Why evaluating the sound of a single speaker is essential

[RexrothPigeon]

If even inferior speakers sound comparable to well engineered ones in a multi speaker setup, why spend more?

I wrote "crappy" not "cheap" in the post you quoted. Crappy could be more expensive than a well engineered one. This question is not primarily about saving money. It's about the findings of the research. Money savings are a potential side effect, but not the main purpose of the question.

see above, the full quote to which i was responding: "why spend more?"

 

[Mnyb]

Uppmixed content is still mostly L & R with ”fillers” from the other channels possibly more from C . And allmost speakers sounds better with subs ( correctly integrated )

But I feel that it really was not about that at all , just a sneaky way off claiming single speaker testing is bad by willfully ignore the nuances in this . It’s not an 1 or 0 situation.

It’s not like if you have more than 1 speaker it’s impossible to discern it’s quality. It’s easier with 1 for development and testing purposes .
And you can get very solid statically valid results which.

I’ll guess for a realy crap speaker it would take seconds to spot the problems in one speaker but in a multichannel setup with a movie to distract you you could probably watch a movie with grooving sense if dissatisfaction that would take a while to realise why

 

[Vuki]

It's funny that no one wants to touch the original question. It was not about money per se. It was about the utility of a well engineered speaker in a multi channel setup. The money question is related, but not that important here. It could be the opposite for all I care. The objectively inferior speakers could be a Martin Logan and the superior speaker could be the Asci Lab F6b at a fraction of the price.
According to the graphic, the Martin Logan would sound every bit as good as the Asci Lab once they were used in a home theater situation with five or more playing at the same time, even though it's a speaker full of wild resonances.

That's what happens when Pink noise is your most revealing track.
Hold your fire - just kidding!

 

[amirm]

Would you say that it's wise to not go for the objectively best speakers if one is putting together a multi channel system, in view of the results above?

No. The masking occurs when the other channels are playing. In many cases such as vocals in movies, they play without other speakers whaling. And of course there is stereo music.

 

[MarkS]

I get that these are well engineered speakers and would do great in a home theater setup. The original question is why bother with well engineered speakers if the fact is that five of them sound subjectively similar to a set of clearly inferior speakers that might cost much less. That's what the graphic in question conveyed.

I would like to know more details about the listening experiments that were done. Was music originally recorded in 2 channel used as a source? What speakers were used and how well did they measure, both the good ones and the less good ones?

 

[amirm]

I would like to know more details about the listening experiments that were done. Was music originally recorded in 2 channel used as a source?

No, they were multichannel DVD-A (Audio) discs that were ripped. From the paper:

"Three musical selections were chosen for these
experiments, all originating from commercially
available multichannel DVD-A and DVD discs (see
Table 2)."

 

[MarkS]

Good to know that 2-channel music was not tested. The results should therefore not be assumed to apply in that case.

 

[Newman]

Good to know that 2-channel music was not tested. The results should therefore not be assumed to apply in that case.

What results don't apply to what?

 

[amirm]

What results don't apply to what?

I *think* he means upmixed stereo was not used but rather, true multi-channel music.

 

[Vladimir Filevski]

The original question is why bother with well engineered speakers if the fact is that five of them sound subjectively similar to a set of clearly inferior speakers that might cost much less.

Because:

Get five Ascilab F6Bs speakers for $350 each. @amirm's conclusion: "Amazing what happens when you follow audio science and engineering to the max: you get a very reasonably priced, gorgeous speaker that is near perfect objectively."

 

[Newman]

I *think* he means upmixed stereo was not used but rather, true multi-channel music.

Could be, but he has a track record of repeatedly insisting that Toole's own interpretation of his own data (in favour of single speaker evaluation) is wrong, and says it really shows that stereo listening is critical. One example. I have tried to address this when it pops up.

Hence, it could be that he is expressing relief that the experiment excludes 2-channel music, so cannot be applied to 2-channel music, so he can stick to his prior interpretation. I would like to know how he currently feels about single speaker evaluation for 2-channel applications.

cheers

 

[Floyd Toole]

Sorry to bump up an old post in this thread, and this might have been brought up earlier, I just didn't read through all 42 pages.
Would you say that it's wise to not go for the objectively best speakers if one is putting together a multi channel system, in view of the results above? If even inferior speakers sound comparable to well engineered ones in a multi speaker setup, why spend more? If one plans to always have all the speakers on, even when listening to stereo, it makes sense to not overspend as the end effect will be subjectively comparable.

Somewhere I have stated the additional fact that this is true only when all channels are active, not just connected. The activity is determined by the recordings and all recordings have intervals of monophonic sound from single channels. In movies the center channel is a dominant factor and it is a mono source. Any hard-panned image is a mono source. So the reality is that neutral loudspeakers still win stereo and multichannel tests--one still needs neutral loudspeakers. Go to YouTube and watch my Toronto AES lecture. In it this is mentioned. Apologies for my coughing at the end.

This is well explained in the 4th edtion.

Fortunately, very neutral loudspeakers can be found at low prices. The spinorama data published by Amir, Erin and others can help you find them. No need to compromise if you put some effort into it.

 

[LevityProject]

One thing I've done to evaluate speakers (and my own mixing) subjectively is to listen to the tonality of Grammy winners over the last 20 years in the engineering category.

These offer excellent fidelity across genres. I personally look at the ones from 1990-present, but looking back further is interesting as well.

Grammy Award for Best Engineered Album, Non-Classical - Wikipedia

en.wikipedia.org

 

[Philbo King]

No, they're absolutely valid. I did not deny the efficacy of measuring a single speaker. I am absolutely stating that no one listens to single speakers when they listen to music, it's half of what the consumer purchased and therefore, is not representative of what someone would experience in their home.

It is therefore not a test 95% of consumers would consider valuable.

If you cannot get off your high horse to understand that basic argument, then that's your issue.

The market for blutooth speakers would be contrary evidence

 

[Hayabusa]

The market for blutooth speakers would be contrary evidence

Many of them still do some stereo..

 

[PGansz]

The market for blutooth speakers would be contrary evidence

Valid point, but different consumer market.

I don't see anyone toting around bookshelf or satellite speakers to the beach.

 

[Floyd Toole]

It is therefore not a test 95% of consumers would consider valuable.

I thought the mono vs stereo issue was dead and buried, but I guess not. I would like to think that 100% of consumers would prefer to have loudspeakers that are free from timbre distorting resonances, and discovering those resonances was found to be much easier when comparing loudspeakers in blind, equal loudness, multiple-loudspeaker (3 or 4 at a time) comparisons in mono, not stereo.

Those loudspeakers rated highest in the mono tests also rated highest in stereo and multichannel tests, but discovering the resonances in those modes was much more difficult and time consuming. The loudspeakers that have been preferred in decades of these evaluations are the technically most accurate products. Now we can identify fundamentally neutral, resonance free, loudspeakers by inspecting anechoic spinorama data alone.

Obviously, the scientific background has not been read and understood.

Nobody, certainly not me, is suggesting that listening in mono is more pleasurable than stereo or multichannel. It was a means to an end and the end has been achieved.

 

[mglobe]

...

Obviously, the scientific background has not been read and understood.

...

Truth

 

[No. 5]

Obviously, the scientific background has not been read and understood.

Indeed. It sometimes appears that home audio enthusiast conflate equipment reviews and scientific investigation.

 

[MattHooper]

I would be surprised if most here had not noticed in their own experience that hard panned voices and instruments (especially voices) seem to be particularly revealing of a loudspeaker’s quality in various respects. (which is consistent of course with Dr. Toole’s writings).

Admittedly, I’ve rarely evaluated a loudspeaker in mono (just a few times), but hard panned content has always been a part of how I evaluate the character of a loudspeaker. If a hard panned voice sounds unnaturally resonant or stuck in the speaker, that’s not a good sign. (And I’ve noticed that when this is the case, soundstaging/imaging tends to take on more of a “U” shape - as objects in the sound field spread towards the left or right they tend to sort of be “ sucked towards the speakers.”)