Art vs Science of Audio Reproduction: Which Team Are You In?

[TheoN]

@PENG you bring up a very valid point with this. I agree that manufacturers should indicated whether the speakers have been tuned with or without the grilles. When I was at the B&W 800 Series Diamond Launch, they made it a point to say that the new 800 Series Diamond tweeters have been tuned with the protective caps on. I do think Gene touched upon some great points in his editorial here.

 

[KEW]

On the flip side, I had 3 inexperienced listeners compare both speakers sighted noting the MA was prettier but still picked the Pinnacle as better sounding b/c of the better bass and less constrained sound. These folks were brand agnostic but they did use their eyes and still picked the uglier speaker as better in a sighted test.

I'm not sure you are making a valid conclusion here. The MA may be the better looking speaker, but the Pinnacle is almost twice as large as the Monitor Audio (1.94 times as large by volume). If people had to guess based solely on appearance which would sound better, I think the Pinnacle would get the vast majority of votes because it is a much larger speaker!
I don't think most people would see beauty as a predictor of performance when the size was such a disparity!

 

[gene]

I'm not sure you are making a valid conclusion here. The MA may be the better looking speaker, but the Pinnacle is almost twice as large as the Monitor Audio (1.94 times as large by volume). If people had to guess based solely on appearance which would sound better, I think the Pinnacle would get the vast majority of votes because it is a much larger speaker!
I don't think most people would see beauty as a predictor of performance when the size was such a disparity!

Oh yes that is the argument by the DBT crowd. Aesthetics is a big bias regardless of size. The MA's also cost more. The MA's didn't look that much smaller, especially since they have a plinth and spikes. They aren't as deep but similar in height from what I remember.

 

[gene]

I understood the grille you referred to was just for the shootout. The word grille simply reminded me of my recent discovery of the difference they made to two of my speakers and not to the other two. Looking at the graphs for the Mirage, they must have bad grilles but good in the sense that they seem to act like a good EQ system that levels the frequency response.

View attachment 16772
View attachment 16770

The Mirage grille on was from 1M whereas the grille off was from 6ft as they were not done on the same day. The R900 one's were done with the mic in the same spot the difference was just grille on vs off. You can see the 5 dB dip at around 13,300 Hz when the grille was on.

I guess I am getting off topic but you have the power to move it to where this belong right?

I've never seen a grille being that lossy. I think you need to remeasure at the exact same position on/off.

 

[shadyJ]

We can scoff at people who feel their speaker cable has made their system 110% better, and we can point to data and measurements that say that is impossible, but I don't think that we appreciate how profound their illusion really is. For those of us who have experienced vivid hallucinations, we can't be rightly told that we only thought we saw them. We definitely saw them- but that isn't to say what we saw was real. I believe that our expectations can color our perceptions just like a hallucinogen.

There is a lot of brain processing that occurs on our sense data before it arrives at our conscious perception. The case may be that someone who is expecting an improvement from their speaker cable may hear that improvement regardless of what the speaker cable does. Furthermore, how do objectivists know they are not being tricked by their own mind in the same way? If we are told this system in this room measures perfectly by trusted authorities, how do we know that when we go in and listen our expectations could be turning a subpar system into a great system? In a sense, objectivists are like subjectivists in that they shape their reality, but by a different set of criteria. However, I would argue they adhere to a more stringent criteria, and that is one reason why I tend toward the objectivist camp in audio. Another reason is it is much less expensive to be an audio objectivist than subjectivist!

 

[gene]

We can scoff at people who feel their speaker cable has made their system 110% better, and we can point to data and measurements that say that is impossible, but I don't think that we appreciate how profound their illusion really is. For those of us who have experienced vivid hallucinations, we can't be rightly told that we only thought we saw them. We definitely saw them- but that isn't to say what we saw was real. I believe that our expectations can color our perceptions just like a hallucinogen.

There is a lot of brain processing that occurs on our sense data before it arrives at our conscious perception. The case may be that someone who is expecting an improvement from their speaker cable may hear that improvement regardless of what the speaker cable does. Furthermore, how do objectivists know they are not being tricked by their own mind in the same way? If we are told this system in this room measures perfectly by trusted authorities, how do we know that when we go in and listen our expectations could be turning a subpar system into a great system? In a sense, objectivists are like subjectivists in that they shape their reality, but by a different set of criteria. However, I would argue they adhere to a more stringent criteria, and that is one reason why I tend toward the objectivist camp in audio. Another reason is it is much less expensive to be an audio objectivist than subjectivist!

great points and agreed!

 

[gene]

@PENG you bring up a very valid point with this. I agree that manufacturers should indicated whether the speakers have been tuned with or without the grilles. When I was at the B&W 800 Series Diamond Launch, they made it a point to say that the new 800 Series Diamond tweeters have been tuned with the protective caps on. I do think Gene touched upon some great points in his editorial here.

Protective caps won't do much in terms of losses or diffraction but the grilles will have a more profound effect and it's usually NEVER better to have grilles on then off.

 

[PENG]

Another reason is it is much less expensive to be an audio objectivist than subjectivist!

Great point!! I guess same goes for why more people want to add an external amp such as an Outlaw or Emo because that's much less expensive and easy to do than to replace their speakers, collection of media and/or a $1500 sub.

 

[PENG]

Protective caps won't do much in terms of losses or diffraction but the grilles will have a more profound effect and it's usually NEVER better to have grilles on then off.

I will redo it with them mic at the exact same location for the Mirage. I did that for the KEF R900 because I got alerted by the dip and wanted investigate the cause. So the 5 dB (more if I apply 1/48 smoothing instead of 1/12) dip at around 13 kHz with the grille on was real and repeatable. Good thing the R900's grille is magnetic so on/off is effortless. The inexplicable part is that the dip looks like a notch filter effect, I wonder if it has to do with the KEF tangerine guide thing. I mean, how can the grille diffraction effect act like a narrow notch filter?

In the case of the Mirage it was done on different days, grille only went back on when I was ready to focus on listening. Then I thought I should measure it again and got the surprise result albeit a nice one as it looked much smoother.

 

[Hobbit]

Agreed. The only issues I find one needs to be careful about when doing blind loudspeaker tests are:
1. the screen losses. If you don't have the speaker right up against the cover screen, it will cause HF losses and diffraction. I've seen and measured this and it caused the less bright speaker to lose a blind test while winning a sighted test.
2. listening fatigue. This can happen with sighted tests too. If you keep switching back and forth for long periods of time (longer than 15 min in my findings) the listener can get overloaded making it difficult to determine what they really prefer. This is why I like doing extensive testing over several days when the results between two speakers are very similar.

On the flip side, I had 3 inexperienced listeners compare both speakers sighted noting the MA was prettier but still picked the Pinnacle as better sounding b/c of the better bass and less constrained sound. These folks were brand agnostic but they did use their eyes and still picked the uglier speaker as better in a sighted test.

Yes I had someone walk into a sighted listening test between a pair of Monitor Audio Silver 5i's (mediocre speaker) and Pinnacle Classic Gold Reference (respectable speaker) thinking the MA would smoke it b/c of all the glowing reviews and brand prestige. He declared it was the better speaker before even listening. So I switched it up on him telling him he was listening to the opposite speaker of reality. The grilles were on both so he couldn't really tell which one was playing. He picked the Pinnacle thinking it was the Monitor Audio. When I showed him, he was quite shocked and cancelled his order.

One of the issues I have with DBT's is that you don't really have a reference. You do not truly know what the music your listening to should sound like. Therefore, a speaker with some coloration may be the one chosen as better. I recall Irv commenting once about playing instruments then seeing which speaker replicates those instruments best. I like this approach. But on that same note, it's why I tend to believe science and measurements trump the ear.

I run into this when adjusting my DSP guitar amp to editing photos in Photoshop. I think, the key word, that some adjustment was great only to come back a few days later and wonder what I was thinking!

 

[mtrycrafts]

One of the issues I have with DBT's is that you don't really have a reference. You do not truly know what the music your listening to should sound like.

Why is that not an issue in a sighted listening as well?

Therefore, a speaker with some coloration may be the one chosen as better.

And you could not choose that same speaker in a DBT? How do you pick that speaker out in a sighted test if not by listening? Will its brand name, size, aesthetics reveal this trait? How so?

I recall Irv commenting once about playing instruments then seeing which speaker replicates those instruments best. I like this approach.

But then, how would one know if that is a speaker issue or a recording issue?
Maybe the recording engineer has a hearing issue?

But on that same note, it's why I tend to believe science and measurements trump the ear.

Oh, the truth teller.

 

[Hobbit]

Why is that not an issue in a sighted listening as well?


And you could not choose that same speaker in a DBT? How do you pick that speaker out in a sighted test if not by listening? Will its brand name, size, aesthetics reveal this trait? How so?


But then, how would one know if that is a speaker issue or a recording issue?
Maybe the recording engineer has a hearing issue?


Oh, the truth teller.

OK, you picked apart my post. I'm not sure where you're going with this? Did I conceive the perfect test. Of course not, that wasn't my intent. Just like I didn't address DBT verses sighted. I also didn't address room correction/equalization, or the recording process et al. Even the presence of a different size speaker, variations in placement, etc, affect what you hear.

What are you advocating? If you're anti what I posted(?, not clear on that either), are you in the other extreme; the Phsycoacoustics corner? I do believe that a speaker can be tuned to appeal to the masses. A lot of science to get the speaker to sound that way too.

On that note, you didn't seem to address my final comments? Which go with the phsycoacoustic approach, and I freely admit I could, and do, fall for it, that I could pick a speaker with some form of coloration over one that's more accurately representing the music. Which of course, is hopefully an excellent rendition of the actual musicians playing. Yet, I'll probably grow tired of the colored version in short order.

Is it live or Memorex! That's how I want my music to sound - as close to how the artists wanted it to. I fully agree there's a chain of events to take us to that point. Many mistakes can be made along the way. I firmly believe however, science offers a better way to get to this point than my, or any ones ears do. It does seem the best speakers out there have a ton a science behind them.

 

[mtrycrafts]

OK, you picked apart my post. I'm not sure where you're going with this? Did I conceive the perfect test. Of course not, that wasn't my intent. Just like I didn't address DBT verses sighted. I also didn't address room correction/equalization, or the recording process et al. Even the presence of a different size speaker, variations in placement, etc, affect what you hear.

What are you advocating? If you're anti what I posted(?, not clear on that either), are you in the other extreme; the Phsycoacoustics corner? I do believe that a speaker can be tuned to appeal to the masses. A lot of science to get the speaker to sound that way too.

On that note, you didn't seem to address my final comments? Which go with the phsycoacoustic approach, and I freely admit I could, and do, fall for it, that I could pick a speaker with some form of coloration over one that's more accurately representing the music. Which of course, is hopefully an excellent rendition of the actual musicians playing. Yet, I'll probably grow tired of the colored version in short order.

Is it live or Memorex! That's how I want my music to sound - as close to how the artists wanted it to. I fully agree there's a chain of events to take us to that point. Many mistakes can be made along the way. I firmly believe however, science offers a better way to get to this point than my, or any ones ears do. It does seem the best speakers out there have a ton a science behind them.

Not sure how one can respond to a lengthy post without addressing specific statements in it as some stands by itself like One of the issues I have with DBT's is that you don't really have a reference.
This isn't a general statement but very specific in my mind. It implies you don't have this issue with non-DBT testing.
The rest follows equally except this But on that same note, it's why I tend to believe science and measurements trump the ear.
which is counter to everything else you posted, at least that is how it registers with me.
So, in essence, no real idea what you were getting at with your post.

 

[Hobbit]

Not sure how one can respond to a lengthy post without addressing specific statements in it as some stands by itself like One of the issues I have with DBT's is that you don't really have a reference.
This isn't a general statement but very specific in my mind. It implies you don't have this issue with non-DBT testing.
The rest follows equally except this But on that same note, it's why I tend to believe science and measurements trump the ear.
which is counter to everything else you posted, at least that is how it registers with me.
So, in essence, no real idea what you were getting at with your post.

It's hard to always be as concise as you'd like in a short answer. I didn't think I implied anything about a DBT vs non DBT. To me, the "reference" is how the music would actually sounds live (preferably in a very good acoustically sound room!). Not necessarily one speaker to another unless I can be convinced one speaker is a true "reference" level speaker.

My point is I liked Irv's way of thinking, and it's a way of thinking I would use to try and make a scientific measurement, e.g. use equipment to get good spectrums of a violin and see how close the speakers can match those. This will never be perfect for many reasons, but I can't repeatedly make the same exact measurement in my lab even using very expensive test equipment. However, it is very close and very accurate. I don't believe my ear can outdo precise acoustic measuring equipment.

 

[mtrycrafts]

It's hard to always be as concise as you'd like in a short answer. I didn't think I implied anything about a DBT vs non DBT. To me, the "reference" is how the music would actually sounds live (preferably in a very good acoustically sound room!). Not necessarily one speaker to another unless I can be convinced one speaker is a true "reference" level speaker.

OK, but then why even mention DBT? It applies to everything and then there is only one case for a reference, playback in the studio if in fact the whole group is present. So, it is next to impossible to have a reference in your context.

My point is I liked Irv's way of thinking, and it's a way of thinking I would use to try and make a scientific measurement, e.g. use equipment to get good spectrums of a violin and see how close the speakers can match those. This will never be perfect for many reasons, but I can't repeatedly make the same exact measurement in my lab even using very expensive test equipment. However, it is very close and very accurate. I don't believe my ear can outdo precise acoustic measuring equipment.

You are right about your ears. or mine for that matter.
Not sure why Toole's approach is not accomplishing this. After all, an instruments spectrum is a bunch of frequencies whether fundamental or harmonics. Measuring a speakers spectrum as he does would be the same. And, he can repeat his.

 

[RHSmith]

One of the things I have come away with over the many years I have been playing with audio (50+ now.. sigh..) is that well thought out and implemented testing often reveals traits that may/may not sound "right" to a listener.. A flaw in a speaker (or any other item within the audio chain for that matter) that 'adds character, is just that .. a flaw... and to some it is an acceptable sound... That is not to say that the flaw is good.. it is more often acceptable because it is a character that is 'familiar and comfortable" to the listener.. and there in lies the subjectivity issue.. We become accustomed to a particular sound and biased towards it and unwilling to accept that it is wrong.. I have found many who consider Bose 901's to be wonderful speakers in spite of the all well known flaws in them and they keep going back to them because they are what they are used to... and no matter how bad they sound next to other speakers they just do not want to hear what they are missing.. You have to be willing to change and accept that what you hear may not be correct in some way and willing to move on to the next phase... Often what you hear when you change the speaker is the new speaker revealing something else in the chain that is not right...

 

[Irvrobinson]

OK, but then why even mention DBT? It applies to everything and then there is only one case for a reference, playback in the studio if in fact the whole group is present. So, it is next to impossible to have a reference in your context.

It is not next to impossible to have a reference, you just make your own recordings in your room. You have a reference of the live sound, which makes the reference sound repeatable, and a good speaker should get as close as possible to fooling you about which is which.

That said, comparing to a reference is not a scientific method; one is still using imperfect human perception to judge accuracy. However, if you are pursuing accuracy, not just sounding good, I know of no better method. Combined with measurements, both professionally measured and in-room, you can begin to get a useful picture of accuracy and relative quality.

As for audio DBTs, I just don't see the value in them at all. I agree with Hobbit, being able to detect differences and "preference" is useless in my mind. Having a preference between two inaccurate sources is irrelevant, IMO.

 

[PENG]

As for audio DBTs, I just don't see the value in them at all. I agree with Hobbit, being able to detect differences and "preference" is useless in my mind. Having a preference between two inaccurate sources is irrelevant, IMO.

I understand you don't see the value but I am sure some do. To me the value is limited to proving whether the often claimed night and day difference to noticeable difference between preamps, power amps, CD quality, 24/96, 24/192, HDCD qualities etc can survive such tests. I am not saying that the claimed difference are real or not real regardless of the test results, but the tests have value for some curious minds depending on what they are curious about.

 

[Irvrobinson]

I understand you don't see the value but I am sure some do. To me the value is limited to proving whether the often claimed night and day difference to noticeable difference between preamps, power amps, CD quality, 24/96, 24/192, HDCD qualities etc can survive such tests. I am not saying that the claimed difference are real or not real regardless of the test results, but the tests have value for some curious minds depending on what they are curious about.

I understand. There are still some people hung up on proving that all electronics sound alike, and I suppose audio DBTs make them feel better about their conclusions. As you know, I'm both a believer that there can be audible differences between some electronics, and I'm also a believer that as a past multiple instance audio DBT participant, I could never reliably detect differences between well-designed electronics in any sort of comparative trial, DBT or not.

But... for speakers, which we know sound different from each other, I still think DBTs are useless. As Hobbit says, you need a reference.

 

[herbu]

I believe it is a matter of education. In the beginning, each of us can listen to different systems and decide which sounds best. But at this point, "best" is invariably just personal preference, often colored by brighter highs and enhanced bass.

Then we begin to learn about accuracy. And if we're lucky, we have someone to teach us. Now we begin to listen for accuracy, and our preference always starts to sway in that direction. The more we learn, the more we sway.

Finally comes the epiphany. It is our money, our system, our ears, and accuracy doesn't necessarily trump personal preference. Now we begin to moderate our bias toward accuracy w/ a portion of personal preference. And this is where the debates heat up.