Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[gorolingtwins]

Future Requests for DSD Dac User in MC18 or Later...

I use Mytek 192 DAC.
I use JRMC17 and HQ Player and Foobar2000...

I'd like to reduce a number of player.

[PCM2DSD realtime conbert] is that I hear naturally sound or Ear-friendly sound.
I want your team to support the conversion PCM2DSD in MC.

Thank you for MC.

[Matt]

Welcome.

MC17 and MC18 can play DSD as high-quality PCM.  More here:
http://wiki.jriver.com/index.php/DSD_Format

[gorolingtwins]

Nice to meet you Matt.

Thank you for reply.

I know DSD2PCM on MC..

I Can Play DSD2DSD for Mytek 192 DAC.
I Can Playe DSD2PCM Multi Channel

I Enjoy DSD Sound on Mytek 192 DAC.

I Can Play [MP3andFLAC(PCM)]2DSD to Mytek DAC on HQ Player..

For Example MP3(320kbps,44.1khz16bit)2DSD64...
[PCM2DSD realtime conbert] is that I hear naturally sound or Ear-friendly sound.
 
That is Strange..

[Matt]

Oh, I understand now -- you want to play regular files as a DSD bitstream.

I'm not sure there's an advantage to this approach.  You can't add information to PCM that isn't there, so I don't see how you could get better quality by pushing the PCM through extra processing.  And when you go to DSD, you lose the super-high frequency information to the 1bit DSD aliasing.

Personally, I've been unable to hear a difference between PCM and DSD as the final bitstream delivery mechanism (once doing level matching; since DAC makers often run the DSD hotter).  It's possible there's something subtle I'm not trained to hear yet, but it's not night-and-day because I have pretty good equipment and quite a bit of experience with listening tests.

To be clear, I think DSD is great if it's the highest quality version of a song you can get.  But that's a bit different than whether there's a reason to convert everything to DSD.

[gorolingtwins]

Sorry to my comment late.

I want to write a little bit about PCM2DSD.

I think you're right.

What about DSD2PCM reversed.
This conversion is carried out in order to hear a lot of people.
It is not good because the sound is performed.

What kind of sound is the sound of our ideal.

Even if the sound of the studio master,
we of course can not be achieved.

I sound different when different CD-Player, for example.
I think there are a variety of sound alone DAC and AMP and PC only.

I said PCM2DSD transformation is implemented in the Playback Designs MPD-3, for example.
 I never heard this. MPD-3 the price is more than $ 10 000.
Is JRiver Media Center?

The others are implemented in software on the HQ-Player.

I think that conversion is the kind of private "Mastering".
I think it is also the same as the Digital Filter and Noise processing, including the Interporation.

The processing power of recent PC is very high.
I think The Software may be making more quality of sound in the software area is not a mere EQ.
What is considered a member of the JRiver Media Center.

If you're in the software to improve the sound, I would hope that you will consider us PCM2DSD.

Thank you Matt.

Sorry my poor English.

[Matt]

I've talked to three different DAC makers about this idea.

I was hoping one would say "Hey, good idea.  That could sound better because _____."

But that isn't really what they've said.  One said the best DACs are PCM internally so this would only hurt them, and another argued that hardware based PCM-to-DSD would be better due to their proprietary algorithm.  One DAC maker did say it would be interesting because a computer might be able to do a better job than hardware.

I've also spoken to Sebastian Gesemann, the leading expert on DSD-to-PCM on a computer.  He was amazingly helpful and educational, but mostly said "that's not a great option."

So I'm not ruling this idea out but I'd like a good technical argument for why it would be worth adding (other than that it could be interesting).

[gorolingtwins]

Thank you very much, Matt.

I have information about one product.

KORG DS-DAC-10
http://www.korg.co.jp/Product/Audio/DS-DAC-10/index.html

It seems for now only in Japanese.

This product has been implemented using the functionality of the app AudioGate PCM2DSD.

I hope you continue to study a digital conversion for good sound.

[Matt]

In a coming build of MC18:
NEW: Added DSD encoder with support for 1x and 2x DSD stored in a DSF file (encoding uses 64-bit data path, JRiver's audiophile-grade upsampler, and 7th order noise shaping + dither).

I wrote this mostly for my own education (and because I needed to play off in the weeds a little).

But the results seemed quite good and instrument tested well, so I thought it might be of general interest.

[gorolingtwins]

Really?
Amazing! !
I'm happy! !

Thank you very much Matt.

So,I have a question.

Are you using the PCM signal after the DSP Effect for output file.

I want to PCM2DSD after using DSP. The reason is because MC great feature of DSP.

I may be converted directly only when I do not want to do anything in the DSP.

I was really good to come here.

[Matt]

The DSD encoder is available now:
http://yabb.jriver.com/interact/index.php?topic=75260.0

You will have to do conversion offline (Library Tools > Convert Format).

We might offer real-time conversion during playback in the future (I actually coded this approach first), but I'm worried about complicating the user experience to support the feature.  Another wrinkle is that it requires a fast computer to do real-time conversion.

[gorolingtwins]

Hi Matt.

Thank you for your support and development.

I think the real-time conversion function may be enabled with caution too.
There is a possibility that the sound is interrupted or, noise is output.
Someone will need a lot of support.

I think advanced and difficult function and anyone frequently used functions should be separated.
But it's difficult.

MC is a very multi-functional. The function I never used is I don't know at all.
However, the well-known features that I want to call easily.
So, I want to arrange a function that I use freely. HaHa. . .

------------------------

I tried to convert some music.

My cheap equipment
PC Core2 1.86GHz(a little old PC)
DAC Mytek Stereo 192-DSD DAC v.1.63
HeadPhone SENNHEISER HD380 PRO

I used both the DSD conversion AudioGate new MC18.
I felt that DSD sounds better when the first friendly.
I felt the texture of percussion DSD is superior.

But. .

When I listen carefully, I felt as if the signal level is changed a little each.
I felt DSF file volume is just a little bigger than the original file.

The difference of this volume is fatal to determine the difference in the sound.

I do not know whether this causes a DAC or File.
I do not know whether the conversion is making a difference in how much sound now.
I am sorry.

I must make a careful evaluation.
Someone, please try to hear the sound quality of the new features.

I noticed that the other...

The conversion speed is fast AudioGate yet unfortunately.
When I played in AudioGate and HQ-Player the converted files in MC18,
there was sometimes a mistake that the total time.

[chrisr]

Matt,

Real time conversion to DSD or especially 2xDSD would be interesting, especally converting on the fly during playback.  I can see where some mfrs might argue that they do a better conversion using hardware, but it would be interesting to try to see if there is a difference.

Cheers

[Alex Peychev]

Hi Matt,

It was nice seeing you on the DSD conference panel at the RMAF!

Most current hardware PCM2DSD implementations use a SONY DSD encoder that must be fed with 8fs (352.8kHz), meaning CD data must filtered and 8x oversampled prior to entering the DSD encoder, meaning an external PCM digital filter device is required. Some use existing DF chips, some use custom DFs programmed to an appropriate logic device. In all cases, I personally doubt that such approach can outperform a well-done computer based PCM2DSD conversion. The other thing is that, although there is no more info added to the PCM signal, a major audio quality difference can be obtained because of the actual "DSD D/A conversion".

This said, as another DAC manufacturer, I am quite interested in PCM2DSD conversion on the fly (both DSD64 and DSD128) because this will allow for using a "DSD Only DAC".

Thanks!

[Matt]

@gorolingtwins

Thanks for the tests.

DSD encoding got much faster in build 64 (and newer) than the first release (build 63).  So please make sure you're testing that.

As for levels, you're right that DSD and PCM may be at different levels.  Levels are a tricky issue with DSD because you have to be careful to avoid overload of the noise-shaping.

Interestingly, the Mytek DAC (and maybe others) play at quite different volumes when delivered PCM vs DSD.



@Alex Peychev

That's a good argument about the 44.1kHz to 352.8kHz resampling.  I learned while writing this encoder that the resampler quality is quite important for this task.  I first started with cheap linear resampling, and there were nasty artifacts in the higher spectrums.  Then I switched to our audiophile-grade resampler used elsewhere in the program and it cleaned all of that up.  It wouldn't be surprising if hardware used a cheap resampler for this step, since it's quite involved to do a really good job.

[Alex Peychev]

Precisely, and that is my point as well - the DF has a great impact on the resulting audio quality, and same applies to PCM processing.

So when we start streaming PCM2DSD on the fly with MC18? 

Thank you!

[Matt]

So when we start streaming PCM2DSD on the fly with MC18? 

I actually coded real-time output before deciding an encoder would be a safer (with regards to user complexity) way to start.

So I think there's a good chance we'll add the option for real-time DSD output at some point.

[Alex Peychev]

So I think there's a good chance we'll add the option for real-time DSD output at some point.

Great, thanks! Mean while, I will experiment with the MC18 DSD encoder, compare it to some other encoders available, and report the outcome, if you are interested to know, of course.

Are you planning on different selectable DSD encoder options as far as processing goes? For example, Korg Audio Gate does not offer such options but the Philips Pro TECH DST encoder does. When "Trellis E 32" is selected, the resulting DSD file sounds very good and considerably better than the same file encoded with Audio Gate. Both files were DSD64.

[Matt]

I will experiment with the MC18 DSD encoder, compare it to some other encoders available, and report the outcome, if you are interested to know, of course.

Yes, please.

I did some instrument testing, but other results would be greatly appreciated.

Are you planning on different selectable DSD encoder options as far as processing goes? For example, Korg Audio Gate does not offer such options but the Philips Pro TECH DST encoder does. When "Trellis E 32" is selected, the resulting DSD file sounds very good and considerably better than the same file encoded with Audio Gate. Both files were DSD64.

A single profile with the highest possible quality is the only goal right now.

A cheap / fast version might be interesting later, but only if there's a clear need.

[Alex Peychev]

Yes, please. I did some instrument testing, but other results would be greatly appreciated.

Sure, will do!

A single profile with the highest possible quality is the only goal right now.

Great, thank you!

[Alex Peychev]

Hi Matt,

Sorry it took me so long but I was traveling. Now I am back to my reference audio system and have done some evaluations. I must say that 44.1/16 PCM converted to DSD with MC18 sounds really nice! Thank you! The only thing is - I feel it will be a good idea offering optional attenuation for the DSD conversion from 0dB to -6dB with 1dB step.

Since it will take a really long time to convert all of the CDs to DSD, it will be absolutely great if you can offer PCM2DSD on the fly.

Regards,
Alex

[preproman]

What DACs are you guys testing this with.  I'm very interested in getting the Mytek stereo192-DSD DAC Black Mastering  http://mytekdigital.com/storeus/index.php/stereo192dsd-dac-bl-mast.html

Has any test been done with this DAC?

[Alex Peychev]

What DACs are you guys testing this with.  I'm very interested in getting the Mytek stereo192-DSD DAC Black Mastering  http://mytekdigital.com/storeus/index.php/stereo192dsd-dac-bl-mast.html

Has any test been done with this DAC?

I am using my own DSD DAC (a prototype) for the evaluation. I think the audiophile who started this thread uses Mytek.

[Alex Peychev]

Yes, please.

I did some instrument testing, but other results would be greatly appreciated.

Happy New Year, Matt! Wishing you and the entire MC team a prosperous new 2013!

Do you think there will be a PCM2DSD on the fly option available soon? The DSD converted WAV files sound absolutely great, but it takes time to covert and it takes too much space on the drive. Would it be possible to include the PCM2DSD DoP option in the output modes under WASAPI, KS, etc.?

Thanks!
Alex

[Matt]

Happy New Year, Matt! Wishing you and the entire MC team a prosperous new 2013!

Do you think there will be a PCM2DSD on the fly option available soon? The DSD converted WAV files sound absolutely great, but it takes time to covert and it takes too much space on the drive. Would it be possible to include the PCM2DSD DoP option in the output modes under WASAPI, KS, etc.?

Thanks!
Alex

Hi Alex.

I think we probably will offer real-time conversion at some point in the future, but it might be a while.

Regards.

[Alex Peychev]

Hi Alex.

I think we probably will offer real-time conversion at some point in the future, but it might be a while.

Regards.

Hi Matt,

Thank you so much, this is great news!

Best wishes,
Alex

[Alex Peychev]

Hi Alex.

I think we probably will offer real-time conversion at some point in the future, but it might be a while.

Regards.

Hi Matt,

I received an email that the MC18 is now stable. Do you have an update about when to expect the PCM2DSD real-time conversion feature? Thanks!

[dean70]

DSD=>DXD (352.8 )=>DSP eq volume etc=>DSD would be nice.

[Matt]

Do you have an update about when to expect the PCM2DSD real-time conversion feature? Thanks!

I think it will be a v19 thing, which is still many months away.

[joelha]

As long as we're on the topic, could someone comment on the advantages and disadvantages of converting from PCM to DSD off-line vs. in real-time?

I'm talking from a qualitative listening standpoint, not a convenience or storage standpoint.

Would the sound likely be better through a MC off-line conversion vs. real-time conversion?

If not through MC, does anyone have a comment to make based on their use of another software package?

Thanks,

Joel

[Alex Peychev]

If not through MC, does anyone have a comment to make based on their use of another software package?

Thanks,

Joel

I am using the ASIOProxyinstall 0.6.1 (for Foobar) working as an ASIO "driver" in MC18 with great success. It converts PCM to DSD real-time with 4 selectable DSM options (I like DSM3 best) but it is limited to 44k and 48k sample rates. Sound is quite nice though, and I personally prefer it over "pure PCM". Now I can't wait for the MC18 or MC19 PCM2DSD real-time conversion because it will also work with sample rates up to 192kHz, I hope.

Regards,
Alex

[Alex Peychev]

The latest ASIOproxy v0.6.1 supports all sample rates x44.1 or x48k, up to 192k/24bit (or maybe more; I haven't tried yet). But, a file converted to DSD64 with the MC18 sounds quite amazing and much better, IMHO! So now I am getting very anxious for the PCM2DSD real-time conversion with the MC. And I do hope that the current filters are still intact, because it sounds best with the 50kHz filter option. :-)

[kstuart]

As long as we're on the topic, could someone comment on the advantages and disadvantages of converting from PCM to DSD off-line vs. in real-time?

I'm talking from a qualitative listening standpoint, not a convenience or storage standpoint.

Would the sound likely be better through a MC off-line conversion vs. real-time conversion?

There is some evidence (blind testing with multiple subjects) that all off-line conversion sounds better than all real-time conversion... but that the difference gets smaller as the PC gets more powerful (and so some sufficiently high powered PC would not be affected).   The testing was done by a DAC designer and manufacturer.

This correlates with anecdotal reports that sound quality improves slightly by removing anything that runs at the time as the playback thread.   Some of the anecdotal reports are by professional mastering engineers, which means that the reports might be more reliable, but also means that the differences being noticed could be quite small.

[JimH]

There is some evidence (blind testing with multiple subjects) that all off-line conversion sounds better than all real-time conversion... but that the difference gets smaller as the PC gets more powerful (and so some sufficiently high powered PC would not be affected).   The testing was done by a DAC designer and manufacturer.

This correlates with anecdotal reports that sound quality improves slightly by removing anything that runs at the time as the playback thread.   Some of the anecdotal reports are by professional mastering engineers, which means that the reports might be more reliable, but also means that the differences being noticed could be quite small.

JRiver believes that conversions in real time are no different than any other conversion, if the machine isn't causing any slowdowns, and any modern machine would not.

[Blaine78]

another vote for PCM -> DSD conversion. Be fun to play with,  Mytek owner here.

[Blaine78]

There is some evidence (blind testing with multiple subjects) that all off-line conversion sounds better than all real-time conversion... but that the difference gets smaller as the PC gets more powerful (and so some sufficiently high powered PC would not be affected).   The testing was done by a DAC designer and manufacturer.

This correlates with anecdotal reports that sound quality improves slightly by removing anything that runs at the time as the playback thread.   Some of the anecdotal reports are by professional mastering engineers, which means that the reports might be more reliable, but also means that the differences being noticed could be quite small.

consistent to what I hear with real time decompression of FLAC. FLAC sounds comparatively dull compared to WAV.

[kstuart]

JRiver believes that conversions in real time are no different than any other conversion, if the machine isn't causing any slowdowns, and any modern machine would not.

It depends on what you call a "slowdown".

PCs were never designed to work in real time.   Delaying any small number of bits by a small fraction of a second can affect the sound.

It's not hard to come up with scientific reasons.

[6233638]

There is some evidence (blind testing with multiple subjects) that all off-line conversion sounds better than all real-time conversion... but that the difference gets smaller as the PC gets more powerful (and so some sufficiently high powered PC would not be affected).   The testing was done by a DAC designer and manufacturer.

You are going to have to cite sources if you are going to make wild claims like that. (hopefully not this forum post again)

It's not hard to come up with scientific reasons.

You mean it's not hard to come up with reasons that sound plausible if you don't understand how the technology works.

As long as you are not using an internal soundcard, it shouldn't make any difference to the audio. And any modern DAC is going to be reclocking the data it receives anyway, if you even think jitter is audible.

[JimH]

I've been thinking about starting a thread on how we should try to communicate about sound quality issues.  The audiophile group has a bias toward some very esoteric descriptions.  I don't think JRiver's forum is a good place to discuss these in any depth.  And I'm concerned that new users will take some of the claims too literally.

On the other hand, people need to be able to state their opinions.  Maybe we just need to label these opinions better.  "In my opinion..."

I may move this to a new thread.

[JimH]

It depends on what you call a "slowdown".

PCs were never designed to work in real time.   Delaying any small number of bits by a small fraction of a second can affect the sound.

They're pretty darn close.  A lot closer than a human being could distinguish.

Again, this is such a loaded subject, that it is probably not one for JRiver's forums.

[Vincent Kars]

AudioAsylum has a nice solution.
Of course, they had their endless and fierce flares about whether audio cables do/don’t make a difference.
In the end, they created a DBT free zone so people could discuss the wonders of a piece of wire in peace.
What about an audiophile corner in the forum?

[Vincent Kars]

They're pretty darn close.  A lot closer than a human being could distinguish.

By design latency is a delay.
All gear including DACs, amps, etc do have a certain latency.
The delay itself is not audible as the sound is not altered, it only arrives a fraction of a second later.

A fluctuating latency is of course a different matter.
In case of PC based audio where everything is buffered a fluctuating latency of the system won’t have any impact as long as the buffer is filled in time.
I do think the concept of software induced jitter is a intriguing one, but up to today, I never have seen any well controlled experiment (that is what I call science) proving its merits.
For those interested;  The hunt for load-induced jitter: http://archimago.blogspot.ca/2013/03/measurements-hunt-for-load-induced.html

[dean70]

Another thing to consider is the timing relationship between channels.

[Alex Peychev]

Another thing to consider is the timing relationship between channels.

Yeah, especially if you are talking about NOS PCM DACs. :-)

[Alex Peychev]

I've been thinking about starting a thread on how we should try to communicate about sound quality issues.  The audiophile group has a bias toward some very esoteric descriptions.  I don't think JRiver's forum is a good place to discuss these in any depth.  And I'm concerned that new users will take some of the claims too literally.

On the other hand, people need to be able to state their opinions.  Maybe we just need to label these opinions better.  "In my opinion..."

I may move this to a new thread.

Sure, that will be great but, IMO, it will detract from the current PCM2DSD subject. FWIW, I have tried converting PCM to DSD with other programs, but found the MC18 converted file to sound best to my ears. So please consider PCM2DSD real-time conversion sooner as my 6TB HDDs are getting full. :-)

Thanks!

[kstuart]

"You mean it's not hard to come up with reasons that sound plausible if you don't understand how the technology works."

I find that both sides mostly use a lot of rationalizing from emotionally chosen viewpoints.

However, I find that - just like the "bits are bits" of the 90's - nowadays it's the "objective" fans that have a cartoonish understanding of technology.

The Wikipedia oversimplifications give people the feeling that they understand how it works.  They then believe incredibly silly things like "we have measurements that correspond to all sound differences".

The big mistake is equating advanced technology with complete understanding.

For example, despite the existence of computer-aided robot surgery, nevertheless only 25% of doctors' diagnostic techniques are based on science.

[kstuart]

I only mentioned the reported difference between real-time and off-line conversion because a) someone asked, and b) if you are an "objective anti-audiophile" person, then you certainly don't believe there is any difference between playing the file as PCM or playing the file as DSD...

[6233638]

I only mentioned the reported difference between real-time and off-line conversion because a) someone asked, and b) if you are an "objective anti-audiophile" person, then you certainly don't believe there is any difference between playing the file as PCM or playing the file as DSD...

Objective is not "anti-audiophile" - it just happens that the objective answer does not align with the audiophile's beliefs a lot of the time.

I personally can't say whether DSD sounds different/better from PCM, as I don't have DSD hardware here. I do know that DSD adds a lot of ultrasonic noise to the signal that needs to be filtered out though, and this is often not done.

Here's an example of a file that was clearly DSD at some stage, then sold as 352.8kHz PCM:

That could certainly sound different to "regular" 16/44, or 24/48 PCM, due to distortion introduced from this ultrasonic noise.

There are a couple of reasons why DSD could sound better in theory (if the DAC is natively a 1-bit device for example) but finding DSD audio that has never been PCM at some point is difficult (you can't edit DSD) and once it's been converted to PCM, it's better staying as PCM. I hear that many engineers prefer DXD (24-bit, 352.8kHz PCM) to DSD these days.

Arguments like "computers were never designed for realtime applications, therefore they mess up the timing of the signal" have been proven wrong many times over, and there is no evidence (i.e. proper double-blind testing) to suggest that realtime vs offline decoding (e.g. FLAC vs WAV) makes any difference.

[Blaine78]

Arguments like "computers were never designed for realtime applications, therefore they mess up the timing of the signal" have been proven wrong many times over, and there is no evidence (i.e. proper double-blind testing) to suggest that realtime vs offline decoding (e.g. FLAC vs WAV) makes any difference.

this statement too is unfounded. there are others that can tell audible differences between processing on / offline.

[6233638]

this statement too is unfounded. there are others that can tell audible differences between processing on / offline.

And yet it is somehow unmeasurable on equipment that can go beyond the limits of human hearing, and no-one seems capable of doing it in a controlled study.

[JimH]

As this has strayed off topic, I'm locking it.