No offense taken here, this might help:
It seems to me you're just a bit disappointed I did not act on your suggestion to try something simple of which you thought might it be interesting enough to try.
Every one that posts here with a strong opinion can expect something back I guess. Were you around this thread when my array broke in half? Have you seen how many ideas and comments (both ugly and constructive) I got at that time? In the end, I was the one who had to decide which course to follow.
I try to be open minded, but also stated in other threads that it's not always easy. Especially because I'm not afraid to form an opinion of my own and go with that. I will not blindly follow anyone for that matter. I will think about it, think some more and make up my own mind. Every time! [Edit: But... I am not afraid to adjust my point of view when I've found enough evidence to convince me I should do that.]
So whatever any of them say on this subject. I absorb, let it cook a while and test it for my own purposes. Including the great Siegfried Linkwitz and Toole, Keely, Earl Geddes, Dave Smith and my personal favourites: Tom Danley and John Dunlavy. Many more that I don't mention. I read what they say. Cross compare, let it cook for a while and draw my own conclusions. If I didn't know any better it would seem logical to me that every one would do it just like that
.So your idea is registered. I already told you I would let it cook for a while:
After that you went on about your EQ-ed line arrays behaving like mine. I don't buy that. I will not even consider letting that cook inside my head.
Been there, seen the difference. The difference is huge! Seriously.
Last edited:
What problem
?I know, not the answer you're looking for. But really, I have found no reason to do that. I do know it actually brings more troubles than it could ever fix. Now adding some kind of super tweeter up high, that I may consider. As well as low support with (multiple) subwoofers. No direct need and certainly no funds to try it just now. But you've got to hear a setup like this to judge or believe it. I haven't found any reason to do as you suggest.
Last edited:
No offense taken here, this might help:
It seems to me you're just a bit disappointed I did not act on your suggestion to try something simple of which you thought might it be interesting enough to try.
Every one that posts here with a strong opinion can expect something back I guess. Were you around this thread when my array broke in half? Have you seen how many ideas and comments (both ugly and constructive) I got at that time? In the end, I was the one who had to decide which course to follow.
I try to be open minded, but also stated in other threads that it's not always easy. Especially because I'm not afraid to form an opinion of my own and go with that. I will not blindly follow anyone for that matter. I will think about it, think some more and make up my own mind. Every time! [Edit: But... I am not afraid to adjust my point of view when I've found enough evidence to convince me I should do that.]
So whatever any of them say on this subject. I absorb, let it cook a while and test it for my own purposes. Including the great Siegfried Linkwitz and Toole, Keely, Earl Geddes, Dave Smith and my personal favourites: Tom Danley and John Dunlavy. Many more that I don't mention. I read what they say. Cross compare, let it cook for a while and draw my own conclusions. If I didn't know any better it would seem logical to me that every one would do it just like that
So your idea is registered. I already told you I would let it cook for a while:
After that you went on about your EQ-ed line arrays behaving like mine. I don't buy that. I will not even consider letting that cook inside my head.
Been there, seen the difference. The difference is huge! Seriously.
Not that you weren't willing to experiment......I understand that's time consuming and can also be costly.
Just the reason on diffraction
I have no doubt your array sounds different than what I built........night and day?......black to white?.......that's certainly a question for me.
Again, you and others have shown that full range arrays are a viable DIY option for some folks and I applaud the efforts. Many, myself included have learned from your efforts and others undertaking similar designs. That's what this hobby and forum are all about!
This isn't a question for me anymore. I've heard both. It really is night and day for me. Only one way to find out
. The non FIR corrected version I ran for a while would turn me nuts over time and would make me consider other solutions.Though it started out as fun to listen to, it was way too wild and diverse to sound good on all the music I want to play. Pick certain songs and you'd be alright. But I want to play anything I like.
So I moved on with DSP (as originally intended back in 2011) and went nuts trying to find answers and share it here. I do believe I have found my personal best answer. The last measurements are from September (except for the ambient speakers) and only small changes have been made to my processing. There isn't a variable I did not play with in DRC-FIR. I cannot begin to tell you what I've tried.
My favourite band of all time is Led Zeppelin. I did not play anything by them until August 2015! The arrays had sound in December 2014...
I wanted to make sure I was happy with the performance before trying anything from them. Am I crazy or what!
My favourite band of all time is Led Zeppelin. I did not play anything by them until August 2015! The arrays had sound in December 2014...
I wanted to make sure I was happy with the performance before trying anything from them. Am I crazy or what!
I can totally relate to this. "In my time of dying" only gets played when all the tinkering is done and dusted. It does not deserve to be played on a less than perfect system.
I love Led Zepp!
Glad I'm not the only one around here that's crazy
One more try for nay sayers as to the power of DSP. No need to agree.
Hope I'm not going to regret this, as this isn't meant as an open invitation to start another senseless stream of posts.
OK here we go,
Posted already on another thread, Wavelet of left speaker, 28 ms window with pré EQ applied:
Basically a visual of the first wave front hitting the listening position (left speaker only)
Throw some DSP magic at it and it changes to this:
Notice the change? Yes, I did hear that difference... again, left speaker only.
Already shown the combined end result of Left and Right:
The summing left and right shows a remarkably coherent first wave front.
If you force this on your speakers with long window DSP or EQ settings it will sound like s##t. Believe me I tried! I try everything, remember?
Except on some material maybe. But it ain't right if you know what I mean.
You need to look what's happening and why, look at what you can fix and what needs to be left alone. It's very easy for me to fix the dip at 72 Hz with only a few setting changes. Believe me I tried that as well and it didn't work.
It did look good though, in plots and graphs. At first sight.
Learning why it didn't work was the valuable lesson. One every one of you need to figure out for yourself before making assumptions. Hope this helps...
But the sound will prove it to you. Force feeding DSP or EQ will get you nowhere fast!
It isn't coincidence I still have 2 dips in my right plot. I should fix that with damping, but definitely not with DSP. Even if it looked way better in plots if I did. If I did do that, I wouldn't have learned anything from these experiments.
DSP cannot fix everything. But it sure is a powerful tool, when used properly. Do I use it properly? I hope so. The music seems to tell me I do. Do I listen to judge? Of coarse. But the clues are there, if you look for them.
One more thing: I falsely stated -20 dB down earlier for the wavelet. I wish... my room isn't that good. -15 dB still looks ok but it looks like this is -10 dB.
My excuses, no intention to mislead any of you. Hence this correction here. I checked the REW plot and confirmed it's "only" 10 dB down on this plot. It is a living room after all.
One more try for nay sayers as to the power of DSP. No need to agree.
Hope I'm not going to regret this, as this isn't meant as an open invitation to start another senseless stream of posts.
OK here we go,
Posted already on another thread, Wavelet of left speaker, 28 ms window with pré EQ applied:
Basically a visual of the first wave front hitting the listening position (left speaker only)
Throw some DSP magic at it and it changes to this:
Notice the change? Yes, I did hear that difference... again, left speaker only.
Already shown the combined end result of Left and Right:
The summing left and right shows a remarkably coherent first wave front.
If you force this on your speakers with long window DSP or EQ settings it will sound like s##t. Believe me I tried! I try everything, remember?
Except on some material maybe. But it ain't right if you know what I mean.
You need to look what's happening and why, look at what you can fix and what needs to be left alone. It's very easy for me to fix the dip at 72 Hz with only a few setting changes. Believe me I tried that as well and it didn't work.
It did look good though, in plots and graphs. At first sight.
Learning why it didn't work was the valuable lesson. One every one of you need to figure out for yourself before making assumptions. Hope this helps...
But the sound will prove it to you. Force feeding DSP or EQ will get you nowhere fast!
It isn't coincidence I still have 2 dips in my right plot. I should fix that with damping, but definitely not with DSP. Even if it looked way better in plots if I did. If I did do that, I wouldn't have learned anything from these experiments.
DSP cannot fix everything. But it sure is a powerful tool, when used properly. Do I use it properly? I hope so. The music seems to tell me I do. Do I listen to judge? Of coarse. But the clues are there, if you look for them.
One more thing: I falsely stated -20 dB down earlier for the wavelet. I wish... my room isn't that good. -15 dB still looks ok but it looks like this is -10 dB.
My excuses, no intention to mislead any of you. Hence this correction here. I checked the REW plot and confirmed it's "only" 10 dB down on this plot. It is a living room after all.
Last edited:
To make up for that mistake, here's -15 dB (also already posted on another thread):
The only room I've seen that does -20 dB comfortably didn't look anything like a living room to me.
6 to 7 ms relates to the wall behind the microphone. The one with my huge (dampening) poster.
The only room I've seen that does -20 dB comfortably didn't look anything like a living room to me
.6 to 7 ms relates to the wall behind the microphone. The one with my huge (dampening) poster.
Last edited:
mayhem13,
TC9 have quite some breakup starting from around 9KHz. It may be easy to correct it (smoothen the SPL) with FIR and parametric EQ but the same may not be that easy with a fixed band 31-Graphic EQ. So first of all, was your frequency response really that smooth as Wesayso has shown? may be it was. In that case, what exactly was your experience with the HF region?
Statements like the arrays didn't work for you due to comb filtering is hugely misleading ( I don't mean intentional from your part) or confusing. If you say so, of course your arrays didn't work for you. Maybe full range drivers are not upto your taste. We all have different taste. Some hates Lowther, some swears by it, just an unrelated example. What I mean to ask you is that: what did you hear ( or did not) & why do you think its comb filtering ( given that generally human ear shouldn't be able to readily detect & confirm such levels of comb filtering in a smooth & corrected FR as in these arrays from around 2m+ ).
A subjective description rather than a one liner might help others confirm if such a problem exists in their own designs or if they have solved it long time ago.
Again, what was your driver to driver distance?
I myself have a line array & can't hear anything that I suspect as comb filtering. Maybe you can help me detect it over my next listening sessions if it exists ;P
TC9 have quite some breakup starting from around 9KHz. It may be easy to correct it (smoothen the SPL) with FIR and parametric EQ but the same may not be that easy with a fixed band 31-Graphic EQ. So first of all, was your frequency response really that smooth as Wesayso has shown? may be it was. In that case, what exactly was your experience with the HF region?
Statements like the arrays didn't work for you due to comb filtering is hugely misleading ( I don't mean intentional from your part) or confusing. If you say so, of course your arrays didn't work for you. Maybe full range drivers are not upto your taste. We all have different taste. Some hates Lowther, some swears by it, just an unrelated example. What I mean to ask you is that: what did you hear ( or did not) & why do you think its comb filtering ( given that generally human ear shouldn't be able to readily detect & confirm such levels of comb filtering in a smooth & corrected FR as in these arrays from around 2m+ ).
A subjective description rather than a one liner might help others confirm if such a problem exists in their own designs or if they have solved it long time ago.
Again, what was your driver to driver distance?
I myself have a line array & can't hear anything that I suspect as comb filtering. Maybe you can help me detect it over my next listening sessions if it exists ;P
Do you really want to learn to listen for it rockk19? Are you sure?
At one time I "recorded" my line arrays with a song. I did that using the left and right impulse as measured and convolved a song with those impulses. I played the result, first on headphones. After that, being bold, I played it back on my line array. Do you get what that means? I doubled the effects of my room. Very hard to "unlearn". Made me very conscious about my room's sound signature.
Just saying
At one time I "recorded" my line arrays with a song. I did that using the left and right impulse as measured and convolved a song with those impulses. I played the result, first on headphones. After that, being bold, I played it back on my line array. Do you get what that means? I doubled the effects of my room. Very hard to "unlearn". Made me very conscious about my room's sound signature.
Just saying
Last edited:
Thats a bold experiment indeed which I would not dare to do with my current room
Regarding comb filtering, I really doubt if we can hear such high Q dips so closely placed over the HF region. If comb filtering is audible, I am already enjoying it. So, if I would have to acknowledge it, I'll learn something new
Regarding comb filtering, I really doubt if we can hear such high Q dips so closely placed over the HF region. If comb filtering is audible, I am already enjoying it. So, if I would have to acknowledge it, I'll learn something new
Sure.......but you'll need to open yourself to the arguement that measurements of simple tone bursts and 5 harmonics aren't nearly a true representations of complex music.
So we take what we know to be true........above a wavelength of the drivers radiating diameter the driver begins to beam. This beaming begins where the signal is radiated not pistonic but from the center of the driver, increasing the effective 'center to center' of the drivers distance and creating a forward lobe in the wave front. In a typical two or three way, you'll have as many or possibly a few more lobes due to baffle layout, etc. In the case of a 24 element line array, you'll have at least 24 of these lobes of varying amplitude due to the relative distance from the center of each to the listeners ears. Now when we measure for response.....on, off axis, whatever........we're basically measuring each tone in relationship to a single moment in time where the driver is directed to operate from steady state to excited and back again........very linear in amplitude. So when we examine the lobing from a measured response, polar, etc.....we 'see' the wavefront as a gentle ebb and flow so to speak which gives way to the notion that multiple drivers when combined in a line create a single cylindrical wave............and for the purpose of that measure it's a fair representation of what appears to the mic and our ears when actually it's comprised of 24 spherical waves or balls each with its own phase interaction at varying points in time all relative to distance to the listener or mic. The arguement from the line array camp is that at a distance, all of these cancellations 'smooth out' and you're presented with a relatively flat wavefront........get up closer to the array and move up and down and you can hear the lobes, farther away less so. In all fairness to the fullrange array camp, the effect is even less so with smaller diameter drivers for freq below the drivers relative diameter as it relates to the wavelength. To this point, we have a generalization of line array theory, yes?
mayhem13,
Thank you for your response.
I have a 25 drivers ( centre to centre distance being 9cms) per channel arrays. Yes, of course comb filtering exist. Its clearly evident in 1/48 graphs measured at a distance of 3m. At a distance of around 1 metre from the arrays if I move my ears up and down, I can hear comb filtering effects. The effect decreases as the distance increases. At 2m, its barely detectable even when playing sine wave tones. At 3m, I could not detect any comb filtering with sine wave tones. I tried this over and over.
And yes, I agree with your line array theory generalization.
Thank you for your response.
I have a 25 drivers ( centre to centre distance being 9cms) per channel arrays. Yes, of course comb filtering exist. Its clearly evident in 1/48 graphs measured at a distance of 3m. At a distance of around 1 metre from the arrays if I move my ears up and down, I can hear comb filtering effects. The effect decreases as the distance increases. At 2m, its barely detectable even when playing sine wave tones. At 3m, I could not detect any comb filtering with sine wave tones. I tried this over and over.
And yes, I agree with your line array theory generalization.
Do you really want to learn to listen for it rockk19? Are you sure?
At one time I "recorded" my line arrays with a song. I did that using the left and right impulse as measured and convolved a song with those impulses. I played the result, first on headphones. After that, being bold, I played it back on my line array. Do you get what that means? I doubled the effects of my room. Very hard to "unlearn". Made me very conscious about my room's sound signature.
Just saying
I'm glad you added the influence the room has as we'll find out just how detrimental beaming can be as opposed to constant directivity.
Not seeing how this is relevant to this discussion. The simple tone burst is used for a very good reason. The actual wave shapes of songs doesn't have anything to do with that. I fail to see why that so called argument has any merit here. If you want to discuss that, please open another thread.
Actually my interpretation of what happens is a bit more complicated than this more generalized version but let's not go there now. Unless there is more demand for it. Read the entire thread and you'll find clues.
But, none of this explains the inability to play a certain kind of music. I am pretty sure it's not the above explanation that is responsible for the bad result.
But often the line array's comb filtering does get blamed for it.
Rockk19, look at what I replied to Halair, if you want to improve your EQ results I'm pretty sure you could get helpful info in that thread.
Plot of the FIR corrected midrange, with comb filter evidence showing as a wavy shape of the frequency response.
Correct your arrays the best you can and look at the waterfall plot with the above settings (or close to it).
Last edited:
