Jeshi/peterbronsson
I cant consciously ignore large SPL peaks either - something I wasn't really clear in articulating.
BUT, the ear compressed SPL as the peak SPL increases (or we more from a dBa to dBc curve). I.e. As SPL increases (average) the further we move from dBa hearing curves, so I wonder what compression the ear adds when drivers exhibit peaks, and play at higher average SPLs?
I have been thinking a while of making a subjective test like XRKs, but to test the audibility of bell curve reverb decay, with varying total time and 'slapback' periods.
Ive also been trained (college trained) as a sound engineer and musician (but I'm only an amateur)
Anyway this decay time and slapback repeat experiment, well its easy to make some test samples in Cooledit. As a sound engineer would be aware, as the reverb period shortens, the combing changes in number of peaks and Q (if i recall my training well).
I'll have to talk less and do more, the results could be surprising.
I cant consciously ignore large SPL peaks either - something I wasn't really clear in articulating.
BUT, the ear compressed SPL as the peak SPL increases (or we more from a dBa to dBc curve). I.e. As SPL increases (average) the further we move from dBa hearing curves, so I wonder what compression the ear adds when drivers exhibit peaks, and play at higher average SPLs?
I have been thinking a while of making a subjective test like XRKs, but to test the audibility of bell curve reverb decay, with varying total time and 'slapback' periods.
Ive also been trained (college trained) as a sound engineer and musician (but I'm only an amateur)
Anyway this decay time and slapback repeat experiment, well its easy to make some test samples in Cooledit. As a sound engineer would be aware, as the reverb period shortens, the combing changes in number of peaks and Q (if i recall my training well).
I'll have to talk less and do more, the results could be surprising.
Let me remeasure it - at one point I cross calibrated my REW excitation gain to actual speaker terminal voltage. There is some attenuation applied here to get the 10F to match the RS225 due to baffle step so the numbers are different. I guess I can set a 1kHz sine wave and use a DMM to measure the AC voltage corresponding to say 96dB at 0.5m. My setup was pretty consistent with driver ratings as I recalled.
So based on that, if the 10F is spec'd as 86.6dB at 2.83v and 1m. If I measure 96dB at 0.5m, that is same as 90dB at 1m. That is 3.4dB above 2.83v, so I would guess it was about 4.2 volts at the driver to do that.
Btw, doesn't anyone find it amazing that the THD at 1kHz is 0.18% for 94dB at 1m? The 1kHz is pretty representative of overall THD too.
That is a clean sounding driver!
It would be interesting to see a comparison of distortion graphs and CSD between the 10F and the TG9. The FR is almost identical, but the price is 4:1. Do we get more than looks for all that extra money?
Throw in the TC9 as well, I bet it can keep up pretty good and is cheaper.
View it like this, the 10F is a little better in every measurement you put it trough.
In an Olympic match, the winner is the one that's just a little better than the rest. 😀
View it like this, the 10F is a little better in every measurement you put it trough.
In an Olympic match, the winner is the one that's just a little better than the rest. 😀
Last edited:
I gave you guys the CSD already. I can take more of these "stress test" distortion measurements and post. The 100dB one really pushes a driver's buttons. We will see if one of them falls apart. The 10F held itself together very well. More importantly, the test showed me that the Dagger is a very low distortion enclosure for how small it is (ca. 1.1 liter). There is a lot of back wave energy that has to be dissipated so that it doesn't resonate and come back out the front for such a small rear chamber.
I have no doubt the TG9 will pretty much keep up to the 10F for 95% of the way. But as Wesayso sais, it's the last 1% that wins races at the Olympics. When you are talking world class performance, it all matters and the 4x price premium is for the diminishing return. If you like smooth glass like vocals of a fiberglass cone, it is tough to beat the TG9 for the price.
Last edited:
Just bought the PS95-8 (and another subwoofer amp to replace my old farting/buzzing one). I'll cut up a few more 19" x 19" baffles and compare to the mild mannered TC9. What fun! And pretty cheap fun too. Thanks X!
It was a sort of analogy : it distors because the output is to high, but it's a bad analogy from me as it's not about spl (so power) but in my mind : spl peak in aera of the FR where the driver distors at high spl ! So samsame but diferent...
But yes defintly : if the distors, clipping, H3 peak, what ever the name of the thing which drilled my ears on clip 3 with the B80 (driver E), I should had listened to the ref clip to verify if it's present on trhe reccording !
If yes : The B80 has higher dynamic capacity than the E I like better (and which seems not distorse on the same peaks !) or 10F (E driver has a similar capacity but shows less odd harmonic by its design ?!
So in fact, some have right : reference clip can be usefull to check if it has default on it ! For me 10F (E) seems to have a better dynamic but is a little dryer (or give the sensation to be lighter : in fact more accurate (imho).
Finally, it could be also a peak on the recording the B80 (F) especially doesn't like (in an aera of weakness on this driver only).
Hard to say : mp3 is maybe changing some harmonics structure also !
Maybe a further match between the two drivers with some other reccording could be usefull ?!
I listened again SYmphony 1 from Tchaikovsky as a member youtubed it above... on the reccording I have there are some huge dynamic passages (high spl and very soft passage: 10;11;12 with both on the low & high FR - so on a wider band for the driver than the peak of the female voice in clip 3)
.... and also triangle clips 7 & 8 iirc) 🙂 ! 🙄.
FLAC ? shared on a Cloud disk ?
Off topic : is there a good of the shelf paper cone treatment (light resine, etc ?) which just add few Mms but damp better than a raw paper ?
Last edited:
First off, thanks for the time and effort it takes to produce something like this.
While I am concerned with the impulse response of the PS95-8 (it probably contributed to my perception that it was more 'detailed' than the other speakers), I could barely tell the difference between it and the B80.
I am looking forward to seeing how it performs in the enclosure I have just started to design.
Jeff
I had actually included the TC9, but then deleted it since it might be asking too much.
Instead of waterfalls, can we see spectrograms? Supposedly they better correspond to what we hear.
I think spectrograms are literally color contour maps of the CSD. At least that is how they look to me when I compare them. I can see the sharp cliff details more easily in the waterfall CSD representation. But I can plot some rainbow color mapped spectrograms.
That volume would not be enough for a TC9 or a TG9 to be at their best. That's why I liked the Nautaloss enclosure better. Same idea but a longer absorption tail and more volume. I think the FR from that enclosure was almost the same for TG9 and 10F. On this last round there were more differences between them.
So for neutrality of the test, without having to go for a true certified size baffle I again vote for the use of the Nautaloss enclosure for future tests or any similar variant you might come up with. Who knows, they way you spit out these boxes something better may come along. 😀
Thanks again so much for all the time you've put into this and the subjective AND objective info you've generated.
As far as the cabinet goes, the shape is of coarse important, but the way it's damped inside is just as important IMO.
I glue 1/2 inch thick felt (or equiv. (upholstery padding lately)) to all inside surfaces (except the front board where the drivers are mounted), then a layer of soft foam rubber (1/2 inch thick in a small cabinet), then the "acoustastuff" fluffy synthetic fiberglass type stuff from PE.
I evolved this technique from what I learned in the Don Davis Audio Engineering Seminar, where Don demonstrated the difference in various padding techniques. It's all glued in place away from the drivers, so they have good and consistent air impedance around them to breath well. I'm confident that you do a great job on this as well.
Very similar to what I've found to work in my damping experiments for my arrays. Good advise! I'd still use fiberglass insulation though instead of the synthetic stuff. It does a better job at the whole frequency range we use.
If you like spending a fortune, mix in some Twaron Angel Hair. I've got to say that works very well too, especially in mid frequencies. A little goes a long way with that stuff.
If it was too disturbing for your ears, it could probably be the problem with your speakers. Do you have impedance measurement of the speakers or of its midrange driver? Look for a blib around 1k4 Hertz.
Have you compared with your other speakers?
Yes, I had the same impression too. This confused me as the 10F in previous round didn't sound like this. The TG9 had better "impact" such that from the sound character alone [not quality] I thought this was the 10F [Too many trebles also with 10F in this round, compared to TG9, don't know where from].
I think Wesayso has found the source of the problem. I didn't realize they have different enclosures, and what's more, the enclosure is just too small [1.1L] in this round.
This TG9 in this test is the 8 Ohm version I believe. So the volume could be fine. May be TG9 is the one that benefit the most out of that small volume, according to my taste. Only TC9 is too underdamped.
Last edited:
They are not generated the same way. The spectrograph(m), if it's the same as a wavelet, should be the better. A CSD trades time resolution or frequency resolution to benefit resolution in the other domain. The wavelet has constant percentage bandwidth and thus has better time resolution in treble frequencies, better frequency resolution in the bass and so better correlates with how we hear.
Just a thought
Last edited: