I had a Alpair 7 moap in the Markaudio bookshelf cab and Kef ls50. I can't get the Alpair to sound good in my room, the sound is a bit tilted to the top and lack of reasonable bass. The LS50 is a much more balanced speaker and I much prefer it over the Alpair. I sold the Alpair and kept the LS50. I got both as a second hand speaker and the price for the LS50 is more than twice the Alpair, to put that in perspective.
And the Kef does need a decent amp to sound good.
And the Kef does need a decent amp to sound good.
AES E-Library >> Intermodulation Distortion Listening Tests
Don't be mislead by the title. The important stuff is in the Appendix which da false prophets Floyd & Olive reference too.
The reason why all this stuff appears in the Appendix was we were going to present 2 papers originally.
We did a lot more work. I know that mid & HF THD is 'relatively' unimportant and there are speakers with high measured THD which are described as 'low distortion' compared to speakers with very low THD in DBLTs.
Does anyone here seriously think measurements tell you how a speaker sound?
When you start designing speakers for real, there is a huge milestone when you can first do proper 'anechoic' frequency responses and you design your first speaker with flat response. Surely this must be the best speaker in the world.
If you are lucky, you might be able to do a direct comparison with a really good commercial speaker eg LS50 .. preferably in a properly conducted DBLT.
What happens is you then spend the rest of your life trying to work out why your speaker with flat frequency response doesn't sound so good
IMHO, only two opinions here are worth listening to ... those who have managed to hear these 2 speakers at home with their own music. Discount ALL other opinions.
When you start designing speakers for real, there is a huge milestone when you can first do proper 'anechoic' frequency responses and you design your first speaker with flat response. Surely this must be the best speaker in the world.
If you are lucky, you might be able to do a direct comparison with a really good commercial speaker eg LS50 .. preferably in a properly conducted DBLT.
What happens is you then spend the rest of your life trying to work out why your speaker with flat frequency response doesn't sound so good
IMHO, only two opinions here are worth listening to ... those who have managed to hear these 2 speakers at home with their own music. Discount ALL other opinions.
Quote:
Does anyone here seriously think measurements tell you how a speaker sound?
I have only heard a few knowledgeable people say they could develop a good speaker without listening and they probably could, but even they still listen extensively before the design is finalized. I think the measurements vs. ears thing is a false dichotomy.
Does anyone here seriously think measurements tell you how a speaker sound?
Measurements are a lot more useful then subjective commentary. I can't tell much from your speakers, your music, your room, your judgment, your priorities. A good measurement is at-least an impartial starting point.
I have only heard a few knowledgeable people say they could develop a good speaker without listening and they probably could, but even they still listen extensively before the design is finalized. I think the measurements vs. ears thing is a false dichotomy.
Measurements do say a lot, and yes you can make a pretty good sounding speaker with measurements alone (especially in an active setup with DSP)
The issue I see with fullrange drivers in general is that a lot of new drivers are engineered/manufactured to people who are nostalgic and like listening to older, compromised drivers. Fostex is a great example of this. Then a lot of other 'audiophile' brands follow this - as opposed to an engineered fullrange driver (like the ones in my 20 year old car) that sound amazing in that application. So much so, that it's hard to find a decently priced upgrade from them
It also doesn't help that a lot of audiophiles with full range drivers listen to very simple almost single instrument material, where the colourations sound favourable to them, but then it turns into a muddy mess when anything slightly busy/challenging is played
The issue I see with fullrange drivers in general is that a lot of new drivers are engineered/manufactured to people who are nostalgic and like listening to older, compromised drivers. Fostex is a great example of this. Then a lot of other 'audiophile' brands follow this - as opposed to an engineered fullrange driver (like the ones in my 20 year old car) that sound amazing in that application. So much so, that it's hard to find a decently priced upgrade from them
It also doesn't help that a lot of audiophiles with full range drivers listen to very simple almost single instrument material, where the colourations sound favourable to them, but then it turns into a muddy mess when anything slightly busy/challenging is played
Last edited:
This is an important point that is often overlooked. It's not that hard to build a fullrange speaker that sounds pleasant with simple music, but with complex music at high volume they usually fall apart.
Why do people always assume that if someone puts a mic in front of a speaker, we are looking for a flat FR?
Even if I have a ruler at home, it doesn't make me a flat-earther!
Without measurements, we are back in the dark ages, trying to design an enclosure without TS params, time alignment and such.
I know what I like. I have my own signature FR that is my own set of coloration that my ears find pleasing, my own Audio recipe. I can look at someone else's FR and if I see emphasis in certain regions, I know I won't like that speaker.
Even if I have a ruler at home, it doesn't make me a flat-earther!
Without measurements, we are back in the dark ages, trying to design an enclosure without TS params, time alignment and such.
I know what I like. I have my own signature FR that is my own set of coloration that my ears find pleasing, my own Audio recipe. I can look at someone else's FR and if I see emphasis in certain regions, I know I won't like that speaker.
I spent much of my previous life from the late 70s trying to devise measurements which had good correlation to the results of DBLTs including KEFplots (CDS spectra), PAFplots, the directivity waterfalls bla bla.
In the late 90s, I said, the very best measurements can tell a good speaker from a bad speaker. But a trained listener (ie not a wannabe Golden Pinnae like most, all?, reviewers) can tell this in 10s flat.
What the most advanced measurements can't distinguish is a good speaker from an excellent speaker. For that you need DBLTs. This hasn't changed in da 21st century.
Designers who work for the big companies like KEF & Celestion use all dem fancy tools & measurements. ALL their designs from the cheapest to most expensive will fullfill da false prophets Floyd & Olive's criteria for good speakers. But it is rare for one of their designs to really excite them.
In da previous Millenium, only the original 104, 104ab and to a lesser extent the original 105 did this for KEF engineers. Lesser for the 105 cos though it had lower THD, more extended LF bla bla ... it didn't have quite the magic midrange which is the biggest factor for good performance in DBLTs ... for ALL listeners.
This century, the best KEF speaker I've ever heard, at any price, size or period, is the 2nd generation Egg in the KHT2001-2.
I have not heard the LS50 but I know the people who worked on it and they are excited.
Size, price, technology doesn't give you any advantage in DBLTs though the best are all the result of a long line of development of a designer or company's prejudices.
Very few full-range units do well in DBLTs and those that do, are sneered at by da 'experts' cos they don't 'measure' well. But most full-range units sound terrible in DBLTs
Last edited:
Does this mean I can design a speaker that outperforms all others in DBLTs?
The best speaker in nearly 2 decades of DBLTs came top in nearly a dozen tests .. usually agains much bigger and more expensive speakers. No other speaker has come anywhere near this record.
It has a lot of technology and features that I had a big hand developing. But I can't put my hand on my heart and say this is what is responsible cos there are other speakers with the same bells & whistles which don't do as well. About all I can say is that I can probably replicate its winning LF response.
The best speaker in nearly 2 decades of DBLTs came top in nearly a dozen tests .. usually agains much bigger and more expensive speakers. No other speaker has come anywhere near this record.
It has a lot of technology and features that I had a big hand developing. But I can't put my hand on my heart and say this is what is responsible cos there are other speakers with the same bells & whistles which don't do as well. About all I can say is that I can probably replicate its winning LF response.
Thanks.
No religion please
It's becoming quite widely accepted that THD is not a useful measure, and not only in speakers
Almost every "please give me advice" thread that pops up veers off into off into these discussions. But that's good....
All speakers are compromised and all listeners are different. You cannot find THE truth in other people's experiences unless you know which version of the truth you are seeking. It's a journey you have to make yourself, you need to find YOUR version of the truth. I'm not saying that guidance or advice is worthless, just that is only relevant if you understand where that guidance comes from, which context to place it in.
My advice to rjbell is to go out and experience stuff, learn which compromises annoy you the most and which annoy you the least. Understand yourself and your environment and that will lead you in the right direction. Making a decision on 10 minutes in Richer Sounds is probably not the best place to start, you need to hear speakers in a room, ideally close in size to yours and even better set up close to how you are going to use them, give at least 20 minutes for your brain to get accustomed to the illusion they are trying to conjure. Then start listening.
All speakers are compromised and all listeners are different. You cannot find THE truth in other people's experiences unless you know which version of the truth you are seeking. It's a journey you have to make yourself, you need to find YOUR version of the truth. I'm not saying that guidance or advice is worthless, just that is only relevant if you understand where that guidance comes from, which context to place it in.
My advice to rjbell is to go out and experience stuff, learn which compromises annoy you the most and which annoy you the least. Understand yourself and your environment and that will lead you in the right direction. Making a decision on 10 minutes in Richer Sounds is probably not the best place to start, you need to hear speakers in a room, ideally close in size to yours and even better set up close to how you are going to use them, give at least 20 minutes for your brain to get accustomed to the illusion they are trying to conjure. Then start listening.
Last edited:
Does anyone here seriously think measurements tell you how a speaker sound?
When you start designing speakers for real, there is a huge milestone when you can first do proper 'anechoic' frequency responses and you design your first speaker with flat response. Surely this must be the best speaker in the world.
If you are lucky, you might be able to do a direct comparison with a really good commercial speaker eg LS50 .. preferably in a properly conducted DBLT.
What happens is you then spend the rest of your life trying to work out why your speaker with flat frequency response doesn't sound so good
IMHO, only two opinions here are worth listening to ... those who have managed to hear these 2 speakers at home with their own music. Discount ALL other opinions.
Surely, you jest.
After establishing a huge body of work and linking objective measurements with subjective preferences, Toole and Olive are not "false prophets," in fact, they are not even prophets. They are good scientists who did hard work over their career so that the rest of us can understand how speakers work and what you must do to improve the sound. Their theory is rooted in data, logic, and sound reasoning.
I'm sorry that designing an anechoically flat frequency response speaker hasn't worked out for you. There are of course other criteria, such as off-axis response and bandwidth, especially down low. And there is the room and placement. You know all this. I am yet to hear a speaker that has as horrendous a response as the Alpair sound good. And this is where Toole and Olive have helped me. Their hypothesis holds true in my observations.
I see what you are saying here. The best of the best in a DBLT had similar measurements to others but it still repeatedly came out on top. There is another thread on this forum somewhere about a DBLT between a JBL M2 and the Revel Ultima Salon2, and the Revel won by some margin despite both having excellent measurements.
However, saying that you cannot win DBLTs against the best solely relying on measurements is completely different from saying designing using measurements is useless. It is easier to agree with the former, but there is no basis for the latter.
1) Using measurements to design will, undoubtedly, result in a better speaker.
2) Speakers that have large swings in the frequency response (+-5db) are not going to sound good in most listening rooms across a wide range of music.
3) Speakers with flat and smooth on-axis and smooth off-axis response sound good across a wide range of music and in most listening rooms.
These statements, which might sound intuitive, have been repeatedly proven both in DBLTs and also for me, personally, in my own observations. And this is the beauty of Toole and Olive's work. It should raise the quality of all speakers sold in the market. Maybe the best of best have many more bells and whistles and we don't fully understand why they beat other superb sounding speakers. But we can all agree what sounds bad.
Worth saying again.
All listeners are different, not only because of genetics, but also due to environment. Did you get all the nutrients you needed when you were young? Did you wear your hearing protection? Did you land on your ear doing a half-gainer with a half-twist? I did. I have a hole in my response 8-10k in my left ear. But i wa slucky to have the “10,000 hrs” of training and the plasticity of the brain allowed me to grow around that. During the training i remember having things pointed out to me in recordings being reproduced, and they stuck.
I have a fairly good track recored of people liking what i like (hardly universal). And i have learned the hard way to be apprehensive about others opinions of stuff. I still get flack from a lotta people from my opionion of the TC9 (it sucks, it is lifeless, fortunetly they were cheap). A good friend went over and listened to the big Fane which his friend loved and he thot atrocious.
As Simon says, you have to experience things yourself.
All speakers are more bad* than they are good so whether a speaker is for you depends on the mix of good & bad that you can live with (in your room, with your ancillary kit). I for instance really value imaging/soundstage and am sensitive to higher frequency crap — hence my attraction to EnABLed drivers and my dislike of the Jordan JX92 and the (very similar sounding) Mark Audio Alpair 10 (heavily influenced by the Jordan). A lot of people love them.
*(if there is perfection, the best we have done with speakers so far is, in my estimate, nearer 20% than 40%.)
In the end, the only thing that matters is whether the hifi connects you with the (emotion of) music. FR measures are a tool (an important one). But in the words of Floyd Toole:
But i have seen too many instances of someone measuring 1st then the review can be seen to mirror the FR measure even if it doesn’t. Trust your ears… they are the only ones that matter. As has been said, measures can tell you bad from good, but where really, really good is compared to good, we do not yet have the tools to quantify those differences. Much of the deta comes down to how the device performs 30-40 dB down. Is the information even there or has it been lost? What is the FR like down there (measured at the same time a full-level signal is playing)? More...
dave
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.