The much awaited Round 2 of the earlier thread http://www.diyaudio.com/forums/full-range/270614-subjective-blind-comparison-3in-5in-full-range-drivers.html now begins!
The contestants are, in no particular order:
1. Vifa TG9FD10-8 (fiberglass cone)
2. Tang Band W4-1320SB (paper cone)
3. Peerless P830986 (aluminum cone)
4. Fostex FF105WK (paper cone)
5. ScanSpeak 10F/8424G00 (fiberglass cone)
6. Mark Audio Alpair A7.3 (silver cone) (aluminum cone)
7. AHE 3in (from Taiwan) (paper cone) http://m.ruten.com.tw/goods/show.php?g=21206258966085
8. A surprise entry whose identity will be revealed after votes are in
Thus we have choices that will be randomly labeled A thru H.
Here is the photo of the entries for Round 2 (not including surprise entry).
Clockwise from 1 o'clock: W4-1320SB, FF105WK, P830986, 10F/8424, AHE, A7.3, and TG9FD (center).
The recordings are of the same sound clips from Round 1, with the difference that a higher recording quality of 96kHz was used and the resulting clips are 320kbps mp3's with 48kHz sample rate. As usual, the same setup as the previous thread was used with respect to the speaker enclosure (trapezoid baffle Nautaloss spiral sealed TL) and XKi dual W5-876SE drivers for the bass. XO was provided by miniDSP with 225Hz LR4 and -3dB baffle step correction applied to all drivers. XY mic of Zoom H4 was set at 1m away. Source was high res files from Audacity played thru laptop for clips 1 and 2, clip 3 was from CD player. It should be noted that one of the drivers here (A7.3) received extra special treatment per the manufacturer's recommendation prior to being tested unlike all the other drivers which were tested as-is, straight out of the box, with no pre-conditioning. The A7.3 was subjected to a 200 hr break-in consisting of clip 1 music at low volumes for first 40 hrs, then progressively louder with clip 3 for 80 more hrs, then ending with moderate volume for 80 more hrs of bass line heavy track by Suzanne Vega ("Headshots").
Here is the first track for all entries (change the extension .asc to .mp3 to listen).
I will post the other 2 tracks later.
Enjoy!
I want to thank all the folks who were so generous in providing me drivers to test - this test and thread would not be possible without you. 🙂
Some of the drivers, I purchased myself this time though...
A reminder to not analyze tracks using software, and if you do, please keep it to yourself.
Answer key given in big reveal after votes were all in.
http://www.diyaudio.com/forums/full...3in-5in-drivers-round-2-a-17.html#post4328938
The contestants are, in no particular order:
1. Vifa TG9FD10-8 (fiberglass cone)
2. Tang Band W4-1320SB (paper cone)
3. Peerless P830986 (aluminum cone)
4. Fostex FF105WK (paper cone)
5. ScanSpeak 10F/8424G00 (fiberglass cone)
6. Mark Audio Alpair A7.3 (silver cone) (aluminum cone)
7. AHE 3in (from Taiwan) (paper cone) http://m.ruten.com.tw/goods/show.php?g=21206258966085
8. A surprise entry whose identity will be revealed after votes are in
Thus we have choices that will be randomly labeled A thru H.
Here is the photo of the entries for Round 2 (not including surprise entry).
Clockwise from 1 o'clock: W4-1320SB, FF105WK, P830986, 10F/8424, AHE, A7.3, and TG9FD (center).

The recordings are of the same sound clips from Round 1, with the difference that a higher recording quality of 96kHz was used and the resulting clips are 320kbps mp3's with 48kHz sample rate. As usual, the same setup as the previous thread was used with respect to the speaker enclosure (trapezoid baffle Nautaloss spiral sealed TL) and XKi dual W5-876SE drivers for the bass. XO was provided by miniDSP with 225Hz LR4 and -3dB baffle step correction applied to all drivers. XY mic of Zoom H4 was set at 1m away. Source was high res files from Audacity played thru laptop for clips 1 and 2, clip 3 was from CD player. It should be noted that one of the drivers here (A7.3) received extra special treatment per the manufacturer's recommendation prior to being tested unlike all the other drivers which were tested as-is, straight out of the box, with no pre-conditioning. The A7.3 was subjected to a 200 hr break-in consisting of clip 1 music at low volumes for first 40 hrs, then progressively louder with clip 3 for 80 more hrs, then ending with moderate volume for 80 more hrs of bass line heavy track by Suzanne Vega ("Headshots").
Here is the first track for all entries (change the extension .asc to .mp3 to listen).
I will post the other 2 tracks later.
Enjoy!
I want to thank all the folks who were so generous in providing me drivers to test - this test and thread would not be possible without you. 🙂
Some of the drivers, I purchased myself this time though...
A reminder to not analyze tracks using software, and if you do, please keep it to yourself.
Answer key given in big reveal after votes were all in.
http://www.diyaudio.com/forums/full...3in-5in-drivers-round-2-a-17.html#post4328938
Attachments
Last edited:
Second sound clip
Here is the second sound clip for all entries.
Here is the second sound clip for all entries.
Attachments
Third clip
Here is the third sound clip for all entries.
Here is the third sound clip for all entries.
Attachments
Thank you for taking the time to do this!
Clips A, E and D sound the best to me personally in that order.
Clips A, E and D sound the best to me personally in that order.
Thanks X,
You're going to add a poll and a deadline like last time?
Don't wait too long on the deadline! 😉
You're going to add a poll and a deadline like last time?
Don't wait too long on the deadline! 😉
For anyone feeling overwhelmed by number of choices you can apply some rigor to the selection process using the tournament method.
Tournament Tree (Winner Tree) and Binary Heap - GeeksforGeeks
Tournament Tree (Winner Tree) and Binary Heap - GeeksforGeeks
Please do listen to the Clip No. 2 - I noticed that folks are downloading 1 and 3 in greater numbers. Metal may not be your thing but this clips tests how the speaker sounds with very dry, non-reverberent source material with large dynamic range.
Doesn't matter what you use to listen as the entire source-amp-speakers-mic-recorder-MP3 encoding-your player-your ADC-your amp-your speaker or headphone, are all the SAME for all tracks except the only difference is the DRIVER.
If it sounds different it MUST BE due to the driver.
But I recommend good headphones and not speakers to listen critically. You can use an iPhone speaker if that is what you normally listen to music with - but probably won't hear a difference in that case.
If it sounds different it MUST BE due to the driver.
But I recommend good headphones and not speakers to listen critically. You can use an iPhone speaker if that is what you normally listen to music with - but probably won't hear a difference in that case.
Last edited:
You might be happy with that but it sounds naive to me, I've a feeling it's a lot more complicated.
It's simple controlled experimental methodology. If you fix everything and change one thing and you detect a difference in the outcome, the difference must have resulted from the one thing that was changed. A lot care went into setup and taking the data to ensure consistency and that the only thing that changed was the driver. Level adjustments between the bass and full range was the one adjustable variable - but that was done with aid of measurement to ensure baseline SPL is flat from bass to mids/treble.
But then it doesn't surprise me that you doubt everything ever presented to you.
But then it doesn't surprise me that you doubt everything ever presented to you.
Last edited:
In case anyone wants to have a blind listening test of the difference in sound between a LR4, LR2, BW1 XO in FAST speaker - I have some sound clips here:
http://www.diyaudio.com/forums/full...-rs225-8-fast-ref-monitor-12.html#post4320047
http://www.diyaudio.com/forums/full...-rs225-8-fast-ref-monitor-12.html#post4320047
Here's a double elimination tournament template if anyone is interested.
Direct Link: https://i.imgur.com/Dd9SxXE.png

Direct Link: https://i.imgur.com/Dd9SxXE.png
Last edited:
The output characteristics of a given driver, or component of any sort for that matter can be positively or negatively effected by the characteristics of the driver or system you play it back through, isn't that why component matching is an issue or does that not matter?
Hope this helps someone else too. Just found this A/B audio testing software:
ABX audio testing tool
It's a java application and runs just fine on my Debian Jessie machine with ALSA. There's a version in the repositories, but it seems to be quite old compared to the "all-in-one package" that I downloaded directly from the website.
ABX audio testing tool
It's a java application and runs just fine on my Debian Jessie machine with ALSA. There's a version in the repositories, but it seems to be quite old compared to the "all-in-one package" that I downloaded directly from the website.
The output characteristics of a given driver, or component of any sort for that matter can be positively or negatively effected by the characteristics of the driver or system you play it back through, isn't that why component matching is an issue or does that not matter?
Let's assume all these drivers can be reasonably driven with a modern 4ohm capable class D amp. None of these are high impedance "tube amp" drivers. The enclosure is broadly applicable for a FAST as it is a sealed tapered TL that measures like an open baffle with very little back reflection. We are not trying to extract bass below 225Hz out of any of them.
This is very interesting...quite a big difference between drivers. Personally still prefer faital from the first round, but in this round A & G are very similar (and also similar to faital), but prefer G.
I could play the clips in the first shootout, but this time it's telling me that Windows can't play the file.
- Home
- Loudspeakers
- Full Range
- A Subjective Blind Comparison of 3in to 5in drivers - Round 2