One of the problems with the blade configuration is the that the side woofers widen the response. The simulations and Blade2 measurements show horizontal pattern control lost by 500 Hz or so. I found by adding a rear woofer and DSP channel I could get a cardioid-like response with directivity controlled down to 100 Hz or so. I think a rear woofer, DSP controlled, will take care of the objections people have had to the side woofers.
Questionable! Your simulations assumes there is no backwall behind the LS !!
And if the ears detecting delayed sound , localisation is corrupted!
Don't mix colourisation with localisation issues!
Sorry - I don't think your "digital improvement" will have a positive impact on how we perceive phantom sources in the "soundcloud" ...
regards , Mr. Phantomas
And if the ears detecting delayed sound , localisation is corrupted!
Don't mix colourisation with localisation issues!
Sorry - I don't think your "digital improvement" will have a positive impact on how we perceive phantom sources in the "soundcloud" ...
regards , Mr. Phantomas
Actually, the simulation can show the effect of the frontwall.
This chart is the Vituix power-DI-PIR chart for the blade clone with rear woofer. The orange curve is the predicted in room. The dashed olive curve is the PIR with reflection from front wall enabled. Vituix drops the level of that trace to separate it from the orange. You can see that the primary effect of the front wall in the bass is to support the bass below 100 hz. The speaker is placed .5m from the front wall.
Not convinced its an improvement? Let's look at what happens without the rear woofer:
The dashed green trace is the in-room response with front wall reflection enabled. The dashed olive trace from the first chart remains in place for comparison. Choose your poison.
This chart is the Vituix power-DI-PIR chart for the blade clone with rear woofer. The orange curve is the predicted in room. The dashed olive curve is the PIR with reflection from front wall enabled. Vituix drops the level of that trace to separate it from the orange. You can see that the primary effect of the front wall in the bass is to support the bass below 100 hz. The speaker is placed .5m from the front wall.
Not convinced its an improvement? Let's look at what happens without the rear woofer:
The dashed green trace is the in-room response with front wall reflection enabled. The dashed olive trace from the first chart remains in place for comparison. Choose your poison.
Last edited:
Sorry , nice simulation but the story was do have front woofers better integration than side-firing ultra-reference-wannabees?
Many listeners claim side firing woofers have drawbacks! (depending of course on more or less advanced filtering/delay)
VituixCad cannot simulate human perception , nor derivate it somehow!
Integration of sounds depends on masking , tresholds over time , HRTF influence , very complex scenario!
At the end - better trust your ears than relying on stupid simulations!
Picking the right test music is king!
good luck with your further career opportunities in becoming a master simulator of the nth degree
Many listeners claim side firing woofers have drawbacks! (depending of course on more or less advanced filtering/delay)
VituixCad cannot simulate human perception , nor derivate it somehow!
Integration of sounds depends on masking , tresholds over time , HRTF influence , very complex scenario!
At the end - better trust your ears than relying on stupid simulations!
Picking the right test music is king!
good luck with your further career opportunities in becoming a master simulator of the nth degree
Thanks for sharing this. It kinda fosters where my thinking has been focused.
I don't think i'll add a DSP controlled rear woofer as a solution though.....i think all fwd facing with the farthest woofers getting a .5 coil instead.
Building this on the cheap with the Dayton DA175 woofers and 2cuft per pair, i can get an in room response below 30hz i think. Efficiency is still going to be limited to the Uni-Q though at around 85db. Active DSP and separate power amps will be needed.
I might just try and do a cardioid enclosure for the Uni Q as i did in my nearfield modded Q150's............Lot's of experimenting on that one to get it to work....and guess what?......i didn't expect to repeat it so i never took notes
Simulatons are not stupid.
Stupid might be someone drawing wrong conclusions.
But I know, you're not entirely serious ...
The all forward firing solution will have a wider forward beam, unless made cardioid-like by any of several methods. DSP controller rear woofer is the least painful but also the least directive. Side vents are the opposite, as you have confirmed, and can control directivity to a higher frequency.
side-firing ultra-reference-wannabees??? what a great turn of phrase!
are they worth the extra cost? Erin apparently thinks so! IIRC he attributed that to the absence of diffraction and didn't take issue with the side woofers.
I don't have personal feelings about side woofers vs front woofers; my feelings form around the resulting directivity characteristics which I predict via simulation. I don't trust simulation over my ears. I use simulation to determine what is worth building so that my ears can make a judgement and so that most times they like what they hear.
what the out of phase rear woofer does is cancel most of the output from the side woofers that goes to the rear and some of what goes to the sides so that you don't hear a too-shortly delayed reflection from the front wall. Without it or some form of cardioid, you may need to pull the speakers further out in the room than the room size permits in order to get the reflection past the Haas limit. This is true for most (non-cardioid) speakers, not just the Blade.
Its the cancellation of "some of what goes to the sides" that I thought might overcome objections I've heard to side woofers. These objections show in erin's measurements as the horizontal side lobes. These are reduced but not completely eliminated (see attached) in my simulation by the effect of the rear woofer and likely steeper crossover slopes. Because those sidelobes aren't completely eliminated, you would not want to place these drivers near a room corner, at least not without absorbing that near sidewall reflection of the side woofers. For really tight quarters, I think you would want to use cardioid technique past 1 khz.
are they worth the extra cost? Erin apparently thinks so! IIRC he attributed that to the absence of diffraction and didn't take issue with the side woofers.
I don't have personal feelings about side woofers vs front woofers; my feelings form around the resulting directivity characteristics which I predict via simulation. I don't trust simulation over my ears. I use simulation to determine what is worth building so that my ears can make a judgement and so that most times they like what they hear.
what the out of phase rear woofer does is cancel most of the output from the side woofers that goes to the rear and some of what goes to the sides so that you don't hear a too-shortly delayed reflection from the front wall. Without it or some form of cardioid, you may need to pull the speakers further out in the room than the room size permits in order to get the reflection past the Haas limit. This is true for most (non-cardioid) speakers, not just the Blade.
Its the cancellation of "some of what goes to the sides" that I thought might overcome objections I've heard to side woofers. These objections show in erin's measurements as the horizontal side lobes. These are reduced but not completely eliminated (see attached) in my simulation by the effect of the rear woofer and likely steeper crossover slopes. Because those sidelobes aren't completely eliminated, you would not want to place these drivers near a room corner, at least not without absorbing that near sidewall reflection of the side woofers. For really tight quarters, I think you would want to use cardioid technique past 1 khz.
Why? At the frequencies the woofers ar eused they are omnidirectional. It would make no difference what direction the woofers face or how they are arranged..
dave
hmm the woofers do indeed go omni, at/around 450 hz but they also form an array and array configuration does indeed affect directivity. If you put them all in a line as in the reference series, fr example, you could hope to get improved vertical directivity in their range. but that is just a quibble.
I should have introduced my post differently. In retrospect, I should have faulted the blade, not the side woofers, for its widening response below 800 Hz or so, setting the stage for showing how that could be substantially improved with an out of phase rear woofer. Its interesting that a rear woofer allows you to have a cardioid-like response working against side woofers as well as front woofers
I should have introduced my post differently. In retrospect, I should have faulted the blade, not the side woofers, for its widening response below 800 Hz or so, setting the stage for showing how that could be substantially improved with an out of phase rear woofer. Its interesting that a rear woofer allows you to have a cardioid-like response working against side woofers as well as front woofers
No they don’t, the DI flat lines at about 150Hz which is consistent with BEM sims of where all directionality is lost. Between 150Hz and 500Hz the DI is rising showing increasing directionality. In this frequency range you will get a different response depending on how the woofers are positioned.
What is best is much harder to say when rooms and boundaries are involved.
Below a quarter wavelength there is no difference between a single driver and array in terms of directivity so it applies to both over most of the frequency range they are being used.
Just me thinking out loud: If anyone wanted a cardioid response , couldn't they place any speaker behind their main and feed it out of phase, filtered programme? Fiddle with delays in bands of frequencies, the overall EQ curve etc .. A spare mid woofer in a cabinet.
Would be similar to what folks do in the PA world with their simulation software.
Would be similar to what folks do in the PA world with their simulation software.
Beolab is still my favorite speaker of all time; the ability to change the beamwidth is a game-changer. It's like having two or three different speakers, all at the same time.
Just occurred to me that an out-of-phase rear woofer could provide something similar. Basically, you could add a switch to the passive crossover that feeds the rear woofer, and then add some resistors to power taper the rear woofer. So with full power to the rear woofer you would have the narrowest beamwidth, and the switches can increase the power to said woofer, which would widen the beamwidth.
Basically, three beamwidths via three power settings for the rear woofer.
Even better, do it with DSP and dedicated amps for each driver, and you can vary the beamwidth with much greater precision, by varying the crossover and time delay.
Yes, you can create a pseudo dipole or bipole using two (ideally) identical speakers, once facing forward, one backward.
In my office I have four of those cheap Monoprice towers, and I've been meaning to mess around with this. Each one is around $150, so it's a cheap/easy way to mess around with designs like you describe.
Add some DSP to the mix, and you can create gradient (in-phase) and end-fire (out-of-phase) cardioid response patterns.
Getting cardioid at low frequency is easy this way, and with high frequencies, the dual tweeters don't interfere with each other very much because high frequencies are short enough that they don't easily wrap around the cabinet.
The super difficult part would be the midrange, because you're right in the range of frequencies where the front-to-back spacing is going to cause a lot of issues. Power tapering on the rear midrange would likely go a long way here.
I imagine if you REALLY wanted to do this right, you could even strap two speakers front-to-back using moving straps, and then you could take the entire thing outside, measure it, and optimize the beamwidth over a full 360 degrees. That would probably make for a heck of a online speaker project, because if you did it with commercial speakers, anybody out there could replicate your results by simply loading up your DSP settings. Very easy to distribute on something like Github.