The making of: The Two Towers (a 25 driver Full Range line array)

It was the easiest way I could think of to realise the inner wavy shape and the outer shape that have way more thinking in them than is obvious to the casual observer.

I could list all of my thoughts on it but may risk being shipped to the asylum in a straight jacket. :D
I've kept them narrow because I found a lot of "stories" on narrow and rounded enclosures having advantages in imaging. I spend a lot of time browsing the net looking for inspiration. I cannot deny that this was one of the sources for my inspiration among others.
The rounded shape is obviously inspired by the Olson paper, linked previously in this thread I believe. There's more, but I bet you guys already think I'm half a lunatic :).
olson-baffleshape-fr.gif

Source of picture: Planet10's HiFi site.

By the way, I should ask for a commission from PE, I bet they sold a lot of the TC9's in large quantities over this weekend. :D
A few others (in pairs) probably went to the Open Baffle speaker lovers.
 
Yes, I agree. for me it does not weigh up to the advantages of the floor to ceiling array in the living room from what I see.



I had way more fun than I expected to have with 4.0 HT sound. And that was only using one SS 10F each for my ambient channels. Arrays for ambient channels would have been better as I do notice the distance to speaker plays a larger role with the point source ambient channels. I didn't exactly have that option and this works fine. But I can only imagine having the room and sources for surround using arrays.
What are your center channel plans? I have thought about it, but not felt any need yet. For music not needed at all, for movies it might have advantages. Although I'd rather go for a bigger screen to fit the stage. If you have that big screen already... :)
Are you thinking of building a HT setup with a projector? Center behind the screen?
Standing center with 25 TG9 behind transparent screen. Had them for a long time. Use some DSP on them to tune them like the TC9 I hope. Probably they sound pretty similiar.
Already have the homecinema with a big screen :) If they (lines) are this good, sell the big (front and center) JBL cinema speakers and keep the 4 18" JBL subs or build 4x12" sub arrays on both sides. Both depending on time and money.
 

ra7

Member
Joined 2009
Paid Member
Every time you post those pictures, it amazes me the amount of work you did. Truly heroic efforts! You should distill your design ideas into simpler plans for diyers to build and sell the plans like Linkwitz. You can even license PE to offer a kit. Cut wood pieces, miniDSP and your custom filter files. Just throwing out ideas :D It would beat the pants off pretty much anything anyone can build or buy.
 
Standing center with 25 TG9 behind transparent screen. Had them for a long time. Use some DSP on them to tune them like the TC9 I hope. Probably they sound pretty similiar.
Already have the homecinema with a big screen :) If they (lines) are this good, sell the big (front and center) JBL cinema speakers and keep the 4 18" JBL subs or build 4x12" sub arrays on both sides. Both depending on time and money.

I don't think it will be a problem having TG drivers do the center channel at all. Especially with DSP involved. I'm kind of getting a little jealous here... :rolleyes:
 
I have a few questions about these speakers. I'm a complete novice, but I did read the whole thread a few months ago.

How did wesayo come up with the 250w power handling? Dave from planet10 posted elsewhere on the forum that line array power handling is calculated by multiplying the power handling of an individual driver by the number of drivers in the array. According to that formula the power handling of these speakers would be 1500 watts! I'm a little confused over this matter.

A line array casts a wide sweet spot. Does that apply to the effects of room correction with these speakers? How would the dsp handle multiple first reflections from the various listening spots?

Lastly, how's the mid bass slam on these speakers? I've read on other threads that line arrays have trouble reproducing bass with small drivers. Is this simply a matter of equalization?
 
Waaaaawwww!!!! That looks super good!!! Ive always loved line array projects!! I think your design is very well thought out.. Very good craftsmanship!! Seeing poeple who still do amazing things like this are awsome.. Like the fact you used cabnite grade plywood. Very good choice for a build such as this one! Good job man! Im sure they sound as good as they look!!!

This is a very nice project havent seen this in a while!
Thanks for sharing... I have an idea now in ly head for my next build haha!!!
 
I have a few questions about these speakers. I'm a complete novice, but I did read the whole thread a few months ago.

No problem, I'll try and give some clear answers :).

How did wesayo come up with the 250w power handling? Dave from planet10 posted elsewhere on the forum that line array power handling is calculated by multiplying the power handling of an individual driver by the number of drivers in the array. According to that formula the power handling of these speakers would be 1500 watts! I'm a little confused over this matter.

The power rating of one driver is usually given for their expected range. For a small driver like this that probably would be 100 Hz to 20 KHz, like it says on the new spec sheet: http://www.parts-express.com/pedocs/specs/264-1062--tymphany-tc9fd18-08-spec-sheet.pdf
Rated noise power would be 30 watt. 25 drivers x 30 watt would give 750 watt if we use the math in your example. (we usually only count one speaker as we use Stereo amplifiers rated for watts per channel). I guess you counted 50 drivers total?

But I don't use the 100 Hz to 20 KHz range. I use ~30 Hz to 20 KHz, and whatever I get below that 30 Hz if I'm lucky. So I took a speaker sim program, put in the TS values and box parameters, and started to simulate EQ out to 30 Hz until one of the variables of the driver maxed out above it's rated specs. The spec that crossed it's max value was x-max. X-max is rated as 2.55 mm. The speaker cone can travel a bit further and wouldn't be harmed immediately but for safety I wanted to get out to x-max at ~30 Hz and no further. The predicted power value at that time was about 240 watt. So that's my self induced limit. The power for one driver would be: 240/25 = 9.6 watt.
So that would be the maximum power a single driver would have to handle.
Does this make enough sense to you?

A line array casts a wide sweet spot. Does that apply to the effects of room correction with these speakers? How would the dsp handle multiple first reflections from the various listening spots?

It really doesn't. I've said it more than once: DSP is no cure for reflections. I use the DSP software to EQ my line arrays. But I use it as a speaker correction for most of the frequency range. You have to be aware of what the speaker does with the signal. I use damping panels to 'take care' of the mayor early reflection points. At lower frequencies I do use the DSP as room correction. But it isn't two separate mechanisms. In order to do it successfully I needed to keep the frequency dependant correction window short.
DSP cannot solve everything. It's way better policy to first correct what you can by more traditional means and use the DSP power to push that last bit.
In my project I combined my need for EQ with a lot of speaker correction, and a little room correction. My damping panels were a vital part to achieve my results. Good question though!

Lastly, how's the mid bass slam on these speakers? I've read on other threads that line arrays have trouble reproducing bass with small drivers. Is this simply a matter of equalization?

I wouldn't say it's just a matter of equalisation. If it were I would have finished this project long ago. I'm convinced you need a smooth phase response from ~50 Hz to at least 200 Hz to experience mid bass slam. The actual slam you hear comes from a higher frequency than most people will let you believe. The slam you hear from the kick that knocks you on the chest is somewhere between 100 and 200 Hz.
The sub frequencies do not sound like a slam by itself. They do create the force that comes with that slam. To really have a fast kick you need to have the timing right. Do I feel the kick? Well, at the levels I listen to I don't experience it.
I've had it in my car by running my sub higher than usual. Although it was fun, for a brief moment, it got in the way of driving. The SPL level was way higher than my average in room listening habits.

Do you want that kind of kick? Or is it enough that you can feel the hits, make the floor and couch move, shake stuff in the room without taking your breath away. If you really want that kind of slam I'd suggest looking at PA speakers. I recently heard a set outside that could do that. But I favour the fast and clean sound I get from my arrays. Where I can actually follow the bass notes.

I have not tried it, but if you wanted to try you'd need to filter at about 40-50 Hz to prevent putting a lot of power to the arrays at sub frequencies. I do feel the low tones, the ones at 20 to 30 Hz are marvellous to feel, more than you hear them. They really give atmosphere to songs that have that. I've felt low bass in my stomach. Especially when I just finished the arrays. I had the low end turned up more and was playing insanely loud, without realising it. Until I tried to talk... But to say it again: I wouldn't recommend this type of arrays for that specific kind of sensation. Even if they could be capable of it. I'd use separate solutions if I wanted that.

Are you by any chance coming from a Car sound background? It seems most people in the SPL category get into ridiculous levels looking for that experience. It wouldn't make you popular to do it in a home environment if you have neighbours. Even in my car I went for quality, not quantity. I hate the one note wonder bass. (and rattling license plates) It's easier in a car though. I just couldn't and wouldn't want to drive like that.

Read: this post again for impressions on this subject.

Waaaaawwww!!!! That looks super good!!! I've always loved line array projects!! I think your design is very well thought out.. Very good craftsmanship!! Seeing people who still do amazing things like this are awesome.. Like the fact you used cabinet grade plywood. Very good choice for a build such as this one! Good job man! I'm sure they sound as good as they look!!!

This is a very nice project haven't seen this in a while!
Thanks for sharing... I have an idea now in my head for my next build haha!!!

Thanks Donny!

Note to myself: why do you need to make such long posts? Is it the language barrier? Problems expressing myself in a short and to the point kind of way? Here I go again (lol).
 
Last edited:
Thanks for your reply!
No, I don't come from a car audio background. I just built my first set of DIY speakers over the winter (audio nirvana). I was actually looking at converting them to open baffle for more visceral bass—with Eminence woofer. I am by no means looking to crack windows or shatter the foundations of my house. I listen to a lot of folk and soul, but also a lot of house and techno. I don't listen to "everything" but I do need versatile speakers.
Mostly I asked that question because of a comment you made equating your 25 drivers to a single larger driver (15" or 18"?). Some comments of people speaking in hypotheticals in an old thread suggested otherwise, so I wanted the opinion of someone with real world experiences.
Your suggestion for PA speakers is interesting, because when I listen to house I really crave that club sound. On the other hand, when I listen to folk I really want to hear the air around the singer's voice and the detail of their finger picking.
Right now I'm thinking of adding woofers to the audio nirvanas, but I'm tempted to experiment with a line array in the future. It all comes down to what space I have available in my bedroom.
And yes, you caught my mistake on point. I multiplied by 50 instead of 25 out of haste.
I have no complaints receiving lengthy replies to my questions from someone who is both unpretentious and knowledgeable.
 
Oh and one more, question... If you had to do it over again would you stick with the aluminum baffle? I read somewhere on the thread that there were detectable resonances

Yes, I'd use it again. After "upgrading" my damping material between both aluminium baffle plates there were no resonances left that bothered me. (and that is saying something, I'm hard to please :))
I try and account for every detail. But that doesn't exclude anyone from doing things different than me and still be successful. For my way of doing things this made a lot of sense, plus I like the look!

By the way, out of sheer interest I play a lot of styles on the array. In fact, I just woke up the family a minute ago with some Dance because I forgot to turn off my amps :eek::eek:.
I'd say these definitely can bring the dance/club sound to your home. I even try heavy RAP like E40, Some dance but also simple hits like "Pitbull - Fireball". It's a lot of fun! (and that last song isn't even a bad recording!)
 
Last edited:
Time for another newsflash/wake up call...

I've been advertising the use of Frequency Dependant Windows. I still support that view, but today learned something new and useful. I figured I'd share it here first. REW's Frequency Dependant Window is a nice addition to an allready usefull tool. But the resolution of that FDW window isn't as high as I'd like it to be.

Out of interest to find out more I made a quick comparison:
FDW-difference.jpg


This might not be the best example, this is a simulated plot of my correction as seen in REW (green) and as seen by DRC (red).
Both are set to show 5 cycles. The same amount of difference in level of detail can be seen in the phase plot.

It's not a small difference. So be warned! I'm glad I caught this. I expected to see a difference, though nothing this large.
So from now on when I talk about Frequency Dependant Windows I'll be using DRC-FIR as the tool to get my hands on it.
I have a better example available but I'd have to ask the owner of the file first if I can show it.
 
Last edited:
Well REW does not offer anything else at this time, no smoothing parameters. Just a setting in the IR box. I caught this today while looking at measurements from someone else. I haven't used it in my project as it wasn't available yet. That deviation at ~70 Hz is strange. For now I'll put my trust in DRC-FIR. Don't make too much of this example, it's only meant as an example. Just know the results from REW are smoothed when using frequency dependant windows. They sure looked that way, but I noticed some bigger deviations or detail changes that made me look and post about it.
My nickname could almost be Mr. FDW thus I wanted to tell you guys as soon as I found out.

P.S. Even the DRC window will be smoothed, the question would be: how much...

Last edit today: one big difference would be: the DRC result is a minimum phase generated view. So to see how it would look in REW, You'd need to export the impulse with "Export Min Phase version of IR" selected, re-import and compare using the FDW. That won't make it more detailed though. Just be warned I guess...
 
Last edited:
Look at my edit above... (and in your mail)

very last edit, one more example courtesy of ra7:
FDW%20difference.jpg


As said, DRC uses a minimum phase generated view. It also does separate windowing for excess phase. Just posted this as is...
You guys discuss while I get some sleep :). I'll make it a quiz, (without a price, I'm broke!) can you guess the used programs belonging to the graphs?
 
Last edited: