Thanks for the kind words CLS! This type of comparison is not something I have seen before because it uses the same rear chamber, baffle, bass unit, amp, miniDSP, etc for all cases with the only thing different is the driver. It's hard to find an implementation that is so flexible and gives a consistent bass result so that we can concentrate on the full range driver and not the box/driver combo.
More drivers to be compared later. The high end version test.
More drivers to be compared later. The high end version test.
FYI, the MP3 algorithm at the quality setting used falls off and then brick-walls at about 17kHz. I'm sure spectral analysis will show the specifics.
Something to keep in mind as I wouldn't conclude on something that wasn't even in the test.
First off, thanks X for the perseverance making all these observations of speaker behavior in action.
The concept of recording for comparison is well regarded by David Griesinger. An excerpt from his paper found at www.davidgriesinger.com/ica2010/paper2.doc:
And so indeed:
I downloaded all the clips. With clip1 A and then B told me this was going to be fun. Onto C, then a jump to F; already a preference for B and C began forming. Not ready to rock I jumped to clip3 with G and didn't stay long before retreating to B, C, and then A was out; E didn't last long either. D was also a short visit.
Next set started with some clip1 and clip3 confirming candidates B and C. Another dash of clip1 and a notion of boredom set in. Turn it up a touch and time to rock by the thimble full. Some alphabet soup with a man all but yelling as only some men can, backed by the harmonic stew of electric guitar and drum kit. Broadband. Soaked up a bit more D and some rest with clip3 before coming to another stop to ponder an ear full of information.
The drivers clearly have frequency response differences and off axis response differences that when repeatedly listened to build a sense of X's recording space.
A late night listen with the yelling man turned down a little and the piano man turned up a little I concentrated on B and C; suspecting that B was the Vifa and wondering about C. With B it seems to me that parts of the piano and male vocal had a more focused slap back from the room; piano runs seemed to step on themselves and male voice lost some articulation relative to C.
For me C sounded more like the piano was in X's room. The yelling man sounded like he was chasing dust out of more corners. So yes, I voted for C.
Please post the REW mdat files with the IR data for the drivers.
C turns out to be a Dayton Audio driver; and their data suggests for me desirable off axis behavior; more so relative to Vifa data.
Both drivers are of interest to me, both as design elements and as common vehicles for exchange of ideas.
Andrew
The concept of recording for comparison is well regarded by David Griesinger. An excerpt from his paper found at www.davidgriesinger.com/ica2010/paper2.doc:
And so indeed:
I downloaded all the clips. With clip1 A and then B told me this was going to be fun. Onto C, then a jump to F; already a preference for B and C began forming. Not ready to rock I jumped to clip3 with G and didn't stay long before retreating to B, C, and then A was out; E didn't last long either. D was also a short visit.
Next set started with some clip1 and clip3 confirming candidates B and C. Another dash of clip1 and a notion of boredom set in. Turn it up a touch and time to rock by the thimble full. Some alphabet soup with a man all but yelling as only some men can, backed by the harmonic stew of electric guitar and drum kit. Broadband. Soaked up a bit more D and some rest with clip3 before coming to another stop to ponder an ear full of information.
The drivers clearly have frequency response differences and off axis response differences that when repeatedly listened to build a sense of X's recording space.
A late night listen with the yelling man turned down a little and the piano man turned up a little I concentrated on B and C; suspecting that B was the Vifa and wondering about C. With B it seems to me that parts of the piano and male vocal had a more focused slap back from the room; piano runs seemed to step on themselves and male voice lost some articulation relative to C.
For me C sounded more like the piano was in X's room. The yelling man sounded like he was chasing dust out of more corners. So yes, I voted for C.
Please post the REW mdat files with the IR data for the drivers.
C turns out to be a Dayton Audio driver; and their data suggests for me desirable off axis behavior; more so relative to Vifa data.
Both drivers are of interest to me, both as design elements and as common vehicles for exchange of ideas.
Andrew
TBH, except for the brightest (harshest) one and the 1 or 2 with obvious rolled-off top end, all the rest (3 or 4) sound very similar (almost identical) to me on my playback equipment - lap top and earbuds. I can't do the comparison into the details even if I blindly trust my lousy playback system. I need much quieter environment and hi-res equipment to evaluate HF, rather than in the noisy office at lunch break as I did it.
So, what I said in the previous post about casual voters etc, is actually pointing myself, oh well. And, as the result, you guys' hearings are better than mine, even though the chain from recording to playback is full of unknown variables.
Give yourselves some applause 😀
So, what I said in the previous post about casual voters etc, is actually pointing myself, oh well. And, as the result, you guys' hearings are better than mine, even though the chain from recording to playback is full of unknown variables.
Give yourselves some applause 😀
I voted for C, with drive D coming as close second. I would probably buy either of them for my next projects, but first need to figure what to do with drivers I already got in stock 😀
Same vote here.
But if I'm going to buy any of them, I'll believe X and his measurements instead of my hearing (through all these recording and playback).
But if I'm going to buy any of them, I'll believe X and his measurements instead of my hearing (through all these recording and playback).
I believe in my own hearing as it's quite subjective on what you find nice and what not. One of the best sound I have ever heard came from a noname 4" fullrange / woofer disassembled from a 80s TV set. And I've heard some really impressive truly flat response speakers after it, including a pair of massive Genelec 1036A with subwoofers to back it up.
I was not questioning your hearing at all, sorry if I didn't make myself clear...
Things I can't trust are the variables in my pc playback stuff.
Actually, I can't trust my own hearing, either. Because it's not consistent. I hear differently all the time, varied by my (physical) conditions, mood, time in a day, music, person(s) nearby if any... etc. I need very long time to sum up a complete impression to the sound of certain hardware.
Things I can't trust are the variables in my pc playback stuff.
Actually, I can't trust my own hearing, either. Because it's not consistent. I hear differently all the time, varied by my (physical) conditions, mood, time in a day, music, person(s) nearby if any... etc. I need very long time to sum up a complete impression to the sound of certain hardware.
Ah, it seems we had a mutual misunderstanding 😀
I agree on the variables you posted as well. I listened to the clips with Koss Porta Pro earphones which are not very hifi, and have likely colored the sound one way or another.
I agree on the variables you posted as well. I listened to the clips with Koss Porta Pro earphones which are not very hifi, and have likely colored the sound one way or another.
As I have mentioned before, the little $23 Dayton PS95-8 is a gem waiting to be discovered. Cast frame, inverted surround, phase plug. Sounds very nice and with a little EQ can tame the HF peak to remove ringing for clean IR.
How does an eardrum probe mic work?
How does an eardrum probe mic work?
Last edited:
.FRD files of all measurements
Barleywater:
Thanks for the detailed comments and your process of decision.
Here are the .FRD files to start with...
Barleywater:
Thanks for the detailed comments and your process of decision.
Here are the .FRD files to start with...
After hearing the reference tracks I would have to say it would come down to between The vifa and the dayton PS95-8 in this particular test for me. It was the only one of the three dayton drivers that I liked though. I rated the RS100P-4 as second worst and the RS100-4 as worst. Both are flatter than the PS95-8, but both sounded mushy and a as though they had a lot more echo (than any of the other drivers) to me. Strangely the second favourite (the faital pro) was my third worst pic, again it seemed to have a little too much echo.
Perhaps even stranger is that the RS100P-4 only got 6 votes compared to the vifa's 23. You could argue that the dayton has the flatter response (albeit with no HF after aprox 12Khz).
Both the vifa and the PS95 have peaking around 13-15Khz, I think my hearing is starting to go by about 13-14Khz, so that probably doesn't factor in my listening tests 😉
Attached is a graph comparing the traced factory curves of the dayton PS95 (black) Dayton RS100P-4 dark blue, and the Vifa TC9-FD yellow/green (I'm colourblind).
I can't see anything in the vifa vs dayton RS100P-4 (apart from the high frequency rolloff) that would make the vifa much more preferred based on the on axis frequency response alone.
I find it most interesting that I vastly prefer the PS95-8 over the RS100P-4 (like chalk and cheese) since the RS100P has a much smoother response. I did prefer the vifa over the PS95-8 though.
Tony.
Perhaps even stranger is that the RS100P-4 only got 6 votes compared to the vifa's 23. You could argue that the dayton has the flatter response (albeit with no HF after aprox 12Khz).
Both the vifa and the PS95 have peaking around 13-15Khz, I think my hearing is starting to go by about 13-14Khz, so that probably doesn't factor in my listening tests 😉
Attached is a graph comparing the traced factory curves of the dayton PS95 (black) Dayton RS100P-4 dark blue, and the Vifa TC9-FD yellow/green (I'm colourblind).
I can't see anything in the vifa vs dayton RS100P-4 (apart from the high frequency rolloff) that would make the vifa much more preferred based on the on axis frequency response alone.
I find it most interesting that I vastly prefer the PS95-8 over the RS100P-4 (like chalk and cheese) since the RS100P has a much smoother response. I did prefer the vifa over the PS95-8 though.
Tony.
Attachments
Last edited:
Two basic flavors of eardrum probe mic: tiny capsule at end of tube, and more often tube attached to larger capsule mounted behind ear.
Referencing is done with free field speaker and playback headphones. Sound recorded is then encoded with personal HRTF. Playback is as 3D as it gets for 2 channel recording.
I've not used in ear microphones, but have experimented by attaching probe tube to ECM8000 microphone via modeling clay. I used referencing by substitution for equalization. Experimentation was done placing probe tip at places normal microphone body prevents placement. Tubing I used was fairly soft wire insulation. Not something I'd bump my eardrums with, but did do some IR measurements with partial insertion into ear canal, and measurements with tip real close to air gap and voice coil former.
I have touched my eardrums with blunt soft probe; slightest contact produces rubbing sound. Tiniest bit more pressure produces intense pain. Not recommended.
pnix appears to have mdat of impulse responses for this data set and I would like an opportunity to work with it to expand upon the work here. Please post.
Thanks again for very stimulating thread.
Andrew
Referencing is done with free field speaker and playback headphones. Sound recorded is then encoded with personal HRTF. Playback is as 3D as it gets for 2 channel recording.
I've not used in ear microphones, but have experimented by attaching probe tube to ECM8000 microphone via modeling clay. I used referencing by substitution for equalization. Experimentation was done placing probe tip at places normal microphone body prevents placement. Tubing I used was fairly soft wire insulation. Not something I'd bump my eardrums with, but did do some IR measurements with partial insertion into ear canal, and measurements with tip real close to air gap and voice coil former.
I have touched my eardrums with blunt soft probe; slightest contact produces rubbing sound. Tiniest bit more pressure produces intense pain. Not recommended.
pnix appears to have mdat of impulse responses for this data set and I would like an opportunity to work with it to expand upon the work here. Please post.
Thanks again for very stimulating thread.
Andrew
I can think of a few reasons Tony, partly a matter of driver design & partly in the response; given that you preferred the TC9 over the 95, but the 95 over the 100P the latter, assuming on-axis, can likely be discounted as the primary factor. However, I rather doubt any such remarks / points will be welcomed on this thread. FWIW, if you're interested, give me a shout via PM.
Barleywater,
Thanks explanation eardrum probe mic sounds interesting, perhaps due to language or I'm not sharp enough i am not quite sure setup, if photos available will probably help 🙂.
Scottmoose,
Sound like interesting knowledge will like to hear, you hint it can be unpopular then suggest write it as open minded you able to.
Thanks explanation eardrum probe mic sounds interesting, perhaps due to language or I'm not sharp enough i am not quite sure setup, if photos available will probably help 🙂.
Scottmoose,
Sound like interesting knowledge will like to hear, you hint it can be unpopular then suggest write it as open minded you able to.
There is no reason why this discussion would not be welcome here. Sounds very interesting to me.
- Status
- Not open for further replies.