It's not really a step, it's more of a dip centred around 3.5kHz.
I also voted for the TC7.
I hated A (the Alpine) within a few seconds of hearing the first track. I immediately scribbled "Not A" on a piece of paper. Like X mentioned, having a a lot of high end probably helps within a car environment, but not in my living room.
I picked E (the FR88EX). I'm surprised I picked a metal cone. Pretty flat as well! I guess I will have to try to have a sample here.
I picked E (the FR88EX). I'm surprised I picked a metal cone. Pretty flat as well! I guess I will have to try to have a sample here.
It could easily be said that the Alpine's loudness made it get the most votes.
If the levels had been set so that overlay looks more like these:
then votes would likely have been more evenly distributed.
No winners or losers at this stage of development. Each would need refinement in XO level matching, and all the drivers except perhaps the Alpine and FR88EX would benefit from some broad cut at about 1kHz.
I haven't seen or used drivers that don't benefit from at least two bands of PEQ. Many need 3 or 4 bands of PEQ before they remotely sound like results achieved with inverse transfer function filtering. Based on my explorations with both of these alignment approaches, once a speaker system is equalized to about ±1.5dB across the spectrum speakers start to sound very much the same. From here differences in polar response, and thus room response become significant.
Round 4 results support this view:
In round 4 vote distribution was much wider. My ±1.5dB criteria is well met by most of the round 4 drivers across much of the spectrum. Polar response cropped up in round 4 with off axis recording of PS95-8. It was well received by voters but room reverberation sounded brash to me. If the on axis response of this driver had some of its high end tamed with EQ for on axis recording, it may have lead the round.
I could have studied the provided FR plots combined with extended listening to identify the drivers by ear, but voted after relatively brief listening based on what seemed a good balance of tone and attack for piano, drum and bass with clip-2. After voting, curiosity led to Photoshop manipulation. This led to simple method for normalizing tracks based on single frequency. Then it was simple to compare frequency responses from same point in each clip-2 recording and deduce the driver assignments. Identical assignments were made repeating the exercise with clip-3 recordings. The reveal indicates 100% identification success.
If the levels had been set so that overlay looks more like these:
then votes would likely have been more evenly distributed.
No winners or losers at this stage of development. Each would need refinement in XO level matching, and all the drivers except perhaps the Alpine and FR88EX would benefit from some broad cut at about 1kHz.
I haven't seen or used drivers that don't benefit from at least two bands of PEQ. Many need 3 or 4 bands of PEQ before they remotely sound like results achieved with inverse transfer function filtering. Based on my explorations with both of these alignment approaches, once a speaker system is equalized to about ±1.5dB across the spectrum speakers start to sound very much the same. From here differences in polar response, and thus room response become significant.
Round 4 results support this view:
In round 4 vote distribution was much wider. My ±1.5dB criteria is well met by most of the round 4 drivers across much of the spectrum. Polar response cropped up in round 4 with off axis recording of PS95-8. It was well received by voters but room reverberation sounded brash to me. If the on axis response of this driver had some of its high end tamed with EQ for on axis recording, it may have lead the round.
I could have studied the provided FR plots combined with extended listening to identify the drivers by ear, but voted after relatively brief listening based on what seemed a good balance of tone and attack for piano, drum and bass with clip-2. After voting, curiosity led to Photoshop manipulation. This led to simple method for normalizing tracks based on single frequency. Then it was simple to compare frequency responses from same point in each clip-2 recording and deduce the driver assignments. Identical assignments were made repeating the exercise with clip-3 recordings. The reveal indicates 100% identification success.
Barleywater,
Thanks your surveys they very educating and you actual did the right reveal at post 118, great done : )
Did you vote and if which did you prefer.
Myself had a battle between TC7 and FR88EX where i ended pick TC7 as Godzilla/peterbrorsson/5th element seems did, because it was so close to reference track listening at my local gear. Having FR88EX as second looks like many had and by that then think it had been real close to have tilted it to win, even Eldam that voted wiggled ALPINE this time seems had the fairly smooth FR88EX as second choice.
Thanks your surveys they very educating and you actual did the right reveal at post 118, great done : )
Did you vote and if which did you prefer.
Myself had a battle between TC7 and FR88EX where i ended pick TC7 as Godzilla/peterbrorsson/5th element seems did, because it was so close to reference track listening at my local gear. Having FR88EX as second looks like many had and by that then think it had been real close to have tilted it to win, even Eldam that voted wiggled ALPINE this time seems had the fairly smooth FR88EX as second choice.
Last edited:
I have a theory that what made Alpine (and TC7 to some degree) sounds natural (others sound like having blanket in front of it) is the tilting response from 6kHz and up. Many speaker builders prefer to do the response like this....
Some other builders (like me) prefer a tilting down response from BSC and up. I have never read people mentioned the reason why, but I know why I did it like that. I even guessed F was TangBand simply from the tilting down response (tho this is only from 1kHz).
Attachments
I voted for G (Vifa TC7FD). But I must confess my hearing is not good. But these looks like nice drivers.
I may be wrong but I have been observing the poll from start. And just near the half way to the poll F and G were voted top most. At the end of the voting A driver got maximum votes. Would a hidden voting so as which driver gets most votes can not be seen possible ? These has been very informative tests.
Thanks for the effort and best regards.
I may be wrong but I have been observing the poll from start. And just near the half way to the poll F and G were voted top most. At the end of the voting A driver got maximum votes. Would a hidden voting so as which driver gets most votes can not be seen possible ? These has been very informative tests.
Thanks for the effort and best regards.
Nice then we have names the 5 of 6 votes for TC7 : ) Hiten/Godzilla/peterbrorsson/5th element/BYRTT
Regarding hidden poll or not it doesn't bother me and just trust people do their best.
What bothers me is what asked at post 118 regarding if voters for ALPINE driver intend to run the device as response is out of box and same as we listen in the sound clips or they intend some electric/DSP/acoustic correction. That's because if we seek a reproducer to be so close to natural real world acoustics as possible it's never gonna happen with the winner uncorrected, take a look at its IR and settling time it adds and extracts to input verse output with simple word colour/distortion. Colour/distortion is no problem if one is aware and want it, but its a problem it one think one have true natural sound.
Last edited:
Oh yes, I see that now! A bit too many traces and shouldn't have looked and posted after late discussion with ex-wife….
Peter
There's a psychology saying that people prefer to be with the winner. Like in politics or football. In audio, there is a psychology where objectivists prefer to prefer low THD, flat FR, etc
(just to be politically, err... scientifically correct) There are many factors defining realness/naturalness/hifi. Even I prefer to consider F (TangBand) as the most "natural". From certain point of view it is the best representing the reference or live sound.
It's as subjective as deciding which one is closer to a male dog: (1) a female dog (2) a male cat.
Last edited:
There's a psychology saying that people prefer to be with the winner. Like in politics or football. In audio, there is a psychology where objectivists prefer to prefer low THD, flat FR, etc
That argument doesn't seems to fit for you Jay on the contrary you never seems in doubt because of experience or constructive thinking, just see below
....that's not bad being strong and having most cut out and experienced but can make you difficult to move even a little bit in a argument : )difficult too choose between all these good drivers
I tested too fastly and now I assume I voted in round 4 & 5 for not too much dynamic drivers because I hear sibilances on dynamic peaks due to the reccording process and maybe my own setups, though I can not hear those sibilances on the three reference tracks provided in the rounds 5 links ! I choosed for the F10 ScanSpeak because of that and also FR88x on the clip 3 (voice), though I prefered piano and cymballs on the Alpine in round 5 (from memory the Scan 10F was better and more neutral on the jazz clip) ! Clip 1 was not tested due to my poor love for this track
This week infew days I receive this stuffs thanks to the good advices from round 4 : http://www.isemcon.net/en/emx-7150.htm ! Have already an EMU external Sound Card with Phantom PS ! I should understand how is the curve of my speakers at listening position as they provide the mic also with a semi-réverbérant calibration file !
If X agree and have time for that between his many speakers projects, it could be fun to do a round between drivers which won and had the votes even in second place (Visaton P80, etc...)
I propose we have a look in our archives CD and find very dynamic reccording choosed in this list : http://dr.loudness-war.info/album/list/dr-max/desc : choosing one rock (I saw Dire Strait in the second of third page in the most dynamic reccording), one classique or world music and one jazz : one of the reccording has to have human voice of course !
My DAC is 16 bits ONLY ! But I surmise it's enough as I can clearly understand Something that drill my ears on dynamics peaks on almost all the drivers (not with the 10F, but with all on the round 5 !). My amp is dynamic (a Chord !) and my speaker are subjectvly fast though not ESL, compression, Array with multiple drivers, magneplanar ! They are wide band between 125 hZ and 2600 hZ then a tweeter : both aluminium but with passive anti glitch devices (acoustical filters).
(PS : did the 5" Aurus Cantum from Continuom by Jeff Bagby were tested previously ?)
regards
I tested too fastly and now I assume I voted in round 4 & 5 for not too much dynamic drivers because I hear sibilances on dynamic peaks due to the reccording process and maybe my own setups, though I can not hear those sibilances on the three reference tracks provided in the rounds 5 links ! I choosed for the F10 ScanSpeak because of that and also FR88x on the clip 3 (voice), though I prefered piano and cymballs on the Alpine in round 5 (from memory the Scan 10F was better and more neutral on the jazz clip) ! Clip 1 was not tested due to my poor love for this track
This week infew days I receive this stuffs thanks to the good advices from round 4 : http://www.isemcon.net/en/emx-7150.htm ! Have already an EMU external Sound Card with Phantom PS
! I should understand how is the curve of my speakers at listening position as they provide the mic also with a semi-réverbérant calibration file ! If X agree and have time for that between his many speakers projects, it could be fun to do a round between drivers which won and had the votes even in second place (Visaton P80, etc...)
I propose we have a look in our archives CD and find very dynamic reccording choosed in this list : http://dr.loudness-war.info/album/list/dr-max/desc : choosing one rock (I saw Dire Strait in the second of third page in the most dynamic reccording), one classique or world music and one jazz : one of the reccording has to have human voice of course !
My DAC is 16 bits ONLY ! But I surmise it's enough as I can clearly understand Something that drill my ears on dynamics peaks on almost all the drivers (not with the 10F, but with all on the round 5 !). My amp is dynamic (a Chord !) and my speaker are subjectvly fast though not ESL, compression, Array with multiple drivers, magneplanar ! They are wide band between 125 hZ and 2600 hZ then a tweeter : both aluminium but with passive anti glitch devices (acoustical filters).
(PS : did the 5" Aurus Cantum from Continuom by Jeff Bagby were tested previously ?)
regards
Last edited:
For those whom want a good dac but cheap, Pedja Rogic is launching a new edition of the AYA 2 2014 (TDA1541A) untiil 7 november ! (30 euros the pcb ) !
Sorry for this off topic and adds but I believe it's friendly DIY and an opportunity ! This is the DAC I use myself with IanCanada stuffs (FiFFO, MCLK, simultaneous mode) in front of it ! Although the AYA 2 2014 edition is a complete project in itself and don't need others stuffs to sing ! http://www.audialonline.com/online-topics/ . It has an external Demning clock circuit than John from ECdesigns used himself with some variations !)
) !Sorry for this off topic and adds but I believe it's friendly DIY and an opportunity ! This is the DAC I use myself with IanCanada stuffs (FiFFO, MCLK, simultaneous mode) in front of it ! Although the AYA 2 2014 edition is a complete project in itself and don't need others stuffs to sing ! http://www.audialonline.com/online-topics/ . It has an external Demning clock circuit than John from ECdesigns used himself with some variations !)
For those whom want a good dac but cheap, Pedja Rogic is launching a new edition of the AYA 2 2014 (TDA1541A) untiil 7 november ! (30 euros the pcb
Sorry for this off topic and adds but I believe it's friendly DIY and an opportunity ! This is the DAC I use myself with IanCanada stuffs (FiFFO, MCLK, simultaneous mode) in front of it ! Although the AYA 2 2014 edition is a complete project in itself and don't need others stuffs to sing ! Online topics . It has an external Demning clock circuit than John from ECdesigns used himself with some variations !)
Another very good yet inexpensive DAC is the UCA202. Read the review by NwAvGuy here:
NwAvGuy: Behringer UCA202 Review
I have one and use it for my headphone amp (need high impedance headphone).
I tested too fastly and now I assume I voted in round 4 & 5 for not too much dynamic drivers because I hear sibilances on dynamic peaks due to the reccording process and maybe my own setups, though I can not hear those sibilances on the three reference tracks provided in the rounds 5 links ! I choosed for the F10 ScanSpeak because of that and also FR88x on the clip 3 (voice), though I prefered piano and cymballs on the Alpine in round 5 (from memory the Scan 10F was better and more neutral on the jazz clip) ! Clip 1 was not tested due to my poor love for this track
This week infew days I receive this stuffs thanks to the good advices from round 4 : http://www.isemcon.net/en/emx-7150.htm ! Have already an EMU external Sound Card with Phantom PS
If X agree and have time for that between his many speakers projects, it could be fun to do a round between drivers which won and had the votes even in second place (Visaton P80, etc...)
I propose we have a look in our archives CD and find very dynamic reccording choosed in this list : http://dr.loudness-war.info/album/list/dr-max/desc : choosing one rock (I saw Dire Strait in the second of third page in the most dynamic reccording), one classique or world music and one jazz : one of the reccording has to have human voice of course !
My DAC is 16 bits ONLY ! But I surmise it's enough as I can clearly understand Something that drill my ears on dynamics peaks on almost all the drivers (not with the 10F, but with all on the round 5 !). My amp is dynamic (a Chord !) and my speaker are subjectvly fast though not ESL, compression, Array with multiple drivers, magneplanar ! They are wide band between 125 hZ and 2600 hZ then a tweeter : both aluminium but with passive anti glitch devices (acoustical filters).
(PS : did the 5" Aurus Cantum from Continuom by Jeff Bagby were tested previously ?)
regards
I am not sure what you are proposing. You would like another round of recordings using songs from your list? Pick top drivers again like top 2 from previous rounds?
I have the AC130F1 (from Continuum) in a speaker design as a mid bass - it's not full range, although close to a full range as it goes up to 10k. I could test it as a 3 way with a good tweeter just to show what a good three way can sound like. My current OB setup with TG9FD mid and Dayton ND25FA tweeter sounds exceptionally nice. I could pair that with the same sealed RS225 woofer.
It's a lot of work but maybe when I don't have a lot of pressing projects. Although I may make a clip of tracks 1 and 2 with my current 3 way OB project and post for you hear.
Hello X,
Forgett the Aurum Cantus, I believed it was FR !
Yes, a sort of round of the best speaker of all the rounds but avaliable in your stock to avoid people to send their own (you seem to have already the best ones in your collection ?). Not my list of song but a more universal proposal with the main three styles (rock, jazz, classic-acoustic) choosed in the link given as known for their huge dynamic gap !
Bad idea ? Maybe, I don't know. People had made their own choice already ?! Although I have to understand from my system why I hear no sibilance in the reference track in MP3 and why I hear sibilance from the reccorded drivers during what I call "dynamic peaks" ! Some don't seem to hear it from both the ref tracks and the drivers tests ! (so some home works for me in the weeks to come !)
regards
Forgett the Aurum Cantus, I believed it was FR
!Yes, a sort of round of the best speaker of all the rounds but avaliable in your stock to avoid people to send their own (you seem to have already the best ones in your collection ?). Not my list of song but a more universal proposal with the main three styles (rock, jazz, classic-acoustic) choosed in the link given as known for their huge dynamic gap !
Bad idea ? Maybe, I don't know. People had made their own choice already ?! Although I have to understand from my system why I hear no sibilance in the reference track in MP3 and why I hear sibilance from the reccorded drivers during what I call "dynamic peaks" ! Some don't seem to hear it from both the ref tracks and the drivers tests ! (so some home works for me in the weeks to come
!)regards
Below are my impressions of the speakers in round 5. I voted for F which is the W2-852SH, as it sounded most like I think a live concert might sound.
I also went back to round 4. There I voted for the 10F. Making a comparison between the 10F and the W2-852SH, I feel the 10F is slightly better.
I immediately rejected driver A and B. I does not surprise me that the DEF and G are so close.
I purchased a UCA202 after I read the NwAvGuy review. I was pleased with it at time. I have since purchased a couple of ES9023 DAC's, and recently a 9018 ASYNCHRONOUS DAC. The ES9023 is a significant improvement over the UCA202, and well worth its small cost.
The improvement of the 9018 DAC over the ES9023 is not so noticable on immediate listening, but on longer listening, there is a greater WOW factor on a well recorded piece of music.
Each of these DACs sound much better through an external headamp than connected directly to headphones.
I also went back to round 4. There I voted for the 10F. Making a comparison between the 10F and the W2-852SH, I feel the 10F is slightly better.
I immediately rejected driver A and B. I does not surprise me that the DEF and G are so close.
I purchased a UCA202 after I read the NwAvGuy review. I was pleased with it at time. I have since purchased a couple of ES9023 DAC's, and recently a 9018 ASYNCHRONOUS DAC. The ES9023 is a significant improvement over the UCA202, and well worth its small cost.
The improvement of the 9018 DAC over the ES9023 is not so noticable on immediate listening, but on longer listening, there is a greater WOW factor on a well recorded piece of music.
Each of these DACs sound much better through an external headamp than connected directly to headphones.
here's a song I'd like for comparisons - not sure of the year maybe 1960 or a bit earlier (?)
https://www.youtube.com/watch?v=S6CMPq5O8tU
https://www.youtube.com/watch?v=S6CMPq5O8tU
Thanks for the feedback - can you get us a link to the ES9023 you got? Also what headphone amp are you using? I was consdering the O2 for a while now but was very happy with the UCA202 so far. There is a mod that NWAvGuy has for the UCA202 that lets it drive low impedance headphones - a new SMT op amp and better electrolytic power caps I believe. Supposedly a $6 in cost for upgrades if you are handy with SMT soldering, which, fortunately, I am.
Should check out the ES9023 though.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.