Who would buy a speaker without listening to it?

[Spocko]

Yep, now I always buy speakers without listening to them. I've done my share of showroom listening before buying speakers in my early days and they've never sounded like what I heard once I got them home because the earlier version of me never realized how much the room and room treatment affected the speaker's soundstage and bass (who knew I was sitting in a 120Hz null??). For critical listening where the objective is the source not the speaker, then I look at measurements that have controlled directivity with dispersion characteristics I want because this is a professional tool.

HOWEVER, for casual listening where it's all about having fun and making the speaker a point of pride for discussion and socializing, then I get whatever gets the conversation started like any of these!

 

[ahofer]

So one question has been whether there are listeners who can translate measurements accurately into sound impressions and vice versa.

Another, raised by this genelec-Neumann discussion, is whether each could be EQ’d so as to be indistinguishable from the other?

 

[Digby]

So one question has been whether there are listeners who can translate measurements accurately into sound impressions and vice versa.

Another, raised by this genelec-Neumann discussion, is whether each could be EQ’d so as to be indistinguishable from the other?

Yes, to me they are quite interesting questions, but there seems to be little interest as to the answers, save from a handful of people. Why the lack of inquisitiveness?

I'll add another question - what is an 'accurate' loudspeaker? If two well measuring speakers of a similar size (say the KH120 and 8030C) are placed within a room at the same point, yet produce a different sound when placed at the same position, which of the two is more accurate?

There is talk of accuracy as a binary yes or no, in that up to a certain ill-defined point a speaker is inaccurate, after this point everything is just "accurate". Surely accuracy can be measured more finely than that? It is a very complicated matter, but if two well measuring loudspeakers are placed at the same position in the same room and produce different sound qualities, it follows logically that they cannot be equally accurate.

Truly equal accuracy would denote the exact same sound qualities in any given space.

As no speaker is totally accurate, maybe discussion should be of greater and lesser inaccuracy in specific parameters (FR, distortion, directivity and so on). Given all the ways in which speakers are more or less inaccurate, then it is likely that some will prefer certain balances of inaccuracies to others. A good example of this is Pearljam5000 finding the KH120 to be 'dark' and Genelec speakers more to his liking. Neither would be considered deficient from the graphs, but whatever inaccuracies he hears in the KH120, is more of a problem for him. For the next person, the opposite may be true and they will prefer Neumann to Genelec.

I don't think there is anything approaching an audio panacea, in that if you just pick an 'accurate enough' speaker, they will all be much of muchness or the listener should just adapt if they are displeased. This seems to me more likely to be a misunderstanding of the function measurements provide.

I'd wager that if the measurements are approached in a blind fashion, in that all speakers beyond the (ill defined) point of 'accurate', are good enough to please anyone, then you'd probably be better with some rules of thumb collected about preference (what directivity do you like? Do you prefer brighter or darker speakers and so on). You may sooner arrive at speakers you like the most this way.

I don't think the measurements are as blunt a tool as is suggested by some. There is lots of information held within - it is up to us, as individuals, to tease out the information that is most pertinent to our preferences and circumstances. This will differ depending on the individual.

Maybe measurements should be approached more as a map, rather than directions to a specific location.

 

[YSC]

Yes, to me they are quite interesting questions, but there seems to be little interest as to the answers, save from a handful of people. Why the lack of inquisitiveness?

I'll add another question - what is an 'accurate' loudspeaker? If two well measuring speakers of a similar size (say the KH120 and 8030C) are placed within a room at the same point, yet produce a different sound when placed at the same position, which of the two is more accurate?

There is talk of accuracy as a binary yes or no, in that up to a certain ill-defined point a speaker is inaccurate, after this point everything is just "accurate". Surely accuracy can be measured more finely than that? It is a very complicated matter, but if two well measuring loudspeakers are placed at the same position in the same room and produce different sound qualities, it follows logically that they cannot be equally accurate.

Truly equal accuracy would denote the exact same sound qualities in any given space.

As no speaker is totally accurate, maybe discussion should be of greater and lesser inaccuracy in specific parameters (FR, distortion, directivity and so on). Given all the ways in which speakers are more or less accurate, then it is likely that some will prefer certain balances of inaccuracies to others. A good example of this is Pearljam5000 finding the KH120 to be 'dark' and Genelec speakers more to his liking. Neither would be considered deficient from the graphs, but whatever inaccuracies he hears in the KH120, is more of a problem for him. For the next person, the opposite may be true and they will prefer Neumann to Genelec.

I don't think there is anything approaching an audio panacea, in that if you just pick an 'accurate enough' speaker, they will all be much of muchness or the listener should just adapt if they are displeased. This seems to me more likely to be a misunderstanding of the function measurements provide.

I'd wager that if the measurements are approached in a blind fashion, in that all speakers beyond the (ill defined) point of 'accurate', are good enough to please anyone, then you'd probably be better with some rules of thumb collected about preference (what directivity do you like? Do you prefer brighter or darker speakers and so on). You may sooner arrive at speakers you like the most this way.

I don't think the measurements are as blunt a tool as is suggested by some. There is lots of information held within - it is up to us, as individuals, to tease out the information that is most pertinent to our preferences and circumstances. This will differ depending on the individual.

Maybe measurements should be approached more as a map, rather than directions to a specific location.

To me the speaker alone is similarly accurate, and the difference comes from the room, mainly from dispersion width and degree of directivity error.

I am not sure about actual beam width, but it seems the KH120 have narrower dispersion compared to 8030C which will affect reflected sound from sidewalls casting different perception above schoreder frequency, and it also seems that on axis, KH120 digs some 10hz deeper than 030C, and in the limited measurement available, KH120 have a slight bass bump before the bass roll off a cliff, all these would contribute to the impression and bass boost from front wall.

measurement arn't perfect in predicting in your room sound as room varies so much and speaker placement would have to be different (the driver surface vs various surface in your room have to be different even you put them on the same stand, as speakers have different sizes and driver height & arrangement), but that is even more of a problem when you listen/audition a pair in anywhere else than your intended room.

So ended up back to the original question: would you buy a speaker without ever listening to one: yes, of course, listening in store is more a wild bet than reading the data (provided those are detailed data)

But if you have a part B question: if yes, would you like to have in your own room audition before final purchase decision?
then my answer is a sure "yes"

it's like if it's a question of listening in some other place vs data I would choose data 10/10 times, but if it's a combination of both, then sure having both is a better way

 

[Digby]

To me the speaker alone is similarly accurate, and the difference comes from the room, mainly from dispersion width and degree of directivity error.

What about outdoors in the open, with no room in the equation. Wouldn't they still sound different then?

So ended up back to the original question: would you buy a speaker without ever listening to one: yes, of course, listening in store is more a wild bet than reading the data (provided those are detailed data)

But if you have a part B question: if yes, would you like to have in your own room audition before final purchase decision?
then my answer is a sure "yes"

How about a part C. Do you think if provided with measurements and the ability to listen in a shop beforehand, that you could make a better choice than just from measurements alone?

Listening in any other room than your own is imperfect, that is true, but does enough of the sound character of a speaker transfer (between different rooms), such that you can get a better understanding of what you would prefer, than from just the measurements?

it's like if it's a question of listening in some other place vs data I would choose data 10/10 times

I know you are quite limited for space, so your speaker choice was informed by this, but imagine you had a room around 20x15x10ft.

Given 5 high performing, similar sized speakers to listen to, blind, in your room - Do you think you'd be able to pick your favourite pair of speakers in the listening test, before any listening was actually done, purely from the measurements alone?

 

[YSC]

What about outdoors in the open, with no room in the equation. Wouldn't they still sound different then?


How about a part C. Do you think if provided with measurements and the ability to listen in a shop beforehand, that you could make a better choice than just from measurements alone?

Listening in any other room than your own is imperfect, that is true, but does enough of the sound character of a speaker transfer (between different rooms), such that you can get a better understanding of what you would prefer, than from just the measurements?


I know you are quite limited for space, so your speaker choice was informed by this, but imagine you had a room around 20x15x10ft.

Given 5 high performing, similar sized speakers to listen to, blind, in your room - Do you think you'd be able to pick your favourite pair of speakers in the listening test, before any listening was actually done, purely from the measurements alone?

for open space, IMO with limited speakers, they actually don't sound that different than in room, mostly can't distinguish for myself though maybe due to the environment noise.

for Part C, I don't think with measurements and able to listen in a shop, I don't think the decision will be better, rather mostly worse or equal, say in a well absorbed room vs a less treated room at home, I would prefer a darker speaker, say slightly sloped downward on axis, and at home with less bass traps, would likely prefer a less bassy speaker, especially when the contenders in the budget isn't perfectly neutral on axis, or maybe in a well dapmed room, prefer a better extended but have significant directivity error, where in room that can cause EQ problems etc.

if it's picking top 5 and played blind in my own room before EQ or so in the same spot, I do think listening could make me confirm the final preferred one, but I would believe I could be happy for any in that list. I listen to music, not the gear nor tinkering them here and there often

 

[ahofer]

I agree that fidelity, once you get to the speaker-room interaction, is a more abstract concept. After all, you are mixing room reflections with reflections recorded in the studio or hall in a unique combination.

However, I think Toole’s research (which involved too few people/trials to be fully satisfactory) suggests that most prefer the more even directivity speaker with flatter frequency response translated to declining in-room response. If we take that as given, we have a model of *preferred* measurement to work with. The first question you asked is whether listeners could reliably identify that speaker, and those that deviate from the model, by looking at the measurements then listening. My hypothesis is that this would require a highly trained listener.

Similarly, if the directivity is even, and we are starting with a flat, clean response (from a high fidelity signal), the user has flexibility to color the FR to their tastes. You can’t fix directivity with EQ, but you sure can improve/tailor FR To your liking. Hence the second question: If you start with two superb measuring speakers (by our model) like Neumann and Genelec, could you alter the EQ so that they were indistinguishable in a blind test.

So, to me, these questions/experiments seem like a vital part of the continuing Audio Science.

Something like this may have been done, I have certainly not read all the literature.

 

[symphara]

audition IMO only matters when the audition is in your own place, own setup, coz that way you have the fixed room effects in play, measurements done 95%+ of decision making considering extension, neutralness, distortion, directivity and SPL capability, audition only helps you to decide between say, top 3 of your own choice and get the best one out. more importantly it makes the educated guess of what you want, so you can then focus on things like connectivity, extra functions, price, WAF etc.

It's a given that by changing the room you change the sound to some extent, but I don't agree with the idea that listening at the dealer's is useless.

This is postulated around here but I don't see why. I think listening at the dealer gives me a pretty good idea about the speaker. If anything, I think all my auditions were in a wider space, with the speakers placed further away from the back wall (people need easier access to move them around, as opposed to the static installation at home), so I get more prominent bass at home but otherwise similar sound.

To put it another way, I never listened to a speaker at the dealer which at home made me say "whooa what crap did I just buy". If anything, my experience has been the opposite.

Not to say it's not ideal to get a home demo, but I think this is more an American advantage, with a huge market and lots of consumer power.

 

[YSC]

It's a given that by changing the room you change the sound to some extent, but I don't agree with the idea that listening at the dealer's is useless.

This is postulated around here but I don't see why. I think listening at the dealer gives me a pretty good idea about the speaker. If anything, I think all my auditions were in a wider space, with the speakers placed further away from the back wall (people need easier access to move them around, as opposed to the static installation at home), so I get more prominent bass at home but otherwise similar sound.

To put it another way, I never listened to a speaker at the dealer which at home made me say "whooa what crap did I just buy". If anything, my experience has been the opposite.

Not to say it's not ideal to get a home demo, but I think this is more an American advantage, with a huge market and lots of consumer power.

I don't think it would be to the extend of "whooa what crap did I just buy", but more like when in the final short list, where any of them would sound great in one's own home and satisified, it would be useless to decide which one is actually best in one's home, given more often than not the placement freedom and room treatments are limited in home.

IME where anechoically neutral speakers, when put in a room, not EQed have quite some huge dips and nulls depending on the degree of treatment done and proximity to shelves, walls etc, as long as directivity is good, the distance to sidewall likely dominate the ended up in room response and should not be too difficult to tune it to liking, maybe it's in my city ppl live in cramped concrete walled apartments with weird geometry, I always found the speaker in showroom vs at the music lover friend's home differ a lot more than signature difference between speaker A vs B in same level of objective performance, with that in mind I think that showroom audition don't make it less of a guess job than buying solely on data.

just to say despite I ended up having the 8030C, I once helped my friend order a pair of Adam T5V and tried it in the same room, I would say if I somehow didn't get the 8030 at first but the Adam, I likely would be similarly happy and don't think it disappointed me

 

[Tatr76]

Just bought some new Focal aria 926 floorstanding speakers for study so have something to listen to while work. Never heard any focal let alone these, bad practise I know but only £1199 so thought I would take a punt. Very few places to audition gear these days sadly let alone take home and hear in your specific room.

 

[killdozzer]

Again, without listening, after reading these two statements:

Ah, what a joy this speaker is to listen to. It has plenty of warm and impactful bass, yet is almost perfectly neutral.

If you want to get a taste of accurate sound production that manages to delight, the JBL 308P MKii is a wonderful entry into this world.

Ordered them and bought them. They've arrived. I'll keep you posted. They're meant for a small home studio which, I must say, never sounded better ever.

Another thing, I'm in the middle of the process of swapping non-recommended LS50 for recommended LS50 Meta after a good review, but this was a safer bet for me, because I already liked LS50 and some improvement never hurt anyone (and I found a very good price).

 

[killdozzer]

Another thing, I'm in the middle of the process of swapping non-recommended LS50 for recommended LS50 Meta after a good review, but this was a safer bet for me, because I already liked LS50 and some improvement never hurt anyone (and I found a very good price).

Holy cannoli!! And I never thought I'd utter such a silly expression, but I have to move this to the deals, deals... thread. These are brand new!!! I didn't expect that.

 

[mj30250]

Virtually every speaker I've owned I bought sound-unheard. Of course, up until more recent years, I generally based my decisions on user and professional reviews (within a given budget), as I had little to no idea what frequency response, directivity, early reflections, etc, meant, and if you'd have asked me to interpret a spinorama during those times to the best of my ability, my response would likely have been as sensible as if you were to ask me to interpret the results of an MRI brain scan.

Having (somewhat) educated myself on the subject, and having auditioned multiple speakers that have been measured in sufficient detail since that time, I'm fairly confident that I could select speakers that I'd be happy with (and would prefer over others) strictly based on their measurements, a large majority of the time (completely arbitrary percentage: 80%). To me, what's more interesting to ponder is how often I'd dismiss a speaker that measures poorly, but, were I to actually listen to it, I'd feel still sounds "good". I bought a particular speaker over two years ago that I quite liked and found no particular issue with. At the time of purchase, no measurements for it were available. About a year later, it was Klippel-measured, and it turned out to have some major objective issues / deviations. By then, I had already sold it off, not because of any specific problems or because I ended up not caring for it, but simply as a result of upgraditis. I wonder if I were to listen to it again, would my knowledge of the measurements suddenly render its sound dramatically worse to me - or would it still sound just as / almost as "good" as I remembered?

 

[RayDunzl]

Speaker coloration can be somewhat EQed (or you might just like it) but distortion cannot.

I might rate distortion correction difficult, but maybe not impossible, from this old experiment:

Using REW, send a 450Hz tone at -20dB to the system (JBL LSR 308 speakers), take an RTA with UMIK-1.

Note high level of 3rd harmonic distortion in the speaker output.

It is clearly audible.

Using the Distortion controls, add third harmonic to the signal at -40dB and (coincidentally) at a -40 degree phase angle.

The resulting tone was audibly pure.

--

Map the distortions a speaker produces with pure tones.

Pay @pkane to create an anti-distortion mapping with a modified version of his Distort software, apply to musical signal and see what happens.

 

[symphara]

I might rate distortion correction difficult, but maybe not impossible, from this old experiment:

Using REW, send a 450Hz tone at -20dB to the system (JBL LSR 308 speakers), take an RTA with UMIK-1.

Note high level of 3rd harmonic distortion in the speaker output.

It is clearly audible.



Using the Distortion controls, add third harmonic to the signal at -40dB and (coincidentally) at a -40 degree phase angle.



The resulting tone was audibly pure.

--

Map the distortions a speaker produces with pure tones.

Pay @pkane to create an anti-distortion mapping with a modified version of his Distort software, apply to musical signal and see what happens.

You did it for one spike of one tone by adding something to the signal. How are you going to do this for music?

BTW was your distortion generated by the speaker, or by electronics? To me it looks like the latter.

 

[RayDunzl]

was your distortion generated by the speaker, or by electronics? To me it looks like the latter.

First picture: The initial distortion was the speaker playing a pure sine wave. The little JBLs sing their own tune readily.

The anti-harmonic-distortion electrical signal was added to the original pure sine signal in REW.

Second Picture: Speaker playing tone with anti-distortion signal added.

 

[RayDunzl]

You did it for one spike of one tone by adding something to the signal. How are you going to do this for music?

Map the distortion products at various levels using a sweep and fancy software.

Map the anti-distortion signals required for frequencies and levels.

Analyze the musical signal for frequencies and levels, and add the appropriate anti-distortion correction on the fly.

I can't do it.

@pkane probably could, if he found it to be an amusing endeavor.

---

Or buy low distortion speakers:

JBL LSR 308

Martin Logan hybrid

 

[symphara]

First picture: The initial distortion was the speaker playing a pure sine wave. The little JBLs sing their own tune readily.

The anti-harmonic-distortion electrical signal was added to the original pure sine signal in REW.

Second Picture: Speaker playing tone with anti-distortion signal added.

That doesn't answer my question. I can see that REW shows the unwanted harmonic, but you don't say whether it's coming from the speaker or generated by your playback electronics (e.g. DAC).

Adding what you call "anti-distortion" signals clearly wouldn't work for music, since you'd need to add a massive amount of stuff all over the place, in real time, and that would of course interact with your actual music.

Unless, of course, your ideea is brilliant and I don't get it, in which case you should just ignore me and patent/develop it further.

 

[RayDunzl]

That doesn't answer my question. I can see that REW shows the unwanted harmonic, but you don't say whether it's coming from the speaker or generated by your playback electronics (e.g. DAC).

The JBL and Martin Logans share everything up to the preamp output. the JBLs are "active", the ML are passive.

Adding what you call "anti-distortion" signals clearly wouldn't work for music, since you'd need to add a massive amount of stuff all over the place, in real time, and that would of course interact with your actual music.

It would be a challenge to code, I agree.

Unless, of course, your ideea is brilliant and I don't get it, in which case you should just ignore me and patent/develop it further.

I'm not that ambitious, nor, even more importantly, skilled with modern coding techniques.

I freely distribute the idea to whoever can make it happen.

---

The JBLs are my daily drivers. TV, HDRadio, whatever.

I don't normally play the JBLs loudly enough to be annoying, but when I do, I don't, and fire up the Krells and the 'stats instead.

 

[pkane]

Pay @pkane to create an anti-distortion mapping with a modified version of his Distort software, apply to musical signal and see what happens.

Mostly done, Ray! I have been testing this as part of Multitone for now. It corrects for a static nonlinearity. Works quite well. Just a few days ago I tested it with an amp that improved THD from around -100dB to -115dB, and of course, also corrected IMD Oh, and yes, of course it also works with music.