tag:blogger.com,1999:blog-7172143857693272648.post3794305123703883662..comments2024-03-18T03:14:54.390-07:00Comments on Audio Musings by Sean Olive: Why Live-versus-Recorded Listening Tests Don't WorkDr. Sean Olivehttp://www.blogger.com/profile/17909033506833141612noreply@blogger.comBlogger12125tag:blogger.com,1999:blog-7172143857693272648.post-48291717493651642062013-07-19T08:39:14.126-07:002013-07-19T08:39:14.126-07:00I'm a degreed EE and AES member who used to re...I'm a degreed EE and AES member who used to review for "Stereophile". I stopped reviewing when I discovered that huge sonic differences heard one day (with matched levels) utterly vanished the next. (Don't read too much into that.) This didn't convert me to an ABX advocate, as that sort of testing has its own problems. (Using a test protocol designed for one form of scientific analysis is not necessarily valid for another.)<br /><br />Though at least one well-known speaker designer expressed his view that the AR/Dyna tests were valid, their validity is beside the point. They are interesting in their own right, simply for giving us an idea of what is or isn't audible, how listeners react when they /know/ there should be audible differences, etc.<br /><br />As an advocate of facsimile reproduction (I couldn't care less about studio recordings that have no acoustical counterpart), I'd like to propose an expensive and difficult project -- an attempt to see just how close one can come to "the original sound" -- the illusion that one is hearing the original performance.<br /><br />This would be a multi-year project, requiring long-term commitment to making live recordings using different mics and mic setups, and playing them through an even wider variety of speakers and speaker arrays.<br /><br />What would we learn from this? I don't know. But that's the point of scientific testing -- seeing what will happen. (As HP used to say... "We never stop asking What If?"William Sommerwercknoreply@blogger.comtag:blogger.com,1999:blog-7172143857693272648.post-39989250690686328542010-11-24T06:40:16.776-08:002010-11-24T06:40:16.776-08:00Steven Sullivan states that we only think MP3s sou...Steven Sullivan states that we only think MP3s sound worse because they measure worse and because we are prejuduced.<br /><br />Sorry Steve, You are flat out wrong here. There is a huge sonic difference. Go to a local high end stereo store and plug your little chip in and take a CD of the same music with you. You will either discover your error or need a hearing test.<br /><br />I listen to a lot of XM radio and enjoy it; not because it sounds as good as CD or a good LP, but because I am used to it. If I listen to CDs first, then it takes a while to ajust to the inferior format When I go from XM to CD, I immediately presented with a vastly superior sound.<br /><br />DanV.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-7172143857693272648.post-31118038437322709232010-07-21T19:35:47.269-07:002010-07-21T19:35:47.269-07:00Dear Dr. Olive,
Thank you for your reply and comm...Dear Dr. Olive,<br /><br />Thank you for your reply and comments. <br /><br />The outdoor photographs of the Fine Arts Quartet -- along with Edgar Villchur, Roy Allison and other individuals -- were taken in Woodstock, NY during the initial recording sessions, and the AR-3s were for monitoring purposes and to check playback levels, etc. Clearly, the AR-3s received unintended publicity in this picture, but AR never used these pictures in any of its advertisements. It should be mentioned that Dynaco and a magnetic-tape manufacturer, Concertapes (as well as Ampex and Sony), were also involved in the LvR concerts. Concertapes had a recording contract with The Fine Arts Quartet at the time, and the Quartet approached Edgar Villchur in 1959 about doing live-vs.-recorded concerts together. <br /><br />In 1956, the New York Audio League (Julian Hirsch) used four AR-1 loudspeakers, along with Bozak midrange drivers and the JansZen electrostatic tweeters, to reproduce the Aeolian-Skinner pipe organ in a Mt. Kisco, NY church, so LvR demonstrations had been done previously using AR equipment. This demonstration was also very successful, and it helped establish Acoustic Research's AR-1 as one of the finer low-frequency reproducers of the era.<br /><br />The Fine Arts Quartet LvR concerts were AR's most successful, in my view, and they were met with almost universal critical acclaim from music critics, equipment reviewers and the public alike. More than seventy-five concerts were performed across the country in most of the largest cities, such as New York, Boston, Washington, DC, Chicago, LA and San Francisco. The ensemble tone of the Fine Arts Quartet -- performed in relatively spacious enclosed spaces -- was reproduced well from accounts from the newspaper reports and popular high-fidelity magazines of the period. <br /><br />Subsequently, AR performed several concerts with classical guitarist Gustavo Lopez and the 1910 Seaburg Nickelodeon; while these demonstrations were quite successful, they lacked the closeness of playback of the Fine Arts Quartet. In the mid-1970s, AR used the then-new AR-10Pi (an updated version of the AR-3 and AR-3a) in a series of live-vs.-recorded demonstrations with jazz drummer Neil Grover. This LvR proved more difficult, as it took over 800-watt peaks to reproduce the drum "rim shots" and other drum sounds through the .5% efficient AR-10s. Amazingly, no AR driver was damaged except for a frozen voice coil caused by dc-offset from a failed Dunlap-Clarke Dreadnaught 1000 amplifier. Overall, I don't believe that these difficult demonstrations were as successful as AR's 1959-1963 Fine Arts Quartet series.<br /><br />ARHPG (Acoustic Research Historic Preservation Group)Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-7172143857693272648.post-39834642239698064642010-07-21T17:37:22.389-07:002010-07-21T17:37:22.389-07:00Dear ARHPG,
Thank you for the correction regarding...Dear ARHPG,<br />Thank you for the correction regarding the Live-vs-Recorded AR outdoor concerts. I think that misconception probably came from photographs on the web showing the Quartet playing outdoors with AR speakers behind them. The speakers must have been added for publicity sake only.Dr. Sean Olivehttps://www.blogger.com/profile/17909033506833141612noreply@blogger.comtag:blogger.com,1999:blog-7172143857693272648.post-54645395399429117792010-07-20T08:56:43.224-07:002010-07-20T08:56:43.224-07:00"During the 1960’s, Acoustic Research (AR), a..."During the 1960’s, Acoustic Research (AR), an American loudspeaker company, performed over 75 live-versus-recorded concerts in cities around the USA featuring The Fine Arts String Quartet, and the AR-3 loudspeaker [5],[6]. To solve the double reverberation problem, the recordings of the quartet were made in an anechoic chamber, or outdoors. Outdoor live-versus-recorded demonstrations had the added benefit that there were no room reflections in either the recording or the live performance. This made the demonstrations less sensitive to off-axis problems in the microphones and loudspeakers. It also relaxed the demands on the recording-reproduction to accurately capture and reproduce the complex spatial properties of a reverberant performing space."<br /><br />This statement is erroneous. The AR Live-vs.-Recorded concerts were *not* conducted outdoors. The recording of the musicians for these demonstrations was performed in anechoic (outdoors in this case) space to avoid "double reverberation," such that the AR-3s and the Fine Arts Quartet (as well as guitarist Gustavo Lopéz and the 1910 Nickelodeon) would be treated to a nearly identical acoustical environment.<br /><br />"To date, most of the live-versus-sighted tests have been performed sighted, where non-auditory cues were available to allow the listener to identify whether they were hearing the live or reproduced sound source. These tests could have been easily made blind via an acoustically transparent curtain; however, scientific validity was apparently not the primary purpose of the test. The visual cues from the musicians (bowing, lip syncing) would also enhance the realism and presence of the reproduction, a well-known <br />cognitive effect observed in research of binaural and virtual reality displays."<br /><br /><br />AR used several methods to avoid "visual clues" in identifying the point at which the musicians stopped playing and the loudspeakers began playing. AR made several recordings of actual performances (both the musicians and the speakers). The method commonly used throughout all of the concerts was to have the musicians "lift" the bows slightly, but continue "playing" while the AR-3s played. However, in the early stages the Fine Arts Quartet played a trick on the audience. Midway through the performance of the first-selected piece, First Violinist Leonard Sorkin stopped the music and asked the audience, "I'm curious, how many of you in the audience could detect when we switched from live music to the recorded music?" There was a show of hands in the audience. "I'm sorry to tell you this, but except for the first two bars, the entire piece was played by the speakers." By the way, there was never any signal equalization used on the speakers -- except to adjust (increase and attenuate) the treble control on the preamp for different acoustical conditions in different concert halls<br /><br />ARHPG (Acoustic Research Historic Preservation Group)Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-7172143857693272648.post-50949586663976028802010-07-19T10:04:01.337-07:002010-07-19T10:04:01.337-07:00jehle, mp3 is not an instance of 'people will ...jehle, mp3 is not an instance of 'people will hear what you tell them to hear', except to the extent that *anti*-mp3 writers tell people that mp3s sound terrible compared to lossless formats. Such claims aren't based on any thoughtful consideration of how mp3s work, or how listening comparisons should be done. It's just bias against the idea of 'lossyness'....and uncritical belief that if something 'measures' better, it MUST sound better.Steven Sullivanhttps://www.blogger.com/profile/01800114865214533118noreply@blogger.comtag:blogger.com,1999:blog-7172143857693272648.post-50032764430993377902010-07-19T09:58:18.038-07:002010-07-19T09:58:18.038-07:00Sean,
If you can ignore its instances of pro-anal...Sean,<br /><br />If you can ignore its instances of pro-analog cheerleading, you might enjoy a book published last year called Perfecting Sound Forever: An Aural History of Recorded Music" by Greg Milner. It too discusses a few of the 'live vs recorded' listening tests, in part to demonstrate how our idea of 'perfect sound' has changed with the times.Steven Sullivanhttps://www.blogger.com/profile/01800114865214533118noreply@blogger.comtag:blogger.com,1999:blog-7172143857693272648.post-35129491192516106292010-07-15T23:41:11.768-07:002010-07-15T23:41:11.768-07:00Hi Anonymous,
I don't disagree with you in th...Hi Anonymous,<br /><br />I don't disagree with you in that recording engineers will make adjustments (e.g. add close mics) to compensate for limitations in 2-channel stereo, hall acoustics, restrictions in where they can place their mics,etc. These adjustments can make the recording sound more natural or pleasing.<br /><br />We are surrounded in life with lots of natural references -human voices, nature,environmental sounds,etc, machinery, instruments, applause- that provide references (if included in recordings) when judging the fidelity of different loudspeakers. Through training listeners become familiar with the programs and how they should sound.<br /><br /> Listeners are also pretty good detectives when it comes to identifying resonances/frequency response aberrations, and distortions in loudspeakers. <br /><br />When doing multiple comparisons among loudspeakers, it becomes easier to separate the distortions in recordings from those in the loudspeakers - more so than a single stimulus test. Any distortions that are constant when switching among loudspeakers are associated with the recording and tend to fall into the background. If the sound quality changes as you switch to a different speaker that sonic character becomes associated with the speaker itself.Dr. Sean Olivehttps://www.blogger.com/profile/17909033506833141612noreply@blogger.comtag:blogger.com,1999:blog-7172143857693272648.post-42866797909224356232010-07-14T11:58:27.242-07:002010-07-14T11:58:27.242-07:00This will sound odd, but I think when the engineer...This will sound odd, but I think when the engineer cranks up the oboe in a Mahler symphony, that's like concert hall listening - actually, something between fulfilling Mahler's intentions and the "creative" listening you naturally are doing at a concert.<br /><br />For sure, can't put Ozawa doing Berlioz' Requiem in Massey Hall into my living room, but something is going right with hifi.<br /><br />I think Vilchur was on the right track looking for a reference - I use an AR-1W in my system. I suggested to my client once they carry around a mechanical alarm clock and a recording of same for reference vis a vis noise annoyance from streetcars.<br /><br />BenAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-7172143857693272648.post-42384801483845077402010-07-11T23:53:15.563-07:002010-07-11T23:53:15.563-07:00Hi Grant,
Thanks for your feedback. I thought it ...Hi Grant,<br /><br />Thanks for your feedback. I thought it was appropriate to historically discuss how the method has been used in the past with some constructive criticism. I find it interesting, and perhaps noteworthy, that the method has been used to fool people into believing something is better than it actually is.It would be interesting to explore more why that is.<br /><br />I did offer some suggestions how the method could be improved upon. However, I don't see myself ever using the method since I don't think it is very sensitive or practical for evaluating loudspeakers. The main problem is that no recordings or recording methods come close to capturing what you perceive in a live performance it seems moot to use this method for evaluating loudspeakers. I think binaural recordings probably come closest to capturing/reproducing the live performance.<br /><br />You are right about the mismatch between the directivity of instruments versus loudspeakers, and how that invalidates these experiments. <br /><br />The directivity of a typical forward facing loudspeaker very much approximates a human voice based on measurements of human voice at NRC. So I think Edison was probably smart in choosing singers for these experiments. The other way to get around this problem is to record and reproduce in outdoors or in an anechoic chamber which is what AR did with the live-vs-recorded demos of the The Fine Arts Quartet.Dr. Sean Olivehttps://www.blogger.com/profile/17909033506833141612noreply@blogger.comtag:blogger.com,1999:blog-7172143857693272648.post-78054068096325926872010-07-10T21:15:23.374-07:002010-07-10T21:15:23.374-07:00Hi, an excellent article as usual, much appreciate...Hi, an excellent article as usual, much appreciated. I am not sure that criticising the way it has been done in the past is a valid reason not to do it properly yourself.<br /><br />I would be interested in your thoughts re: playback-of-recording-of-live-performance vs playback-of-recording-of-loudspeaker-playing-recording-of-live-performance. How close does this come to assessing the loudspeaker vs the live performance?<br /><br />Having said that, I cannot see any way around the different polar patterns of live music compared to any particular loudspeaker. If a piano and a violin, playing the same note, have different polar patterns, a loudspeaker cannot play back the same note with different polar patterns for each instrument. Hence live-vs-speaker is invalid.<br /><br />GrantGrantnoreply@blogger.comtag:blogger.com,1999:blog-7172143857693272648.post-47791731462627816822010-07-10T12:48:59.955-07:002010-07-10T12:48:59.955-07:00That's great Sean, I love the "people wil...That's great Sean, I love the "people will hear what you tell them to hear." Hence the rise of mp3 and Bose...<br /><br />Thanks, Bill Jehleleafhttps://www.blogger.com/profile/16752714502763399326noreply@blogger.com