Instead of saying that I read that the "tweeter" in this LS has narrow vertical dispersion, I should have said "treble".
Apparently, this type of two-way MTM configuration requires careful crossover design to avoid "holes" in the response above and below the axis.
'Faster' transient response translates to wider bandwidth. But even the ears of sound and young people are bandwidth limited to 20kHz. So any tweeter that can do 20kHz within reasonable level tolerances is 'faster' than our ears. As pointed out earlier however, your wish for 'faster' most likely originates from you lacking a bit sound pressure level in high frequencies, either by hearing loss, by the loudspeaker or by acoustics in your living room (that is why I suggested moving closer to your speakers btw). Raising the sound pressure level of high frequencies by altering the crossover or by eq-ing the source material is a viable alternative.
A resistor placed before the XO would appear not to tilt down the treble response, whereas one placed after the XO does.
That is why the suggestion of putting a 0.47 uF capacitor in parallel with a resistor which is placed after the XO has been made - this would lift the upper treble response.
Wrong. The current in any case passes through that resistor, and the power loss is always the same, proportional to the square of the current and the resistance.
Adding a capacitor parallel to this resistor raises the frequency range according to the decreasing reactive impedance of that capacitor at higher frequencies that bridges the resistor.
Adding a capacitor parallel to this resistor raises the frequency range according to the decreasing reactive impedance of that capacitor at higher frequencies that bridges the resistor.
Last edited:
There is only one resistor. So it is possible to bridge it with a resistor and raise the whole part that does the tweeter, or with a capacitor where only a part of the highest frequencies will be raised depending on the size of the capacitor.
If it is not possible to measure the frequency range, only the rehearsal and listening method remains until it is set properly.
If it is not possible to measure the frequency range, only the rehearsal and listening method remains until it is set properly.
I think there are other factors to take into consideration, such as frequency response phase, as well as the impedance seen by the amplifier.
Re-stating things, you are after a top end with more "sparkle", that is more "realistic", "faster" and more "agile" and with better "transients".
Three solutions come to mind:
1. The issue could be a matter of quantity, so not enough SPL in the "sparkle" frequencies somewhere probably above about 6 or 7kHz, either on-axis or off or both. Since you know how to use REW, try actually taking some measurements of the speakers to identify if this is a source of your problem. Reducing the tweeter resistance or adding a super tweeter as have been mentioned are potential fixes. Possibly replacing caps might show some improvement too and even tweaking the xo could be an option as well.
2. The issue could be one of tweeter quality. Technically, it is correct to say that there are no "faster" drivers because if one cone is moving faster than another then they won't be playing the same note any longer. Right? Maybe this is what giralfino meant earlier today (and I do believe without any insult intended) but I will disagree with him and say that in my experience there are definite differences in tweeter quality, specifically in the transient response and in the sense of a cleaner response. So in the former case, not that one cone is moving faster than the other but that one is doing a better job of starting and stopping sooner while moving at the same speed and playing the same notes and in the latter case, that one has lower harmonic distortion than the other.
So it helps that we can measure these things too.
The first can be examined by a Cumulative Spectral Decay measurement (aka a waterfall plot) and the second by the Harmonic Distortion measurement, both available with REW btw.
So here is the CSD of the D2604:
And then of the D2904:
So those are actually very similar responses with maybe the more expensive tweeter having very, very slightly faster decay times above about 3kHz but not really by much. (I'm kind of assuming that you know how to read these or that you are smart enough to figure them out, but if anything isn't clear, just ask.)
And then here is the Harmonic Distortion for the 2604:
And for the 2904:
It's the 3rd harmonic in blue that is usually considered the worst offender and again, both drivers are remarkably similar - both 3rd harmonics are down around the -80dB mark above about 3Khz. Below 3Khz, the 2904 is doing much better but that's not really where your problem lies is it? But the 2904 does show a better response in terms of 4th and 5th harmonics so there is that. But just in general, the quality differences on these 2 major metrics are actually very small, which is just reinforcing what others have said in that the 2604 is considered to be a very high value little tweeter.
3. And lastly, your issue might obviously be a combination of both quantity and quality.
So if you were looking for a new tweeter, I might be looking at some of the top end SB/Satori tweeters if they fit or can be modded to fit because of the high quality and because many of them have a rise in the HF response up above about 8-9kHz which would seem to be what you are after. Some kind of xo change or re-design would most likely be required though. Maybe you can get that done successfully here in this thread or here is someone who does it professionally if you're willing to ship your speakers although he's not exactly in your neck of the woods (Texas, I think): https://gr-research.com/
Three solutions come to mind:
1. The issue could be a matter of quantity, so not enough SPL in the "sparkle" frequencies somewhere probably above about 6 or 7kHz, either on-axis or off or both. Since you know how to use REW, try actually taking some measurements of the speakers to identify if this is a source of your problem. Reducing the tweeter resistance or adding a super tweeter as have been mentioned are potential fixes. Possibly replacing caps might show some improvement too and even tweaking the xo could be an option as well.
2. The issue could be one of tweeter quality. Technically, it is correct to say that there are no "faster" drivers because if one cone is moving faster than another then they won't be playing the same note any longer. Right? Maybe this is what giralfino meant earlier today (and I do believe without any insult intended) but I will disagree with him and say that in my experience there are definite differences in tweeter quality, specifically in the transient response and in the sense of a cleaner response. So in the former case, not that one cone is moving faster than the other but that one is doing a better job of starting and stopping sooner while moving at the same speed and playing the same notes and in the latter case, that one has lower harmonic distortion than the other.
So it helps that we can measure these things too.
The first can be examined by a Cumulative Spectral Decay measurement (aka a waterfall plot) and the second by the Harmonic Distortion measurement, both available with REW btw.
So here is the CSD of the D2604:
And then of the D2904:
So those are actually very similar responses with maybe the more expensive tweeter having very, very slightly faster decay times above about 3kHz but not really by much. (I'm kind of assuming that you know how to read these or that you are smart enough to figure them out, but if anything isn't clear, just ask.)
And then here is the Harmonic Distortion for the 2604:
And for the 2904:
It's the 3rd harmonic in blue that is usually considered the worst offender and again, both drivers are remarkably similar - both 3rd harmonics are down around the -80dB mark above about 3Khz. Below 3Khz, the 2904 is doing much better but that's not really where your problem lies is it? But the 2904 does show a better response in terms of 4th and 5th harmonics so there is that. But just in general, the quality differences on these 2 major metrics are actually very small, which is just reinforcing what others have said in that the 2604 is considered to be a very high value little tweeter.
3. And lastly, your issue might obviously be a combination of both quantity and quality.
So if you were looking for a new tweeter, I might be looking at some of the top end SB/Satori tweeters if they fit or can be modded to fit because of the high quality and because many of them have a rise in the HF response up above about 8-9kHz which would seem to be what you are after. Some kind of xo change or re-design would most likely be required though. Maybe you can get that done successfully here in this thread or here is someone who does it professionally if you're willing to ship your speakers although he's not exactly in your neck of the woods (Texas, I think): https://gr-research.com/
thanks for the plots.
yeah I can read them and have REW.
im assuming you got those from the scanspeak website under 3D-models?
i can’t open those.
can you tell me how to open them?
also yeah I’m intrigued that the 2904 doesn’t have a faster start stop time.
are metal dome tweeters better in that area?
thanks again
That is what I thougth at reading the thread. But having had a Proac D15 I find power response of Proacs very flat already... maybe this particular tweeter.(6k to 10 kHe a little notched on this unit or loudspeaker ?)
Also what are the brands of the caps on the op's Proac? If Proac still use Solen there is perhaps something to do. I would measure each cap carefully and mark them for the rigth channel for comming back. With the uF value I would try a Jantzen Sup or Silver. I do not like changing crossover but perhaps a Panasonic Radial mkp FE serie will do the trick too with keeping the genuine caps. Try adding 0,1 uF ? Still third order filter use at Proac ?
Worths the try first as cheap. A first price hp cable, the cheap Cardas blue one sold on turet, may add some ligth and energy in the top end as well but expensive for a test. Hard to say without listening your systeme.
Last edited:
No, graphs are from http://www.zaphaudio.com/, Tweeter Mishmash. It always helps when the tests are done by the same individual or under the same conditions and with the same equipment in other words. It gets a little trickier when comparing across sources, but if it helps, here are some other 3rd party measurement sources:
https://hificompass.com/index.php
http://www.dibirama.altervista.org/
http://www.audioexcite.com/
https://www.erinsaudiocorner.com/driveunits/
http://www.justdiyit.com/category/tests/
So what is a faster start stop time?
Regarding your ProAcs, the enclosure isn't braced at all. Baffle and other panels will radiate some unwanted frequencies. Also the damping of the cavity, well, it could be that foam is reasonable, but it is applied quite sparsely and won't absorb enough interior sound. Standing waves in the interior will contribute to coloration. I can think of a few tweaks that really make a better speaker here.
Check. Bitumen mass plate doesn't do very much, be it it's better than nothing, especially at higher modes of the panels.
You can read CSD, so:
The bare stuff. Think of this: your midwoofers are mounted on a bare panel which will resonate like @#%%!.
MDF + 'bitumen', 7mm (!) thick. Most of it is 4mm and far less effective. But the higher order panel resonance is under control.
11mm (!!) mass plate actually, but let's call it 'bitumen' plus braces on 18mm MDF. It doesn't get much better than this.
Still lots of room to improve.
Pics courtesy of Martin van der Hoff
You can read CSD, so:
The bare stuff. Think of this: your midwoofers are mounted on a bare panel which will resonate like @#%%!.
MDF + 'bitumen', 7mm (!) thick. Most of it is 4mm and far less effective. But the higher order panel resonance is under control.
11mm (!!) mass plate actually, but let's call it 'bitumen' plus braces on 18mm MDF. It doesn't get much better than this.
Still lots of room to improve.
Pics courtesy of Martin van der Hoff
Attachments
thanks for the info. Appreciated.
here is a pic which shows the thickness of the bitumen. it is at least 7mm.
yeah it’s not the best dampening and cabinet construction, but a lot of British speaker makers don’t like internal bracing.