This is great work! I am keeping up with high interest. Having used the designer's previous driver, the Scanspeak Revelator, I am concerned about the 1.3kHz midrange dip. The Revelators are lacking in clarity at the frequency of the dip, compared to Seas Excels, Peerless HDS, or the midrange of a smaller 4" driver. It's not muddy, but it's a bit dark sounding.
But the Revelators had a distortion bump at that 1.3kHz dip, and the Satori does not. So it will be interesting to see how the Satori sounds subjectively compared to the other high-end drivers.
The designer of the Revelator and Satori said that while the dip can be suppressed with stiffer surrounds, it's not caused by the surrounds. That dip (with its corresponding impedance blip) is caused by the cone transitioning from pistonic behavior to decoupled in the center at higher frequency.
But the Revelators had a distortion bump at that 1.3kHz dip, and the Satori does not. So it will be interesting to see how the Satori sounds subjectively compared to the other high-end drivers.
The designer of the Revelator and Satori said that while the dip can be suppressed with stiffer surrounds, it's not caused by the surrounds. That dip (with its corresponding impedance blip) is caused by the cone transitioning from pistonic behavior to decoupled in the center at higher frequency.
I send you a PM!
Do you know if the 4ohm version in the future will be available to the public?
I guess so since you are constructing a kit for them.
/Göran
cotdt, at that frequency everything comes together. The highest diameter is at the cone edge and there the breakup starts too. Metal or ceramic cone drivers have an advantage there. Nevertheless a well designed paper cone has advantages too. One is low mass that leads to better efficiency, the other extended range without horrible breakup.
I got some 4 ohms because that way i may not need a big resistor before the tweeter.
I do not know if they make that unit available over the usual distributors but they can make them.
I got some 4 ohms because that way i may not need a big resistor before the tweeter.
I do not know if they make that unit available over the usual distributors but they can make them.
What I really want to know is if they will be cleaner than my ER18RNX drivers to my ears. The Seas requires a larger box and is not without flaws, but is the best all around woofer I have used. The HDS and Usher need larger boxes than I like. At three times the cost, I hope for more than a slightly smaller box, but this is in the range where big bucks provide small improvements. There is nothing really wrong with the 17NRX for a third of the price! That leaves me torn between stepping up to the big boys game, or a set of the lowly SB's that may be better than I can hear anyway. (I never build the same speaker twice)
Have not picked a tweeter for my next project. Eliminated quite a few.
I have been playing with the SB29RDC. It is quite good and easy to use, but so far not the match for the 27TBFC in my last build. I have only tested it on an active crossover with no compensation, so judgement is not final. I have not seen much on the Satori version you are working with. I have not quite come to terms with the HDS yet, but getting there. XT25 has not impressed me. The D2904 is always viable but a big price jump.
Have not picked a tweeter for my next project. Eliminated quite a few.
I have been playing with the SB29RDC. It is quite good and easy to use, but so far not the match for the 27TBFC in my last build. I have only tested it on an active crossover with no compensation, so judgement is not final. I have not seen much on the Satori version you are working with. I have not quite come to terms with the HDS yet, but getting there. XT25 has not impressed me. The D2904 is always viable but a big price jump.
I have the ER18RNX on hand. It is a special version with copper in the gap though and 4 Ohm.
I can measure it in comparison but i am nether a marketing guy for SEAS nor for SB.
SEAS makes nice tweeters, sure.
They made me a tweeter after my wishes too that i think is better then the 27TBFC.
Again a comparison can be set up.
I know the HDS since many years. Peerless did always sell more woofers then tweeters when i do not count the early efforts from the 80th, so something had to be done.
That was before Tymphany took over.
The XT can be difficult to make sound good. One problem has to do with the poor mechanics. If you make sure that this tweeter in in a good environment, mechanical wise, you do not know what it is capable of.
What do you spend usually on the rest of your equipment ?
I do really not think that the prices of the SB are high in comparison of what the are and what they do. Some customers of mine spend ten times as much on the wiring.
I can measure it in comparison but i am nether a marketing guy for SEAS nor for SB.
SEAS makes nice tweeters, sure.
They made me a tweeter after my wishes too that i think is better then the 27TBFC.
Again a comparison can be set up.
I know the HDS since many years. Peerless did always sell more woofers then tweeters when i do not count the early efforts from the 80th, so something had to be done.
That was before Tymphany took over.
The XT can be difficult to make sound good. One problem has to do with the poor mechanics. If you make sure that this tweeter in in a good environment, mechanical wise, you do not know what it is capable of.
What do you spend usually on the rest of your equipment ?
I do really not think that the prices of the SB are high in comparison of what the are and what they do. Some customers of mine spend ten times as much on the wiring.
I agree on the value. SB seems to have solid design, good construction, quite reasonable price and quite a lot cheaper than the ScanSpeak, Morel, or several other exotic brands.
I guess I don't know what you mean by putting the tweeter in a good mechanical environment. I have not moved past simple flush mounting. My one attempt at de-coupling did not seem to make any difference. I found the XT resonance to be an issue if crossed below 2500 at less that 4th order, and the top to be rather bland. The HDS in my test box seemed a bit harsh, but I know it can do much better. Being quite happy with both the metal and cloth dome mid-price Seas, time to try and get the best of the SB29. It looks quite easy on paper and sounded pretty good on the bench.
I don't have big bucks to spend. As a reference my main system has a pair of Parasound 1200's for amps. A Creek in one room, an old Denon AVR in another. Practical. Mostly bought used. I typically use generic coils and Dayton poly caps. I tried Carity last time but did not hear a difference I could trust. That gives you an idea where I am coming from.
I guess I don't know what you mean by putting the tweeter in a good mechanical environment. I have not moved past simple flush mounting. My one attempt at de-coupling did not seem to make any difference. I found the XT resonance to be an issue if crossed below 2500 at less that 4th order, and the top to be rather bland. The HDS in my test box seemed a bit harsh, but I know it can do much better. Being quite happy with both the metal and cloth dome mid-price Seas, time to try and get the best of the SB29. It looks quite easy on paper and sounded pretty good on the bench.
I don't have big bucks to spend. As a reference my main system has a pair of Parasound 1200's for amps. A Creek in one room, an old Denon AVR in another. Practical. Mostly bought used. I typically use generic coils and Dayton poly caps. I tried Carity last time but did not hear a difference I could trust. That gives you an idea where I am coming from.
Was the drivers relative acoustic center off-set measured at on-axis?
Usually it differs some when you go off-axis. I usually optimize my cross-overs around the 15deg off-axis measurements, where the baffle diffraction is less pronounced and where the acoustic center off-set is more consistent among other things.
Typically I've found out that the acoustic center off-set is about 2-3mm less in the 15-30deg measurements than the on-axis measurement.
Yes, it's not much, but if you strive for perfection and optimal phase tracking between the drivers it could make a difference.
/Göran
I will go to 30mm.
tvrgeek, i will post the XT tweeter i use. It has a much more rigid mechanical construction. Unfortunately that special version is not made any more. I have secured a batch of them.
I have not tried the HDS so i can not say much about it but i understand that you are on a budget and this is fine. I also find that many boutique passive parts are grossly overprices. When i want a really good cap i usually use Epcos MKV. They are not cheep but they are really well damped and have very low loss.
I measured the SEAS reed cone against the satori. Give me some time and then i come with more information.
tvrgeek, i will post the XT tweeter i use. It has a much more rigid mechanical construction. Unfortunately that special version is not made any more. I have secured a batch of them.
I have not tried the HDS so i can not say much about it but i understand that you are on a budget and this is fine. I also find that many boutique passive parts are grossly overprices. When i want a really good cap i usually use Epcos MKV. They are not cheep but they are really well damped and have very low loss.
I measured the SEAS reed cone against the satori. Give me some time and then i come with more information.
For example this ones :https://www.buerklin.com/datenblaetter/D102150_TD.pdf?ch=27274
They are the lower grades, they make some that have even less loss.
They are the lower grades, they make some that have even less loss.
My budget is mostly driven by only paying for something that makes a difference. That's probably why I can afford to play in this hobby anyway. Well, that and selling my TVR and Morgan. I have decided my next build will be the SB17/SB19 set mostly so I can try to come to terms with SoundEasy again. "Easy" it is not. I really want to hear the SB17NRXC35-8 against the Seas. By then I will get to see the fruits of your efforts here and probably go for the Salori woofer. I'll never know if they are worth the difference if I don't try them.
Epcos. I'll look into them. I don't remember seeing them in Mouser of Newark.
Time for another coat of poly on my sub cabinents. Three more to go.
Epcos. I'll look into them. I don't remember seeing them in Mouser of Newark.
Time for another coat of poly on my sub cabinents. Three more to go.
I doubt such a small changes can be easily audible in a real room, but it's certainly measurable and I see no reason to completely ignore it.
Use your usual method to measure the acoustic center of-set at the on-axis measurement as well as the 15, 22.5, 30 and 45deg and I would be surprised if you didn't see a small difference. The shape of the dome and the cone will make the acoustic center to shift somewhat between different measurement angles.
Usually my 15, 22.5 and 30deg measurements are the most consistent.
My personal preference is to optimize the cross-over within a window somewhere between e.g. 10-20deg and don't use toe-in for the loudspeaker.
Sometimes this could give a more irregular on-axis response an some elevated high frequency response, but when going off-axis it gives a smooth power response and great sense of 3D dimensions. However some voicing and fine-tuning is always needed in order to get the proper balance.
As always there are hundreds of factors to consider and compromise between and what measures the best doesn't always sound the best.
/Göran
Interesting I will read it tomorrow.
I usually sit around 2.5m and I guess the room setup will influence the best listening setup. I think it's important to have an open mind and test what's optimal for each loudspeaker + room.
/Göran
As Goran says, doing the measurements to allow you to accurately simulate/design for the phase relationship may sound like something concrete, but in reality it's anything but. Very small changes in microphone height, cone/dome geometry, inaccuracies in your turntables motion for making off axis measurements, will mean that two sets of measurements, done on different days, can easily end up giving you a couple of mm difference in this regard.
This isn't really all that much of a problem though and for a couple of reasons. The first is that any change in listening height, be it up or down by only a couple of cms will have a similar effect in as much as the acoustic centres don't change, but you've effectively done the same thing by changing the relative path differences between the drivers to your ears. This of course is what causes lobing in the vertical axis so it's important you pay attention to this to a certain degree.
Now when you're in the design phase, what you should be doing is paying attention to the phase tracking and observing the reverse nulls to make sure that everything remains reasonably stable even if you adjust the relative offset by a few mms or if you alter the listening height. The lobing patterns around the crossover frequencies allow you see this rather well. If you've got a large/wide primary lobe then the design is going to be rather robust vs listening height, any off axis listening and also importantly component value tolerances/ variations. The latter is something that you shouldn't ignore because often you get 10% variation and depending on how well you've designed your xover these variations could have very little effect, or could have some quite severe (read unacceptable) consequences.
What this means though, is that if you''ve got a good design, all the little 'errors' you could call them, ie from the slight changes in measurement conditions, or in listening habits, tend to become enough of a non issue so that you don't need to obsess over them. Of course, as Goran says, it doesn't hurt to optimise a design for perfection, when your head happens to be optimally aligned at that one given listening distance, but really the important thing is to make sure that it will function well over a range of conditions.
Sorry for the long OT!
This isn't really all that much of a problem though and for a couple of reasons. The first is that any change in listening height, be it up or down by only a couple of cms will have a similar effect in as much as the acoustic centres don't change, but you've effectively done the same thing by changing the relative path differences between the drivers to your ears. This of course is what causes lobing in the vertical axis so it's important you pay attention to this to a certain degree.
Now when you're in the design phase, what you should be doing is paying attention to the phase tracking and observing the reverse nulls to make sure that everything remains reasonably stable even if you adjust the relative offset by a few mms or if you alter the listening height. The lobing patterns around the crossover frequencies allow you see this rather well. If you've got a large/wide primary lobe then the design is going to be rather robust vs listening height, any off axis listening and also importantly component value tolerances/ variations. The latter is something that you shouldn't ignore because often you get 10% variation and depending on how well you've designed your xover these variations could have very little effect, or could have some quite severe (read unacceptable) consequences.
What this means though, is that if you''ve got a good design, all the little 'errors' you could call them, ie from the slight changes in measurement conditions, or in listening habits, tend to become enough of a non issue so that you don't need to obsess over them. Of course, as Goran says, it doesn't hurt to optimise a design for perfection, when your head happens to be optimally aligned at that one given listening distance, but really the important thing is to make sure that it will function well over a range of conditions.
Sorry for the long OT!
Making loudspeakers is a fuzzy art, sure.
Here is another one on imagining :http://www.essex.ac.uk/csee/researc.../C74 Localization experiments in 3D audio.pdf
Here is another one on imagining :http://www.essex.ac.uk/csee/researc.../C74 Localization experiments in 3D audio.pdf
Here is a piece on the audibility of phase distortion.http://www.essex.ac.uk/csee/researc...ubdocs/C23 Audibility of phase distortion.pdf
To my knowledge we made the only experiment where the phase distortion of a L/R4 could be switched in and out without changing the radiation pattern or the frequency response so we could listen to the phase distortion in isolation to other parameters. I call that good science. Only change ONE thing at a time.
To my knowledge we made the only experiment where the phase distortion of a L/R4 could be switched in and out without changing the radiation pattern or the frequency response so we could listen to the phase distortion in isolation to other parameters. I call that good science. Only change ONE thing at a time.
Someone here on DIYaudio did that very same thing but slightly differently. They basically setup a DSP within a PC, fed it a stereo signal, split it through a LW filter, then summed the split signal. This was all handled in the digital realm from loaded sound files so it was all done computationally out of real time. At the end you had the original file, then a number of files each one done with a different filter order, the only differences between the files was the change in phase that would have occurred because of the filters. The net effect was that you couldn't hear it.
Of course this isn't quite the same as when you're making a loudspeaker, but at least it was helpful in illustrating that if you need to go forth order to make things work properly, that you really aren't going to be losing out on anything as far as the phase rotation through the crossover is concerned.
Of course this isn't quite the same as when you're making a loudspeaker, but at least it was helpful in illustrating that if you need to go forth order to make things work properly, that you really aren't going to be losing out on anything as far as the phase rotation through the crossover is concerned.
Did they hear over headphones ?
That way you loose the time difference and only have pressure difference.
What happens when you do not sit in the middle ( sweet spot ) but listen from the side ?
When they listened over speakers was it a phase linear one ( Quad 63, Manger MSW ) or was it a speaker that already had an all pass ?
That way you loose the time difference and only have pressure difference.
What happens when you do not sit in the middle ( sweet spot ) but listen from the side ?
When they listened over speakers was it a phase linear one ( Quad 63, Manger MSW ) or was it a speaker that already had an all pass ?