Hi RHosch,
I don't have enough experience yet, comparing the straight-line in a room with flat ceilings vs sloped ceilings, to give you any decent impressions.
I've been going back and forth between two such rooms, but they are so different acoustically, it's hard to say what role the different ceilings are playing.
At this point, my guess is that the biggest effect is how well the very bottom end is controlled.
As far as the cbt having a vertical focal point 1/3 to halfway up the array, that doesn't vary with ear height or listening distance....
I just tried putting shading and delay on a straight-line to mimic the cbt.
And got the same phenomenon. It wasn't as rock solid stable as the physical cbt, when varying ear height and distance to array, but it was close.
I also inverted the shading and delay, running from top to bottom, to maybe mimic an upside down cbt. The image center moved upwardly..the inversion seemed to work. This was in the sloped ceiling room.
I think you are right about the cbt having better, 'self correcting', horizontal coverage. Wesayso appears to think the same, unless we do dedicated room tuning.
I also find the arrays to have a larger than life sound, but I just attribute that to the wider horizontal dispersion of the drivers in them, than I'm used to listening to with horns. I'm an admitted fan of more direct, less reflected sound.
Also, I keep thinking... aren't line arrays all about vertical control?
Isn't horizontal control for a line the same-old well known axioms....re the size of the driver/horn?
One thing I've been enjoying that I totally didn't expect...
I've got my room set up with a single point source stack on the long wall center, and a pair of straight arrays in the end corners.
The arrays make the best damn surround channels i think I've ever heard.
Their wide horiz pattern, and somewhat diffuse high end (compared to the horns) make for a really smooth blend in with the ps stack.
I got sick of surround many years back, but this is kinda cool 🙂
Maybe trying to tune a punch of point sources to a single spot was what pushed me away from surround?😕
I sat on the fence between straight array and CBT for a long time before sending plans for a straight array out to a CNC shop. This discussion gives me pause because I think a lot depends on how the array integrates with the room and indications are CBT is more tolerant of room boundaries and doesn't depend on ceiling reflections.
My straight array differs from Mark's in that its much less volume and designed to fit tightly into corners or on a wall. Because of its smaller size, it can fit tighter than Mark's moving the boundary interference higher where it can be treated with smaller thickness of absorber, if necessary.
If not taking advantage of tight wall corner integration, and with sloping ceiling, CBT could well work better. But ultimately, one has to build and see because available simulation tools only give hints. Then, fixing the problems one sees with DSP and room treatments, one fulfills the prophecy one makes in choosing one array over the other.
My straight array differs from Mark's in that its much less volume and designed to fit tightly into corners or on a wall. Because of its smaller size, it can fit tighter than Mark's moving the boundary interference higher where it can be treated with smaller thickness of absorber, if necessary.
If not taking advantage of tight wall corner integration, and with sloping ceiling, CBT could well work better. But ultimately, one has to build and see because available simulation tools only give hints. Then, fixing the problems one sees with DSP and room treatments, one fulfills the prophecy one makes in choosing one array over the other.
Hi nc535, fwiw I wouldn't take pause with your decision.
I don't think there will be any significant sonic difference between them, once the corner array is fully dialed in. Maybe the straight will end up with a slight bottom end edge, and the cbt with a top end edge ...best forecast guess so far.
Abd the straight line straight surely provides more opportunity for experimenting, with frequency shading and delay steering.
If I weren't planning on such experimenting, I'd probably choose cbt vs straight simply on which design physically fit the room layout the best. (again, given my limited experience so far)
I don't think there will be any significant sonic difference between them, once the corner array is fully dialed in. Maybe the straight will end up with a slight bottom end edge, and the cbt with a top end edge ...best forecast guess so far.
Abd the straight line straight surely provides more opportunity for experimenting, with frequency shading and delay steering.
If I weren't planning on such experimenting, I'd probably choose cbt vs straight simply on which design physically fit the room layout the best. (again, given my limited experience so far)
The CBT array I'll be building is a 68" arc and about 60" vertical from floor. Along with vertical and horizontal power response, I'll be paying close attention to perceived imaging height and midbass capability.
Theater speakers don't necessarily have to be at screen mid height. Ear level is a valid choice at around 42" from floor. The comment about psychoacoustic effects pulling the sound up to match the visual is correct. However, if its too far off the effect can be fatiguing over extended periods.
The option to build a taller CBT is always an option if I find I like everything else but imaging height. Could denote the first pair to surround duties I suppose.
Now have to what material to make the face plate from. I wish I had friends with a CNC. Just too pricey to pay for a pair of milled faces with ~100 holes/recesses each. So I designed an aluminum router guide template that covers everything needed and can be stepped up the face aligning from the holes just made. That was only ~$100 and is being CNC now.
Theater speakers don't necessarily have to be at screen mid height. Ear level is a valid choice at around 42" from floor. The comment about psychoacoustic effects pulling the sound up to match the visual is correct. However, if its too far off the effect can be fatiguing over extended periods.
The option to build a taller CBT is always an option if I find I like everything else but imaging height. Could denote the first pair to surround duties I suppose.
Now have to what material to make the face plate from. I wish I had friends with a CNC. Just too pricey to pay for a pair of milled faces with ~100 holes/recesses each. So I designed an aluminum router guide template that covers everything needed and can be stepped up the face aligning from the holes just made. That was only ~$100 and is being CNC now.
re' CNC price.
I actually had a friend with a CNC but, like me, he got old and retired. Last year he sold his machine and bought a Harley. Now I'm breaking in a new shop and its costing me $450 to do the pair of line arrays from just a single sheet of ply. There was a guy in Kansas willing to do it for half that price using a Shaper CNC and bevels on a table saw
Custom Speaker Designs and Kits | Speaker Hardware
Using a template, you will eat a lot of sawdust and spend lots of time. Just ask Wesayso.
Only 100 holes each? Are you not counting the driver mounting holes?
My baffles have 160 holes including 4 pilot holes per full range driver
I actually had a friend with a CNC but, like me, he got old and retired. Last year he sold his machine and bought a Harley. Now I'm breaking in a new shop and its costing me $450 to do the pair of line arrays from just a single sheet of ply. There was a guy in Kansas willing to do it for half that price using a Shaper CNC and bevels on a table saw
Custom Speaker Designs and Kits | Speaker Hardware
Using a template, you will eat a lot of sawdust and spend lots of time. Just ask Wesayso.
Only 100 holes each? Are you not counting the driver mounting holes?
My baffles have 160 holes including 4 pilot holes per full range driver
Hadn't seen that speakerhardware site before, thx.
Happy to see he says BB only, NO MDF EVER lol
You guys are making the driver baffles out of wood, right?
I guess it depends on the driver, and whether you are recessing them flush, but I found cutting the all the holes (120) for the TC9's really quick and easy. (front mounted non-flush)
Just plunge routed 3 1/16" driver holes every 84mm, then stuck a couple of drivers in the holes to force square alignment for drilling mounting bolt holes...used the drivers as a drill guide....went down the line..done.
Took a couple of hours per baffle..
Think I have about $400 in all four line-array boxes, including mounting hardware and paint. (albeit only 2 baffles, and of course excludes driver cost)
Kinda why I was willing to make four boxes...not that much $ considering what could be learned.
nc535, what driver(s) are you using? SBC65WAC25? can't find that with search...
100 was a rough estimate of time required. There will be 48 tweeters and 16 woofers per side. Woofers get recessed for flush mounting, tweeters are sans faceplate and just single thru hole. That's 80 counting holes and recesses. Then 4 mounting screws per woofer and another dozen or so for mounting faceplate to.box. but those are all drilled and much quicker.
I'm using Dayton ND105-8 for woofer to gain some displacement and extension for theater application. We will see if it is enough. Peerless OX20SC00 19mm softdomes for tweeters. Hope they can keep up!
I think I will.use plywood for the faceplate, at least for this first prototype pair. Would like true Baltic birch but am longer than the 60" dimension and don't want a joint I'm curved panel. MDF machines well but more.brittle for curving. Thought about black ABS or polypropolene donut could just be brushed or buffed for final finish, flexible and machines well... but cost is not trivial!
I'm using Dayton ND105-8 for woofer to gain some displacement and extension for theater application. We will see if it is enough. Peerless OX20SC00 19mm softdomes for tweeters. Hope they can keep up!
I think I will.use plywood for the faceplate, at least for this first prototype pair. Would like true Baltic birch but am longer than the 60" dimension and don't want a joint I'm curved panel. MDF machines well but more.brittle for curving. Thought about black ABS or polypropolene donut could just be brushed or buffed for final finish, flexible and machines well... but cost is not trivial!
You can get 4'x8' BB marine plywood; just as nice as the 5x5 BB and with marine glue.
Nice tweeter for line array. No mounting screws needed!?! Glue it in or clamp or?
be sure to ask for special pricing from PE on your quantity, they were nice to me
I would be tempted to rear mount those woofers to avoid diffraction/reflection from the big half roll suspensions
Nice tweeter for line array. No mounting screws needed!?! Glue it in or clamp or?
be sure to ask for special pricing from PE on your quantity, they were nice to me
I would be tempted to rear mount those woofers to avoid diffraction/reflection from the big half roll suspensions
Thanks for the tip on marine ply. I have thought about rear mounting. The router template is fine for either side so I'll play with it and see what I think. Tweeters mostly friction fit with a couple of dabs of silicone, hot glue or other easily removeable glue once they are in place.
Hi guys,
thought i'd pass on some more observations...
I've been concentrating solely on the straight arrays, before I switch them both to CBTs.
Experiments have fallen into 3 categories:
a. corner vs out in the room,
b. different amounts of measurement averaging for the response that gets corrected, and
c. applying CBT-like shading
a. I wholeheartedly agree with all the folks who have built corner line arrays, and talk about the need for absorptive side wall treatments. There is plain and simple a whole lot of horizontal reflecting going on.....
nc525, I think you are going to be happy that you have smaller and excellent drivers to sit deeper in the corner, but I imagine you will still need plenty of good absorption.
That said, I can see that there may be people that prefer the more reflected sound, and might want to try diffusion panels instead. I think I mentioned earlier that the corner lines make great surrounds imo. If diffusion can dampen the comb filtering from side wall reflections, that would be cool to try i think. Any good, tall diffuser ideas, that could run parallel to the arrays, anybody?
I like the sound of the arrays moved out of the corners the best, more open sounding. Like Nvphotos said his 'White Whale' thread.
Measurements are much nicer to work with too, as I haven't yet used enough absorption in the corners to tame tuning needs.
I would keep the lines out of the corners but bass simply becomes too anemic. Sustained bass notes are ok, but dynamics aren't there.
x-max rears it's head when trying to eq bass response up to the needed level. RHosch, I think you are going to be happy with your ND105-8 inclusion.
b. averaging has really worked.
I've settled in on a 9 position average, comprised of on-axis and +/- approx 30 degrees, taken at about 1/4, 1/2, and 3/4ths of the array height. About 8ft mic distance....mic distance doesn't matter a lot to the average (within +/- a few feet.)
A raw average of the speaker out in the room is below, one that I'm currently using for the FIR file correction.
I'm only using full octave correction, which keeps sounding the best.
IIR would really work fine imo, no real need for FIR other than the ease of automatic EQ built in the FIR generator.
c.using the same EQ, the same FIR files, as the straight array has whether in the corner, or out in the room....
So far, amplitude shading and delay offsets have not yielded as nice a sound as the normal line, in either the corner or out in the room.
I'm thinking I need to build new EQ files for when in 'CBT' mode....either that or trying to mimic a CBT's delay offsets with groups of 4 drivers at a time is just too course compared to geometric curvature..
thought i'd pass on some more observations...
I've been concentrating solely on the straight arrays, before I switch them both to CBTs.
Experiments have fallen into 3 categories:
a. corner vs out in the room,
b. different amounts of measurement averaging for the response that gets corrected, and
c. applying CBT-like shading
a. I wholeheartedly agree with all the folks who have built corner line arrays, and talk about the need for absorptive side wall treatments. There is plain and simple a whole lot of horizontal reflecting going on.....
nc525, I think you are going to be happy that you have smaller and excellent drivers to sit deeper in the corner, but I imagine you will still need plenty of good absorption.
That said, I can see that there may be people that prefer the more reflected sound, and might want to try diffusion panels instead. I think I mentioned earlier that the corner lines make great surrounds imo. If diffusion can dampen the comb filtering from side wall reflections, that would be cool to try i think. Any good, tall diffuser ideas, that could run parallel to the arrays, anybody?
I like the sound of the arrays moved out of the corners the best, more open sounding. Like Nvphotos said his 'White Whale' thread.
Measurements are much nicer to work with too, as I haven't yet used enough absorption in the corners to tame tuning needs.
I would keep the lines out of the corners but bass simply becomes too anemic. Sustained bass notes are ok, but dynamics aren't there.
x-max rears it's head when trying to eq bass response up to the needed level. RHosch, I think you are going to be happy with your ND105-8 inclusion.
b. averaging has really worked.
I've settled in on a 9 position average, comprised of on-axis and +/- approx 30 degrees, taken at about 1/4, 1/2, and 3/4ths of the array height. About 8ft mic distance....mic distance doesn't matter a lot to the average (within +/- a few feet.)
A raw average of the speaker out in the room is below, one that I'm currently using for the FIR file correction.
I'm only using full octave correction, which keeps sounding the best.
IIR would really work fine imo, no real need for FIR other than the ease of automatic EQ built in the FIR generator.
c.using the same EQ, the same FIR files, as the straight array has whether in the corner, or out in the room....
So far, amplitude shading and delay offsets have not yielded as nice a sound as the normal line, in either the corner or out in the room.
I'm thinking I need to build new EQ files for when in 'CBT' mode....either that or trying to mimic a CBT's delay offsets with groups of 4 drivers at a time is just too course compared to geometric curvature..
Attachments
Thanks, Mark. Timely post. Glue is drying on my first array as we speak. Planning on 2" of absorption, hope that is enough. I'm tight enough into the corner (and one out along the wall) that it should be.
Next big commit/decision for me is how to wire it. Allow for shading or not? 4 parallel strings of 8 four ohm drivers in series for a net of 8 ohms nominal or 8 groups of 4 in series brought out to barrier strips - like you did? Barrier strip is the easiest and most flexible. Can hide the barrier strips in the absorber...
Jack
Hi Jack, sorry for the typo on your nc535 name..
Wow, I wish i had any strong feelings to share on how worthwhile the shading effort may end up turning out.
I'm beginning to doubt shading can produce better results than running all drivers together. But gosh knows, I'm still in the infantile stage of learning.
I can say however, I'm loving the learning...seeing for myself...it helps me evaluate and see deeper into all the posts from guys previous builds/experiments/accomplishments.
I guess if I were in your shoes, I'd either skip shading entirely, or wire for as many groups as possible.
Hi TNT, I don't have a fully installed treatment yet.
Been trying lot's of things.
The straight columns are pretty close to the Murphy design, with 4.25" sides going into the room corners.
Simply flaring reflective wings to the wall produces deep unacceptable notches.
So I've been hanging all kinds of absorptive crap in all kinds of ways.
Best has been heavy quilted blankets draped on pieces of plywood about 3 ft wide, running height of line. Temporary as heck, but clearly a step in the right direction.
If I keep on with the corner tuning, next stop is a HVAC supply shop to get some rigid fiberglass.
Well most of the parts are here with the exception of the CNC'd router guide. Should be here Thursday. I hope to get a chance to visit a real lumber yard and cabinet supply shop tomorrow (vs big box store) and pick up some quality wood tomorrow. I have a 14 channel amp available and important parts for another 42 channels if needed. Thought about playing around with shading real time but only have 10 channels of dsp available. Will probably make front panel removeable though in case I want to tweak or rewire groupings.
OK - thanks! Are these drapes tucked all the way close to the drivers or sit a bit out on the wall?
//
//
Just a couple of notes, once you get the arrays out of the corner, into the room but have them close to side walls with absorption on that side wall you can keep the openness and retain bass response.
There is a marked difference between FIR and IIR. IIR changes the frequency response and it will be a hump or dip that stays the same over time. FIR can make changes over time. The difference is important, if you want to reconstruct the output to match an ideal IR shape. This is the goal of the software like Audiolense, Acourate and/or DRC-FIR. It needs a time window to try and correct the speaker, not the reflections.
Just mere changes in the properties of that time window can make huge perceptional changes without changing the target FR shape.
The straight array with shading isn't going to act as the CBT does unless you first remove it's reflections. Hence my warning about parallel planes to the array. Once you've done that, maybe you will get different or even better results more alike a true CBT.
I'd have my doubts about the rigid fiberglass panels and it's absorption at higher frequencies. I have no experience with it, but I kinda expect the fluffy materials to do better at the higher frequencies. It's hard to beat wool felt there, so that's on the outside of my damping panels with the fluffy fiberglass 'house insulation' in between (another layer of wool felt on the back). It serves another purpose to keep the nasty stuff inside if for whatever reason the panel cover might deteriorate over time. (which it did once we got a kitten 😉. The side wall cover is a terry cloth bed sheet which has made a nice climbing rack for her).
Hi Wesayso,
Yep, with regards to room placement out of the corners, but close to rear wall. I've already started the process of shuffling stuff in my largest room to try the arrays up against the the long wall, more towards the middle of it. That should greatly reduce corner reflections and get back some bass response.
I'd like to tighten up my vocabulary in this thread with regards to FIR and IIR.
By IIR, more precisely, I meant to be saying minimum phase, whether implemented by IIR or by FIR. Where freq and phase are invariably linked.
By FIR, I've been meaning the ability to adjust frequency and phase independently of each other, as you just hit on.
I think the full range TC9 array can be tuned solely with minimum phase.
As we know, any truly correctable frequency response deviation has an associated phase response deviation. Fix one, and you fix both. If we get frequency response smoothed out with minimum phase EQ, phase will be flat too.
Since these minimum phase frequency response deviations are the only ones that can be truly fixed,
any further phase correction is inappropriate other than for the goal of tuning to a specific small spot. But most all of us know all this...
I think what we forget is what that implies for a full range system using a single type driver (which is without any non-minimum phase crossovers).....
to me it implies not only can we use solely minimum phase, it also means that's all we should use.
So either conventional IIR EQ which is always minimum phase, or FIR with embedded minimum phase EQ, should work equally well and be fully sufficient.
I think the greatest effect I am hearing when I use a FIR filter to do the line array tuning has more to do with the finer minimum phase EQ adjustments being made by the FIR generator than anything else.
And again, heck, if phase is being adjusted ...should it be ???
Make sense?
Yep, re relections mucking up CBT simulation. Hope to try the emulation outside on the deck today.
Also hope to stand the straight line on top of a 4ft sub and see if I can steer the beam downward.
I've used the rigid compressed fiberglass with great success for higher frequencies.
I used to have tables that showed absorption coefficients for various materials though the frequency spectrum..I remember they looked good for the rigid fiberglass. I had a large, overly live room to tame.
I tucked them all the way back to where the speaker meets the wall, which still leaves the drivers protruding some 2-odd" from drapes.
If I remember correctly, I could move them about 6" away from the speaker, along the wall without much difference. Who knows if that was material dependent...