Denon DCD-SA1 Review (CD/SACD Player)

[mocenigo]

HiRes PCM 24-bit / 96Khz files I bought from HDTracks.

They sound clearly better than CD with more "air" and depth similar yet different from DSD.

Perform the following test:
1) with sox, ffmpeg, or other good software, downsample a 24/96 track to 16/44,1 — or, even better, a DSD track to 24/96 and to 16/44,1
2) perform a double bling listening test.

Let us know if you can actually hear differences.

 

[mocenigo]

High Res is usually not worse sounding that 44.1kHz, but in some cases I actually prefer the 44.1kHz.
The mastering techniques have more impact on the final sound than whether its 192kHz or 44.1kHz.

A LOT of hires is just upsampled, and I have observed several artifacts in hires sold as such. Like images “appended” to multiples of the original frequency, or inordinate amount of out of band noise, or just nothing above 44.1 or 48. Many of these cases are outright fraud. I can post some spectrograms (on vacation at the moment).

In fact, a sound engineer told me to downsample everything to 44,1 or 48 for optimal sound quality…

 

[bigpilot]

It is probably confirmation bias (the frequency is so high!) or an artifact of the particular DAC used.

Especially with older devices that had sub-par PCM conversion, direct DSD would have been superior because of the simpler decoding path.

Anyone that can't hear the difference between CD and SACD needs to get their hearing examined. Tests have shown that most humans can hear a clear difference between PCM (16-bit/44.1Khz) and DSD.

I haven't (yet) tested a conversion from DSD to HiRes PCM.

 

[krabapple]

The best way to test is to switch between the CD and DSD layer on dual-layer discs. I wouldn't know how to test between pure DSD and HiRes PCM since the formats are difficult to transcode.

I hope you realize by now that this is really not the way to demonstrate the sonic superiority of DSD.

And transcoding is easy. Really. DSD to 88.1 is common.

 

[krabapple]

Anyone that can't hear the difference between CD and SACD needs to get their hearing examined.

LOL. OK you went there.

Tests have shown that most humans can hear a clear difference between PCM (16-bit/44.1Khz) and DSD.

Which tests? And what about 'hi rez' PCM?

I haven't (yet) tested a conversion from DSD to HiRes PCM.

You haven't even properly tested the same mastering presented as DSD and PCM. Which is crucial.

 

[krabapple]

The best way to test is to switch between the CD and DSD layer on dual-layer discs. I wouldn't know how to test between pure DSD and HiRes PCM since the formats are difficult to transcode.

You could do that though it's rather slow, and difficult to do 'blind' . And you'll have to make sure the output levels are very carefully matched.
Assuming of course the mastering on the two layers is the same -- which it usually is, but isn't always. Famous case in point is the best-selling SACD of all time: Pink FLoyd's Dark Side of the Moon.


Anyway: there is zero good scientific evidence that DSD and PCM are routinely audibly different. Nothing to support a claim that hearing no difference means you 'need to get your hearing checked', i.e. there is some obvious difference available to anyone with good hearing. That is a complete myth, as far too many audiophile 'difference is obvious' claims are. . It doesn't belong anywhere on a place calling itself Audio Science Review.

 

[mocenigo]

Anyone that can't hear the difference between CD and SACD needs to get their hearing examined. Tests have shown that most humans can hear a clear difference between PCM (16-bit/44.1Khz) and DSD.

I haven't (yet) tested a conversion from DSD to HiRes PCM.

Provide the proper bibliographic references of peer reviewed studies.

 

[krabapple]

I hear a cavalry of dead horses advancing over the horizon...again.

 

[Kremmen]

Provide the proper bibliographic references of peer reviewed studies.

This information should be useful:

Audio Engineering Society​

A Meta-Analysis of High Resolution Audio Perceptual Evaluation​

Over the last decade, there has been considerable debate over the benefits of recording and rendering high resolution audio beyond standard CD quality audio. This research involved a systematic review and meta-analysis (combining the results of numerous independent studies) to assess the ability of test subjects to perceive a difference between high resolution and standard (16 bit, 44.1 or 48 kHz) audio. Eighteen published experiments for which sufficient data could be obtained were included, providing a meta-analysis that combined over 400 participants in more than 12,500 trials. Results showed a small but statistically significant ability of test subjects to discriminate high resolution content, and this effect increased dramatically when test subjects received extensive training. This result was verified by a sensitivity analysis exploring different choices for the chosen studies and different analysis approaches. Potential biases in studies, effect of test methodology, experimental design, and choice of stimuli were also investigated. The overall conclusion is that the perceived fidelity of an audio recording and playback chain can be affected by operating beyond conventional resolution.

Author (s): Reiss, Joshua D.;
Affiliation: Queen Mary University of London, London, UK (See document for exact affiliation information.)
Publication Date: 2016-06-06

The 16-page downloadable pdf from the following webpage should help to clarify the paragraph above.

E-library page - AES

AES E-Library ← Back to search

aes.org

 

[krabapple]

Ah yes, charging onto the scene right on schedule: Reiss and his 2016 meta-analysis. Certainly one of the horsemen in the cavalry. (Oohashi riding shotgun.)

1) it is about CD rate vs 'hi rez', not DSD vs PCM

2) consider that meta-analysis is a method that depends crucially on what published paper data are admitted and what are not.

3) if you need a meta-analysis to unearth a 'small but statistically significant' effect, it's pretty, er, well-hidden

Which is to say, its controversial findings don't support for the claim that the difference between CD rate and hi Rez (much less CD vs DSD) is obvious enough that those who can't hear it need their hearing checked.

Since we're revisiting golden oldies from 2016, here's an ASR thread about DSD vs PCM...from 2016.

And a serious ASR post about Reiss (2016) from 2020

Everything old is new again!

 

[Kremmen]

Ah yes, charging onto the scene right on schedule: Reiss and his 2016 meta-analysis. Certainly one of the horsemen in the cavalry. (Oohashi riding shotgun.)

1) it is about CD rate vs 'hi rez', not DSD vs PCM

2) consider that meta-analysis is a method that depends crucially on what published paper data are admitted and what are not.

3) if you need a meta-analysis to unearth a 'small but statistically significant' effect, it's pretty, er, well-hidden

Which is to say, its controversial findings don't support for the claim that the difference between CD rate and hi Rez (much less CD vs DSD) is obvious enough that those who can't hear it need their hearing checked.

Since we're revisiting golden oldies from 2016, here's an ASR thread ABOUT THE SAME TOPIC ...from 2016.
I note several of the subjectivist-friendly 'but still I hear it' types ('you guys demanding science on ASR are so mean and predictible!!') on that thread have fallen silent on ASR since then.

I thought I was helping out!

 

[krabapple]

More on Reiss (2016) , this time from Archimago's site in 2016.
Ahh, this really brings me back!

 

[NTTY]

And a serious ASR post about Reiss (2016) from 2020

More on Reiss (2016) , this time from Archimago's site in 2016.
Ahh, this really brings me back!

Fantastic readings, thanks!

 

[mocenigo]

This information should be useful:

Audio Engineering Society​

A Meta-Analysis of High Resolution Audio Perceptual Evaluation​

Thank you. While this is definitely an important (and well known) data point, besides the fact that it is based on studies which make significant mistakes (like downsampling without proper dithering), it also misses one crucial aspect: whether any audible differences may be consequence of the DACs themselves behaving in different ways with different formats.

Up to about 10 years ago, I could hear differences between different formats (and reconstruction filters) on the same DAC. Even downsampling hires and comparing the original with the downsampled version. On the AKM side this stopped way before the first 4497 and on the ESS side with the 9018, IIRC. There are two possibilities: either they have improved their conversion on the “lo-res” side or they have made their conversion of hi-res worse. I would humbly submit that the second is unlikely.

During the last 10-15 years DAC chips have improved tremendously, and internally they can compute much more precise reconstruction filters than earlier, more computationally limited chips. This is also the reason for the advanced filter in players like HQplayer, which today make no more sense except for bragging rights (like having a DAC with a SINAD of 123.8 instead of 123.6): once, there was a need for that. To put this in a better perspective, almost all DACs apply some form of oversampling, and there are different levels of approximation involved when upsampling a 44,1kHz and a 96kHz signal.

And the metastudy by necessity can only cover studies that are 10+ years old.

There are some aspects that *might* be improved, but they are outside the conversion circuitry itself and do not depend on the resolution of the source material. I still think that hysteresis effects due to anything ferromagnetic around signal conductors are understudied.

Also, presence of ultrasonics HAS been correlated with higher enjoyment of music even on people with frequency limited hearing. It has not been proven that it must be correlated with the original signal and it is perception via bones or eyes (yes, well...). If proven it would justifiy adding spurious frequencies but not the use of a more voluminous transport format.

 

[krabapple]

Thank you. While this is definitely an important (and well known) data point, besides the fact that it is based on studies which make significant mistakes (like downsampling without proper dithering), it also misses one crucial aspect: whether any audible differences may be consequence of the DACs themselves behaving in different ways with different formats.

Up to about 10 years ago, I could hear differences between different formats (and reconstruction filters) on the same DAC. Even downsampling hires and comparing the original with the downsampled version. On the AKM side this stopped way before the first 4497 and on the ESS side with the 9018, IIRC. There are two possibilities: either they have improved their conversion on the “lo-res” side or they have made their conversion of hi-res worse. I would humbly submit that the second is unlikely.

Another guy who says he could routinely hear differences between DACs. Amazing.

There are some aspects that *might* be improved, but they are outside the conversion circuitry itself and do not depend on the resolution of the source material. I still think that hysteresis effects due to anything ferromagnetic around signal conductors are understudied.

Also, presence of ultrasonics HAS been correlated with higher enjoyment of music even on people with frequency limited hearing. It has not been proven that it must be correlated with the original signal and it is perception via bones or eyes (yes). If proven it would justifiy adding spurious frequencies but not the use of a more voluminous transport format.

And cites Oohashi.

A week of surprises on ASR.

 

[NTTY]

I bought 20 SACD this week to enjoy playing them with the big Denon.

…

I can’t tell the difference between the SACD and PCM layers of each, and I was amazed by the dynamic range, this is so good.

Fun fact is that the Denon will ALWAYS start from the PCM layer (why????) and that’s how I discovered my pleasure is equal between the two.

When the master is well done, CDA can deliver.

PS: don’t trust my old ears, test yourself.

 

[mocenigo]

Another guy who says he could routinely hear differences between DACs. Amazing.

Another person with heavy reading comprehension.
Do you think that DACs have *always* been perfect? I spoke of DACs designed *at least* 10 years ago, maybe significantly more.
I have had a Gustard X20, where the differences between the filters was very subtle (I could not care less) but identifiable–and I was younger, hence better ears.
Where did I say that DACs are not a solved problem now?

And cites Oohashi.

Well, yes. Unless you have a study that refutes it, it is peer-reviewed research that should be taken seriously or confuted with arguments.
And I did not say I agree with the conclusions. I have said "has been correlated" and "If proved"...

So, yeah, you do not understand English.

A week of surprises on ASR.

I am not surprised. Most people do not understand what they read In fact, I was careful to imply that, even if were all true, it does not imply that there are intrinsic differences between formats if properly converted.

 

[NTTY]

So, yeah, you do not understand English.

Not my conversation, but hey, seriously?

 

[mocenigo]

Not my conversation, but hey, seriously?

Yes, they left out 90% of what I have written, commented on a couple of sentences out of context, and made fun of an articulated discourse instead of arguing it. I prefer to attribute this to lack of understanding of written text rather to lack of intellect or lack of intellectual honesty. Anyway I am blocking them and if the moderators want to ban me they are free to do. I was once banned by diyaudio for pointing out the intellectual dishonesty of a forum donor, so I don’t care.

 

[mocenigo]

I bought 20 SACD this week to enjoy playing them with the big Denon.

…

I can’t tell the difference between the SACD and PCM layers of each, and I was amazed by the dynamic range, this is so good.

For many years my CD/SACD player has been a Marantz SA-7001 KI with an upgraded clock (Tentlabs XO), with its own power supply. Great player. But I could never the difference between the two layers.

Fun fact is that the Denon will ALWAYS start from the PCM layer (why????) and that’s how I discovered my pleasure is equal between the two.

When the master is well done, CDA can deliver.

PS: don’t trust my old ears, test yourself.