The AA 6.5 looks like the new value leader to me. I've been drooling over the Vifa XG18 for a long time now, and this driver seems to exceed it's performance. Low order harmonics are slightly worse, but it's 3db more sensitive, and the CSD looks better.
As a bonus, the aluminum phase plug should complement the Seas 27TBFC/G quite nicely.
John/Zaph - any chance someday you will move towards higher resolution CSD plots? I recently measured my Dayton RS125S and found some problems at 900hz and 450hz that didn't show up in your tests. Link to nearfield CSD w/ crossover (EDIT: Nevermind. My cabinet is 14.5 inches tall internally, which is 950hz.)
Dan
As a bonus, the aluminum phase plug should complement the Seas 27TBFC/G quite nicely.
John/Zaph - any chance someday you will move towards higher resolution CSD plots? I recently measured my Dayton RS125S and found some problems at 900hz and 450hz that didn't show up in your tests. Link to nearfield CSD w/ crossover (EDIT: Nevermind. My cabinet is 14.5 inches tall internally, which is 950hz.)
Dan
Mr Zaph,
could you elaborate a little how to read the waterfall plots?
For me -as a beginner- the plots from the Seas P18 and de Scanspeak do look quite similar, when differences in distortion are large, so I must be looking the wrong way at the plots...
Thanks in advance!
could you elaborate a little how to read the waterfall plots?
For me -as a beginner- the plots from the Seas P18 and de Scanspeak do look quite similar, when differences in distortion are large, so I must be looking the wrong way at the plots...
Thanks in advance!
Paulus1981 said:
Hi,
No you are not. The CSDs are very similar. The factors that affect
distortion and the factors that affect CSD are very different.
🙂/sreten.
Other issues on test conditions
Again, zaph has done us all a great service through his labor. What one must must must be aware of, and can easily ignore: These tests aren't done at what people necessarily listen at, or in 'typical use'. The fostex are a great example. Full range drivers aren' t meant to go tremendously loud. To get the high efficiency they usually have, they are generally low MMS, very low excursion motor systems. When you hornload them, or use them in very mild conditions, the results could be very different indeed.
Nothing new to zaph or to shinobiwan, but to an inexperienced diyer, this could make the difference between giving a driver they may love a shot and completely discounting it.
There's no substitute for trying something in an enclosure, and fiddling with it to yank the best performance out.
Of course... one could always build an affordable, great performing system directly from zaphs page and save themselves the hassle........
I'm thoroughly impressed by his designs- many are extremely affordable, and the crossovers seem very competently designed. I haven't heard any of them of course, being a builder of my own 'style' of system.
Again, zaph has done us all a great service through his labor. What one must must must be aware of, and can easily ignore: These tests aren't done at what people necessarily listen at, or in 'typical use'. The fostex are a great example. Full range drivers aren' t meant to go tremendously loud. To get the high efficiency they usually have, they are generally low MMS, very low excursion motor systems. When you hornload them, or use them in very mild conditions, the results could be very different indeed.
Nothing new to zaph or to shinobiwan, but to an inexperienced diyer, this could make the difference between giving a driver they may love a shot and completely discounting it.
There's no substitute for trying something in an enclosure, and fiddling with it to yank the best performance out.
Of course... one could always build an affordable, great performing system directly from zaphs page and save themselves the hassle........
I'm thoroughly impressed by his designs- many are extremely affordable, and the crossovers seem very competently designed. I haven't heard any of them of course, being a builder of my own 'style' of system.
Paulus1981 said:
Hi Paulus,
Sorry not Zaph here, but my 2 cents worth.
The way I read the waterfall plots is by interpreting the rate which the amplitude falls versus time. Ideally they would drop like a wall from the initial signal. But due to things related to energy storage, cone breakup, and other electro/mech resonances, they don't. Look for lack of ringing and general smooth behavoirs as other signs of quality.
Hi John,
I went back and listened to her debut, Call Off the Search, which I really enjoy too.
Currently, the CD that is getting the most rotation is Alan Jackson's Like Red On a Rose. Recommended!
I started writing a little summary on each of the drivers in the 6.5" roundup, so I could "compare notes" with yours later, but stopped halfway through, after realising that I didn't know many nice things to say about that Fostex FF166, and started to get depressed about all the great drivers that costs more to ship, than to buy, to Australia.
I could probably manage better to do it if I put my comments into headings like-
Great
Good
Bad
Starting to sound like all those hi-fi rags...
My vote for - Best performance/dollar- In the affordable catagory<$98/driver
Best in Canada - Mach 5 MLI-65 ($17) cost of shipping doubles the price here in USA.
Nice waterfall plots, HD not too bad. Not sure if it holds up under high levels. More of a mid than woof on this one.
Best in USA - Dayton DA175-8 ($19)
Nice waterfall plots, HD not the worst, cone breakup needs attention. Should be all around midwoof workhorse.
Worst Seas L18 (H1224) ($70)
Horrible waterfall plots, HD OK to poor, maybe something broken on this driver.
I didn't pay too much attention to FR on any of these.
Best in Canada - Mach 5 MLI-65 ($17) cost of shipping doubles the price here in USA.
Nice waterfall plots, HD not too bad. Not sure if it holds up under high levels. More of a mid than woof on this one.
Best in USA - Dayton DA175-8 ($19)
Nice waterfall plots, HD not the worst, cone breakup needs attention. Should be all around midwoof workhorse.
Worst Seas L18 (H1224) ($70)
Horrible waterfall plots, HD OK to poor, maybe something broken on this driver.
I didn't pay too much attention to FR on any of these.
Big BIG thanks for Uncle Zaph for taking the time out to measure all those drivers.
😎 😎 😎 😎 😎
Most certainly interesting stuff...
Mark
😎 😎 😎 😎 😎
Most certainly interesting stuff...
Mark
so according to your expert analysis
( you beeing every member here of course )
Wich driver is as the best performance ?
( no cost comparison ..just pure performance )
i personally cannot choose a winner out
alot of expennsive drivers are of what seems to be very equal quality
Zaph : would it be possible for you to do this test again at a higher output spl ?
could be with only the top 5 drivers ..
( if one can determine wich are in the top5 😛 )
( you beeing every member here of course )
Wich driver is as the best performance ?
( no cost comparison ..just pure performance )
i personally cannot choose a winner out
alot of expennsive drivers are of what seems to be very equal quality
Zaph : would it be possible for you to do this test again at a higher output spl ?
could be with only the top 5 drivers ..
( if one can determine wich are in the top5 😛 )
John/Zaph
What a great wealth of info on your website...excellent job!
Any chance you can export the driver response and impedance plots to a text file so we can play with them in some different design programs. It would also be really nice if you could give us the sealed box impedance plots and the volumes of the test boxes. I realize that you have given us all of the necessary TS parameters, but the files would still prove very useful.
Thanks again for the fantastic service to the DIY community!
G.
What a great wealth of info on your website...excellent job!
Any chance you can export the driver response and impedance plots to a text file so we can play with them in some different design programs. It would also be really nice if you could give us the sealed box impedance plots and the volumes of the test boxes. I realize that you have given us all of the necessary TS parameters, but the files would still prove very useful.
Thanks again for the fantastic service to the DIY community!
G.
Re: What distortion measurements are best
I agree with on most points you made. IMD measurements with multitones or just 2 tones would be quite useful and should be more of a standard linearity test. (at least it is in the communication industry) This also should be applicable here in the audio community.
My main point I would like to make is, it would be REALLY useful to have more compression tests using IMD to find out at what power levels degrade a certain drivers linearity. This could be accomplished by looking at the rate of change of the harmonic levels as they are driven harder. I mean most drivers would sound really good with a fundamental level at 92 dB SPL with HD artifacts down 40-50 dB. But where to they start to fall apart would be the best indicator for a particular driver selection parameter.
Zaph, I want to personally thank you for all time and effort you put into creating all this data. It's really good stuff and keeps getting better all the time. I don't want you to see this as a critical jab, just something to think about for the future.
ucla88 said:
I agree with on most points you made. IMD measurements with multitones or just 2 tones would be quite useful and should be more of a standard linearity test. (at least it is in the communication industry) This also should be applicable here in the audio community.
My main point I would like to make is, it would be REALLY useful to have more compression tests using IMD to find out at what power levels degrade a certain drivers linearity. This could be accomplished by looking at the rate of change of the harmonic levels as they are driven harder. I mean most drivers would sound really good with a fundamental level at 92 dB SPL with HD artifacts down 40-50 dB. But where to they start to fall apart would be the best indicator for a particular driver selection parameter.
Zaph, I want to personally thank you for all time and effort you put into creating all this data. It's really good stuff and keeps getting better all the time. I don't want you to see this as a critical jab, just something to think about for the future.
There was a typo in John's measurements. It's 92dB@1/2m.
I think this is an adequate test level. It's moderately loud, but not very loud, and probably represent "average" typical listening levels in the home environment.
92dB @1/2m= 98dB@1m= 101 dB@1m for stereo pair
Seated listening distance 9ft = 92dB at listening position.
Too high and people start to complain that's it's unrealistic real-world SPL level, too low and people say that it's not enough to differentiate between the better drivers.
<sigh>
Of course in the perfect world we'd do every test imaginable, but this is one man show, and John has done a superb job.
Even with the current core set measurements, it seems to me that a lot of people, including some of those particularly vocal on the DIYAudio/HTGuide forums, are already having some difficulty interpretting the data.
No wonder some still prefer to simply describe what they hear with their ears.
I think this is an adequate test level. It's moderately loud, but not very loud, and probably represent "average" typical listening levels in the home environment.
92dB @1/2m= 98dB@1m= 101 dB@1m for stereo pair
Seated listening distance 9ft = 92dB at listening position.
Too high and people start to complain that's it's unrealistic real-world SPL level, too low and people say that it's not enough to differentiate between the better drivers.
<sigh>
Of course in the perfect world we'd do every test imaginable, but this is one man show, and John has done a superb job.
Even with the current core set measurements, it seems to me that a lot of people, including some of those particularly vocal on the DIYAudio/HTGuide forums, are already having some difficulty interpretting the data.
No wonder some still prefer to simply describe what they hear with their ears.
tktran303 said:
yep, very nice sing and tunes !
notice , however, the crappy FX unit they used on some tracks, particularly obvious on track 6 and 8.
92 dB@0.5m = 86 dB@1m = 80 dB@2m, or 86 dB@2m for a stereo pair. Still an ok value for average listening levels (movie theater reference level is 83 dB average), though it does not tell you what the driver will do at a +20 dB peak as might occur especially with classical music (up to 20 dB peak to average ratio; with pop this is seldom reached due to compression).
MBK,
Thanks for correcting my typo.
<sigh>
I thought a pair of speakers give extra 3dB?
ie. 83dB for stereo pair?
Where's that SPL meter gone...
Thanks for correcting my typo.
<sigh>
I thought a pair of speakers give extra 3dB?
ie. 83dB for stereo pair?
Where's that SPL meter gone...
Well, a pair will output double the SPL of a single speaker, so that would be +6 dB. Incidentally the power consumption will only double, and in power dB this means +3 dB power consumption for +6 dB radiated power... or in other words, a doubling of power efficiency. Linkwitz has a nice discussion of this on his FAQ page
In a sense you are both right.
SPL would double +3dB for connecting another channel of a stereo pair playing a mono source. (radiating surface is doubled)
But if you connect another driver in parallel to the same channel of a constant voltage PA then you get an additional +3 dB on top. Because the impedance is halfed with the same voltage.
From the SL link.
"In summary, when two identical drivers are connected in parallel and driven with constant voltage, then twice the electrical power is consumed (+3 dB), the radiated acoustic power is increased by a factor of four (+6 dB), and the free space sound pressure level is doubled (+6 dB) at a given distance."
SPL would double +3dB for connecting another channel of a stereo pair playing a mono source. (radiating surface is doubled)
But if you connect another driver in parallel to the same channel of a constant voltage PA then you get an additional +3 dB on top. Because the impedance is halfed with the same voltage.
From the SL link.
"In summary, when two identical drivers are connected in parallel and driven with constant voltage, then twice the electrical power is consumed (+3 dB), the radiated acoustic power is increased by a factor of four (+6 dB), and the free space sound pressure level is doubled (+6 dB) at a given distance."
infinia said:
Hi, so this is a waste of time ? http://www.zaphaudio.com/audio-speaker17.html
🙂/sreten.
- Status
- Not open for further replies.