A Subjective Blind Comparison of 3in to 5in Full Range Drivers

[xrk971]

I wanted to provide a reference for what the differences between drivers sound like given identical setups and conditions with the only variable being the top driver. I will use a FAST to offload all the bass duties which is really enclosure and alignment dependent so that we can concentrate on the intrinsic differences between the drivers. Think of it almost like a wine tasting. There are no winners or losers here - just "subjective" differences and preferences. So let's try to keep the conversations civil, friendly, engaging, and open-minded. Let's not go down the parochial path of insisting one way is better or the best.

I request that you do not use a program or DAW to de-construct the clips so that you can 'see' the response - let's just use our ears. If you do so, please do not post "objective" results of your analysis or use the analysis as the basis of your comment. Finally, I would suggest you listen using a high quality set of 'reference' or 'studio' headphones that will give the least coloration.

Here are the drivers that will be part of this blind listening test:

- Vifa TC9FD (3.5in paper cone 8 ohm)
- Faital Pro 3FE22 (3in paper cone 16 ohm)
- Dayton Audio RS100-4 (3.5in aluminum cone 4 ohm)
- Dayton Audio RS100P-4 (3.5inpaper cone 4 ohm)
- Dayton Audio PS95-8 (3in paper cone 8 ohm)
- MA CHN-70 (4in paper cone 8 ohm)
- PRV 5MR450NDY (5in 'midrange' pro audio paper cone 8 ohm)

Edit March 21, 2015: Update on poll results - answer key for blind test:
A=5MR450NDY
B=TC9FD
C=PS95-8
D=3FE22r16
E=RS100P-4
F=CHN-70
G=RS100-4


Here is a photo of the palette of cheese, err I mean drivers to taste. (Remember that this is all for fun - like cheese or wine tasting):

I setup an identical FAST system for the test using an XKi dual driver bass unit and the Nautaloss sealed spiral TL top with a trapezoidal baffle as described previously here:An Objective Comparison of 3in - 4in Class Full Range Drivers - diyAudio

The XO was provided by a miniDSP and set at 225Hz 24dB/oct LR for all cases, a -3dB high-shelf Q=0.7 baffle step correction was applied at 750Hz for all cases (note correction here from previous Q=0.5, 850Hz, -5dB - my mistake). Only level matching between the sub and top was adjusted between cases. The mic is a Zoom H4 recording at 48kHz and 32bits and converted to 192kbit mp3 files in Audacity so that a 40 second clip can fit under the 962kB upload limit. In Audacity, I adjusted the level so that the final output file has a maximum peak at -1dB.

Here is a photo of the Nautaloss-XKi FAST used for this test (shown with PRV 5MR450NDY, not included in above photo):

3 test tracks were selected - source for 1 and 2 were high res 48kHz 32bit wav files played from Audacity via built in cheapo DAC to headphone out in my laptop, and third was straight from original CD via old Philips Carousel CD player. Just so you get an idea of the signal source and chain. The amp after the miniDSP is a custom TPA3116D2 using member DUG's v0 PCB with quality Panasonic FM 560uF power rail caps, 1uF 250V EPCOS polyester film input caps, and quality Bourns 10uH inductors.

1. tests the tonality and texture for simple female vocals and instruments
2. hard rock classic that was recorded in a very acoustically 'dry' environment with almost no reverb or reflections audible
3. has a lot of dynamic range and detail

Edit March 7, 2015: here are the source tracks in 320kbit mp3 format for use as reference.
A Subjective Blind Comparison of 3in to 5in Full Range Drivers - Page 25 - diyAudio

The drivers will be posted and grouped according to song. The letter designations are random amongst drivers but consistent across different test tracks. That is, driver A is the same for all tracks.

Enjoy!

Please provide your *subjective* comments and opinions of the sound you hear. You are welcome to try to guess the driver but I won't disclose until most people have had a chance to listen to the tracks and decide for themselves. if you make a prediction, we don't want to sway opinion, so I suggest you put your predictions in a txt file and post as an attachment so that people need to click on it for the 'spoiler'. - Thanks.

Post #509 and #510 contain sound clips of some drivers that are no longer made provided to my by Godzilla. You may want to have a listen. There is no voting on these as you can't get them anymore so they are there for fun.

A Subjective Blind Comparison of 3in to 5in Full Range Drivers - Page 51 - diyAudio


Edit March 19, 2019:

Index of all Blind Virtual Audition threads:

Subjective Blind Testing of Fullrange Drivers by xrk971

Round 1:
A Subjective Blind Comparison of 3in to 5in Full Range Drivers - diyAudio

Round 2:
A Subjective Blind Comparison of 3in to 5in drivers - Round 2

Round 3:
http://www.diyaudio.com/forums/full...ind-comparison-2in-4in-drivers-round-3-a.html

Round 4:
http://www.diyaudio.com/forums/full...ind-comparison-2in-4in-drivers-round-4-a.html

Round 5:
http://www.diyaudio.com/forums/full...d-comparison-2in-3-5in-drivers-round-5-a.html

Round 6:
http://www.diyaudio.com/forums/full...-blind-abx-test-enabled-ff85wk-round-6-a.html

Testing (virtual audition) of Amplifiers:
http://www.diyaudio.com/forums/solid-simple-quasi-mosfet-amstate/295286-virtual-audition-very-p.html

 

[xrk971]

Test Track No. 1

Here is test track number one for cases A through G. I suggest adding all tracks to a playlist so you can quickly go between them to do A/B'ing.

 

[xrk971]

Test Track No. 2

Here is the next set chosen for its 'dry' recording environment.

 

[xrk971]

Test Track No. 3

Final set here with some jazz to test detail and range.

 

[perceval]

Thanks X!

That's a lot of work!

Even though it is mighty subjective, it is interesting for me as I do not have access to so many drivers for building or just plain listening!

So it's a fun way to spend a rainy afternoon. I'll give your files a go a little later.

Quick question: Are the drivers always at their respective lettering or did you mismatch them between the clips also? Meaning, is driver A always the same driver playing the 3 tracks?

 

[xrk971]

Perceval,
Drivers are consistently labeled. A is always A etc.
I hope you enjoy it and that is another reason for putting this out is for folks who have not had a chance to listen to a bunch of different drivers can get an idea of the variation in the character.

 

[nesha]

Hi X, have you compared sound of original music file with the one you recorded? I can only compare sound of recordings... Sound that you hear have to travel through your recording equipment and my playback chain... I don't expect you and I to like the same driver. Right now they all sound the same to me... With better phones, maybe... With Nora Jones I would pick E and G as slightly better.Guessing the drivers would be too wild for me.

 

[xrk971]

Pull the song from YouTube or iTunes etc if you want to hear source. It will be different of course - it has gone through a DAC, miniDSP, amp, speaker and my room. However, everything is constant except the top full range driver. Thus, any difference you hear is the effect of the driver itself.

Your choice of drivers is very interesting but I can't comment yet. Please post your top picks in a text file and upload next time so we don't sway people's opinions. Trying to avoid popularity begets popularity behavior in humans.

 

[CLS]

Scanned all of them quickly by my lap top & (tiny earplug-type) earphones, I found C & D are my favorite - with compromise fitting my preference.

It's a very preliminary observation, though. It's a mixed bag. The one with sparkling cymbal is too harsh on the electric guitar, or the one with rich voice maybe somewhat nasal or too dark on top... etc. I'd want to grab the parts I like and 'remix' them if I could. Maybe that's why my main system is multiway, haha...

 

[nesha]

I don't say it 's impossible to hear difference in this test, but question is would I like the same drivers if I make my choice there in your room.

 

[ra7]

X, thanks! Can you post a blind poll at the top of the thread? I know polls can be done on diyaudio but not sure if it allows a blind poll. A blind poll is where the results are not disclosed until the poll is closed. We don't want to add a bias before votes are cast.

 

[Bigun]

Good thread !

How many hours on these drivers before you recorded from them ?

 

[xrk971]

I don't say it 's impossible to hear difference in this test, but question is would I like the same drivers if I make my choice there in your room.

Effect of room above 225Hz is not as much as below 225Hz. So differences really will be driver related rather than room related.

@Bigun, no break in considered. Some drivers like 5MR450NDY or TC9FD I have had for 1 to 2 years so maybe lots of use. Most other drivers were newly purchased with 2hrs to 50hrs. Break in usually affects the lower frequencies rather than higher.

 

[planet10]

It should be evident that the clips are all only valid in the context of the system & room they are in.

And that they are all 1st convolved by the mic, mic preamp and ADC. Then on the other end by the system that it is played back thru,

Other questions that arrise:
1/ how good is the foam-core baffle?
2/ the baffle is small enuff and the XO low enuff that the sound field will be coloured by the wrap-aroud the baffle.
3/ how much information is lost converting the files to MP3s?

Fun, but of limited applicability.

dave

 

[planet10]

Break in usually affects the lower frequencies rather than higher.

That raises another question: How valid is the above subjective assertion?

It is usually clues much higher up that i use as clues to evaluate driver breakin.

dave

 

[xrk971]

Foam core baffle is good enough to support weight and vibration of driver with harnonic distortion and frequency response provided in previous thread. But then that is getting quantitative and objective - which is not the point of this thread.

Wrap around sound field is identical for all drivers tested here - so effect at least is uniformly applied.

True info is lost - I had to make a decision on trade off of having a 5 second long lossless FLAC vs a 40 second long compressed sound clip that has no content above 16kHz. I think most of the differences for these drivers is below 16kHz anyway. I can't judge sonething on a 5 second clip.

 

[ErnieM]

There were two that stood out as detailed but fairly neutral to me. The rest were either too dull or had peaks in the high/mids that were irritating.

Thanks for putting this together. It's nice to hear equal comparisons of a variety of drivers. Art (weltersys) did something similar with compression drivers.

 

[planet10]

Wrap around sound field is identical for all drivers tested here - so effect at least is uniformly applied.

But does further limit the applicability of the results.

dave

 

[planet10]

True info is lost - I had to make a decision on trade off of having a 5 second long lossless FLAC vs a 40 second long compressed sound clip that has no content above 16kHz. I think most of the differences for these drivers is below 16kHz anyway. I can't judge sonething on a 5 second clip.

You don't have sufficient storage space for full-rez files? With the availability of storage these days clip size seems a pretty artifical constraint.

dave

 

[BYRTT]

Hi planet10, think you too serious.

As xrk971 write "There are no winners or losers here - just "subjective" differences and preferences".
Perceval write "Even though it is mighty subjective, it is interesting for me".

They recorded under same conditions, will you give it a try or just shoot on flaws, i tell you think i hear microne capsule HD but doesn't complain it's same for all the recordings.