Sorry I can't find it either, it might only exist as a historical snippet. It was played for me >40yr. ago by a music professor from a Boston university. The surviving wartime recordings are different.
You keep saying this even though this was demonstrated and presented as an oversight from almost day 1 and accepted as a mistake in the standards decades ago
George: The Dutton version is much cleaner. Cedar really does a good job on noise removal (which it ought to for the money)! I Can send you a copy.
I'll listen again to both and decide which I prefer. I wasn't just referring to time, but the initial thing that leaps out is that it a lot darker at the slightly reduced pace before you start digging further in. Rachmaninov may of course had to go faster than he wanted to in order to fit on disc!
I'll listen again to both and decide which I prefer. I wasn't just referring to time, but the initial thing that leaps out is that it a lot darker at the slightly reduced pace before you start digging further in. Rachmaninov may of course had to go faster than he wanted to in order to fit on disc!
In a recent project the amplifiers used were class D switching amplifiers. They packaged 2 stereo amplifiers in one box. Thus there were two clock frequencies. The clock noise on the loudspeaker cables ran around 15 volts P-P.
The isolation between twisted pairs of loudspeaker cable is -46 dB/100M. The run was 165M. The maximum amplifier voltage was 224 volts P-P. The difference frequency of the clocks was 10,000 hertz.
With conversion efficiencies that would produce a heterodyne noise 81 dB down from the peak level of 115 dB. Don't you think folks would complain about a constant whistle at 34 dB?
It would take noise of around 60 dB ish to mask that.
Music has dynamic range artifacts produced at maximum level may hang around until it is no longer maximum level and become an issue.
The hang around time does not just come from reverb it also comes from issues such as driver resonances.
So a static measurement often does not reflect perception.
(BTY Derfy you might have missed meeting JJ by not visiting my semi local project!)
Last edited:
I kept saying it because that has been my understanding. If it is incorrect, then I will stop saying it.
However, I can hear it. And I can hear the difference between pristine hi-res source material and 16/44 files made from it, with or without dither. Now, I'm not saying I can hear it on all systems, or with all source material because I don't know about that.
I do know on my system I can hear it and so can other people. And I can also say that the difference is pretty small. Most people wouldn't care even if they could hear it if they listened carefully enough.
I can also say that hearing small differences like that is easier using the DAC-3 than the DAC-1. The DAC-3 is not hugely better, but it is a little more clear and detailed.
So, when I have said that undithered 16-bit audio is audible, I was also saying it from the perspective of having heard it myself.
The only issue as I see it is whether it remains controversial or not in terms of verification by academic research. Since no one wants to pay for research now that we have better test equipment such as improved DACs, I guess we will just have to continue arguing about it from time to time, or maybe agree to stop talking about it altogether.
But Mark, if you are presented with a 24/96 file with no provenance can you tell if it was recorded and mixed at 24/96 or if it was upsampled after the event? This to me is the key question as to whether its worth the bother.
What to me would be interesting is if you could eq you headphones to be the same as your speakers and see if the listening results transfer. The FR deviations that your speakers have may conspire to artificially boost midrange and make certain things easier to detect. Certainly of interest to know.
What to me would be interesting is if you could eq you headphones to be the same as your speakers and see if the listening results transfer. The FR deviations that your speakers have may conspire to artificially boost midrange and make certain things easier to detect. Certainly of interest to know.
Ed at times you seem to be fighting a private battle dealing with things that most of us don't. I'm pretty sure your description of your listening environment puts you at least 3 sigma beyond the norm. You've listed those three or four references numerous times but the way you combine them I have only seen from you. I would like to see some buy in from industry experts (in print) that 26/500k is it.
Once you make Gerhard's A/D and build a custom interface to stream and store the data (I can give you some names at LHC) you still need a DAC, amplifiers, and speakers. But in the end you are up against a sleight of hands like MQA.
I'm not saying it's incorrect, just that no one does it (at least not knowingly). A professor at MIT showed me on commercial CD's maybe when CD's were a year or two old. The industry got it and maybe by year 5 or 6 they were no more. The process of dither in digital systems was well known before CD's.
Furtwaengler/Bethoven 9th
What is on youtube now :
May 1, 1937 London
YouTube
March 22-24, 1942 Berlin
YouTube
YouTube
April 19, 1942 Berlin
YouTube
July 29, 1951 Bayreuth
YouTube
https://www.youtube.com/watch?v=VhyNCnqtmV4
1951 Wien
https://www.youtube.com/watch?v=41aLtyd2NOk
March 1952 Wien
https://www.youtube.com/watch?v=47QKizv6L2s
1953 Wien
https://www.youtube.com/watch?v=jl4mW2Mc9is
1954 Lucern
https://www.youtube.com/watch?v=32LMghURtKI
1954 Bayreuth
https://www.youtube.com/watch?v=DR80Mo0rXWA
-------------------------------------
The first post-war Ninth (1949 Vienna Philharmonic) and two Berlin Philharmonic performances recordings (1950 & Sep 5, 1951) have not survived .
There are also an Aug 7 and an Aug 31, 1951 Salzburg recordings
I have the Aug 31 on CD. Maybe it’s the one Scott mentions?
George
What is on youtube now :
May 1, 1937 London
YouTube
March 22-24, 1942 Berlin
YouTube
YouTube
April 19, 1942 Berlin
YouTube
July 29, 1951 Bayreuth
YouTube
https://www.youtube.com/watch?v=VhyNCnqtmV4
1951 Wien
https://www.youtube.com/watch?v=41aLtyd2NOk
March 1952 Wien
https://www.youtube.com/watch?v=47QKizv6L2s
1953 Wien
https://www.youtube.com/watch?v=jl4mW2Mc9is
1954 Lucern
https://www.youtube.com/watch?v=32LMghURtKI
1954 Bayreuth
https://www.youtube.com/watch?v=DR80Mo0rXWA
-------------------------------------
The first post-war Ninth (1949 Vienna Philharmonic) and two Berlin Philharmonic performances recordings (1950 & Sep 5, 1951) have not survived .
There are also an Aug 7 and an Aug 31, 1951 Salzburg recordings
I have the Aug 31 on CD. Maybe it’s the one Scott mentions?
George
Attachments
Attachments
So to prove you are incorrect, one would have to prove that you (and perhaps others) are not hearing it.
Since it is impossible to prove a negative, you are safe in keeping saying it. Nice try for a classical logical fallacy.
Scott: I mention 16-bit without dither audibility not to say there is currently a problem with most commercial CDs, but rather in response to people who say that there is no way anybody could possibly hear anything down at the 16th bit. The reason I mention dither is because it was a problem and it is accepted that people could hear quantizing noise down at that level.
I understand that it just might be uniquely noticeable, but I don't think so. My personal opinion is that some people can also hear very small distortion of certain types also down at very low levels. However, that remains unproven in the research literature so far as I am aware. All we really know for sure is about dither or it's absence at 16-bits, so I mention it so that I will not be accused of imagining things that are unproven. That's all.
There is of course the whole other issue of who cares if there is a little low level distortion almost nobody notices? It seems that a few people are attuned to microscopically focus on little tiny distortion artifacts and it apparently bothers some of them a lot. When people focus that way a little thing may seem like it is much larger than it really is. But the displeasure associated with it can be quite real. To such people it may matter if they have ultra low distortion systems.
Unfortunately, since we don't have good research in this area, it's very easy for people who don't hear or focus on little imperfections to be in disbelief that hearing such things is possible. To greatly complicate the situation, people can easily imagine they hear things that really aren't there. However, what people usually imagine is not very small distortion. It's much easier to imagine things that are harder to pin down, like depth for example. What's that? How do you measure it? At least we know how to measure distortion pretty well. And I suspect claims of being able to hear it are more likely to be true than hearing something called depth. Of course, that is speculation on my part only.
One thing about cross-over distortion and music I would like to mention, with a test sine wave zero crossings are equally spaced in time, so cross-over distortion might be expected to sound like buzzing. In real music, zero crossings may be very irregularly spaced, even for one solo instrument. Zero crossings may even be pretty uncorrelated with musical notes, since note harmonics, beat notes, changing notes, and multiple instruments playing at once all affect zero cross timing. In that case, the distortion may be less audible to listeners focused mostly on the music and not so much on little tiny distortion artifacts that other's may focus on a lot. In such cases, there may also be disagreement about audible distortion due the nature of focusing as described in this paragraph.
Bill: I though the reason I didn't hear distortion as well in my headphones might have been because the DAC-1 headphone amp wasn't nearly as clean and low-noise as the line outputs. I will have to try again with the DAC-3 which has a better headphone amp, but still not as good as the line outputs.
I understand that it just might be uniquely noticeable, but I don't think so. My personal opinion is that some people can also hear very small distortion of certain types also down at very low levels. However, that remains unproven in the research literature so far as I am aware. All we really know for sure is about dither or it's absence at 16-bits, so I mention it so that I will not be accused of imagining things that are unproven. That's all.
There is of course the whole other issue of who cares if there is a little low level distortion almost nobody notices? It seems that a few people are attuned to microscopically focus on little tiny distortion artifacts and it apparently bothers some of them a lot. When people focus that way a little thing may seem like it is much larger than it really is. But the displeasure associated with it can be quite real. To such people it may matter if they have ultra low distortion systems.
Unfortunately, since we don't have good research in this area, it's very easy for people who don't hear or focus on little imperfections to be in disbelief that hearing such things is possible. To greatly complicate the situation, people can easily imagine they hear things that really aren't there. However, what people usually imagine is not very small distortion. It's much easier to imagine things that are harder to pin down, like depth for example. What's that? How do you measure it? At least we know how to measure distortion pretty well. And I suspect claims of being able to hear it are more likely to be true than hearing something called depth. Of course, that is speculation on my part only.
One thing about cross-over distortion and music I would like to mention, with a test sine wave zero crossings are equally spaced in time, so cross-over distortion might be expected to sound like buzzing. In real music, zero crossings may be very irregularly spaced, even for one solo instrument. Zero crossings may even be pretty uncorrelated with musical notes, since note harmonics, beat notes, changing notes, and multiple instruments playing at once all affect zero cross timing. In that case, the distortion may be less audible to listeners focused mostly on the music and not so much on little tiny distortion artifacts that other's may focus on a lot. In such cases, there may also be disagreement about audible distortion due the nature of focusing as described in this paragraph.
Bill: I though the reason I didn't hear distortion as well in my headphones might have been because the DAC-1 headphone amp wasn't nearly as clean and low-noise as the line outputs. I will have to try again with the DAC-3 which has a better headphone amp, but still not as good as the line outputs.
- Status
- Not open for further replies.