and not about upsampling filters. But I have tested/measured my modulators with different DAC chips, and designed those to work well, that's my part of the job. And for the discrete (non-chip) DSD solutions I have the recommendations on what to use.

 

You're all over it, I'm not, the more I look at filter selection, gives me the heebee jeebees, where's that crucifix. It begs the question then to introduce profiles for different types of music, gosh even down to the album (?), so the filter has an optimum, or better chance of working.

 

Doing these things with software is possible. While DAC manufacturers try to pretend that such doesn't even exist, because the DAC cannot know what genre the music is about. All you get is either "one size fits all", or couple of filter options to choose from. Unfortunately they are usually quiet about how you should be selecting between those filters. While I at least try to formulate a logic on how (IMO) the filter selection could be done.

 

Even order harmonic distortion is a good thing then, music has gobs of harmonics, so makes sense to add it in, rather than being clinically cold.... any plans to add it as a "filter"? A+ does, I believe it's Mode 2.

 

No, I don't like those kind of ideas. If the recording is clinically cold, it should sound like clinically cold. But the old fashioned digital stuff used to make even warm sounding recordings sound "digital", "clinically cold" or "hash in the HF" due to those HF artifacts. It's like some of my first CD players with SAA7220 digital filter and TDA1541A DAC chip - making everything sound digital - the sound of 80's. Like the plastic Casio keyboards.

 

Computer audio does make selection of music very easy, rather than shuffling through discs, but to get the optimum out of it, is a lot of work hardware wise and software wise to get it sound less digital. Makes me wonder and why bother with it all.

 

Primary problem is RedBook content. Some engineers really wanted to push storage requirements to bare minimum in 70/80's and really followed the theory to the extremes. Making it very hard to properly go back to analog. If you'd have content in 192/24 PCM or DSD you'd have much less problem to begin with.

 

Luckily they didn't have processing power to do psycho-acoustically lossy codecs at that point (in a commercially feasible way for the playback side)!

[Jud]

Even order harmonic distortion is a good thing then, music has gobs of harmonics, so makes sense to add it in, rather than being clinically cold.... any plans to add it as a "filter"? A+ does, I believe it's Mode 2.

 

 

Where did you get that from? "Mode 2" isn't even a filter, let alone designed to add gobs of harmonic distortion. If you want that, don't oversample or use any filters at all.

[One and a half]

Where did you get that from? "Mode 2" isn't even a filter, let alone designed to add gobs of harmonic distortion. If you want that, don't oversample or use any filters at all.

 

On the run typing there, I did put the word "filter" in, should maybe used a different word, plugin perhaps, Mode 2 does add warmth to the sound, from the A+ manual.

 

Mode 1 that brings the highest transparency, and soundstage depth

Mode 2 that is more on the warm side

 

Adding even harmonics does create the warm sound, in an attempt to re-create this situation, experimented with Tube plugins, so far not settled on anything for a workable solution. Mode 2 is a good compromise. It's a preference.

 

I did read somewhere here at CA, that the Nadac has added distortion to make it sound audiophile...whereas the pro components HAPI and Horus don't.

[Jay-dub]

P.S. I have for example TEAC NT-503, it has lot of features, but the only thing I use from it is DSD256 input over USB. Everything else is unnecessary extra, but I don't mind since it doesn't interfere with the stuff I want to do. (the DSD upsampling they have built-in is horrible quality and nobody should use it)

 

By horrible DSD upsampling, I take it you mean that the filter on its chip (the AK4490) to lowpass DSD input (for volume adjustment and remodulation) is low quality. Does it alias a lot of DSM noise into low frequencies? From what I've read, the modulator on the 4490 is better than on most chips, so it should be OK for high-rate PCM input. You've also posted that the Chord Mojo's treatment of DSD input is not good. The problems would, I presume, be of similar nature in both cases. I find this disturbing. DSD input is such a small niche that it doesn't make sense for a manufacturer to implement the feature unless they're going to do it well, and lowpassing DSD shouldn't be a particularly difficult thing to do – at least if it's not going to be downsampled, which would seem to be avoidable in these DACs.

[semente]

On the run typing there, I did put the word "filter" in, should maybe used a different word, plugin perhaps, Mode 2 does add warmth to the sound, from the A+ manual.

 

Mode 1 that brings the highest transparency, and soundstage depth

Mode 2 that is more on the warm side

 

Adding even harmonics does create the warm sound, in an attempt to re-create this situation, experimented with Tube plugins, so far not settled on anything for a workable solution. Mode 2 is a good compromise. It's a preference.

 

I did read somewhere here at CA, that the Nadac has added distortion to make it sound audiophile...whereas the pro components HAPI and Horus don't.

 

In my experience warmth comes from exaggerating the upper bass range or a response tilting down from lows to highs, whilst even lower order harmonics add a sense of reverberation or space.

 

R

[One and a half]

In my experience warmth comes from exaggerating the upper bass range or a response tilting down from lows to highs, whilst even lower order harmonics add a sense of reverberation or space.

 

R

 

If AC power Harmonics enter the DC supplies, they are usually odd order and create reverse (inverse) problems.

[Miska]

By horrible DSD upsampling, I take it you mean that the filter on its chip (the AK4490) to lowpass DSD input (for volume adjustment and remodulation) is low quality.

 

No, I mean the DAC has it's own separate PCM-to-DSD converter you can enable/disable at will. I'm not sure if it's made using AK4137EQ chip or if they have their own implementation in DSP/FPGA.

 

The 0 - 22.05 kHz sweep looks like this with the built-int PCM-to-DSD conversion enabled and 44.1/24 source:

 

With DSD inputs the DAC behaves just fine. With PCM inputs (using the on-chip digital filters) it behaves like most other DACs, with images around multiples of 352.8/384k because the digital filter can do max 8x.

[YashN]

Actually' date=' nearly all [i']finished DAC products[/i] for sale are superfluous to your preference of upconverting to DSD256 ? In the sense that you'd rather have no further processing, just a low-pass filter ?

If so, what good are finished DAC products ? What audience, for what purposes are they marketed towards ?

 

The better questions are:

 

- Are commercial DACs optimised for the D/A process, or as optimal as they could be, given today's home technology availability?

 

- If not, can I build a DAC at home for such a purpose?

 

Having tested HQ Player, using A+ usually and liking high-rate DSD, I wish DACs were optimised solely for that D/A, and that manufacturers stop adding tons of 'features' and processing there, affecting the D/A process.

 

Thing is, consumers love buying and getting boxes with tons of features, manufacturers play the number game among themselves.

 

I'd rather have a DAC which does native DSD512+ optimally and that's it. Then I'd use HQ P + NAA to send it the highest rate it can receive. Just that D/A in the purest and simplest way possible.

[tedwoods]

I agree with Yashn. My system is based on such a principal and I couldn't be happier with the results.

[Jud]

Mode 1 and Mode 2 are "integer modes" in Audirvana Plus. These have to do with avoiding OS audio processing rather than harmonics or the topic of this thread, upsampling.

[徐中銳]

As I've said :

In short' date=' some of you are misleading with your approach and conclusions to the topic.[/font']

 

DAC, digital-to-analogue-converter. Analogue is the goal ? Amplification of the-analogue-signal out ?

...

 

You know I empathise with your high ideals of DSD256+. But let's consider most people, the average readers of this Tread.

 

Generalised drumming about ever higher iterations of DSD won't get you critical mass in brotherhood. Because DSD (like everything) has to be implemented well within a product's design (preferably with exhaustive listening tests before public release) or else it's but a catchcry feature that sloppy manufacturers throw-in but sounds substandard, even awful.

 

Start by recounting your personal experience with whatever specific products you esteem and we can examine if it can be applied, adopted by others.

[Miska]

Start by recounting your personal experience with whatever specific products you esteem and we can examine if it can be applied' date=' adopted by others.[/quote']

 

We've discussed quite a number of products before. Like especially the T+A DAC8 DSD, iFi iDSD series, TEAC UD-501/UD-503/NT-503, exaSound DACs and Mytek DACs. And also others like Marantz, Resonessence Labs, Sony and Lampizator DACs. With those, DSD works great.

 

And then there are category of DACs where DSD is sort of primary function, like EMM Labs/Meitner, Playback Designs and dCS.

 

Then there are DACs where it is not so great like Chord DACs and with those you are better going with the highest supported PCM rate.

 

But overall, as long as the DAC uses a DAC chip, the amount the DAC device manufacturer can do about it is pretty limited. Both good and bad.

[Dan Moroboshi]

Mytek

 

Sent from my HTC One M9 using Computer Audiophile mobile app

[徐中銳]

Your superficial defence of discussions amounts to but cursory generalities.

 

For example, youself can give more substance of your operation of a TEAC NT-503, detailing not least equipment chain (which isn't listed on your Profile) and listening-room layout, plus what music sourced, fed.

 

Each person who advances a position should do the same.

We've discussed quite a number of products before...

 

You gotta proof better than simply saying :

DSD works great.

[Miska]

For example' date=' youself can give more substance of your operation of a TEAC NT-503, detailing not least equipment chain (which isn't listed on your Profile) and listening-room layout, plus what music sourced, fed.[/quote']

 

I have so much equipment in different configurations that I don't see any point in listing all those.

 

You gotta proof better than simply saying :

 

I've been posting quite a bit of my own measurement data and such. But I have not seen such from you. Have you made any measurements that proof the contrary?

 

Are you trying to say that the equipment I listed works bad in DSD? I'm pretty sure everybody would be happy to know what is the input format that works best with those.