I think I disagree. The impulse response graph contains both the properties of the tweeter and the sound card. What you see, is the impulse response of the tweeter (for the sake of this argument assume this is perfect, so a dirac delta), convoluted with the impulse response of the sound card (which usually has a steep low pass filter, for example a sinc function). In this example, the result is a sinc function. If you sample the result and draw straight lines between the points, you can end up with a ragged looking graph, with the raggedness being strongest near the impulse peak.
I do not say that this applies here, just that this is possible.
The frequency response graphs go to 45-50 kHz and have a steep low pass characteristic at 43 kHz, which suggests a sample rate of 96 kHz and a 'ringing' period of 0.023 ms.
I do not say that this applies here, just that this is possible.
The frequency response graphs go to 45-50 kHz and have a steep low pass characteristic at 43 kHz, which suggests a sample rate of 96 kHz and a 'ringing' period of 0.023 ms.
Last edited:
It has been a while since I looked at these things so I'll defer to you on that.
As for the speaker choice, I think the Satori MW16TX-8 is overkill for a mid that is going to be high-passed at 250 Hz. Most of the 17mm coil height will be dead weight, adding resonances and distortion.
As for the speaker choice, I think the Satori MW16TX-8 is overkill for a mid that is going to be high-passed at 250 Hz. Most of the 17mm coil height will be dead weight, adding resonances and distortion.
All I can say from experience is that for a halfway correct picture of a tweeter's step (or impulse) response you need
The last two points (min-phase ADC filter and min-phase display) gain more importance when lower sample rates have to be used.
It's basically the same strategy that is used in good digital oscilloscopes where you also don't want to see things "happening before the fact" : https://cdn.teledynelecroy.com/files/whitepapers/group_delay-designcon2006.pdf (Fig.3)
- a mic that has 50kHz...100kHz bandwidth
- 192kHz or higher sample rate of the ADC, and selecting a minimum-phase digital filter for anti-aliasing, preferably one with a gentle roll-off.
- use a software that does a decent reconstruction of the waveform display using a sinc approximation. Adobe Audition is a good example. Ideally, one would best use a minimum-phase display reconstruction to avoid any chance of getting "pre-ringing" artifacts (Gibbs Ringing) which is not actually present in reality. One way to emulate this is by proper upsampling with a minphase filter so that the actual display algorithm is less important, not dominating.
The last two points (min-phase ADC filter and min-phase display) gain more importance when lower sample rates have to be used.
It's basically the same strategy that is used in good digital oscilloscopes where you also don't want to see things "happening before the fact" : https://cdn.teledynelecroy.com/files/whitepapers/group_delay-designcon2006.pdf (Fig.3)
I wouldn't choose that driver for midrange duty either, especially because there is already a midrange version of it:
https://sbacoustics.com/product/6½-satori-mr16tx-8-textreme/
Last edited:
Thank you for your response, to be honest, sometimes I can listen the music the at high volume and my amplifier is 100 watt RMS, so if I cut the mr16tx-8-textreme at 250Hz/2.5Khz, around 30% or more power go to the midrange that is only 30 Watt. so I don't want to risk getting close to the limit.
If the Satori MW16TX-8 is not a right choose, I need to find something else I can cut at about 250Hz/2.5Khz.
If the Satori MW16TX-8 is not a right choose, I need to find something else I can cut at about 250Hz/2.5Khz.
Last edited:
No problem at all.
Satori MR16TX-8 in 7L closed box, xo at 150Hz/2.5KHz, xmax is 2.1mm with 100W (105 dB spl max.)
Satori MR16TX-8 in 7L closed box, xo at 150Hz/2.5KHz, xmax is 2.1mm with 100W (105 dB spl max.)
The SATORI MW16TX-8 got a quite large voice coil of 36mm diameter. That results typically in a power handling of 100W. You are using the driver only for midrange, where the crest factor is quite high (low average power with higher peaks), so it will be able to take a lot more power within 250-2,5k and can dissipate the heat from the higher peaks easily, no need to worry about that at all.
Oh, the driver knows when to stop, that's when the input signal is stopped. 😉
Actually, no. The sensitivity is also dependent on the active surface. And the required speed strongly depends on the frequency. Resonances can increase the level but if they are out of phase, they can reduce it and I suspect that this happened with the Bliesma on the impulse response. The impulse answer will change with the xo too, more or less. And you can have high frequency resonances (which almost all hard dome drivers have). For a high frequency the dome has to move faster too, that's how frequencies work. The resonances of the Bliesma are at a much lower frequency than on the Satori, you can read that by the distance (=time, =frequency) of the peaks.
Differences are minimal re xmax and max power/spl. Qtc is 0.51 instead of 0,61. Consider that both cabinets are completelly filled with stuffing so the volume is actually higher. Even with 7L the Fb is about 83Hz so crossing it over at 250 is perfectly fine.