Home |
Search |
Today's Posts |
#1
Posted to rec.audio.high-end
|
|||
|
|||
AES article: hi-rez more like analog?
Actually it's from a 2007 AES conference, not the JAES, so I'm not sure it's peer
reviewed...but anyway: "Which of the Two Digital Audio Systems Best Matches the Quality of the Analog System?" http://www.hitech-projects.com/hera/...s/aar07pu4.pdf It appears that the formats compared to a live feed (analog) were DXD (353.8 kHz/24) and 44.1/24, both in surround, using a blind comparison protocol. Two additional listening conditions were tested : one where the A/D signal bandwith was 100 kHz (thanks to special microphone 'super-tweeters') and the other where bandwidth was limited to 20kHz. The authors say their results show that listeners 'more often than not' identify the hi-rez audio (and not the 44.1 kHz audio) as being similar in quality to the analog feed...but only when the bandwidth is limited! In other words, only the initial sampling need be done in hi-rez, the listening can (and SHOULD) be done in 'standard rez', to achieve the analog-like effect. Their test setup and signals must be seen/read to be believed (the pdf includes photos). They're not your typical listener setup, to say the least (compare to, say, Meyer and Moran's setup for their SACD vs CD test). Also, I'm having trouble making heads or tails of their statistics. Also being discussed on HA.org http://www.hydrogenaudio.org/forums/...howtopic=55966 ___ -S "As human beings, we understand the world through simile, analogy, metaphor, narrative and, sometimes, claymation." - B. Mason |
#2
Posted to rec.audio.high-end
|
|||
|
|||
AES article: hi-rez more like analog?
Here's the first paragraph of their Conclusion section:
"The results of this test indicate that listeners more often than not identify high-resolution audio as being similar in quality to the unprocessed analog audio. This conclusion, based on listening to the audio scene captured and reproduced with microphones and loudspeakers limited to 20 kHz bandwidth, indicates that high-sampling conversion system seems to be more transparent and provides a higher degree of fidelity to the analog reference." Here's what their data says: Out of 54 trials with that limited bandwidth, subjects chose the high-sampling-rate system as closer to the original analog feed 31 times. That's not even statistically significant at the 90% confidence level. So by their own data, they can't reject the possibility that their subjects could not tell the difference between 352.8 kHz sampling and 44.1, even under the best scenario. What's odd is that, when they used mikes and speakers with a bandwidth extended to 100 kHz, the high sampling rate did substantially worse (23/60). That result is highly counterintuitive, and the authors twist themselves into pretzels trying to explain it, with no success. So they're making claims their own data don't fully support and they can't explain. I think there's no there there. bob |
#3
Posted to rec.audio.high-end
|
|||
|
|||
AES article: hi-rez more like analog?
Steven Sullivan wrote:
Actually it's from a 2007 AES conference, not the JAES, so I'm not sure it's peer reviewed...but anyway: "Which of the Two Digital Audio Systems Best Matches the Quality of the Analog System?" http://www.hitech-projects.com/hera/...s/aar07pu4.pdf nice ref! thank you very much It appears that the formats compared to a live feed (analog) were DXD (353.8 kHz/24) and 44.1/24, both in surround, using a blind comparison protocol. Two additional listening conditions were tested: one where the A/D signal bandwith was 100 kHz (thanks to special microphone 'super-tweeters') and the other where bandwidth was limited to 20kHz. The authors say their results show that listeners 'more often than not' identify the hi-rez audio (and not the 44.1 kHz audio) as being similar in quality to the analog feed...but only when the bandwidth is limited! In other words, only the initial sampling need be done in hi-rez, the listening can (and SHOULD) be done in 'standard rez', to achieve the analog-like effect. Their test setup and signals must be seen/read to be believed (the pdf includes photos). They're not your typical listener setup, to say the least (compare to, say, Meyer and Moran's setup for their SACD vs CD test). if there is an open on-line ref to this, i'd very much appreciate someone providing it; thanks in advance Also, I'm having trouble making heads or tails of their statistics. Also being discussed on HA.org http://www.hydrogenaudio.org/forums/...howtopic=55966 very nice 2nd ref! only 5 people, but still very interesting i'm going to have to start reading www.hydrogenaudio.org on a regular basis imho, multichannel sound is not driven by audio, it is driven by movies. of the recent 2 formats (HD-DVD and Blu-ray), which really has the better multichannel sound? imho#2, better sound isn't going to make any diff on which wins. or are they both too similar (on sound) to be worth worrying about it? where is Kalman when we need him? fwiw this past few days i've gotten a Toshiba HD-DVD A35 player, with an open order for a copy of Magic Flute (arriving after Xmas) http://www.amazon.com/Mozart-Zauberf...7563048&sr=8-1 the 4 current customer reviews give it high marks on sound. read them for yourself of course SACD doesn't have video, but is the multichannel sound on Magic Flute comparable to SACD? all ears, bill "As human beings, we understand the world through simile, analogy, metaphor, narrative and, sometimes, claymation." - B. Mason Never confuse movement with action. Ernest Hemingway (1899-1961) |
#4
Posted to rec.audio.high-end
|
|||
|
|||
AES article: hi-rez more like analog?
willbill wrote:
Steven Sullivan wrote: Actually it's from a 2007 AES conference, not the JAES, so I'm not sure it's peer reviewed...but anyway: "Which of the Two Digital Audio Systems Best Matches the Quality of the Analog System?" http://www.hitech-projects.com/hera/...s/aar07pu4.pdf nice ref! thank you very much It appears that the formats compared to a live feed (analog) were DXD (353.8 kHz/24) and 44.1/24, both in surround, using a blind comparison protocol. Two additional listening conditions were tested: one where the A/D signal bandwith was 100 kHz (thanks to special microphone 'super-tweeters') and the other where bandwidth was limited to 20kHz. The authors say their results show that listeners 'more often than not' identify the hi-rez audio (and not the 44.1 kHz audio) as being similar in quality to the analog feed...but only when the bandwidth is limited! In other words, only the initial sampling need be done in hi-rez, the listening can (and SHOULD) be done in 'standard rez', to achieve the analog-like effect. Their test setup and signals must be seen/read to be believed (the pdf includes photos). They're not your typical listener setup, to say the least (compare to, say, Meyer and Moran's setup for their SACD vs CD test). if there is an open on-line ref to this, i'd very much appreciate someone providing it; thanks in advance Meyer and Moran's paper isn't available online AFAIK, but it's in circulation... Also, I'm having trouble making heads or tails of their statistics. Also being discussed on HA.org http://www.hydrogenaudio.org/forums/...howtopic=55966 very nice 2nd ref! only 5 people, but still very interesting i'm going to have to start reading www.hydrogenaudio.org on a regular basis Also discussed on AVSforum http://www.avsforum.com/avs-vb/showthread.php?t=953640 I'm kinda guessing it WON'T be hashed over in Stereophile, though even thougb John Atkinson is citing this paper as proof that Redbook sample rate is inadequate for home audio. '"In October I will write more on Peter's ideas about why this should be. I will end this month's essay by quoting, from a paper given at the conference, the results of experiments on the audibility of high sampling rates: "To achieve a higher degree of fidelity to the live analog reference, we need to convert audio using a high sampling rate even when we do not use microphones and loudspeakers having bandwidth extended far beyond 20kHz. Listeners judge high sampling conversion as sounding more like the analog reference when listening to standard audio bandwidth." (footnote 2) So that's that, then. " http://stereophile.com/asweseeit/907awsi/ Atkinson used to post here, maybe he can be induced to comment? of course SACD doesn't have video, but is the multichannel sound on Magic Flute comparable to SACD? all ears, bill Surely DVD-A is comparable to SACD (I'd wager that in a blind test, DTS 24/96 is indistinguisbale form them too). So why not the new 'hi rez' formats? ___ -S "As human beings, we understand the world through simile, analogy, metaphor, narrative and, sometimes, claymation." - B. Mason |
#5
Posted to rec.audio.high-end
|
|||
|
|||
AES article: hi-rez more like analog?
bob wrote:
What's odd is that, when they used mikes and speakers with a bandwidth extended to 100 kHz, the high sampling rate did substantially worse (23/60). That result is highly counterintuitive, and the authors twist themselves into pretzels trying to explain it, with no success. There is a stock answer to this, which may or not be right, but always needs to be tested. That is, with nonlinear speakers, inaudible high frequencies can cause audible intermodulation distortion. This can of course be tested if you have a really really good high frequency test system (super linear high frequency microphone and digitizer) to aim at the speakers. Doug McDonald |
#6
Posted to rec.audio.high-end
|
|||
|
|||
AES article: hi-rez more like analog?
Steven Sullivan wrote:
willbill wrote: Also discussed on AVSforum http://www.avsforum.com/avs-vb/showthread.php?t=953640 I'm kinda guessing it WON'T be hashed over in Stereophile, though even thougb John Atkinson is citing this paper as proof that Redbook sample rate is inadequate for home audio. '"In October I will write more on Peter's ideas about why this should be. I will end this month's essay by quoting, from a paper given at the conference, the results of experiments on the audibility of high sampling rates: "To achieve a higher degree of fidelity to the live analog reference, we need to convert audio using a high sampling rate even when we do not use microphones and loudspeakers having bandwidth extended far beyond 20kHz. Listeners judge high sampling conversion as sounding more like the analog reference when listening to standard audio bandwidth." (footnote 2) So that's that, then. " http://stereophile.com/asweseeit/907awsi/ i've got the hard copy but hadn't thought to look at the on-line, which makes it easy to look at his refs in earlier editions Atkinson used to post here, maybe he can be induced to comment? fwiw, he did some minor posting in one of the other rec.audio... newsgroups a week or two ago of course SACD doesn't have video, but is the multichannel sound on Magic Flute comparable to SACD? all ears, bill Surely DVD-A is comparable to SACD (I'd wager that in a blind test, DTS 24/96 is indistinguisbale form them too). So why not the new 'hi rez' formats? i've got the player (OPPO), but so far have only bought SACD disks bill |
#7
Posted to rec.audio.high-end
|
|||
|
|||
AES article: hi-rez more like analog?
"bob" wrote in message
Here's the first paragraph of their Conclusion section: "The results of this test indicate that listeners more often than not identify high-resolution audio as being similar in quality to the unprocessed analog audio. IOW, they mostly fail to hear a difference. This conclusion, based on listening to the audio scene captured and reproduced with microphones and loudspeakers limited to 20 kHz bandwidth, indicates that high-sampling conversion system seems to be more transparent and provides a higher degree of fidelity to the analog reference." ???????????????? Here's what their data says: Out of 54 trials with that limited bandwidth, subjects chose the high-sampling-rate system as closer to the original analog feed 31 times. I sense a conclusion that they were randomly guessing coming on. That's not even statistically significant at the 90% confidence level. So by their own data, they can't reject the possibility that their subjects could not tell the difference between 352.8 kHz sampling and 44.1, even under the best scenario. So much for the hypothesis that the audio CD format is somehow deficient. What's odd is that, when they used mikes and speakers with a bandwidth extended to 100 kHz, the high sampling rate did substantially worse (23/60). Random guessing sometimes works out that way. Reams have written by ignorant audiophiles about the meaning of worse results than random guessing. In the end, worse than random guessing is probably best interpreted as being an indictment of the experiment. If people are truely producing independent results, then their results will converge to random guessing when they can't hear a difference. When they are guessing wrong more often than 50%, the hypothesis that they are producing independent results can be questioned. That result is highly counterintuitive, and the authors twist themselves into pretzels trying to explain it, with no success. They don't want to admit that their results are critical of the over-all quality of the experiment. So they're making claims their own data don't fully support and they can't explain. I think there's no there there. Agreed. |
#8
Posted to rec.audio.high-end
|
|||
|
|||
AES article: hi-rez more like analog?
"Steven Sullivan" wrote in message
Actually it's from a 2007 AES conference, not the JAES, so I'm not sure it's peer reviewed Conference papers aren't reviewed at all - only the title and abstract have been seen by the session chairman. ...but anyway: "Which of the Two Digital Audio Systems Best Matches the Quality of the Analog System?" http://www.hitech-projects.com/hera/...s/aar07pu4.pdf It appears that the formats compared to a live feed (analog) were DXD (353.8 kHz/24) and 44.1/24, both in surround, using a blind comparison protocol. It appears that the program material is best described as being "sound effects". Also, every time some high resoution advocate cites this paper, ask them to reflect on how closely it resembles listening to music for pleasure in a residential listening room. ;-) Two additional listening conditions were tested : one where the A/D signal bandwith was 100 kHz (thanks to special microphone 'super-tweeters') and the other where bandwidth was limited to 20kHz. The authors say their results show that listeners 'more often than not' identify the hi-rez audio (and not the 44.1 kHz audio) as being similar in quality to the analog feed...but only when the bandwidth is limited! Let's cut to the chase and look at the raw data on page 19. As I read it, there were 10 subjects and 6 listening sessions for which individual responses were required. IOW, 60 trials. I pasted their matrix into Excel and tried to do some quick sums. I came up with Test Condition 1 = 23/60 and Test Condition 2 = 31/60. IOW, one test produced an outcome that was worse than random guessing, and the other was random guessing. Results that are worse than random guessing may cause some head scratching, but they are not all that unusual in experiments like this where communication between the listeners can affect the outcome. The most probable explanation for worse-than-random results is that that some of the listeners were basing their results on their perceptions of what other listeners were perceiving, and the total number of independent responses was far less than what you get from a naive count of the actual responses. IOW, the actual situation was not 23 independent responses of 60 trials, but maybe more like 4 independent responses of 10 independent trials, so the true numbers were so small that statistics doesn't really apply. I think these results would pretty well explain themselves to just about anybody, were they reproduced any place within the actual paper. Short answer - the outcome was random guessing, and both the test itself and the statistical analysis were greviously flawed. I guess we can chalk this paper up as yet another "Proof by complex statistical analysis" which defies common sense. |
#9
Posted to rec.audio.high-end
|
|||
|
|||
AES article: hi-rez more like analog?
On Dec 13, 10:15 pm, Doug McDonald
wrote: bob wrote: What's odd is that, when they used mikes and speakers with a bandwidth extended to 100 kHz, the high sampling rate did substantially worse (23/60). That result is highly counterintuitive, and the authors twist themselves into pretzels trying to explain it, with no success. There is a stock answer to this, which may or not be right, but always needs to be tested. That is, with nonlinear speakers, inaudible high frequencies can cause audible intermodulation distortion. But wouldn't these inaudible frequencies be present in the original analog feed, as well as the high-sampling rate digital conversion? That should make the two sound *more* similar, and less like the 44.1 conversion, which would filter those frequencies out before they reached the speakers. bob |
#10
Posted to rec.audio.high-end
|
|||
|
|||
AES article: hi-rez more like analog?
Arny Krueger wrote:
"Steven Sullivan" wrote in message Actually it's from a 2007 AES conference, not the JAES, so I'm not sure it's peer reviewed Conference papers aren't reviewed at all - only the title and abstract have been seen by the session chairman. ...but anyway: "Which of the Two Digital Audio Systems Best Matches the Quality of the Analog System?" http://www.hitech-projects.com/hera/...s/aar07pu4.pdf It appears that the formats compared to a live feed (analog) were DXD (353.8 kHz/24) and 44.1/24, both in surround, using a blind comparison protocol. It appears that the program material is best described as being "sound effects". Also, every time some high resoution advocate cites this paper, ask them to reflect on how closely it resembles listening to music for pleasure in a residential listening room. ;-) Two additional listening conditions were tested : one where the A/D signal bandwith was 100 kHz (thanks to special microphone 'super-tweeters') and the other where bandwidth was limited to 20kHz. The authors say their results show that listeners 'more often than not' identify the hi-rez audio (and not the 44.1 kHz audio) as being similar in quality to the analog feed...but only when the bandwidth is limited! Let's cut to the chase and look at the raw data on page 19. As I read it, there were 10 subjects and 6 listening sessions for which individual responses were required. IOW, 60 trials. I pasted their matrix into Excel and tried to do some quick sums. I came up with Test Condition 1 = 23/60 and Test Condition 2 = 31/60. IOW, one test produced an outcome that was worse than random guessing, and the other was random guessing. Results that are worse than random guessing may cause some head scratching, but they are not all that unusual in experiments like this where communication between the listeners can affect the outcome. The most probable explanation for worse-than-random results is that that some of the listeners were basing their results on their perceptions of what other listeners were perceiving, and the total number of independent responses was far less than what you get from a naive count of the actual responses. There was science news article recently -- can't find it right now, but it was in the last week or so -- reporting a study demonstrating the influence of having someone else in the room, on perception of *comedy* (in this case, the probe signal was sketches from Saturday Night Live). Just seeing the back of someone else's head, as they watched, was enough to influence subjects to find the presentation funny or not. ___ -S "As human beings, we understand the world through simile, analogy, metaphor, narrative and, sometimes, claymation." - B. Mason |
#11
Posted to rec.audio.high-end
|
|||
|
|||
AES article: hi-rez more like analog?
On Dec 14, 1:01 pm, "Arny Krueger" wrote:
I pasted their matrix into Excel and tried to do some quick sums. I came up with Test Condition 1 = 23/60 and Test Condition 2 = 31/60. To be precise, Condition 2 is 31/54, since one subject's results were not available. (It happens.) IOW, one test produced an outcome that was worse than random guessing, and the other was random guessing. Actually, both could be random guessing, since neither hits even the 90% confidence level. Results that are worse than random guessing may cause some head scratching, but they are not all that unusual in experiments like this where communication between the listeners can affect the outcome. Not the case here, where each subject was tested individually. Also, unlike an ABX test, there is no "wrong" answer here. Either conversion could be judged better. (Unless, of course, you're Philips and you sell hi-rez converters.) The most probable explanation for worse-than-random results is that that some of the listeners were basing their results on their perceptions of what other listeners were perceiving, and the total number of independent responses was far less than what you get from a naive count of the actual responses. IOW, the actual situation was not 23 independent responses of 60 trials, but maybe more like 4 independent responses of 10 independent trials, so the true numbers were so small that statistics doesn't really apply. Again, this is an inaccurate description of the actual test. The responses were independent. I think these results would pretty well explain themselves to just about anybody, were they reproduced any place within the actual paper. Short answer - the outcome was random guessing, and both the test itself and the statistical analysis were greviously flawed. The test itself was not grievously flawed. The only obvious problem was the one Scott mentioned--using different mikes for the two conditions. But that only matters if you're comparing the results of under the two conditions. Looked at individually, the two test conditions tell us nothing. The statistical analysis is another story. I guess we can chalk this paper up as yet another "Proof by complex statistical analysis" which defies common sense. Agreed. bob |
#12
Posted to rec.audio.high-end
|
|||
|
|||
AES article: hi-rez more like analog?
"bob" wrote in message
On Dec 14, 1:01 pm, "Arny Krueger" wrote: I pasted their matrix into Excel and tried to do some quick sums. I came up with Test Condition 1 = 23/60 and Test Condition 2 = 31/60. To be precise, Condition 2 is 31/54, since one subject's results were not available. (It happens.) IOW, one test produced an outcome that was worse than random guessing, and the other was random guessing. Actually, both could be random guessing, since neither hits even the 90% confidence level. Results that are worse than random guessing may cause some head scratching, but they are not all that unusual in experiments like this where communication between the listeners can affect the outcome. Not the case here, where each subject was tested individually. Perhaps not that individually. Also, unlike an ABX test, there is no "wrong" answer here. Sure there is. Inconsistency is wrong. Either conversion could be judged better. (Unless, of course, you're Philips and you sell hi-rez converters.) The most probable explanation for worse-than-random results is that that some of the listeners were basing their results on their perceptions of what other listeners were perceiving, and the total number of independent responses was far less than what you get from a naive count of the actual responses. IOW, the actual situation was not 23 independent responses of 60 trials, but maybe more like 4 independent responses of 10 independent trials, so the true numbers were so small that statistics doesn't really apply. Again, this is an inaccurate description of the actual test. The responses were independent. But how independent? I think these results would pretty well explain themselves to just about anybody, were they reproduced any place within the actual paper. Short answer - the outcome was random guessing, and both the test itself and the statistical analysis were greviously flawed. The test itself was not grievously flawed. It acted that way. A good test of perception is either random or correlated with the stimulus. When the correlation is negative, something went badly wrong. The only obvious problem was the one Scott mentioned--using different mikes for the two conditions. But that only matters if you're comparing the results of under the two conditions. Looked at individually, the two test conditions tell us nothing. That's a sign of a flawed test - it tells us less than we expected to find out. The statistical analysis is another story. I guess we can chalk this paper up as yet another "Proof by complex statistical analysis" which defies common sense. Agreed. |
Reply |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Forum | |||
IEEE Article | Pro Audio | |||
Did anyone see the article . . . . | Vacuum Tubes | |||
For analog fans: BOSS DM-3 DM3 ANALOG DELAY PEDAL VINTAGE RARE | Tech | |||
Whole article | Pro Audio |