When you work with DSP:s you don't have those pesky problems with driver impedances, that you have to work around with passives. Anywhoo, I prefer to not play with DSP:s. Not my modus operandi. Currently at least.
Anyway, here's the response that I ended up with an afternoon's work. I only listened to it in mono, so can't say anything definitive, but it didn't sound half bad. Smoothing is 1/6th octave, and it's measured in my living room, so take it with a pinch of salt. I can't measure outdoors like saurama, too much ambient noise here.
Anyway, here's the response that I ended up with an afternoon's work. I only listened to it in mono, so can't say anything definitive, but it didn't sound half bad. Smoothing is 1/6th octave, and it's measured in my living room, so take it with a pinch of salt. I can't measure outdoors like saurama, too much ambient noise here.
Attachments
That area from 680 to 2.3 kHz needs to be flattened first by about 3 db.
Then listen to it and see if the result is more pleasing (..and do so from different angles).
(..I'm discounting the freq. response below 300 Hz.)
..btw, the measured results of the speaker(s) that saurama produced were pretty close to B&C's - which is nice!
Then listen to it and see if the result is more pleasing (..and do so from different angles).
(..I'm discounting the freq. response below 300 Hz.)
..btw, the measured results of the speaker(s) that saurama produced were pretty close to B&C's - which is nice!
Last edited:
3dB sounds like a lot, but you might be right. Still, I'll first give it a thorough listening like this, but in stereo. And that'll be tomorrow, since I've made a deal with my neighbor that no music after 9pm. We share a common wall, and the bass goes through it like nothing.
Which actually makes me a little bit concerned, since even in these small prototype enclosures the bass is quite strong. These are only 100 litres, 180 litres would be closer to optimal. But they do pack a punch, even with my flea-powered 8W single ended tube amp.
Which actually makes me a little bit concerned, since even in these small prototype enclosures the bass is quite strong. These are only 100 litres, 180 litres would be closer to optimal. But they do pack a punch, even with my flea-powered 8W single ended tube amp.
Hmm, might try an open baffle approach with a wide baffle on the floor. It's remarkably good at removing bass from other rooms - even when "boosted" on the low-end (..which I know wouldn't really be the case here).
Concept:
Check out the "Musical Affairs" design here (..though thinking about it without the upper driver/waveguide):
https://www.youtube.com/user/MoizAudio/videos
Concept:
Check out the "Musical Affairs" design here (..though thinking about it without the upper driver/waveguide):
https://www.youtube.com/user/MoizAudio/videos
I had open baffles before. They have their strengths acoustically. But they're not what I'm after this time. The simplicity of a two way is what I'm after, and IMHO OB's need at least three.
But now I have a pair of crossovers. Waiting for tomorrow to have a good listen. The crossover is still quite simple, second order electrical on the bass/midrange and second order electrical and an L-pad on the tweet. 6 parts in total per speaker currently. I like keeping it as simple as possible, but no simpler.
Edit: Oh, I forgot to mention, I still have OB's on my desk. 10" SLS and a Tangband 4" fullranger. Actually quite nice for computer speakers, driven by a 5W SET tube amplifier...
But now I have a pair of crossovers. Waiting for tomorrow to have a good listen. The crossover is still quite simple, second order electrical on the bass/midrange and second order electrical and an L-pad on the tweet. 6 parts in total per speaker currently. I like keeping it as simple as possible, but no simpler.
Edit: Oh, I forgot to mention, I still have OB's on my desk. 10" SLS and a Tangband 4" fullranger. Actually quite nice for computer speakers, driven by a 5W SET tube amplifier...
Last edited:
You might be surprised..
You might even be surprised at what you can do acoustically as a net result with an open baffle of a particular size and the floor loading. (Think: reduced filter complexity.)
Try modeling a 900mm wide by 800 mm high baffle with the 380mm driver centered on the baffle but placed vertically closer to the top edge (at least 350mm's from the bottom edge of the baffle for the lowest point of the driver's frame) - and don't forget the floor gain you'll get up to the lower mid-band (that usually isn't factored-into the model).
Alfred, your plot is too low resolution in the vertical, very hard to judge any speaker from such a plot.
So, get to a 1 or 2 dB grid.
Then a good way to proceed as a first step is to lower the horizontal resolution (as much smoothing as your program gives you, 1/1 for Arta is good for this purpose). This measurement will give you a good sense of the relative energy going to the different drivers and will allow you to get the alignment roughly right.
Then go to less and less smoothening and solve problems as you encounter them, step by step, increasing the resolution until you are at wits end.
So, get to a 1 or 2 dB grid.
Then a good way to proceed as a first step is to lower the horizontal resolution (as much smoothing as your program gives you, 1/1 for Arta is good for this purpose). This measurement will give you a good sense of the relative energy going to the different drivers and will allow you to get the alignment roughly right.
Then go to less and less smoothening and solve problems as you encounter them, step by step, increasing the resolution until you are at wits end.
OK, now I've had some time to give these prototypes some listening time, and I'm starting to think that they actually might not be a lost cause. Which is good having paid a substantial amount of money for them. Actually I'm quite starting to like them. That peaking at the top of the spectrum does mean that overly bright recordings will make you aware of them, but a suitable toe in might mostly cure that.
Drums sound effortless, but cymbals might be a tad too hot still. Then again, having played drums in past life I'm quite aware that when you hit a metal object with a wooden stick, it makes a very very loud noise. Something that dome tweeters really can't portray right. A suitable anecdote would be like comparing a black and white photograph of a microwave pizza to a steak dinner in a really good restaurant (In Finland we affectionately call those prefabricated pizza-like things 'clutch plates').
Vocals sound already quite balanced, both male and female. I actually could live with this, but since it can be made even better, it should be.
Drums sound effortless, but cymbals might be a tad too hot still. Then again, having played drums in past life I'm quite aware that when you hit a metal object with a wooden stick, it makes a very very loud noise. Something that dome tweeters really can't portray right. A suitable anecdote would be like comparing a black and white photograph of a microwave pizza to a steak dinner in a really good restaurant (In Finland we affectionately call those prefabricated pizza-like things 'clutch plates').
Vocals sound already quite balanced, both male and female. I actually could live with this, but since it can be made even better, it should be.
That's one thing I've been wondering about; do compression drivers have some kind of break in period? I don't remember noticing anything like that with previous drivers I've had (Eminence, Selenium, BMS). I do accept that PA cone drivers with impregnated fabric surrounds might be a little stiffer at first.
A few years ago I came across a blog where one guy actually contacted a number of driver manufacturers about 'break in'.
The ones who responded all agreed that the mechanical part of a driver certainly experiences some bedding in/softening before coming up to spec.
Also all suggested that a few minutes at Xmax will sort that out. Less excursion and it will take a bit longer. ATC responded that that is taken care of during their testing of every driver before shipping. Not everybody does test that rigorously.
I recently used some 5" FaitalPros and they did not change noticeably while some 6.5" Audax I used years ago changed quite a bit but most of that was during the first 15 minutes or so. Almost as if you very slowly turn up the bass.
The ones who responded all agreed that the mechanical part of a driver certainly experiences some bedding in/softening before coming up to spec.
Also all suggested that a few minutes at Xmax will sort that out. Less excursion and it will take a bit longer. ATC responded that that is taken care of during their testing of every driver before shipping. Not everybody does test that rigorously.
I recently used some 5" FaitalPros and they did not change noticeably while some 6.5" Audax I used years ago changed quite a bit but most of that was during the first 15 minutes or so. Almost as if you very slowly turn up the bass.
I felt I heard a big improvement with a pair of Eminence Beta15A's after I blasted them at their resonant frequency for a good while some years ago. Note: felt, not measured.
The whole issue of break in is somewhat problematic anyway. If there's something perceived but not measured, is it really some difference in sound (I don't discount this possibility) or rather your ears getting used to new things. Might be both, but with what ratio?
The whole issue of break in is somewhat problematic anyway. If there's something perceived but not measured, is it really some difference in sound (I don't discount this possibility) or rather your ears getting used to new things. Might be both, but with what ratio?
Some drivers change alot, some only a little. I'm not talking about t&s parametres only.
All materials including metals and plastics in compression drivers "relax" when they are bent, which is what all speaker drivers are doing when they conduct and radiate all sorts of vibration. When materials relax they behave somewhat differently, for example the cone resonance Q would get slightly lower, Fs would drop etc. This takes much more time than a few minutes, and it actually never stops, the changes just become slower and slower after the initial "burn in". The initial perioid could be for example 200 hours of normal listening, which has safe margin to take account all sorts of drivers. It could be less. Vice versa, if driver is left unused for long perioid of time it experiances a "anti burn-in", it's suspesion starts to get stiffer etc, and it "burns in" again when it's used again.
Driver that doesn't change absolutely at all versus time, is fiction imo. The material of the dome, voice coil and voice coil former, suspension, even the surrounding motor and chassis of the driver would have to be infinitely hard and resist all possible deformation under all circumstances. Only that would make a driver that does not change during time... because it cannot.
All materials including metals and plastics in compression drivers "relax" when they are bent, which is what all speaker drivers are doing when they conduct and radiate all sorts of vibration. When materials relax they behave somewhat differently, for example the cone resonance Q would get slightly lower, Fs would drop etc. This takes much more time than a few minutes, and it actually never stops, the changes just become slower and slower after the initial "burn in". The initial perioid could be for example 200 hours of normal listening, which has safe margin to take account all sorts of drivers. It could be less. Vice versa, if driver is left unused for long perioid of time it experiances a "anti burn-in", it's suspesion starts to get stiffer etc, and it "burns in" again when it's used again.
Driver that doesn't change absolutely at all versus time, is fiction imo. The material of the dome, voice coil and voice coil former, suspension, even the surrounding motor and chassis of the driver would have to be infinitely hard and resist all possible deformation under all circumstances. Only that would make a driver that does not change during time... because it cannot.
Last edited:
Our (yours and mine/Adolf's) listening enviroments are very different. Typical livingroom in Finland is about 20 square meters. In addition to this walls are thick (harsh winter, up to -30C even in southern parts), many times made of stone. So the rooms boost low frequencies heavily. That's why we need to think directivity when balancing the loudspeaker. It's basic balance is very different in free-air measurement tower and in normal room.
And of course I didn't mean trying to "eq out" the dip made by floor reflection. I mean compared to free-air, the floor boost bass even if the woofer is 60cm from it.
But I agree, using woofer very close to floor is a good way to eliminate the 100-300Hz dip. But in this case it is no option, Adolf want's a strict 2-way loudspeaker.
Oops. Just noticed something today. Since I got a new laptop a few months ago, my previous measurements lacked the microphone calibration file... So that previous plot can be disregarded.
But some new ones I have, and this time with the calibration file. The green plot is what I ended up last thursday. It still sounded a little "grainy" in the treble, and something I felt was missing.
The red plot is probably the best I could come up with today, but I felt it sounded a little too bright on some records. Then again with Infected Mushroom's BP Empire it was even maybe a tad too polite.
The blue one is basically the same crossover as the red, but with a couple dB more padding for the HF driver. It doesn't sound as bad as it looks. Not yet sure, and listening for today is over, since it's over 9pm.
But some new ones I have, and this time with the calibration file. The green plot is what I ended up last thursday. It still sounded a little "grainy" in the treble, and something I felt was missing.
The red plot is probably the best I could come up with today, but I felt it sounded a little too bright on some records. Then again with Infected Mushroom's BP Empire it was even maybe a tad too polite.
The blue one is basically the same crossover as the red, but with a couple dB more padding for the HF driver. It doesn't sound as bad as it looks. Not yet sure, and listening for today is over, since it's over 9pm.
Attachments
Room construction differences aren't going to make that large of a difference above 100 Hz. Room size can, but that's also true of any room with solid walls (..which would include sheet rock).
Basically the range were are most interested is above 100 Hz. The mid-bass driver is essentially omni-directional to 500 Hz. That the two drivers in combination are more directional above 500 Hz would not benefit more from a reduced bass response relative to freq.s greater than 500 Hz.
Thinking "directivity" as you indicate below, will result in an erroneous result. (..it is counter-intuitive.
)
Last edited:
The Red looks good, but you need to drop the level between 900 Hz and 2.8 kHz down to the level at 900 Hz. The upper freq. response beyond that should be pretty much the way it is (..even if the peak at 4.2 kHz is a bit troubling). Frankly though, I'd be looking to drop the range on a broader scale down to 450 Hz as well (..the upper midrange, at least as represented in the freq. plot, is still to "prominent").
Last edited:
As Adolf's measurements are made in a room at listening setup of obviously both the speaker and the mic, I must agree with Scott. Tonality has too much middle weight, it must sound too clean/bright.
Lowest bass will get some help from the larger box, but these PA drivers could work better at home as sealed because room modes give boost to 50Hz... Acoustic piano and orchestra have lots of "sound/tonality material" below 40Hz! Perhaps not so relevant question for rock?
A stand-mounted speaker always has the floor bounce problem around 2-300Hz, I believe that this is a major difference between them and well-designed 3/4-ways.
Lowest bass will get some help from the larger box, but these PA drivers could work better at home as sealed because room modes give boost to 50Hz... Acoustic piano and orchestra have lots of "sound/tonality material" below 40Hz! Perhaps not so relevant question for rock?
A stand-mounted speaker always has the floor bounce problem around 2-300Hz, I believe that this is a major difference between them and well-designed 3/4-ways.