which quality USB cable to consider?

[mansr]

I don't recall ever buying a USB cable. Nevertheless, I have a crate full of them.

[mansr]

2 hours ago, kilroy said:

These differences are obvious and easy to hear on any sufficiently resolving system if you know what to listen for.

 

It is sometimes also the case that differences are not always immediately discernible, protracted listening and comparison (days or even weeks) is sometimes required.

How can something be "obvious and easy to hear" while also being "not immediately discernible"? You'd be more (though still not particularly) believable if you stuck to one of those.

[mansr]

5 minutes ago, CuteStudio said:

I understand you feel there are two classes of listener, the incapable, lower class with cloth ears who can't hear anything and those superior listeners who can tell - which would be you.

It's funny how the same gene that gives you golden ears also makes you gullible beyond belief.

[mansr]

Just now, Ralf11 said:

You guys are torquing pretty hard on his religion.  Let's all hope he is not an Audio Jihadist.

It's fine. I have asbestos underwear in case of flame wars.

[mansr]

13 minutes ago, CuteStudio said:

I did find the platinum wire good in the garden though, it doesn't go rusty so it's good for the runner bean poles, so it's not been a total waste.

Do the beans taste better?

[mansr]

6 minutes ago, CuteStudio said:

I'm not a fan of DSD for the reasons stated in this article:

http://positive-feedback.com/Issue66/dsd.htm

The author of that article has severely misunderstood how DSD and digital systems in general work.

[mansr]

1 hour ago, CuteStudio said:

The author is Lynn Olson!

Is that supposed to mean something?

 

1 hour ago, CuteStudio said:

I'm interested in your assertion, could you explain what he has got wrong?

Too much to detail. If he genuinely understands the subject, he's done an abysmal job explaining it, or he's being deliberately misleading. Either way, it's a terrible article.

[mansr]

6 minutes ago, CuteStudio said:

Yes. Lynn Olson

http://nutshellhifi.com/

If you do nothing else this week: read his site. Decades of accumulated knowledge and experience.

More knowledge about HiFi, audio engineering and sound than almost anyone.

I've seen that site and had another look just now. There's not a word about anything digital there.

[mansr]

1 hour ago, CuteStudio said:

4) You still haven't specified any fault with the article he wrote about digital music and DACs.

 

http://positive-feedback.com/Issue66/dsd.htm

 

Please point out an error in the article. Is the graph wrong? Is the maths wrong?

Fine, I'll give a few examples.

 

Quote

I understand that noise-shapers are used at the encode and decode end, but what of the system as a whole? There's no way to wrap feedback around a recording made in 2001 and played back in 2013, unless we invent time travel.

Looking at the bitstream coming out of the encode ADC/noise-shaper, and the bitstream going into the decode DAC/noise-shaper, I see a string of DSD samples at the 64fs rate.

A sigma-delta ADC has a noise shaper. A DSD DAC (or "decoder" as he calls it) does not. That's a fundamental misunderstanding right there, and we've barely made it past the introduction. The talk about feedback around the whole record/playback chain makes no sense at all.

 

Quote

I have a question: what is the smallest signal this system can represent at 3.675kHz, and what would it look like in the DSD bitstream coming out of the ADC? I've picked 3.675kHz since it is close to the peak sensitivity of the ear, and a convenient divisor of 44.1kHz.

The smallest possible 3.675kHz signal would be nothing more than a single "one" that replaces a zero in the following string:

… 0101 0111 0101 0101 …

All the rest of the string is the usual tedious … 0101 0101 … pattern mentioned above. Except for that lone "one", all the rest come out as analog zero. How often does the rogue "one" appear? That's easy: 3,675 times a second. (Yes, I know that creates a small DC offset, but let's ignore that for now. Besides, practical DACs don't usually pass DC anyway, to protect the power amplifier in case the recording has an accidental DC offset.)

S/N ratio? Again, easy. The DSD samples are clicking by at a rate of 2.8224MHz, and we're interested in our one little guy at its 3.675kHz rate. Our "one" appears once every 768 DSD samples. The smallest possible modulation at 3.675kHz is one part in 768, or 9.5-bits, or -57dB below full-scale modulation.

This is not how it works. It just isn't. With sigma-delta modulation, you get a much better SNR, as is easily demonstrated. It doesn't really matter if the author actually believes in this naive presentation or is intentionally misrepresenting DSD in order to discredit it. Either way, he can't be trusted.

 

Quote

In the absence of noise-shaping, the resolution of DSD is a function of frequency; the higher you go, the less there is. At what frequency do we get 16-bit resolution? Well, 16-bits represents 65,536 possible levels; let's divide 2.8MHz by 65,536, and we get 43Hz. With every octave above 43Hz, the resolution drops by another bit, topping out at 1 bit 2.8MHz.

This is nonsensical. Without noise shaping you'd have 1 bit (~6 dB) of dynamic range, end of story. Where he got the idea of dividing the sample rate by some resolution figure I have no idea.

 

Quote

The most likely source of coloration is on the DSD end; when DSD is converted to (very) high-resolution DXD/PCM, noise-shaping algorithms are required in the conversion process.

Wrong. Converting DSD to PCM does not involve noise shaping. Noise shaping is used when reducing the sample precision (bit depth), not when increasing it.

 

Quote

DSD-Wide (2.8MHz/8-bit) is interesting professional format; unlike DSD-Narrow, it doesn't require noise-shaping, thanks to abundant dynamic range.

More nonsense. 8 bits does not provide "abundant" dynamic range unless noise shaping is used.

 

I could go on, but I have better things to do. It seems clear that the author, for whatever reason, is on a mission to discredit all things digital, and DSD in particular, based on a confused and rudimentary understanding of the processes involved.

[mansr]

1 hour ago, CuteStudio said:

Reading articles (e.g. this: http://www.mojo-audio.com/blog/dsd-vs-pcm-myth-vs-truth/ and the referenced articles) does indicate it's not been a particularly smooth path given the need for PCM for anything other than backing up tapes.

You sure know how to pick them. That site is a real hoot.