Thread: I Hate Direct Stream Digital

Posts: 82
Page: prev 1 2 3 4 5 6 7 8 9 next

Post by Zammo December 23, 2011 (51 of 82)
AmonRa said:

If the ultrasonicl overtones in DSD recordings were higher in level than the noise shaping noise hump, there would be no hump there at all,it would be buried under the signal. Think about that for a second. Then look at the frequency graphs of any SACD.

I showed this post to my dog. He just shook his head and left the room mumbling something about human's intellect being blinded by ultraviolet light or some such.....

Post by DSD December 23, 2011 (52 of 82)
AmonRa the noise hump is always there, that is how DSD works it pushes noise up to the ultrasonic. So that the audible range has the lowest noise and widest dynamic range. And what difference does it make to you since you don't believe in reproducing audio above 20kHz anyway?

Here is more information on the graphs.

"These are all plots using a 40kHz fundamental input frequency, around which you can see the noise floor."

As you can clearly see the 40kHz fundamental frequency reaches to 0dB, this chart of the Sony SCD-X780 clearly shows the second harmonic of 80kHz above the -80dB noise floor at -60dB. Also it is important to note that most SACD players use the optional 50kHz filter and that is another reason the second harmonic 80kHz is so low in level as they start rolling off the high frequencies at 50kHz. In all pictures the 40kHz funtamental is at 0dB.

The Pioneer DV-656A player in the previous graph was to show the difference in noise between SACD and DVD-Audio and that player does use the 50kHz filter so it should not be expected to reproduce the 80kHz second harmonic but as that plot clearly shows the 40kHz fundamental was at 0dB, but even after you seen that twice you still incorrectly claimed that SACD has no frequencies above 30kHz.

There are other players compared as well. http://www.oregonunwired.com/noise.html

Post by korowa December 23, 2011 (53 of 82)
AmonRa said:

That is not a problem if we are analyzing the frequency spectrum, it is true. It is not a ripped signal by the way, but digitized from the analog out. So that is what actually comes out of the SACD player. All in all there is nothing new in that blog article. DSD does have a lot of HF noise.

It was the author, Stephen Dawson, who said he ripped the PCM signal;
"So I ripped the audio out of this as six 24 bit, 96kHz PCM files"
I just found it problematic that one could draw a conclusion about DSD from the PCM signal of a bluray disc.
His blog entry was based on a number of assumptions..

FWIW If his blog entry was to provide a better understanding of the issues.. I feel it had failed. It created more question than answers.
I don't expect anyone to have all of the answers but the holes in this entry illustrate the level of research undertaken.

Post by korowa December 23, 2011 (54 of 82)
AmonRa said:

Instead of ad hominem attacks how about analyzing the DSD signal comparison to waveform I criticized also, and tell were I went wrong there, as I "know nothing"? None of you have been able to successfully contend the facts, only attack me in person. Which clearly shows I am winning.

As I said earlier, I read somewhere (and unfortunately I can not find the article), that, unlike PCM, the DSD process captures the complete analog wave or signal. From this, I can understand why Sony & Philips chose it as an archival tool.
Are you claiming that this is not true?

Also, whilst looking for the article in question; I came across an article, which I found improved my understanding of the DSD process and Hybrid discs.
I'm glad to hear others also found it interesting.

As far I can see, its only you trivialising the article and stating that the information is incorrect.. and its not unreasonable for anyone to request an explanation why it is so.

Post by korowa December 23, 2011 (55 of 82)
Generally speaking, there will always be a subjective nature to all of this. What one can hear (and in many cases) what one doesn't hear. As I said elsewhere, I have come to SACD through my passion with jazz. I only have a modest playback system. Although I'm more than happy with the investment I've made so far. The outlay has definitely improved my listening experience.
(I actually had to import the player)
A friend suggested I get a player that directly converts the DSD signal to analogue. this maybe an issue when assessing SACD as I've been told that many players convert the DSD signal to PCM first then analogue.

Post by AmonRa December 23, 2011 (56 of 82)
DSD said:

the noise hump is always there, that is how DSD works

The noise hump starts from 23 kHz in that graph, and like you can see in the OP graph #1, buries the musical signal from 30 kHz upwards. That is why you can see the hump also when analyzed from a real disk, not just a test tone like here. The hump is much louder than the music signal at those frequencies, so it is false to claim DSD HF extends to 100 kHz. Noise yes, music no. This is my point.

Post by AmonRa December 23, 2011 (57 of 82)
tailspn said:

Instead of trying to educate all of us ignoramuses,

That erroneous picture showing DSD signal and waveform incorrectly aligned and inaccurately explained is from Sony marketing department, made probably by in-house graphic artist. How it escaped the notice of the engineering department escapes me, too. Why is it so hard to admit that is not right, it does not need to be a professional secret. Or maybe: "graphs in the marketing material are misleading, what else is there bending the truth? Better ignore this!"

I have been thinking about getting the Korg, does not cost all that much.

Post by Jann December 24, 2011 (58 of 82)
In the above graph (about the output of a 40khz tone with a Sony SCD-X780), the signal/noise ratio is very good : near 90 db at 40kHz. I think perhaps that in normal music content frequencies of these kind are of lesser amplitude in regard with content around 2kz, for example. Picture should then not be so good. To know it exactly, you have to compare on real music the content before and after DSD encoding.

Subjectively, I have to say that I am very satisfied with many SACD that I have about high frequencies reproduction and I am using supertweeter to extend sound reproduction until 50-60khz.

Post by Fitzcaraldo215 December 24, 2011 (59 of 82)
korowa said:

As I said earlier, I read somewhere (and unfortunately I can not find the article), that, unlike PCM, the DSD process captures the complete analog wave or signal. From this, I can understand why Sony & Philips chose it as an archival tool.
Are you claiming that this is not true?

Not trying to defend anyone or to offend anyone's beliefs, but the notion that "the DSD process captures the COMPLETE (my emphasis) analog wave or signal" is erroneous. No digital process captures all of it. Rather it, by definition, samples the waveform at discrete time intervals for eventual reconstruction as analog signal upon playback. That is true of both DSD and PCM. So, the question remains is anything meaningful or audible, subject to the limits of human hearing, lost or distorted in the process? That debate will, no doubt contine to rage for some time, often pointlessly.

Note that I am not trying to argue the perfection of analog for the capture, storage, transmission or playback of music. To me, hi rez digital is demonstrably superior in that regard. Personally, I find either DSD or PCM comparable to the point that engineering differences in recordings are what are even more significant. Either digital system is capable of delivering a high quality, highly enjoyable replica of the live musical experience. I have no distinct preference.

Post by AmonRa December 24, 2011 (60 of 82)
korowa said:
the DSD process captures the complete analog wave or signal.

Nothing can capture the complete signal in the scientific sense, but the culprit in not necessarily the digital process itself, but the limitations of the analog components in the system.

DSD has better impulse response, but high sample rate 24 bit PCM systems have both higher and flatter high frequency extension (no DSD noise shaping problems) and higher dynamic range both in theory and in practice (best PCM converters have 127 dB DR, best DSD converters 117 dB). So both have their strengths and weaknesses (not to mention real life usability).

"Capturing the complete signal" can be viewed from the practical viewpoint also: can the listener hear if the signal has been digitized with a good converter? No. In that sense the complete signal has been captured, both DSD and PCM are good enough to pass this test. Analog recording methods fall badly short of this goal.

Page: prev 1 2 3 4 5 6 7 8 9 next

Closed