ANOTHER Example of Why I HATE DSD and Why Customers Who Bought Sony's Boloney Are So Annoying

[Miska]

Dont both CD5XS and CDX2 CD players use the Burr-Brown 1704 also ?

 

Yes, at least CD5XS seems to be. I've been mostly checking DAC-V1 (PCM1791A), ND5XS (PCM1791A), NDX (PCM1791A), UnitiLite (PCM1793), NaimUniti2 (PCM1793) and UnitiQute2 (PCM1793?).

[Miska]

How about extracting 1 bit DSD from the same 6 bit D-S modulator? Is that "perfect"? Isn't worse?

 

That process doesn't require running any filters at all...

[Miska]

Even in DSD, you cannot have better analog resolution, using the same stages. If any, you will have worse THD in bypass mode, because you don't use the multilevel D-S. Using higher DSD modes will NOT bypass the limitations imposed by the analog stages inside DAC's (or ADC's). is just wasted bandwidth.

 

1-bit converter has less THD than multi-bit because it doesn't have non-linearities of multi-bit. DSD is also very bitrate efficient. DSD128 has bitrate of 352.8/16 PCM.

 

Another reason why DSD sucks is lack of mixing ability.

 

Wrong, you can

1) mix in analog domain

2) mix multiple DSD streams

[Miska]

But since there will be no aliasing into the audio band with signals of less that 170 kHz, I don't think this will be a problem. No musical instruments or microphones have output at these frequencies so it is a non-issue.

 

It is issue because:

1) Even those aliases that are above audio band will cause correlated intermodulation products down to audio band.

2) Even your analog stages seems to have quite high THD at ultrasonic frequencies, and mic amps may too have fairly high THD at 20+ kHz. So 100 kHz input will generate THD harmonics at 200, 300, 400, 500 kHz. These THD harmonics will alias back to the Nyquist band due to low stop-band attenuation in your anti-alias filter.

 

Try out 100 kHz at 2V with something like 1% THD and check if 0 - 88.2 kHz stays completely clean down to -144 dB at least.

 

You can check out what instruments produce:

http://www.cco.caltech.edu/~boyk/spectra/spectra.htm

(test your system by putting that Sanken microphone 60 cm from a crash cymbal and start smashing - then you know one example what closei-mic'ed rock would produce)

[Miska]

Well, 1 bit PCM has only 2 levels. Using a RZ encoding, you can add pulses one next to another and voala... you have the DSD. With all that 'nice' DC component.

 

Only the delta-sigma modulation is missing there... So it sounds like white noise.

[Miska]

The question is of the two widely available digital high-resolution formats 192/24 and DSD (64), which sounds better. That's all.

 

192/24 PCM sounds better when played back converted to DSD128.

[Miska]

*AND* a multichannel DAC/streamer (like the Oppo) and leave volume control out completely.. two different products for two very different users. Cover both types of computer audiophiles.

 

No, THAT would be bad option. For analog volume control, there are plenty stereo pre-amps and integrateds with decent analog volume control. However, try to find a pure audio machine (not AV pre-pro) one with 7.1 channel decent analog volume control and more than one analog multichannel input! Not many options, even less if you'd like to have balanced inputs.

 

I spent plenty of time trying to find such thing in the past, and the only reasonably priced I found was

Product Lines > Halo > P 7 Preamplifier

 

I found couple of other devices mainly targeted for studios with $$$ price tags.

[Miska]

Linn

Altough not multi-channel, the akurate dsm seems close to requirements, with both digital and analog inputs, including HDMI, and with a good analog volume control (not sure if good in the Charles's terms, but good to my ears).

So it's no longer one source, it's a multisource option.

 

...I was specifically after something that has two or more 7.1 channel balanced XLR analog inputs. Preferably without any digital inputs or outputs. Just to connect multiple 8-channel DACs, have a source switch and analog volume control. Swapping eight XLR cables manually to switch source is kind of tiresome.

 

Maybe another thing I just have to build myself, but technically kind of boring project so I haven't got myself to do it.

 

This is one of the closest I've found, but I think I don't even want to know the price:

EMM Labs - Switchman MkII Multichannel Controller Switch

Or the audiophile variant:

EMM Labs - Switchman 3 - 6 Channel Preamplifier

[Miska]

However, none of them was as good as taking out the preamp and using a good digital volume control upstream. Other MCH preamp contenders are the Audio Research MP1 (which Kal now owns) and the Conrad Johnson MEP1 (tubes).

 

I was looking for it before I had digital volume control for DSD, I didn't want any PCM conversion. I also found the AR MP1 at the time, but it was out of my budget. And although my Sony SACD player has delay and level adjustments for channel balance (in DSD), it doesn't have proper master volume.

 

I don't have capability to rip my SACDs, so I need to keep using SACD player.

 

Onkyo PR-SC5509 has DSD Direct conversion for HDMI but then it lacks delay and level adjustments, which I have in player software but it doesn't help with my SACDs... So I'm stuck to Sony and analog outputs when it comes SACD. AVR pre/pro's can of course serve as analog pre, but way too much money goes to the video sections I don't need or use at all...

[Miska]

Of course, your software aside, any AVR/prepro that can implement delay and level adjustments with DSD is converting it to PCM in order to do so.

 

That was precisely what I found unacceptable, as well as extra A/D/A conversion. Sony players with analog outputs are the only SACD players I've found that can actually do it in DSD...

[Miska]

I guess by implication, if his Sony SACD player does just that (delay and channel trim), it has to convert to PCM as well, and his believe he has been listening to DSD direct all this time is the mother of all placebo effects.....

 

It has that "e-chip" discussed earlier inside (what ever huge chip with large DSD logo on top). I'm pretty sure of it, because it has three stereo DAC chips inside, and only the one responsible for front channels has PCM pins (I2S) connected, for CD playback. The two other chips have only DSD pins connected. I have verified this both from the PCB and from the schematics.

 

So the player's center, LFE and rear channels physically only capable of DSD.

 

The part I don't like in the player is that it's bass management makes it sound really bad, so that part cannot be used.

[Miska]

Modded Oppo digital out (DSD to PCM conversion) into Trinnov processor

 

I hope this doesn't mean Mediatek's DSD to PCM conversion? That's (IMO) horribly bad quality.

 

Every step represented an improvement, some bigger than other. Some was DSD direct, some DSD converted to PCM.

It is also conceivable that DSD to PCM conversion running through the Trinnov still sounds better than MCH DSD direct (depending on room).

 

I'm now just running all the delay, level and DRC for DSD in software, but that works only for DSF/DSDIFF files.

 

SACD was still kind of unresolved in this context for me. However I just replaced entire speaker system, so now I need DRC only for the subwoofer and no bass management. And I now subwoofer is connected through DSPeaker Anti-Mode 8033S-II so I can easily switch between it's correction and bypass. So things have been solved in a way I'm happy with.

[Miska]

I found actually good presentation about Sony SCD-XA9000ES (IMO, best model they've produced so far) and the difference of "multi-bit" (PCM) and "multi-level" (SDM) conversion as graphs:

Sony SCD-XA9000ES - Printer Friendly version

 

I never bothered to draw such pretty pictures, just trying to explain. Maybe these pictures describe more than thousand words?

 

P.S. Although there's one error in the multi-bit graph, the value for second state there should be 5 and not 10.

[Miska]

Yes it does require filtering...

 

I have code to do it without filtering, could you explain why it would require filtering?

[Miska]

Categorical statements about inherent superiority of one format or conversion method over another are meaningless.

 

OK, I should have said with the DACs I have and DACs that are within my price range of interest.

 

I could be more interested on MSB's 10k$ DAC if it would accept their advertised 3 MHz 24-bit PCM data. But it still costs too much, I could make pretty crazy discrete SDM-DAC myself by purchasing 10k$ worth of electronic components.

 

I'm especially interested on squeezing maximum performance out of 500 - 2000 € range.

[Miska]

2) Performing signal manipulation of DSD signals is inherently more difficult than equivalent operations with PCM (I am willing to bet that Miska will argue this one).

 

No I won't!

It is more challenging! I love challenges, PCM is kind of boringly easy.

 

3c) The sonic "quality" of DSD is strongly related and/or caused by the noise-shaping filter used in the feedback loop of the modulator.

 

Same applies to any SD ADC and DAC, which is 99.9% of current market. Bad side is that it is hard-coded into the hardware. So it is not easy to switch it or update it through software updates. I love possibility to software-upgrade DAC performance and sound quality every couple of months to a even high level!

 

DSD is a vehicle to get past that part in DACs.

 

4) One could (probably correctly) conclude that the best sounding converters are yet to come, but they likely won't be in the form of "all-in-one" chips favored by mass-market manufacturers.

 

That I could agree with too...

[Miska]

All DRC engines work in the PCM domain. Decoding DD/DTS gives you PCM so yes this true, but only partially, because "true DRC" with all codecs that are decoded natively to PCM (including say Flac). DSD needs to be converted to PCM in order to run through a fully capable DSP engine, which is causing some of the fine audiophiles on this forum a lot of heartache, and is even making life unbearable for some.

 

I have two completely separate equally capable DSP engines, one for PCM and one for SDM (DSD)...

[Miska]

I don't know this can be done..and I find it very hard to follow the steps (are there any) in the setup of HQ player as a network player..

 

It is very simple at the HQPlayer side, you just start it in network mode and use it the same way as you would do in local playback mode. Only difference is that instead of choosing local audio endpoint from settings, you select of the network endpoints.

 

Setting up a NAA requires suitable Linux already configured and then installing one single software package there. No configuration needed, although you can change name of the particular NAA from HQPlayer side if you wish.

 

Much simpler than any UPnP and works much better at least for me.

 

NAA and UPnP shouldn't be compared because there's almost nothing in common.

 

Why I don't have UPnP support? Because it's obviously "designed by committee" with over complex protocol, missing required features while getting almost everything possible wrong way.

[Miska]

I anticipated this response. If your software can do in DSD domain what say Trinnov or Dirac can do with PCM you need to talk to a venture capitalist and bring this to market quickly.

 

It has been on the market for quite a while already. And there's no venture capital in Finland. And if I ever have learned anything in my past career it is that I wouldn't let anybody have any control over my stuff.

[Miska]

Actually I was referring to his hardware products, which he has said many times are not for sale and only used for his software development.

 

My hardware is pointless without the software, because it moves as much as possible from hardware to software. Specifically any kind of processing.

 

I've been thinking that maybe I should begin to release my schematics and PCB gerbers under Creative Commons license for DIY people...

[Miska]

Good to hear this . My CS4398+Amanero project needs maybe some help.

(Kysessa on jossa DSD inputtin pääse PCM signaali, onko see vaaralista tai tule vain hirvea sahina? Minulla ei ole täälä hetkella niin sanottu I2S switchia. Uskon etta ensimaiseksin katselun riita 1 kpl muuntimen, jos see on tärkea. Ja anteeksi, en ole koskaan oppinu suomea...)

 

Sounds like a good project. I also have a proto 2xPCM4202 + CS4398 A/D/A with DSD input where I can connect Amanero and other similar interfaces (or just SDIF-3). Originally the device was for recording and comparing 192/24 PCM and DSD128 simultaneous with same analog stages and two ADCs running in parallel with two outputs and monitor DAC that can switch between the two. Board ended up pretty big 12x12" because all components that can be thru-mount are such for better quality components and easy changing.

 

Väärä inputti todennäköisesti kuulostaa aika karmealta, minä käytän CS4398 DSD inputtia erillisten pinnien kautta jolloin DSD ja I2S ei sekoitu niin helposti. MCU ohjaa output mutea, CS4398:n rekistereitä sekä muxia jonka kautta inputit menee. Tällöin moodin vaihto on paremmin hanskassa. Tosin se ei ole niinkään ongelma, koska haluan lähinnä käyttää DACia kiinteästi Direct DSD moodissa 6.1 MHz samplekellolla.

 

Nyt ei puutu kuin aikaa, turhan monta rautaa tulessa.

[Miska]

46 minutes ago, Slaphead said:

I mean is DSD really still a thing? I thought it died ages ago like MQA did.

 

I play everything as DSD to the DAC which operates with bit-perfect conversion to analog... No matter what is the source.

 

How does your DAC operate?

[Miska]

2 hours ago, lucretius said:

What do you mean by bit-perfect conversion to analog?

 

That the data stream goes directly as-is to the actual D/A conversion section without any intermediate DSP. IOW, that the actual D/A conversion section sees the exact same bits that leave the player.

 

[Miska]

30 minutes ago, Ryan Berry said:

Unfortunately, yes.  In a way, it's worse than before as people continue to redefine DSD to give it new life with standards never defined by Sony.

 

Because SDM (DSD is a marketing name for one flavor) is so much more modern than PCM and can be native to many actual D/A conversion sections unlike PCM. Making a good native PCM DAC - R2R ladder is very hard and expensive. And to make such perform well it still needs some similar technologies as already utilized for SDM.

 

16x oversampling filters don't get yet anywhere close to rates needed by modern delta-sigma modulators (> 10 MHz), so rest of the oversampling and delta-sigma modulation is up to tiny DSP on the DAC chip. I'm not so big fan of such all-in-1 solutions...

 

Instead I prefer DAC to be just that - digital to analog converter. Converting input bits to analog bit-perfect, as good as possible, without touching the input data. DSP is better performed elsewhere, outside of the box where sensitive analog signals are.

 

P.S. Quick calculation is that for example my EC modulators perform about 6.5 billion calculations per second per channel for "DSD256" output. This is modulator alone, without 256x oversampling filters... So quite a bit of calculations.

 

[Miska]

27 minutes ago, sandyk said:

Outside of forums like this, and readers of the Hi Fi printed media, most members of the General Public wouldn't even know what DSD was . Even BluRay appears to be struggling these days.

 

Let's keep two things separate. Format on distribution media and what you send to the DAC. There's no reason to assume these two have anything to do with each other.

 

Most distribution media is 44.1/16 PCM and it's portion is not going to change soon in a big way.