Let me expand a little bit on your comment about digital sound. The Nyquist theorem is mathematical, and although I'm probably not qualified to judge is said to be mathematically proven. No science yet. Then engineers try to build equipment to implement the Nyquist theorem in recording and reproduction, and various ADC's and DAC's are built, along with storage media and methods to transfer data to and from storage. Again, no science yet. Now, we can try to compare the input (from microphones, analog tape, or whatever source you choose) to the digital system with the output; if a study is designed well, this might, for the first time in this process, generate some scientific data. AFAIK, some listeners claim that hi-res digital can be indistinguishable from the source, but most claim it isn't quite there. I have not heard of any well-designed studies actually testing this; if there aren't, then there is no scientific information about the accuracy of digital audio recording/reproduction.
When does art become science?
- Thread starter Gregadd
- Start date
Perhaps that's what you meant. What you said was that DBTs couldn't even tell the difference between violins.
Tim
The birth of digital audio was for communication. It was invented by Bell Labs in 1930s to provide a secure communication channel between US government and that of Europe because the analog system was compromised. That was an invention. And proven to work in life and death situation.
Usage of digital audio for storage and playback of music did not come for some 40 years later. That was also an invention and proven to work.
Let's not confuse our online and print battles of 16/44.1 (i.e. CD) with real history of digital audio that predates it.
Usage of digital audio for storage and playback of music did not come for some 40 years later. That was also an invention and proven to work.
Let's not confuse our online and print battles of 16/44.1 (i.e. CD) with real history of digital audio that predates it.
How consistent were the subjects? Of the fifteen who chose new violins more often than old in Part 1, seven later chose old violins to take home. Against this, five subjects who chose old violins more often in Part 1 later chose new violins to take home (see Supporting Information). By this measure, just nine of 21 were consistent – though this seems unsurprising, given the way preferences shifted as time was spent with individual instruments (see Supporting Information). What was consistent through Parts 1 and 2 was a preference for new violins, and a specific dislike for O1. Abstract Supra.
Yes the could tell the difference .However they were not very good at it.
As advertised ("forever"), the original digital recordings are still available for audiophiles to listen to, and gasp at their imperfections. You could choose the vinyl pressings available at the time (digital masters having been chosen by the labels over the analogue masters made simultaneously), or the digital versions now available on CD. But the weird thing is: you can play them today and they sound fine! I saw a classical music forum where someone was rating an early recording (1976, I think!) as one of his favourite versions of one particular piece.
It is unequivocally true that digital recording has been improved in the meantime, but I am not convinced that humans could hear the difference. There was already an anti-digital campaign going on even before CD arrived, so I think that expectation bias could have played a large part in audiophiles failing to embrace it. Audiophiles never seemed to notice when the evil BBC scientists started feeding their FM transmitters with 13 bit/38 kHz digital audio in the 1970s.
In fact, all of the audio and video we have had access to over the years has been mediated by scientists and engineers, and yet we have most definitely witnessed pure art via it, I would say.
Amir,
Can you point me some information about the invention at Bell Labs in the 30's? I thought that Alec Reeves, the British scientist who invented PCM for voice encoding, was credited as the creator of digital audio in 1937. It is known that Bells Labs used it in the early 40's during World War 2 to encode audio in transatlantic transmissions between the allies, but as far as I remember at that time they ignored the preceding work of Reeves.
In the 80s and 90s, I spent a lot of time in studios recording human voices -- radio spots, TV and film voiceovers. The first time I heard digital playback was stunning. It captured the human voice better than any tape machine I've ever heard. Now, mind you, that was a straight voiceover track. No background, no processing, no mastering. Just straight from the same microphone I'd recorded to tape with a couple of weeks earlier, through the digital recording system, then through the same monitoring system into the same room. It wasn't "CD," but it was early digital.
Of course it could have been expectation bias.
Tim
Sorry, but I think we are in a diversion here. Also, we may be having semantic issues as to what constitutes "science", "applied science", "technology", etc. My point was there would have been no A-bomb at all if there had not been basic theoretical science that predicted nuclear fission was possible and what was necessary to cause it to occur, e.g., specific isotopes of Uranium or Plutonium forced into critical mass.
The bomb devices themselves still had to be engineered to work, which is where applied science or engineering came into play. And, while not all effects of the detonations were known in advance, some things were from theory, such as the predicted blast magnitudes in kiloton equivalents of TNT. But, experiments still had to done to more fully understand this emerging technology and its effects, which was in its infancy. And, like all technology, the understanding of how to make better bombs evolved and improved, whether or not we agree with the nuclear arms race that ensued. Unfortunately, many innocent people were killed in those two early "experiments" over Japan.
Incidentally, early in my career, I was involved in a minor way in the analysis of electromagnetic effects from the last series of US above ground nuclear tests over the Pacific.
Nothing is ever "known" from theory. Einstein's Theory of Relativity took a long time to be proven; many other theories, such as Einstein's Unified Field Theory, not only are never proven true, they are often proven false.
As with a rigged test, you would have to explain how that outcome was forced. And how you can run the blind test differently to generate different results. Mere criticism that something is sales oriented means nothing without that. Every new drug is a sales opportunity. By your definition the blind testing can be cooked to always deliver on that goal. For our own health, let's hope the world of blind testing is not remotely that corrupt
I don't think you want to use the drug industry as your gold standard. To be polite they have a checkered past at best. You don't have to rig a test .Certain methodologies can be relied on to produce certain results. You argue against sighted tssts all the time. time. Is a sighted test rigged, or just flawed methodology in your opinion?
An imperfect past is not the same as checkered. I don't worry when I am taking a prescription drug that I am going to suffer badly from a side effect that came through because the tests were rigged to just make money. I would however if I bought some drug from overseas factory with no oversight as we have in US.
The latter enables the former. And unfortunately that enablement is endemic.
Reeves theorized PCM as a way to encode speech and patented that idea. He did not build such a system. If he had tried, he would have realized that voice compression had to be applied to make it work over radio communication. That is what AT&T did. They combined digitization, voice compression and encryption in an end to end system. So it was the first instantiation of digital audio system. Here is the reference on Wiki on their system, SIGSALY: https://en.wikipedia.org/wiki/SIGSALY
"SIGSALY has been credited with a number of "firsts"; this list is taken from (Bennett, 1983):
The first realization of enciphered telephony
The first quantized speech transmission
The first transmission of speech by pulse code modulation (PCM)
The first use of companded PCM
The first examples of multilevel frequency shift keying (FSK)
The first useful realization of speech bandwidth compression
The first use of FSK - FDM (Frequency Shift Keying-Frequency Division Multiplex) as a viable transmission method over a fading medium
The first use of a multilevel "eye pattern" to adjust the sampling intervals (a new, and important, instrumentation technique)"
That said, this may be one of those things where the history told on each side of the Atlantic varies
.(...) That said, this may be one of those things where the history told on each side of the Atlantic varies
OK, but once someone develops the theory and patents something a few years early before the achievement of others, for me one version gets more credit ...
(...) And though Reeves' extraordinary patent of 1937 showed how this might be done in theory, the valve-based technology of the time was not up to the job. Pulse Code Modulation could not be implement economically until the invention of the transistor decades later. But economy was not always a priority. PCM was first used by Bell Labs for the complex and cumbersome radio system on which Churchill and Roosevelt talked in total secrecy for much of World War 2.
We can read the patent here http://www.google.com/patents/US2272070
From the patent:
"According to the present invention, a signaling system for transmitting complex wave forms, for example speech, wherein the wave form is scanned at the transmitter at predetermined instants, and at these instants signal elements are transmitted to the receiver is characterised in this, that the amplitude range of the wave form to be transmitted is divided into a finite number of predetermined amplitude values according to the degree of fidelity required. The instantaneous amplitude value of the wave form to be transmitted at each predetermined instant being transmitted in a signal code representing the I nearest predetermined amplitude value above or below said instantaneous amplitude value.
The number of elements in the code may equal the number of predetermined values of the amplitude range of the complex wave, or may be equal to' a number such that the total possible number of combinations of the code elements is greater than the minimum required for repro ducing the complex wave form within the desired limits of accuracy.
In accordance with another feature of the invention, the amplitude of the complex wave form is transmitted at a given moment by a predetermined combination of m separate signaling channels, each having n difi'erent signal elements, the resultant possible total number of signal combinations, that is to say n being greater than the minimum number required for the transmission of the information within the desired limits of precision.
In accordance with another feature of the invention, thev combinations of signals are automatically transformed at the receiver into a correct reproduction of the instantaneous amplitude at the moment in question of the wave form to be communicated.
In accordance with another feature of the invention, the number of possible signal elements transmitted in each channel is reduced to a value at which the undesired circuit noise in each of these channels, namely, a peakvalue not exceeding a predetermined limit, is incapable of affecting the occurrence or non-occurrence in the re.- ceiver of correct and definite series of operations associated with the particular combination of signals transmitted at that moment."
We still have many disputes about who was the first in inventions. BTW, do you know what was the first music to be recorded in history?
Last edited:
The former for sure if the facts were on my side.
Let me give you a more direct answer Greg: appearance of bias, is not bias. You cannot dismiss the results of a test because there was a motivation toward that outcome by the people who conducted the test. You must, must demonstrate with real data that the test was not a fair one. The mere implication has no value whatsoever.
If we throw that rule out we should also throw out everyone who says the audio equipment they own is the best. Or any other positive attribute for that matter. After all, they are biased to to defend what they own.
Nowhere is this more apparent than in law. I was once called to be a jury. While in the waiting room, the said the next case was a high-profile case with the person accused of a ton of bad things (I forget the details). Due to that reason, instead of going through 20 or so jurors to pick the 12, they were going through something like 120! Our group was brought in last. I go to the courtroom. I see a ton of lawyers, all dressed up in their suits as always. Then there is the defendant. A black gentleman, dressed like he is walking the dark streets of the city, sitting in his chair with no care in the world and zero respect for the system. If you wanted a case where "guilty was written all over it," this was it. Here is the remarkable part though: not one person on either side, nor the judge seemed to care one bit about that. They were conducting themselves as if they had the most upstanding citizen sitting in that chair. Appearance of guilt had not changed how they went about their work. They presumed innocence and took no action on appearance of facts.
The above gave me hope that heaven forbid, if I am ever sitting in that chair, I would likely receive a fair trial than a lynching based on appearance. Likewise here Greg. I don't care if the person conducted the test had a motive for an outcome. That simply is not material to me. What is material is that you study and understand the test so well that you point out the flaws in it that caused that outcome. Everything else is appealing to lay emotions as opposed to having a discussion of science.
And oh, for nearly a decade, we were in a fight in streaming media that was exactly like coke and pepsi as we were the two leading companies. Showing better fidelity in audio/video compression was the ticket to superiority. They would hire a company to run tests to show they were better. Immediately I would get call from press. I had to give a real reason as to why they were not better. No way could I just say, "well, they are our competitors so of course they did this test to claim they are better." One answer by the way was new data in the form of us hiring another company to run the same test and showing different results because we could discover what they had done to advantage themselves. This is what you do, you fight things with facts.
I can't stop you for judging the book by its cover. But please don't tell me to do the same. It is core to who I am and I will not go there. I hope you understand
.
Well that is a burning question in the law. How do we determine what goes on in the human mind. Or how we determine intent? Generally speaking a person is presumed to intend the natural consequences of his acts. Bias results indicate bias intent. Conflict of interest is a different matter. That is to say that a person may benefit financially or professionally from a particular outcome.
Of course the Pepsi oke challenge indicates a test blind or otherwise may be flawed merely by the way you pose the question. In the former parti in the test would have to chose Coke eight out of ten times to prove Coke superior. In the latter they would only have to fail to pick Pepsi at least eight of ten times.
This is the premise behind double blind tests instead of single. In double blind test the proctor, the person conducting the test, doesn't know which is which so can't cook the results that way.
To be fair, flaws remain in any tests and we need judgement as to whether they were sufficient to invalidate the results.
Are you avoiding prescription medication Greg due to that concern? I am trying to figure out if you are giving me advice that you have taken yourself.
Similar threads
