This was something that you were provided with, given the files to listen for yourself. The point was it doesn't matter what you listen to the files on, headphones, loudspeakers, whatever, but headphones make the most sense as they generally have the highest resolving power.
One sound file has the sound as it appears on the CD and then the others have had the sound split and then recombined with the associated filters phase shift applied to it. Obviously the summed amplitude is perfectly flat, only the phase altering effects of the filters was added.
This wasn't really a test of which one is better, just a test of can you tell the difference, in other words is the phase shift really audible. I couldn't tell any difference, I did the test over headphones. Maybe I would have detected small changes spatially if I'd done this via loudspeakers, but what I took from this was that the differences between say a 2nd order LW or a 4th order were negligible at best and that it's foolish to opt for a shallower filter just for the sake of using a shallow filter.
Some designers appear to choose using a shallow slope at the expense of all other design parameters, due to the simple belief that the shallower xover will sound better because it introduces less phase shift. That test helped to show that that way of thinking is tenuous at best and that it is always better to use a steeper filter if it brings about improvements in other areas.
It is nice though when you get a pair of drivers that work well enough together so that you have the option of using either 2nd to 4th order slopes. Usually this requires drivers such as the ones you are using here though so you're probably going to have a lot of fun trying out different crossovers
One sound file has the sound as it appears on the CD and then the others have had the sound split and then recombined with the associated filters phase shift applied to it. Obviously the summed amplitude is perfectly flat, only the phase altering effects of the filters was added.
This wasn't really a test of which one is better, just a test of can you tell the difference, in other words is the phase shift really audible. I couldn't tell any difference, I did the test over headphones. Maybe I would have detected small changes spatially if I'd done this via loudspeakers, but what I took from this was that the differences between say a 2nd order LW or a 4th order were negligible at best and that it's foolish to opt for a shallower filter just for the sake of using a shallow filter.
Some designers appear to choose using a shallow slope at the expense of all other design parameters, due to the simple belief that the shallower xover will sound better because it introduces less phase shift. That test helped to show that that way of thinking is tenuous at best and that it is always better to use a steeper filter if it brings about improvements in other areas.
It is nice though when you get a pair of drivers that work well enough together so that you have the option of using either 2nd to 4th order slopes. Usually this requires drivers such as the ones you are using here though so you're probably going to have a lot of fun trying out different crossovers
As always a very nice and detailed sum-up from 5th element!
As Joachim says, designing loudspeakers is like making art.
There so many details to consider when making loudspeakers. Some are small and some are crucial for a successful design. I'ts easy to be obsessed by details that in the big picture doesn't count for much, but puzzling each and every bit together is what makes it an art in my opinion.
Of course, each and every constructor builds in it's own sound preferences/philosophy into the design. It's hard not to make a subjective sounding loudspeakers. Perhaps professionals in the industry is better in doing more objective designs that fits a broader public, I don't know?
Often when I listen to loudspeakers at Hifi shops and shows I end up with liking some parts of the design, but seldom the whole design. It inspires me though in my own search for the "holy grail" of sound and makes me jealously over the loudspeaker cabinet design and finish.
Sorry for my babbling!
Regards
/Göran
I agree, when we switched between an L/R4 with and without phase correction the difference was small. I could hear differences with artificial signals like a needle ( Dirak) impulse. The phase corrected version sounded "tick, tick, tick" whereas the non corrected sounded "tock, tock, tock". On music the difference was much harder to tell.
Flat amplitude response and good radiation pattern are the most important goals.
That is also the reason i will design a L/R4, just for comparison. It has also the advantage that out of band artifacts like distortion is well suppressed and the drivers overlap only over narrow range. I found another problem about
that i talked shortly before. A passive L/R4 can be quiet complex and passive components have losses.
Flat amplitude response and good radiation pattern are the most important goals.
That is also the reason i will design a L/R4, just for comparison. It has also the advantage that out of band artifacts like distortion is well suppressed and the drivers overlap only over narrow range. I found another problem about
that i talked shortly before. A passive L/R4 can be quiet complex and passive components have losses.
Up to 3K they are very similar, I guess that the baffle is making most of the contribution to the irregularities in the response curve, though the dips in the 1-2K region are definitely more pronounced in the Satori, so I assume that the driver is also contributing in this area. It's interesting as I have a similar 10db Suck out in my Morels but at around 2.1Khz That I've never been able to work out the cause of.
Could you do a < 1cm nearfiled measurement of the Satori? I'd be very interested to see that
Tony.
The Seas is less extended because the membrane is stiffer and the motor has much more inductance. The small wrinkles come from the baffle but the dip over 1khz is real in both drivers. I have made a near field response of the Satorin in 10cm. If that is enough i can post it again.
What it also obvious is that the Satori is much lower in distortion. Second harmonic is half and third really low. This is an exceptional extended and low distortion driver.
What it also obvious is that the Satori is much lower in distortion. Second harmonic is half and third really low. This is an exceptional extended and low distortion driver.
That's better from the Satori, crank up the drive level and the third harmonic drops as the SnR increases. Now we get to see more of what you're paying for.
I agree that a Satori nearfield would be interesting, but mainly if done right near the surround. This would let us see when the driver actually stops being pistonic, which would be interesting considering its extremely wide on axis bandwidth.
I agree that a Satori nearfield would be interesting, but mainly if done right near the surround. This would let us see when the driver actually stops being pistonic, which would be interesting considering its extremely wide on axis bandwidth.
Member
Joined 2009
Indeed the Satori is a notch above the SEAS. As long as the 2nd harmonic is around 55-60dB below the fundamental, I tend to ignore it. What I do pay attention to though is the level of 3rd harmonic, I am no driver designer so I don't know exactly how 3rd harmonic is generated in the motor, but it's quite clear that there tend to be two types of driver out there.
The first is like the ER18, it has low third harmonic at lower frequencies, but this then rises slowly as frequency increases. All of SEAS drivers tend to show this but the better excel drivers do manage to keep the level lower throughout the upper region of their range. The second set of drivers are like the Scan revelators/illuminators, the Vifa NE range and the Satori, where the third harmonic starts off low and remains low, often decreasing in level as frequency increases.
I am not sure what causes this difference, I do know that SEAS use T shaped pole pieces with copper rings, the other drivers most certainly use copper in the gap, but might use a different shape of pole to linearise the field through the gap and maybe it's this.
Clearly the latter is better for midrange performance, whereas the former, tend to give slightly better performance before the rise occurs. Of course you do need to compare apples with apples and although the ER18 is a nice driver it isn't quite up to the standard of the excel line.
The first is like the ER18, it has low third harmonic at lower frequencies, but this then rises slowly as frequency increases. All of SEAS drivers tend to show this but the better excel drivers do manage to keep the level lower throughout the upper region of their range. The second set of drivers are like the Scan revelators/illuminators, the Vifa NE range and the Satori, where the third harmonic starts off low and remains low, often decreasing in level as frequency increases.
I am not sure what causes this difference, I do know that SEAS use T shaped pole pieces with copper rings, the other drivers most certainly use copper in the gap, but might use a different shape of pole to linearise the field through the gap and maybe it's this.
Clearly the latter is better for midrange performance, whereas the former, tend to give slightly better performance before the rise occurs. Of course you do need to compare apples with apples and although the ER18 is a nice driver it isn't quite up to the standard of the excel line.
The Excel line is a bit harder to use well though. Low crossover and still need to notch the breakup. Great candidate for active crossovers. I think I would try the Satori first.
Need to check, but I don't think I had much of a dip at 1K. Could be my baffle being a different size and radius. I do have something not quite perfect, as Joni Mitchel is a tiny bit off. I finished redoing my sub boxes so I will go back and measure in place.
It is clear from the measurements, the Saroti is a driver to be taken seriously. The Seas is a very good driver. Try it for a budget built, you may be impressed as I am. Every step in price does not always give a step in performance. (I could name a few that fail.) This one seems to.
You don't always get what you pay for, but you sure well won't get it if you don't.
I was measuring the SB29 today. Strange big offset in the resonance peak, and even after two days break-in, the Fs is still closer to 700; spec at 600. Not impressed yet. The published impedance curve does not show this. The back? I only have one, so nothing to compare it to.
Need to check, but I don't think I had much of a dip at 1K. Could be my baffle being a different size and radius. I do have something not quite perfect, as Joni Mitchel is a tiny bit off. I finished redoing my sub boxes so I will go back and measure in place.
It is clear from the measurements, the Saroti is a driver to be taken seriously. The Seas is a very good driver. Try it for a budget built, you may be impressed as I am. Every step in price does not always give a step in performance. (I could name a few that fail.) This one seems to.
You don't always get what you pay for, but you sure well won't get it if you don't.
I was measuring the SB29 today. Strange big offset in the resonance peak, and even after two days break-in, the Fs is still closer to 700; spec at 600. Not impressed yet. The published impedance curve does not show this. The back? I only have one, so nothing to compare it to.
Second comes from unsymmetrical distortion like the suspension that can be stiffer in one direction then the other and the "zero" is not where the driver is at rest. Because it is mainly from mechanical origin is rises more or less smooth with level. Third is symmetric distortion and i think the Satory has lower third because the impedance rise is extremely low because of a well designed motor with lots of copper. Scan Speak is a pioneer here so the drivers they make have also low third. The Exell from SEAS have also more copper in the gap and the T-Pole. Some of Scan Speak drivers are underhang so the variation of impedance over level is low at the price of less sensitivity unless a really huge magnet is used. The material of the magnet playes a role too because it effects the amount of distortion in the current that is fed back into the amplifier.
Here ferrite is not that good and Alnico and field coil best. You do not see that type of distortion in the voltage so it does not show in acoustic harmonic distortion measurements. You can learn more at Home.
I belong to the Klippel advanced users group and actually i bought the first Klippel system made straight from the table of Prof.Klippel more then 10 years ago.
Here ferrite is not that good and Alnico and field coil best. You do not see that type of distortion in the voltage so it does not show in acoustic harmonic distortion measurements. You can learn more at Home.
I belong to the Klippel advanced users group and actually i bought the first Klippel system made straight from the table of Prof.Klippel more then 10 years ago.
Last edited:
When you want to spend even more then the Satori tweeter, here is an option.
Diamond has other disadvantages though. Mainly low efficiency in the upper regions because it does not benefit from the bending wave effect that decouples parts of the membrane and makes the moving mass lower the higher it goes. Pick your poison.
Diamond has other disadvantages though. Mainly low efficiency in the upper regions because it does not benefit from the bending wave effect that decouples parts of the membrane and makes the moving mass lower the higher it goes. Pick your poison.