Lots of great comments and questions here in past 12hrs! I will see if I can answer them.
Wesayso is right about setup improving each time so comparing between rounds is not fair. Round 1 was not even 320kbit MP3. Round 3 had RS225 woofer and Harsch XO. In hindsight, maybe I should have frozen configuration? But there were improvements to be had and being a DIY tinkerer I couldn't leave it alone.
If you all are not interested in a blowout round, then I cannot do more tests without more drivers. Perhaps the one person that needs all of our gratitude for his generosity is Byrtt who provided a substantial stable of drivers (in pairs and for me to keep!): TG9FD, PS95-8, 3FE22r16, RS100-4, RS100P-4, FF105WK, 10F/8424, SB65WBAC25-4, there may be more that I missed. But you see, without this kind of a kick start we would not have had any of this. Other folks have also stepped up and lent drivers and some have bought a driver or two and all are greatly appreciated. But for a non blowout final round 4 that is new drivers - I am tapped out of new ones to evaluate.
The suggestion of letting them run full range sans FAST woofer is interesting. It will allow a heavy bias for drivers with low fs, high Qts, and high xmax because that is how you get bass in a sealed cabinet. The IMD is something that will become apparent with this sound clip though. It is easy enough to do with a click of the mouse to disable the bass via miniDSP.
There seems to be overwhelming requests for mono source and mono recording. However my goal was to let one hear about what you would hear if you were in my lab. Would you stick an earplug in your left ear in order to listen with just the right ear to the right channel on mono? That would not sound very good - kind of unpleasant actually. But perhaps putting mono recorded onto both channels so folks listening with headphones don't feel unbalanced. If you listen with mono there should be no image - only perhaps pseudo image from balanced presentation with perfectly centered soundstage but no spatial localization of performers.
I think the blowout round 4 is important from standpoint of letting us hear if the $12 TC9 and $22 TG9 can be selected by a large portion of listeners. In a non blind test, people will tend to pick the more expensive system. This was shown in a study by Sean Olive when he was manager for Subjective Evaluation at Harman International. When he made blind tests, the cheap speaker got more votes. It would be good if we can get substantially higher numbers of participants to get a higher level of confidence in the statistics. With 50 votes assuming random, the votes for any favored "winner" needs to exceed the next closest by 7 votes on average for it to make statiscal significance assuming Poisson distribution (zero is minimum). You statistics experts out there can correct me if I have this wrong.
There is no doubt in my mind that the 10F is perhaps the highest performing driver of bunch. But is it really that audible when blind tested with other rather high performing drivers as well?
Wesayso is right about setup improving each time so comparing between rounds is not fair. Round 1 was not even 320kbit MP3. Round 3 had RS225 woofer and Harsch XO. In hindsight, maybe I should have frozen configuration? But there were improvements to be had and being a DIY tinkerer I couldn't leave it alone.
If you all are not interested in a blowout round, then I cannot do more tests without more drivers. Perhaps the one person that needs all of our gratitude for his generosity is Byrtt who provided a substantial stable of drivers (in pairs and for me to keep!): TG9FD, PS95-8, 3FE22r16, RS100-4, RS100P-4, FF105WK, 10F/8424, SB65WBAC25-4, there may be more that I missed. But you see, without this kind of a kick start we would not have had any of this. Other folks have also stepped up and lent drivers and some have bought a driver or two and all are greatly appreciated. But for a non blowout final round 4 that is new drivers - I am tapped out of new ones to evaluate.
The suggestion of letting them run full range sans FAST woofer is interesting. It will allow a heavy bias for drivers with low fs, high Qts, and high xmax because that is how you get bass in a sealed cabinet. The IMD is something that will become apparent with this sound clip though. It is easy enough to do with a click of the mouse to disable the bass via miniDSP.
There seems to be overwhelming requests for mono source and mono recording. However my goal was to let one hear about what you would hear if you were in my lab. Would you stick an earplug in your left ear in order to listen with just the right ear to the right channel on mono? That would not sound very good - kind of unpleasant actually. But perhaps putting mono recorded onto both channels so folks listening with headphones don't feel unbalanced. If you listen with mono there should be no image - only perhaps pseudo image from balanced presentation with perfectly centered soundstage but no spatial localization of performers.
I think the blowout round 4 is important from standpoint of letting us hear if the $12 TC9 and $22 TG9 can be selected by a large portion of listeners. In a non blind test, people will tend to pick the more expensive system. This was shown in a study by Sean Olive when he was manager for Subjective Evaluation at Harman International. When he made blind tests, the cheap speaker got more votes. It would be good if we can get substantially higher numbers of participants to get a higher level of confidence in the statistics. With 50 votes assuming random, the votes for any favored "winner" needs to exceed the next closest by 7 votes on average for it to make statiscal significance assuming Poisson distribution (zero is minimum). You statistics experts out there can correct me if I have this wrong.
There is no doubt in my mind that the 10F is perhaps the highest performing driver of bunch. But is it really that audible when blind tested with other rather high performing drivers as well?
Last edited:
Apologies, xrk971. Here I am saying I'd like to see a greater variety of drivers while forgetting that these things don't just grow on trees. Thanks to Byrtt and any others who helped supply some of the drivers for these tests. Your generosity is greatly appreciated.
Now that we've been reminded about these 'minor' logistical details, I say do whatever you feel is interesting with whatever you happen to have available. After all, it's your show.
Personly very interested for a finale where the best are setup under the newer better system setup so we can compare them under these conditions.
X could also choose to just release recorded sound clips reference tracks and objective measured data at day one at same time without the poll competition, but then think we probably pick what please our eyes plus personal bias and miss the horse race competition at the top.
Personally have learned a lot participating in all three rounds, it is stuff as learn to concentrate for a critical listening and what to listen for, its also stuff as having to eat what please my eyes and ones pre bias is needed to be open minded. Looking back the three rounds my voted ranking will still be the same if listening sound clips at exactly same system setup as i did before send my vote even now knowing what driver the clip belongs.
Thanks xrk971 and all diy'ers here, think we learn a lot from each other even we some times go into small shoes for our critics or praise but its good enough discussions and can open our minds.
Regarding the propose to have drivers tested without the FAST setup, think problem would be that every driver would need its own special box setup be it sealed ported or horn of a kind to get the best low extension for each one. This would create a impossible work load for X and probably critics would come that hey that driver can go lover in a Z box instead of the created Y box. As is now the setup is absolute the same only variable is DUT. If wanted to not to listen the woofer one can alwas play the tracks in Foobar or Jriver and set example a high pass filter with a certain slope and frq point, then compare the sound clips again.
X could also choose to just release recorded sound clips reference tracks and objective measured data at day one at same time without the poll competition, but then think we probably pick what please our eyes plus personal bias and miss the horse race competition at the top.
Personally have learned a lot participating in all three rounds, it is stuff as learn to concentrate for a critical listening and what to listen for, its also stuff as having to eat what please my eyes and ones pre bias is needed to be open minded. Looking back the three rounds my voted ranking will still be the same if listening sound clips at exactly same system setup as i did before send my vote even now knowing what driver the clip belongs.
Thanks xrk971 and all diy'ers here, think we learn a lot from each other even we some times go into small shoes for our critics or praise but its good enough discussions and can open our minds.
Regarding the propose to have drivers tested without the FAST setup, think problem would be that every driver would need its own special box setup be it sealed ported or horn of a kind to get the best low extension for each one. This would create a impossible work load for X and probably critics would come that hey that driver can go lover in a Z box instead of the created Y box. As is now the setup is absolute the same only variable is DUT. If wanted to not to listen the woofer one can alwas play the tracks in Foobar or Jriver and set example a high pass filter with a certain slope and frq point, then compare the sound clips again.
Missing (so far) are the similar woven-cone Seas full and mid range equivalents (FU10RB/MU10RB). I'd love to hear the TG9, 10F and FU10 side by side as representatives of the "damped cone" design perspective, compared to the MA7.3 (at least) as representative of the "other school" . . .
Measurements of the FU10RB can be found on Zaphs Blog. Long story short it doesn't measure nearly as good as the F10
As for comparing drives full range vs FAST. I agree that I personally would prefer full range. IMO though, this would be best done open baffle/dipole to eliminate all independent variables in box configuration. Panel resonances, different sealed enclosed for same alignment, ect.
The best way/ most scientific way to compare drivers would be open baffle by far. Not to mention much less work.
Regarding OB as test enclosure - in the original "Objective Measurements" thread that got shut down I demonstrated that the Nautaloss with trapezoidal baffle approximates an OB without the big unwieldy size of the OB. Here is comparison of frequency response of both with the hotly debated CHN70.
The lack of box coloration was the main reason I went with the Nautaloss.
I'll go with "doesn't measure as good" . . . but the "nearly" is a bit of an exaggeration. The "glitch" around 1.5kHz is a bit concerning . . . some 4" drivers have it and some don't and I don't know why (or if it matters).
But the question is how they sound, not how they look on paper . . .
After Wesayso comments and yours byrtt, I can definitely see the value in a final shoot-out round for these drivers with the RS225 woofer FAST system. It will put all the top-pick drivers onto the same system with a nice set of recordings to allow for comparisons, and it will bring some closure to this series of tests. X's FAST testing has made it much easier to compare the drivers in the same box, this is for certain.
Maybe in the future another series of tests can be done without the FAST woofer, maybe on an open baffle, or custom box sizes. But yes I agree now that this series of tests should be finished off with a shoot-out with the RS225 FAST system (like in round3). I am not sure it will change my personal "want to buy" driver list, but let's see.
X- It blows my mind that you successfully designed an enclosure that can achieve that. That's amazing!
Dew- I have to agree it's the sound that ultimately matters. It would be my hypothesis the F10 would do better in blind tests by the measurements, but ultimately you're right. Its how they perform in blind tests that is the ultimate 'measurement'.
My biggest concern with the Seas though is its really high QTS. I have this hypothesis that I can't prove that a QTS of ~.5 is important for sound quality. This is just observational, as time and time again, the very best measuring/rated drivers usually have a Q around .5.
"A system with an intermediate quality factor (Q = ½) is said to be critically damped. Like an overdamped system, the output does not oscillate, and does not overshoot its steady-state output (i.e., it approaches a steady-state asymptote). Like an underdamped response, the output of such a system responds quickly to a unit step input. Critical damping results in the fastest response (approach to the final value) possible without overshoot. Real system specifications usually allow some overshoot for a faster initial response or require a slower initial response to provide a safety margin against overshoot."- Wiki
Of course I understand that enclosure effects Q. Its just one of my theory's that goes along with.. "You can polish a turd, but it's still a turd".
Dew- I have to agree it's the sound that ultimately matters. It would be my hypothesis the F10 would do better in blind tests by the measurements, but ultimately you're right. Its how they perform in blind tests that is the ultimate 'measurement'.
My biggest concern with the Seas though is its really high QTS. I have this hypothesis that I can't prove that a QTS of ~.5 is important for sound quality. This is just observational, as time and time again, the very best measuring/rated drivers usually have a Q around .5.
"A system with an intermediate quality factor (Q = ½) is said to be critically damped. Like an overdamped system, the output does not oscillate, and does not overshoot its steady-state output (i.e., it approaches a steady-state asymptote). Like an underdamped response, the output of such a system responds quickly to a unit step input. Critical damping results in the fastest response (approach to the final value) possible without overshoot. Real system specifications usually allow some overshoot for a faster initial response or require a slower initial response to provide a safety margin against overshoot."- Wiki
Of course I understand that enclosure effects Q. Its just one of my theory's that goes along with.. "You can polish a turd, but it's still a turd".
BRYTT,
I totally agree. Much is learned from each other despite the wide range of opinions and personal interpretations of drivers, methods, theories. Even disagreement is ultimately productive.
Whilst there are some imperfections in the test methods and its evolution, I really don't believe that much can be improved upon, besides my earlier comments regarding my desire for a mono recording of a mono speaker, I have no doubt 320 kbps is fine for this test, and whilst the purist in me dislikes the multiple DA and AD conversions, convolutions and DSP 'magic' I certainly dont feel short changed at all, and I recognise the large efforts XRK has made.
I also don't believe P10 should feel he has to participate, although I honestly believe it would be a brilliant opportunity to showcase measured improvement made by his modifications. Whilst I don't believe Enable is a patentable technique, I.e. Correctly distributed mass, absorption or slits can help improve drivers, though there is no reason to think enable only way or best way.
The trouble with belief is fact can shatter it. Some are happy with what they believe, and what they hear. Sometimes its like Monty Python in here, belief takes over and ignites conflict.
Whilst I have a reasonable ear, I know how easily I can be deceived, and data is always welcome. Those with commercial.interest should really consider the DIY nature of the forum and that education is the aim, not sermon. Plenty have made their relative fortune here by shameless plugging, it saddened me to be honest.
Any commercial venture I participate in, will never (and have never) been mentioned here.
Anonymity is a blessing
Above all we all share one thing: Passion for what we listen to.
Sometimes we all find a need to step back and remember that, especially me
Peace and thanks for all the good work X, do not be discouraged.
(and if I ever come across as short or harsh, it isn't actually intended. My.mind is just an odd place)
I totally agree. Much is learned from each other despite the wide range of opinions and personal interpretations of drivers, methods, theories. Even disagreement is ultimately productive.
Whilst there are some imperfections in the test methods and its evolution, I really don't believe that much can be improved upon, besides my earlier comments regarding my desire for a mono recording of a mono speaker, I have no doubt 320 kbps is fine for this test, and whilst the purist in me dislikes the multiple DA and AD conversions, convolutions and DSP 'magic' I certainly dont feel short changed at all, and I recognise the large efforts XRK has made.
I also don't believe P10 should feel he has to participate, although I honestly believe it would be a brilliant opportunity to showcase measured improvement made by his modifications. Whilst I don't believe Enable is a patentable technique, I.e. Correctly distributed mass, absorption or slits can help improve drivers, though there is no reason to think enable only way or best way.
The trouble with belief is fact can shatter it. Some are happy with what they believe, and what they hear. Sometimes its like Monty Python in here, belief takes over and ignites conflict.
Whilst I have a reasonable ear, I know how easily I can be deceived, and data is always welcome. Those with commercial.interest should really consider the DIY nature of the forum and that education is the aim, not sermon. Plenty have made their relative fortune here by shameless plugging, it saddened me to be honest.
Any commercial venture I participate in, will never (and have never) been mentioned here.
Anonymity is a blessing
Above all we all share one thing: Passion for what we listen to.
Sometimes we all find a need to step back and remember that, especially me
Peace and thanks for all the good work X, do not be discouraged.
(and if I ever come across as short or harsh, it isn't actually intended. My.mind is just an odd place
)
Last edited:
That applies at/around driver resonance, but makes no nevermind when you're crossing off of it several octaves above that (FAST).
It is patented, althou i believe the patent is now expired. Bud did give it to diyers the world over, and he has gotten royalties from me.
dave
I was under the impression it applies to everything, as it effects transient response.
"The total system Q of the woofer combined with cabinet is a ratio of values (with no units) that defines both the shape of the frequency response curve roll-off. It also determines the amount of damping to oscillation or ringing after the signal stops. This is also spoken of as “transient response”. QTC is the term for sealed cabinets, and the similar QTS term is for ported reflex cabinets." - Swed, Audioholics
An externally hosted image should be here but it was not working when we last tested it.
From my understanding. Q < .5 can begin to compress dynamics because it does not reach the zero point/ starting point as fast as a Q greater than or equal to .5. Although it does not ring, which is good.
'As the quality factor decreases the slower decay mode becomes stronger relative to the faster mode and dominates the system's response resulting in a slower system."- Wiki
Q=.5 Results in no oscillations, no overshoot, and the fastest response possible back to zero point/starting point without oscillations.
Q>.5 Reaches zero point/starting point fast, but oscillates.
A low Q is usually refered to as "Dry" sounding, while a high Q is usually "Boomy". So if you want a Dry sound.. go for a low Q driver, if you want a Boomy sound, go for a high Q driver. If you want an accurate sound, go for a driver with a Q as close to .5 as possible.
Everything is subjective of course, and some people find that Q=.5 is still too "dry" for them. Maximally flat amplitude vs frequency occours at Q=.707, so in my opinion, good designs will fall in this region. There is no reason to design a speaker with a Q higher than .707, it just adds unnecessary ringing. A speaker driver that is designed with a Q < .5 isn't good IMO as well, as it will have compressed dynamics in a ported enclosure, and not enough LFE in a sealed enclosure.
Again Xaborus I agree.
Q of 0.5 is the critically damped control term, like Bessel is in a filter. Same thing. I think...
Often with smaller drivers I agree that I find a Qts of 0.4-0.5 a good model for good bass range in a ported box when also the Vas is low/medium value. (the box isn't tiny but also not huge)
Q=0.5-0.7 can also work closed box, with similar Vas. (larger than bookshelf)
Q of 0.5 is the critically damped control term, like Bessel is in a filter. Same thing. I think...
Often with smaller drivers I agree that I find a Qts of 0.4-0.5 a good model for good bass range in a ported box when also the Vas is low/medium value. (the box isn't tiny but also not huge)
Q=0.5-0.7 can also work closed box, with similar Vas. (larger than bookshelf)
Last edited:
Spent much of the day listening to the Vifa on OB on top of H-frame with Alpha 15. It’s a superb setup. I am coming off about a month of listening to the Tangband 1808 on top of the same H-frame. These are both very enjoyable systems with different personalities. The Tangband is brighter and more forward. At first listen the Vifa sounds dull in comparison. But once my ears adjust the Vifa is a very balanced driver. It’s less lively sounding than the Tangband and less efficient so it needs a more powerful amp.
If you like a laid back presentation that’s relaxed and a bit rolled off at the top (I angle the Vifa’s about 5 degrees off the listening position, the Tangband’s about 30 degrees) then the Vifa is an excellent choice. I could listen to this speaker all day.
For home use I believe I could argue these systems (as well as similar ones) are among the best options for incredible sound.
If you like a laid back presentation that’s relaxed and a bit rolled off at the top (I angle the Vifa’s about 5 degrees off the listening position, the Tangband’s about 30 degrees) then the Vifa is an excellent choice. I could listen to this speaker all day.
For home use I believe I could argue these systems (as well as similar ones) are among the best options for incredible sound.
Attachments
I have a 36"X36" open baffle with interchangeable mounting holes in my workshop. All my full range drivers spend at least several weeks (months) playing mostly fm radio. I have a pair of ff85wk's on tap at the moment, sounded a little squawky at first but I'm really impressed with how they sound now!
Larry
Larry
Thanks would you share how they sound compared to other FR that has sit in the OB and what's their brand/type.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.