I only became respectable after I "was" an old duffer!
As my former German Professor once told me: "I will begin giving good advice when I can no longer set a bad example."
Best Regards,
TerryO
A Peabody Chromatic Inverter. I don't know. Ask Barry Diament. It's his opinion. I'm sure he'd love to have a debate with you.
Yeah, it resonates, man. 🙂
My favorite part:
FWIW, I never did hate CDs, and even tho I didn't hear a CD until about 1986, I was lucky enough to hear it first on very high end equipment. I never thought it superior to LP in all ways, just in some.
My favorite part:
Very much in line with what I and others have experienced.
FWIW, I never did hate CDs, and even tho I didn't hear a CD until about 1986, I was lucky enough to hear it first on very high end equipment. I never thought it superior to LP in all ways, just in some.
The ADCs I referred too, and occasionally measured, all use massive oversampling. That's how the world works these days. So, even with oversampling true 24 bit performance remains far from trivial. The problem is thermal noise at the front-end.
--
The MSB ADC is interesting, even though obviously a brute-force approach. What's next? Cryogenic cooling?
I would also take their specification with a lump of salt. For starters their SNR figures are A-weighted (giving at least 3dB advantage), and they have a history of rather aggressive marketing literature. Or, putting it in another way, if I realised a true 24 bit convertor I would run to the IEEE and AES journals and conferences.
Last edited:
"Observation" implies scientific or systematic study. As in a doctor saying "We'll keep him in for observation". You can't just imagine something and then say you "observed" it - even though people do it in these forums all the time. As I said a few posts back, simple folk-intuition about how things work means that even a sceptic like me finds it hard to dissociate the physical form of audio equipment from what they're hearing. I once fooled myself with my little CD/Minidisc experiment and so I may be the only person here who has direct experience of how worthless an "observation" can be. Had I not realised my error, I might have come onto these forums and shouted my mouth off about how I had observed the "lack of definition" of Minidisc - and even boasted of the objectivity of my experiment - I know this; that's why I mention it!
What you seem to be saying Abraxalito is that if there is an audio equivalent of the placebo effect, then people shouldn't be denied it. That's fine, except they then come on here and dress it up with pseudo-technical language with such conviction that, slowly and surely, it is becoming a 'given' that vinyl is 'better' than CD. The title of this thread is an example. How do I know that there won't come a time when certain releases are only made on vinyl, or that there will routinely be a 'loud' master for rubbishy old digital and only an 'audiophile' version for vinyl? Will manufacturers abandon the outer reaches of digital because there is no demand for it? All based on folk-intuition that a physical object like a turntable simply must make better sound than a small plastic chip?
(Edit: There may be a genuine lack of definition with Minidisc compared to CD, but the point is that I didn't know, because I hadn't been listening to Minidisc!)
Last edited:
Its only the hard-line objectivists (such as yourself) who are claiming that these people's observations are indeed 'imaginings'. Yet there's no evidence that they are 'imaginings'. As I've pointed out before, in medical science doctors who observe patients getting better by means of placebos do not say 'they're imagining themselves better'. They accept the patient's response to placebo is a real one. So what gives?
Yes - my point about not observing earlier was referring to this - that you had not observed your headphones were not plugged in to the correct socket.
Actually a bit more nuanced - I'm not saying what people should or should not do, just pointing out that denying them this is unreasonable. People are quite free to be unreasonable, but let's call a spade a spade 😀
But we're all free to choose our own meaning for 'better' in this context. Some choose 'better at signal fidelity' and others choose 'better sound'. What's to complain about?
I wonder, although we might agree that high level content is digitised unharmed... has it yet been demonstrated that low level ambient detail that cannot be successfully quantised, is infact missed if it isn't there?
If it cannot be successfully difgitized, it would have to be below -90db level or so. I'm not sure about the exact dynamic range of vinyl, but I believe that the LP recording and pressing process is losing low level detail at a point where the digital recording and pressing process can still include it.
jan didden
How low is 'low level'?
Did you ever listen to the 4-bit (!) music clips I published to demonstrate the effects of naive quantisation, proper dithering, and noise-shaping?
That Dither Thing - [English]
Because they're not choosing 'better sound', they're choosing magical thinking.
Of course I don't have a problem with that as such, but if, every time I am having an interesting discussion about the finer points of audio - perhaps with a bit of a newcomer to the subject - someone pops up and says with great conviction "Neodynium cable spacers are best for opening up the soundstage, and they quicken the bass." I find it very irritating.
As in all walks of life, there are people making large amounts of money by preying on the naive. I hate those TV programmes that expose rogue builders, and dodgy plastic surgeons. And sometimes the customers are happy and don't know there's anything wrong, but I still can't help but feel indignant on their behalf. It is not a feeling I want to be reminded of every time I visit a specialised technical forum!
Edit: And I'm not such a "hard line objectivist" that I am not riddled with self-doubt. A problem I have always had is that I am impressed by people who hold views with strong conviction. To me, everything is complex, multidimensional, and painted in shades of grey. I am not able to dismiss their views out of hand and I think it always takes me some mental effort to filter out the noise.
Last edited:
But the converse is not true, i.e., signals WELL below the quantization noise floor can be resolved easily. The concept of dither is still not well understood by much of the audiophile set, nor the difference between resolution and noise. Werner's demonstration should be a required listen before anyone trots out that million-times-discredited and overly tired duck about losing resolution for signals approaching the LSB.
Drop 'quantisation', as it is not required in this sentence. And drop 'floor', as it leads to misunderstanding.
1) Undithered digital systems should be outlawed. It is moronic to quantise and yet not to dither. Discussing such systems is a waste of time. They have no relevance in our world.
2) Dithered quantised systems become equivalent to analogue systems with a specific noise content. Hence my earlier remark to drop 'quantisation'. Both analogue and digital systems will have their resolution limited by their noise.
The main difference is that the spectral shaping of the noise will differ, compare e.g. magnetic tape, LP, and digital.
3) There is no magic involved in hearing below the noise. In fact, we do not (or hardly) hear below the noise. Standard auditory masking applies (JJ can step in and correct me if necessary). We can hear a spectrally-narrow signal (e.g. a 3kHz sine) as long as it is not being masked by something else. In a quiet channel with just the noise floor only that part of the noise that is spectrally adjacent to our 3kHz tone has masking properties. All the other noise has not. But when we discuss noise and SNR and such, including the fabled 90dB or 96dB of 16 bit, we are discussing all of the noise integrated over the full 20kHz bandwidth. So when we can hear a sine at 120dB below full scale in a channel with 90dB SNR, it appears as if our auditory system can drill down 30dB into the noise. It cannot. Apples and oranges.
There is no difference. Noise is a measure for uncertainty. When the amount of uncertainty in a signal reaches the level where an observer process cannot distinguish the payload signal anymore, the limit of resolution is reached.
For the human auditory system this happens (apologies for simplification) when the level of the payload signal sinks below the level of the noise integrated over a small band that sits around the signal of interest.
Digging deeper requires a-priori knowledge of the signal, or long-term averaging. Both do not apply to music.
Werner, I am intrigued. Your profile summarises your interests thus:
Are you interested in turntables for historical reasons? as the objects of desire they undoubtedly are? Or do you feel they offer some genuine merit in sound reproduction that digital fails to provide..?
Are you interested in turntables for historical reasons? as the objects of desire they undoubtedly are? Or do you feel they offer some genuine merit in sound reproduction that digital fails to provide..?
Panasonic Mash, Sony, Pioneer and other oversampling/ditherings sound different...to my ear some are ok and some drive me out of the room.
Phono preamp noise spectrum shaped dithering anyone ?.
Eric.
It is not required, but it is nonetheless true and addresses directly the beloved and off-tossed Cairina that somehow "detail" is lost in 16 bit systems because of quantization limits, or that the quantization limit somehow causes reverberation or instrument decay to be suddenly cut off.
I agree on all points (as to point 1, has anyone actually ever done this on a commercial recording?) other than the last, but I suspect that our terminology may differ. I'm again addressing the sadly prevalent misbelief that changes in a signal that fall between quantization levels are somehow lost. Generally, this assertion is accompanied by a drawing of a quantized, stair-stepped sine wave with an arrow or two between two of the steps. Lots of pointing to the steps at zero crossing as well. There will then follow a sine wave with some bumps, the bumps falling between the quantized voltage levels and, again, "lost." As your demonstration dramatically shows, the magic of dither makes a digital and analog noise floor equivalent, though the digital system's noise may be much more easily manipulated for best psychoacoustic noise performance.
- Status
- Not open for further replies.