A Subjective Blind Comparison of 2in to 4in drivers - Round 3

Select the driver that you think sounds best here.

  • A

    Votes: 1 3.6%
  • B

    Votes: 9 32.1%
  • C

    Votes: 1 3.6%
  • D

    Votes: 5 17.9%
  • E

    Votes: 2 7.1%
  • F

    Votes: 4 14.3%
  • G

    Votes: 2 7.1%
  • H

    Votes: 4 14.3%

  • Total voters
    28
  • Poll closed .
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Founder of XSA-Labs
Joined 2012
Paid Member
Please do not post your opinions or endorsements of any driver for easy public viewing in this thread until the reveal. If you want to post your opinions, put it in a txt file and attach so that people can read if they want a spoiler. Thanks!

After receiving the sets of drivers submitted by generous members yesterday, I was able to complete collecting data for the Round 3 of the Subjective Blind Comparison thread. This test will include drivers as small as 2in class and not go above 4in. Originally, Round 3 was going to be the final round to have a showdown between the top contenders of the earlier rounds. However, more entries arrived. In fact, I have 7 new drivers and a request to put the A7.3 back in the race again since the contestants now include an A7P (paper cone).

In case you are new to this, more info can be found in previous threads here:
http://www.diyaudio.com/forums/full-range/274092-subjective-blind-comparison-3in-5in-drivers-round-2-a.html

http://www.diyaudio.com/forums/full-range/270614-subjective-blind-comparison-3in-5in-full-range-drivers.html

The hardware setup is a little different this time with a sealed 24 liter cabinet loaded with a Dayton RS225-8 woofer running with a Linkwitz transform, and I will be using the new S. Harsch crossover topology (http://www.diyaudio.com/forums/multi-way/88135-filler-driver-ala-b-o-26.html#post4399135) that provides near transient perfect response with a steep 4th order Butterworth low pass on the woofer and a 2nd order Bessel high pass on the mid/tweet. The electrical frequency is 350Hz, but resulting acoustic crossover is near 500Hz. A FAST implementation is use once again, to offload the bass duties from the drivers so that their midrange and upper range can be compared in an apples-to-apples fashion. Some of these drivers will have superior bass capability vs others in a properly designed cabinet (the A7.3 for example has extension down to 40Hz in a MLTL). However, that would require different cabinets for different drivers and then we cannot isolate the effect of the listening test to just the driver vs the driver and cabinet combination. The amp will be a Ybdz TPA3116D2 modified with Panasonic OSCON 330uF SEPF caps and the 330pF/10R bootstrap snubber mod on the mid/tweet. Amps are driven with standard laptop 19v 4.65amp smps bricks. The DSP crossover is a 2x4 miniDSP running advanced 2-way XO as usual. Source is a Phillips 5 disc carousel CD player using the headphone outs to drive the miniDSP inputs. Music is taken from original CD's and trimmed to 45 second clips and placed on one long track so that all song clips are recorded in one continuous stream without touching recorder setup. This really improved the workflow and made taking data much faster, and more consistent between songs.

The measurement mic is now a miniDSP UMIK-1 calibrated by a third party (Cross Specrtrum Labs). The measurement mic and Zoom H4 digital recorder share the same stand and are swapped in and out. The clips are recorded as 96kHz 24bit raw .wav files and then trimmed and level adjusted in Audacity for a max -1.0dB FS output. Audacity is also used to convert the .wav files to 320kbit mp3 (48kHz sample rate). The setup is fixed with the only thing variable is the driver under test and its adapter baffle plate to mount it to the Nautaloss infinite baffle like sealed rear spiral TL. The center-to-center distance between the driver under test and the woofer is 7.75in. The mechanical and physical locations of the mic stand, speaker cabinets, etc are all identical between cases. The ONLY thing that changes is the driver between tests, and the ONLY adjustment electrically is the level matching gain on the mid/tweet (up to -5dB depending on driver sensitivity) and a very small 20 to 80 microsecond delay adjustment to compensate for driver offset due to bezel thickness differences.

Here is an example of the XO frequency response taken from the 10F/8424 and RS225-8 Ref Monitor but using the same Harsch XO topology with a XO frequency of 400Hz:
495043d1437630819-10f-8424-rs225-8-fast-ref-monitor-harsch-10f-rs225-xo.png


This is the absolute phase of the system (fairly flat for a 4th order BW and 2nd order Bessel at 400Hz - no phase wraps):
495044d1437630819-10f-8424-rs225-8-fast-ref-monitor-harsch-10f-rs225-phase.png


More info on the S. Harsch XO topology here:
http://petoindominique.fr/pdf/phase.pdf

Here is a photo of the test setup with a driver being measured for frequency response (mic at 0.5 m distance, shorter distance used to avoid floor bounce cancellation and to minimize room reflections). Note that the ScanSpeak 10F/8424 visible in the same cabinet as the woofer is not connected or used. (More info about the woofer cabinet can be found in the 10F/RS225 Ref Monitor thread):

495455d1437850169-subjective-blind-comparison-2in-4in-drivers-round-3-round-3-meas-mic-setup.png


The new drivers for Round 3 are:

1. SB Accoustics SB65WBAC25-4 2.5in aluminum cone Nd magnet
2. Peerless P830983 2in aluminum cone Nd magnet
3. Tang Band W3-1364SA 3in bamboo cone Nd magnet
4. GR Research LGK 3in paper cone ferrite magnet
5. Mark Audio Alpair A7P 4in paper cone ferrite magnet
6. Visaton B80 3.5in paper cone Nd magnet
7. Visaton FRS5X 8ohm 2in paper cone ferrite magnet
8. Mark Audio Alpair A7.3 aluminum cone ferrite magnet (previously tested in Round 2)

The reference sound files for the test tracks are attached below. Note that the sound clips during tests are recorded in stereo from the Zoom H4 but the speaker is mono playing only the right channel. So when listening to the reference, keep that in mind as it may sound different. Change .asc extension to .mp3 to listen.

Here is a photo of the line-up for Round 3. From the top and going clockwise A7P, W3-1364SA, LGK, P830983, FRS5X, SB65WBAC25-4, B80 (center),and A7.3 not shown:

495454d1437850169-subjective-blind-comparison-2in-4in-drivers-round-3-round-3-lineup.png


Here is a photo showing the Zoom H4 recording the sound clips in the same position as the measurement mic:

495456d1437850169-subjective-blind-comparison-2in-4in-drivers-round-3-round-3-recorder-setup.png


The sound clips will be uploaded in later posts.

*** Results posted here: http://www.diyaudio.com/forums/full...2in-4in-drivers-round-3-a-18.html#post4416564 ***
 

Attachments

  • Round-3-Lineup.png
    Round-3-Lineup.png
    417.4 KB · Views: 3,741
  • Round-3-Meas-Mic-Setup.png
    Round-3-Meas-Mic-Setup.png
    339.1 KB · Views: 3,708
  • Round-3-Recorder-Setup.png
    Round-3-Recorder-Setup.png
    351.4 KB · Views: 3,638
  • Clip-2-Reference.asc
    1.7 MB · Views: 205
  • Clip-4-Reference.asc
    1.7 MB · Views: 127
  • Clip-5-Reference.asc
    1.7 MB · Views: 116
  • Clip-6-Reference.asc
    1.7 MB · Views: 122
Last edited:
Founder of XSA-Labs
Joined 2012
Paid Member
Sound Clip #2 for all drivers

Drivers are randomly labeled A thru H, but naming convention is consistent for all sound clips.

Sound Clip #2 (pop - male vocals)

Note, change file extension from .asc to .mp3 in order to listen.
 

Attachments

  • A-Clip-2.asc
    1.7 MB · Views: 294
  • B-Clip-2.asc
    1.8 MB · Views: 317
  • C-Clip-2.asc
    1.8 MB · Views: 222
  • D-Clip-2.asc
    1.8 MB · Views: 234
  • E-Clip-2.asc
    1.7 MB · Views: 219
  • F-Clip-2.asc
    1.8 MB · Views: 231
  • G-Clip-2.asc
    1.7 MB · Views: 220
  • H-Clip-2.asc
    1.8 MB · Views: 245
Last edited:
Founder of XSA-Labs
Joined 2012
Paid Member
Sound Clip #6 for all drivers

Sound Clip #6 (pop - female vocals)
 

Attachments

  • A-Clip-6.asc
    1.7 MB · Views: 166
  • B-Clip-6.asc
    1.7 MB · Views: 202
  • C-Clip-6.asc
    1.7 MB · Views: 158
  • D-Clip-6.asc
    1.7 MB · Views: 176
  • E-Clip-6.asc
    1.7 MB · Views: 154
  • F-Clip-6.asc
    1.8 MB · Views: 167
  • G-Clip-6.asc
    1.7 MB · Views: 173
  • H-Clip-6.asc
    1.7 MB · Views: 177
Interesting line up.

Can you say why the FRS5 was chosen as opposed to the FRS5X which is the better variant? I use the latter, and despite being SPL limited it is very natural sounding indeed, especially considering the cost

If you weren't so far from me I would've happily shipped over my 2" neo.magnet TBs for this test :(
 
Last edited:
Founder of XSA-Labs
Joined 2012
Paid Member
Interesting line up.

Can you say why the FRS5 was chosen as opposed to the FRS5X which is the better variant? I use the latter, and despite being SPL limited it is very natural sounding indeed, especially considering the cost

If you weren't so far from me I would've happily shipped over my 2" neo.magnet TBs for this test :(


My mistake, it is indeed the FRS5X 8ohms. I have it simply because member Timpert threw it in when he provided me his well broken-in A7P's.

The distance didn't stop Timpert or many others in Eurozone or even East Asia from sending me their drivers :D
 
Last edited:
interesting this test

some tracks are far more revealing of sound than others....

I have completely different preferences between tracks 2 & 4, and tracks 5 & 6

this is a tough one....:scratch:

interesting also that only one track is more complex than guy/girl and guitar...I was looking forward to some rock not a ballad!

:violin:
 
Last edited:
The only criticism of this test I would make is that:

It is a very tough job for the 2" drivers - without giving anything away, I already am very very certain my preference is not a 2" driver.

(In my little experiments with 2" for FAST stuff, I have found that using them with a HP filter lower than 600Hz BW2 rather than BSL2 is really pushing them too much, where most 3-4" drivers are happy at lower crossover points.)
 
Last edited:
I was looking forward to some rock not a ballad!

:violin:

Agreed.

The only criticism of this test I would make is that:

It is a very tough job for the 2" drivers - without giving anything away, I already am very very certain my preference is not a 2" driver.

The SB is more than capable at 85dB. It's distortion is pretty much state of the art and remains extremely low even when driven a lot harder. With a decent high pass in place this driver will be thermally limited and with a suitably powerful amplifier will melt before its distortion goes high enough to be objectionable.

It will be interesting to see how the results pan out!

I am missing the AC/DC track though. It was perfect, in my opinion, for driver assessment, which is what you want for something like this, music that highlights weaknesses.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.