Over time you unlock the mystery of what
really goes on with live audio systems.
And what was previously only heard or assumed
what was going on.
for generic example say " 500 watt"
15 inch driver.
500 watts is the thermal rating
before the voice coil melts.
Linear excursion or Xmax is how far it can move
forward/back before 10% distortion.
Excluding " super woofers"
or whatever people call them.
With maybe 7mm to 12mm xmax.
5mm to 7mm is rather typical for
many live sound drivers. Many have
450 watt to 500 watt thermal ratings.
For the most part below 100 Hz
it wont be more than 150 to 180 watts
to hit 10% distortion or xmax
So a amplifier with " only"
250 watts is more than enough
plus dynamic overhead to reach
full SPL for a " 500" watt speaker
Keep in mind speakers get very very
hot in operation. So as a high SPL
sound engineer. Thermal rating is more
a timeline. If a speaker is operating for
a live sound performance say 3 to 4 hrs or
more. The thermal rating has to handle
that because of heat over a period of time.
Having done numerous under ground or
renegade electronic music festivals.
With up to 3 or 4 DAYS of constant music.
I have seen plenty of " 500" watt speakers
get evaporated to their graves being operated
at no more than 100 to 160 watts of power
per driver.
really goes on with live audio systems.
And what was previously only heard or assumed
what was going on.
for generic example say " 500 watt"
15 inch driver.
500 watts is the thermal rating
before the voice coil melts.
Linear excursion or Xmax is how far it can move
forward/back before 10% distortion.
Excluding " super woofers"
or whatever people call them.
With maybe 7mm to 12mm xmax.
5mm to 7mm is rather typical for
many live sound drivers. Many have
450 watt to 500 watt thermal ratings.
For the most part below 100 Hz
it wont be more than 150 to 180 watts
to hit 10% distortion or xmax
So a amplifier with " only"
250 watts is more than enough
plus dynamic overhead to reach
full SPL for a " 500" watt speaker
Keep in mind speakers get very very
hot in operation. So as a high SPL
sound engineer. Thermal rating is more
a timeline. If a speaker is operating for
a live sound performance say 3 to 4 hrs or
more. The thermal rating has to handle
that because of heat over a period of time.
Having done numerous under ground or
renegade electronic music festivals.
With up to 3 or 4 DAYS of constant music.
I have seen plenty of " 500" watt speakers
get evaporated to their graves being operated
at no more than 100 to 160 watts of power
per driver.
Perhaps we go back a little further than you do so our perspectives are probably different; how much did a 1kW amplifier cost in 1970? (Hint: a Crown 190W amp was £325, about £4500 in today's money).
No mystery.
To improve low end you turn up the bass knob.
Equalization is nothing new.
Changing the name to DSP isnt a magic device.
It is same filters being run with a different
application.
Why it is good to not spend hours or months
of your life in simulation trying to get " more bass"
Since more bass, specially in actual full space.
Is the same old " bass boost" or shelf filter.
You can compare pretty graphs for a live
sound 15" to a 18"
and many times -3dB response be the same old
45 to 55 ish Hz....What gives here
Generic speaking the 15 tuned to 45 Hz
compared to a 18 tuned to 30 to 20 Hz
when you boost the shelf filter / EQ / " DSP"
magic button. To " get more bass"
Your not boosting into the fluffy high tuned
port with a low tuned 18"
Most the " magic" to a reflex system is a port
way way below where music actually is.
Or likewise why many go to sealed to eliminate it
completely.
The suspension of the speaker is so loose or so tight.
You just need correct volume to make it work happily.
That is it.
Why simulation is a " transfer function" of a filter.
It is all the T/S theory is. Suspension found to work
well when it matches common filter cutoff rates.
Hence the word "Alignment" like QB3 or QB4
your just match or Align theoretical transfer function of
the speaker to transfer function of a filter.
People get lost trying to find 2 dB of port fart.
When active EQ can be 3 or 6 or 10 or 12 dB
of actual bass. No magic fart tubes needed.
The suspension is this tight or this loose, the box
is this big...the end, Resonant frequency is this many Hz
Like the hi-fi shops, turn up the Eq to have any real bass authority and definition and by the time some appears, you can no longer turn up the volume to enjoyable levels. That all sounds lo-fi
But that's no different from many mobile PAs. I experienced a rental JBL(?) box in a ~500sqm school hall one time. Ear-popping disco levels, have to yell at the person standing right next to you so they can sort-of hear what you're trying to say, that sort of thing... The speaker box had a built-in amplifier rated at over 3kW, but it was just 2x 15's + compression tweeter. By my guesstimate, the real power heating the voice coils was about 10W, with occasional peaks of a couple hundred watts for small fractions of a second.
For comparison, a floating line array at an outdoor concert had vastly better sound quality, but it's like it's some dirty secret that with the high efficiency and controlled dispersion, the whole array likely used even less power than the 2 PA woofers, despite being outdoors.
For comparison, a floating line array at an outdoor concert had vastly better sound quality, but it's like it's some dirty secret that with the high efficiency and controlled dispersion, the whole array likely used even less power than the 2 PA woofers, despite being outdoors.
For anyone interested. It took some time to actually build this first project because of plywood availability and other reasons, but I managed to finish it.
I still need to get into the weeds regarding measuring and tweaking the DSP settings but I'm definitely happy with the results for a first build.
Eventually I went with these components:
High: BMS 4550 in a B&C ME45 waveguide
Mid: 18 sound 8NMB420
Low: Faital pro 12RS430
Hypex Fusionamps FA503
This combination is somewhere between hifi and PA. I get quite a bit of clean output out of them. My idea was to make it a modular system where I could use it in a 2.2 but also in a 2.1 configuration if necessary (to save some space). Made two types of doors for the tops backs. One closed and one with ports. The ported version makes it possible to have them crossed over at around 90hz with this mid driver. So this could even give the posibillity to hang them from the sealing. For now I just use the tops directly on top of the subs, in that setup a crossover at around 300hz with the closed back gives nice and clean results.
I've been looking at the Fatial Pro 8RS340 as an alternative low/mid driver. For one because the aestathics of the cone match the sub driver better. But also becasue this could make the setup even more modular, as this driver can be tuned down to around 37hz at F10. Making it possible to use the tops stand-alone. In that case the efficiency of the system would be lowered a bit though.
For now I tried to tune them a towards the Harman Curve for a start.
Had a lot of fun building the project and I think it worked out really well.
Thanks for everyone’s input!
I still need to get into the weeds regarding measuring and tweaking the DSP settings but I'm definitely happy with the results for a first build.
Eventually I went with these components:
High: BMS 4550 in a B&C ME45 waveguide
Mid: 18 sound 8NMB420
Low: Faital pro 12RS430
Hypex Fusionamps FA503
This combination is somewhere between hifi and PA. I get quite a bit of clean output out of them. My idea was to make it a modular system where I could use it in a 2.2 but also in a 2.1 configuration if necessary (to save some space). Made two types of doors for the tops backs. One closed and one with ports. The ported version makes it possible to have them crossed over at around 90hz with this mid driver. So this could even give the posibillity to hang them from the sealing. For now I just use the tops directly on top of the subs, in that setup a crossover at around 300hz with the closed back gives nice and clean results.
I've been looking at the Fatial Pro 8RS340 as an alternative low/mid driver. For one because the aestathics of the cone match the sub driver better. But also becasue this could make the setup even more modular, as this driver can be tuned down to around 37hz at F10. Making it possible to use the tops stand-alone. In that case the efficiency of the system would be lowered a bit though.
For now I tried to tune them a towards the Harman Curve for a start.
Had a lot of fun building the project and I think it worked out really well.
Thanks for everyone’s input!
@Andersonix my knowledge about speaker design was almost literally zero. So used AI kind of like a mentor.
Have been going back and forward with google Gemini mainly, and had conversations about anything regarding the build. Ranging from questions regarding best materials to use (explored birch alternatives), to how to use the router without mutilating myself, to essential things to look out for in the modeling stage in winisd. Like port velocity, xmax etc and how different variables effect each other. It also helped interpret measurements graphs that I just screenshotted from REW.
Quite interesting to see how much better these models got over a longer project like this. Especially when it comes to calculations.
Can definitely see a time in the future where you could model speakers fully conversationally without having to use winisd. Having these models plot graphs based on t/s parameters and user preferences.
Have been going back and forward with google Gemini mainly, and had conversations about anything regarding the build. Ranging from questions regarding best materials to use (explored birch alternatives), to how to use the router without mutilating myself, to essential things to look out for in the modeling stage in winisd. Like port velocity, xmax etc and how different variables effect each other. It also helped interpret measurements graphs that I just screenshotted from REW.
Quite interesting to see how much better these models got over a longer project like this. Especially when it comes to calculations.
Can definitely see a time in the future where you could model speakers fully conversationally without having to use winisd. Having these models plot graphs based on t/s parameters and user preferences.