Funniest snake oil theories

Status
Not open for further replies.
Jump to Latest
J

jcx

Member
Joined 2003
another who doesn't understand dither??

noise shaped dither specifically delivers more than the nominal 16 bit 96 dB dynamic range "at midband"

or the performance of US$150 soundcards

347990d1368275595-using-ad844-i-v-juli-.png


shows fine perfonmance in midband of underlying ADC, DAC - the math for proper dither would preserve the distortion performance, A-weighted noise could exceed 110 dB with RedBook
 
Last edited:
T

thoriated

Member
Joined 2003
Originally Posted by thoriated
These plus the brick wall pre-ringing artifacts would seem to be sufficient reasons for CD's woeful high frequency performance.

???? ... you're staring hard in the face of the classic digital distortion that I talk of, to say that. The first time you hear how staggeringly good the treble of Redbook really is, you'll be able to drop the perception that there is a "problem" with the format, rather than within most playback setups, where the real issues lie ...
Click to expand...

Can't say the Redbook high end is 'staggeringly' good. Envelope distortion is another result of pushing hard up against the Nyquist limit and results in 3% or more level instability above 10khz with CD. Some of the most noticeable effects of this is image degradation and coloration.
 
T

thoriated

Member
Joined 2003
I have a test CD, the classic Denon, with classical snippet at full volume, then next track, the same at -20db, then -40 and finally -60dB. The last is so telling, even with the volume pegged at maximum, and putting my ear right next to the speaker driver, it's only just audible. So what does it sound like? Well, it sounds like just like the full volume version, only noisy! Like talking to someone on the phone, and hearing their stereo playing in the background. The content is all there, the instruments sound normal, it doesn't sound a mess! With the old Yamaha, for this -60dB version, I can hear digital glitching during the quiet spaces, but this is subjectively masked once there is some volume in the signal; the recent DACs I've tried this on are dead clean, the noise doesn't sound digital at all even at the quietest points.
Click to expand...

So what? I doubt you could pick out anything more than generic information at the lower levels. I occasionally listen to MP3. It 'sounds like' a digital master, but that doesn't mean they're anywhere close to equivalent in sound quality.
 
T

thoriated

Member
Joined 2003
Dither is capable of producing a modest improvement in LF to midband apparent detail, perhaps from 8 - 15 db, but the more aggressive noise shaped approaches deteriorate the HF starting as low a 2Khz.

More to the point, most CDs don't use it, so no need to assume if intelligent dithering is universally applied.
 
Last edited:
J

jcx

Member
Joined 2003
I believe Meyer/Moran have an open invitation to any who can travel to their labs, demonstrate this ability to hear "noticeable effects of this is image degradation and coloration"

do you have a cite for "most CD don't use dither" - my impression was the opposite - nearly universal for nearly a decade now - available since 1995 at least in HDCD

UV22 used more in regular CD releases

The History of UV22HR

First introduced to the recording community in 1993, the UV22 dither algorithm was first adopted by all of the major mastering studios and record labels and became the industry standard for encoding high resolution, 24-bit audio on to the standard compact disc. UV22 was surveyed to be employed on 8 out of 10 hit CD releases by artists as diverse as Bruce Springsteen, The London Philharmonic Orchestra and Donald Fagen, to name a few. Countless recordings have benefited from this process, allowing mastering studios and record companies to release product on the CD format retaining all of the sonic integrity of the original master mixes.
Click to expand...
 
Last edited:
T

thoriated

Member
Joined 2003
Someone named 'Christopher Hicks' wrote the following:

"Much of the preceding text on quantisation and requantisation noise
assumes the signal to be random. For many signals this is not the
case, and the result is that the quantisation noise, rather than being
white, is found to be highly correlated with the signal. This
manifests itself as very nasty-sounding level-dependent distortions
which become more prominent as the signal is decreased in amplitude."


Considering that none of you deniers can hear any of what Mr. Hicks is discussing, but I can, I have doubts there is any standard of proof that you would accept.
 
D

DF96

R.I.P.
Joined 2007
myweb.tiscali.co.uk/g8hqp/
Is it possible that there is some confusion here between ordinary dither and Bitstream type methods? Dither is like adding some polish to an already fairly smooth surface - it swaps a little distortion for a little noise. The emphasis is on 'little'.

thoriated said:
Someone named 'Christopher Hicks' wrote the following:
Click to expand...
Reference - so we can look at the context of these remarks?
 
Last edited:
M Gregg

M Gregg

Disabled Account
Joined 2010
I don't like music,

I just listen to the HIFI...:D

Dither..

Shall I use polypropylene or paper in sugar water?
Will a piece of fools gold glued to a PCB absorb RFI???
Ahhh but if it has a diamond stylus..will it sound better than a sapphire :D
Hang on we can reflect for a moment...it must be the sap in the wood..
I'll give it a shake...it will sound better...

Regards
M. Gregg
 
Last edited:
SY

SY

Ex-Moderator
Joined 2002
SYclotron.com
I would guess that Mr. Hicks would be highly impressed with the quote mining, not so much with the understanding.

I've posted links a few dozen times to Werner Ogiers's excellent demonstrations. Perhaps you've let those slip out of your mind?

Still no cites, I assume?

Here's one, from Mr. Hicks, ca. 1995:
Addition of suitable dither is eff ective at eliminating quantisation-related distortion, with minimal
degradation of the signal to noise ratio.

from http://www.benchmarkmedia.com/sites/default/files/documents/ditherap.pdf
Click to expand...
 
T

thoriated

Member
Joined 2003
I checked his site, failed to see anything related to what you describe from Ogier - maybe you can provide a more specific reference?

Dither is fine on a CD - properly applied, it obscures objectionable digital artifacts and does improve LF and MF detail - I'd say less than half of the CD's I have have used well applied dithering, by listening to them. However, it in no way improves the CD recording to what is commonly referred to as 'high definition' for either digital or analog recording. Most of my SACD's and DVDA's and even some Blu Ray audio tracks would consign their CD equivalents to a resale shop, quality wise.
 
Last edited:
J

jcx

Member
Joined 2003
also Lukin's dither examples - as low as 8 bit quatization - I can clearly can hear the correlated quantization artifacts with truncation/rounding - before dither is applied - the dither decorrelates the quantization artifacts - rendering them noise like to the ear

Homepage of Alexey Lukin

it obscures objectionable digital artifacts
Click to expand...

is a common misunderstanding or misleading characterization - dither doesn't "obscure" quantization artifacts in the sense of masking/being louder - it destroys the correlation


it is curious that some people want to be so adamantly wrong when the info, examples are easily found to trivially show the opposite is true


most CD musical recordings will be "naturally dithered" by recoding mic, venue background noise - tape hiss does a good job too for anything that started out analog
 
Last edited:
scott wurcer

scott wurcer

R.I.P
Joined 2004
jcx said:
I believe Meyer/Moran have an open invitation to any who can travel to their labs, demonstrate this ability to hear "noticeable effects of this is image degradation and coloration"

do you have a cite for "most CD don't use dither" - my impression was the opposite - nearly universal for nearly a decade now - available since 1995 at least in HDCD

UV22 used more in regular CD releases
Click to expand...

The early test CD's were not dithered, at least my Denon "Super Audio Check" CD is not.
 
SY

SY

Ex-Moderator
Joined 2002
SYclotron.com
Yes, it took the audio world a few years to figure out what was actually best practice once digital got going (it was obvious to the guys in the instrumentation world who had been doing this stuff for years). But that's pretty ancient history- EVERYONE uses dither, at least on any music CD I've seen since the late '80s. It's a standard part of all production software. Don't know about test CDs...
 
Status
Not open for further replies.
Top Bottom