Kindhornman said:
@Kindhornman,
only omni is omni.
Some of the "radial" designs may even reach rather high DI due to beaming in the vertical plane.
Thus you have sound power decreasing with frequency in most "radial" designs like you see it too in almost all "conventional" speaker designs (now more looking at wavelength in relation to driver size).
Same with "radial reflectors" / "radial horns" e.g. Duevel ...
http://www.tnt-audio.com/gif/enterprise_woofer_graph.gif
Catastrophic inconsistency here in vertical radiation and also high DI in highs and especially at top end.
Last edited:
@FrankWW
I do not think so:
We are talking "direct sound (on axis) vs. sound power" here in real loudspeaker systems, even those "commonly believed to be omni".
This is the stuff (e.g. real transducer behaviour) that the "impressive inroom measurements" (by DRC) are made of in reality:
No one has shown up to now, how the direct sound looks like, when the inroom curve has massively been altered to look smooth.
In case you had a poor speaker anyhow and the anechoic response was ragged already, you have of course only little to loose when applying "room correction" (as it is called) even in the mid-to-high range.
Last edited:
@Pano:
We need to know, whether the flat and smooth anechoic response even of a decent speaker may stay as flat and smooth after applying DRC Product XY at > 400Hz which optimizes for "sound power" (averaged over positions) or "inroom response" (at listening seat) .
At least this is what i am interested in. But i know the answer in advance:
"Leave a decent speaker alone > 400Hz and try your DRC System with the crap available on the market:
Then you won't get audible worsening. And due to improvements at the low end (where "room compensation" can have a meaning), you will feel better afterwards anyhow ..."
We need to know, whether the flat and smooth anechoic response even of a decent speaker may stay as flat and smooth after applying DRC Product XY at > 400Hz which optimizes for "sound power" (averaged over positions) or "inroom response" (at listening seat) .
At least this is what i am interested in. But i know the answer in advance:
"Leave a decent speaker alone > 400Hz and try your DRC System with the crap available on the market:
Then you won't get audible worsening. And due to improvements at the low end (where "room compensation" can have a meaning), you will feel better afterwards anyhow ..."
Last edited:
Lets start with the "primary" listening Window and frequency range say >400Hz:
Directly "on axis" and some just small "off axis" angles (anechoic, very little smoothing).
With vs. without "room correction".
Important is using a speaker, which has a flat and smooth anechoic response in the primary listening window already, which is also fairly consistent under angles.
Otherwise it is no real fun ...
Directly "on axis" and some just small "off axis" angles (anechoic, very little smoothing).
With vs. without "room correction".
Important is using a speaker, which has a flat and smooth anechoic response in the primary listening window already, which is also fairly consistent under angles.
Otherwise it is no real fun ...
Last edited:
Pano,
Myself, definitely not speaking for LineArray, Wonder what is happening off axis that is not also affected on axis if you do some of this correction? I don't see how you can do one without affecting the other? Total in-room power response I understand, and changing that I also can follow, but if you have a speaker that measures correctly and doesn't need any on axis changes what happens to that when you try to correct for the room response?
Myself, definitely not speaking for LineArray, Wonder what is happening off axis that is not also affected on axis if you do some of this correction? I don't see how you can do one without affecting the other? Total in-room power response I understand, and changing that I also can follow, but if you have a speaker that measures correctly and doesn't need any on axis changes what happens to that when you try to correct for the room response?
OK, I was thinking right.
Fortunately I have a hypo-echoic chamber to test in. (a lava cave under my house)
Unfortunately my speakers aren't all that flat and smooth.
And I'm leaving tonight for a 2 week business trip 4900 miles away. So no new measurements for a fortnight.
But I'm sure that several people here can show on and off axis results for their rooms and speakers with DRC. I'll do so when I return, but by then the dust will have settled. I'll follow here to see what plots get posted. I'm interested to see!
Fortunately I have a hypo-echoic chamber to test in. (a lava cave under my house)
Unfortunately my speakers aren't all that flat and smooth.
And I'm leaving tonight for a 2 week business trip 4900 miles away. So no new measurements for a fortnight.
But I'm sure that several people here can show on and off axis results for their rooms and speakers with DRC. I'll do so when I return, but by then the dust will have settled. I'll follow here to see what plots get posted. I'm interested to see!
Just for Reference
Below shows the FR of my left speaker - at the listening position - which is 25' (7.6m) from the speaker. It is also off axis, being at the point of the stereo triangle, the center position.
Blue is the uncorrected response, 1/12 octave smoothed. You can see the baffle peak and some other nasties. Red is with DRC in 4 cycle mode, 1/12 octave smoothed.
I will get some measurements away from that position as soon as I return. I am curious to see them.
Below shows the FR of my left speaker - at the listening position - which is 25' (7.6m) from the speaker. It is also off axis, being at the point of the stereo triangle, the center position.
Blue is the uncorrected response, 1/12 octave smoothed. You can see the baffle peak and some other nasties. Red is with DRC in 4 cycle mode, 1/12 octave smoothed.
I will get some measurements away from that position as soon as I return. I am curious to see them.
Attachments
Hopefully on topic and of interest?
OP, Please do tell me to shut up if you feel I am going off topic!
But I think Pnix questions and points deserve a response.
First let me re "transmit" my point....Its simple biology of how our ear / brain (HAS) decodes sounds....It is not disputed by Dr Toole, Daisuke or Lipshitz ( the man not a sub group of phase doubters...!) our Human Auditory System (HAS) uses time based factors to locate sounds and frequency based factors to identify what sounds are. Fact, no grey area or debate.
I posted a 2014 paper by John Watkinson which details my points about HAS....
Your link to "Mark from the Audioholic forum" giving his opinion on a selection of 15 to 50 year old studies has no relevance to my points or Johns research.....
On a wider point.....
Whilst much of Dr Tool's research (all funded by his employer, Harman international.....Largely dependant on sales of passive crossovers & ported boxes!) is of interest to DIY speaker builders, you wont find any current research (less than 5 years old) on HAS in his papers.....The results wont favour Harman Internationals products.....
Yes I am synical about any commercially funded research!
Lastly, on an even wider point....
Moores Law is our friend, research done in the last 5 years or so has 1,000 times the computing power available than research in the 80's....
I am not dismissing the "old school" ....Penicillin was discovered before computers and it still works!
But my point is modern medicine, engineering, acoustics, aviation, physics , chemistry.....Every research discipline on Earth is advancing and dispelling old myths and discovering new frontiers .....Loudspeaker design (as John states in his 2014 paper) is dominated by geriatric thinking and dinosaur incumbent manufacturers.....
We in the DIY world should be free thinkers and open to new ideas, not still donning our cardigans slipping out to the garden shed to read the 2016 edition of "loudspeaker design cookbook" which still promotes the 1916 "two way 8 inch bass mid & dome tweeter in a ported box"....Ok now in 2016 its got diamond and ceramic coatings and "advanced" motor design....Give me a break!!
Sorry to ramble but its all from the heart and based on current research....Hope its of interest and hey, happy listening whether you agree or not!
Cheers
Derek.
OP, Please do tell me to shut up if you feel I am going off topic!
But I think Pnix questions and points deserve a response.
First let me re "transmit" my point....Its simple biology of how our ear / brain (HAS) decodes sounds....It is not disputed by Dr Toole, Daisuke or Lipshitz ( the man not a sub group of phase doubters...
!) our Human Auditory System (HAS) uses time based factors to locate sounds and frequency based factors to identify what sounds are. Fact, no grey area or debate.I posted a 2014 paper by John Watkinson which details my points about HAS....
Your link to "Mark from the Audioholic forum" giving his opinion on a selection of 15 to 50 year old studies has no relevance to my points or Johns research.....
On a wider point.....
Whilst much of Dr Tool's research (all funded by his employer, Harman international.....Largely dependant on sales of passive crossovers & ported boxes!) is of interest to DIY speaker builders, you wont find any current research (less than 5 years old) on HAS in his papers.....The results wont favour Harman Internationals products.....
Yes I am synical about any commercially funded research!
Lastly, on an even wider point....
Moores Law is our friend, research done in the last 5 years or so has 1,000 times the computing power available than research in the 80's....
I am not dismissing the "old school" ....Penicillin was discovered before computers and it still works!
But my point is modern medicine, engineering, acoustics, aviation, physics , chemistry.....Every research discipline on Earth is advancing and dispelling old myths and discovering new frontiers .....Loudspeaker design (as John states in his 2014 paper) is dominated by geriatric thinking and dinosaur incumbent manufacturers.....
We in the DIY world should be free thinkers and open to new ideas, not still donning our cardigans slipping out to the garden shed to read the 2016 edition of "loudspeaker design cookbook" which still promotes the 1916 "two way 8 inch bass mid & dome tweeter in a ported box"....Ok now in 2016 its got diamond and ceramic coatings and "advanced" motor design....Give me a break!!
Sorry to ramble but its all from the heart and based on current research....Hope its of interest and hey, happy listening whether you agree or not!
Cheers
Derek.
On the same page!
Hiya Kindhornman,
Thanks for your reply and glad we are on the same page, I think you are on the right tracks and I really like the way you listen to a very wide variety of opinions and do your market research....I'm sure you will do well with your commercial product and I wish you good luck!
I'm totally with you on the " small single driver full range speakers suck", well ok you put it far more diplomatically than I do!
But one key to massively improving a small drivers performance is to use a big cone up to approx 300Hz to 400Hz and then cross in the little guy.
Currently we are running 10, 12 and 15 inch bass / mids crossed over to our new 4.5 inch RED (Rapid Energy Decay) driver we developed.....We ended up changing the BMR so much it was no longer functioning as a Bending Mode Radiator!
All the best
Derek.
Hiya Kindhornman,
Thanks for your reply and glad we are on the same page, I think you are on the right tracks and I really like the way you listen to a very wide variety of opinions and do your market research....I'm sure you will do well with your commercial product and I wish you good luck!
I'm totally with you on the " small single driver full range speakers suck", well ok you put it far more diplomatically than I do!
But one key to massively improving a small drivers performance is to use a big cone up to approx 300Hz to 400Hz and then cross in the little guy.
Currently we are running 10, 12 and 15 inch bass / mids crossed over to our new 4.5 inch RED (Rapid Energy Decay) driver we developed.....We ended up changing the BMR so much it was no longer functioning as a Bending Mode Radiator!
All the best
Derek.
OP, Please do tell me to shut up if you feel I am going off topic!
But I think Pnix questions and points deserve a response.
First let me re "transmit" my point....Its simple biology of how our ear / brain (HAS) decodes sounds....It is not disputed by Dr Toole, Daisuke or Lipshitz ( the man not a sub group of phase doubters...
I posted a 2014 paper by John Watkinson which details my points about HAS....
Your link to "Mark from the Audioholic forum" giving his opinion on a selection of 15 to 50 year old studies has no relevance to my points or Johns research.....
On a wider point.....
Whilst much of Dr Tool's research (all funded by his employer, Harman international.....Largely dependant on sales of passive crossovers & ported boxes!) is of interest to DIY speaker builders, you wont find any current research (less than 5 years old) on HAS in his papers.....The results wont favour Harman Internationals products.....
Yes I am synical about any commercially funded research!
Lastly, on an even wider point....
Moores Law is our friend, research done in the last 5 years or so has 1,000 times the computing power available than research in the 80's....
I am not dismissing the "old school" ....Penicillin was discovered before computers and it still works!
But my point is modern medicine, engineering, acoustics, aviation, physics , chemistry.....Every research discipline on Earth is advancing and dispelling old myths and discovering new frontiers .....Loudspeaker design (as John states in his 2014 paper) is dominated by geriatric thinking and dinosaur incumbent manufacturers.....
We in the DIY world should be free thinkers and open to new ideas, not still donning our cardigans slipping out to the garden shed to read the 2016 edition of "loudspeaker design cookbook" which still promotes the 1916 "two way 8 inch bass mid & dome tweeter in a ported box"....Ok now in 2016 its got diamond and ceramic coatings and "advanced" motor design....Give me a break!!
Sorry to ramble but its all from the heart and based on current research....Hope its of interest and hey, happy listening whether you agree or not!
Cheers
Derek.
Frustrating. You obviously missed my point completely. I've never disputed that small timing errors can lead to huge localization errors or why do you think I've posted this? But localization is different from audibility of monaural phase errors. Here huge phase distortions are largely inaudible. In other words you can introduce large phase errors and have no localization issues IF the phase errors are THE SAME in BOTH stereo channels.
In order to put everything into context and move forward one needs to look at how stereo works. I recommend reading Blumlein's original "stereo" patent or Lipshitz, "Stereo microphone techniques: Are the purists wrong?"
In a stereo field intensity differences in the recording translate to phase differences at the ears hence frequency response is important to get the time domain right.
Simple slogans like "phase is more important than frequency response" might help sales but don't help understanding things any better. The same is true for "my guru knows more than your guru".
P.S. I urge you to read Schnupp, "Auditory Neuroscience" and you'll notice that what you think is just black and white becomes a vast grey area.
Last edited:
Pano,
i know you being on a longer journey right now, but let me add something also for others, who may post measurements ...
Using an "N-Cycle mode" (and having N small) is at least going to approximate what i was proposing as being useful:
Going towards a "speaker correction only" at say >400Hz (or even prefereably lower according to setup) and correct just for the anechoic response.
Having an interval of 4 cycles to be averaged, the pathlength of sound ( N x Wavelength) will be
approximately:
400Hz....3,4m
800Hz....1,7m
1600Hz...0,9m
This means at 1600Hz in many room steups just bottom reflection an a reflection from a close sidewall might come into play and >3KHz in many cases there will be no reflection from "structures away/outside the speaker" having influence, so we are arrived at "pretty anechoic" condions then and wat whe do is mainly "speaker correction".
But i'd like to remind posters and readers kindly:This is not the method proposed by the OP.
Last edited:
Pnix - May I link this post for you from member kessito: http://www.diyaudio.com/forums/multi-way/283068-mini-synergy-horn-experiment-25.html#post4554771
I share the above view and you can find that if you look at my posts on the subject. Kessito worded it a lot better than I could.
For that to happen you definitely need to keep the room in mind though. It has been stated early reflections enhance spaciousness. Very true. But they do mess with the goal of getting the phase right if they are too early.
Though often we speak of linear phase, I've come to the conclusion I need minimum phase throughout the bandwidth of my speakers. Meaning no phase wrap of crossovers.
You are right that you still get imaging and placement as long as both speakers give the same phase rotation. But it isn't the same sound.
There are differences in sounds, the one that follows minimum phase throughout the entire bandwidth sounds more real, subjectively.
I have experimented with true linear phase. As if my speakers had output down to DC. A true flat line group delay to about 20 Hz. To me that sounds hasty, made me restless. Did not sound natural. But phase following the frequency response sounds natural and right.
I've done a lot of experiments only to please myself. You cannot convince me otherwise with papers from others as I've heard that difference in my room. I prefer the minimum phase without phase rotation. The timing is right for the signal I hear.
If I had to chose one thing, phase or frequency, to be right I'd pick Frequency. I just prefer to have both right.
In my 2D APL_TDA graphs you can see what I mean:
You do see that phase isn't linear here, it is minimum phase throughout the bandwidth.
The above graph is made with the demo version of APL_TDA, which hopefully brings this thread on topic again.
No amount of digital signal processing could have gotten me this without using room treatment. I treated the first reflections. With the treatment I did loose a sense of spaciousness. Adding later reflections, between 15 ms to about 30 ms can give you back that sense of spaciousness without being detrimental to imaging or phase behaviour. Still it is not perceived as echo or a separate sound source.
This trick is often called a Haas kicker.
While Toole embraces early reflections to get that sense of spaciousness I tend to side up with Dr. Geddes to avoid them as much as possible.
Introduce them later than the mentioned 15 ms (still may be called early) and they make me "believe" I'm in a bigger room.
Even Linkwitz played with that: http://www.linkwitzlab.com/surround_system.htm
My implementation is similar, though I do this within JRiver. Look in this post: http://www.diyaudio.com/forums/full-range/273971-group-delay-questions-analysis-50.html#post4581047 for another example of a Haas kicker but done in a passive way. This is a result of an excellent room with speakers that have second order filters and ported mains and subs. Still the hand off of the signal is exceptional. And much better room behaviour that I will ever get in my living room.
You can clearly see the Haas kicker at ~24 ms. It is there on purpose.
In my measurement it was turned off as I "fake" that signal, much like Linkwitz described.
I share the above view and you can find that if you look at my posts on the subject. Kessito worded it a lot better than I could.
For that to happen you definitely need to keep the room in mind though. It has been stated early reflections enhance spaciousness. Very true. But they do mess with the goal of getting the phase right if they are too early.
Though often we speak of linear phase, I've come to the conclusion I need minimum phase throughout the bandwidth of my speakers. Meaning no phase wrap of crossovers.
You are right that you still get imaging and placement as long as both speakers give the same phase rotation. But it isn't the same sound.
There are differences in sounds, the one that follows minimum phase throughout the entire bandwidth sounds more real, subjectively.
I have experimented with true linear phase. As if my speakers had output down to DC. A true flat line group delay to about 20 Hz. To me that sounds hasty, made me restless. Did not sound natural. But phase following the frequency response sounds natural and right.
I've done a lot of experiments only to please myself. You cannot convince me otherwise with papers from others as I've heard that difference in my room. I prefer the minimum phase without phase rotation. The timing is right for the signal I hear.
If I had to chose one thing, phase or frequency, to be right I'd pick Frequency. I just prefer to have both right.
In my 2D APL_TDA graphs you can see what I mean:
You do see that phase isn't linear here, it is minimum phase throughout the bandwidth.
The above graph is made with the demo version of APL_TDA, which hopefully brings this thread on topic again.
No amount of digital signal processing could have gotten me this without using room treatment. I treated the first reflections. With the treatment I did loose a sense of spaciousness. Adding later reflections, between 15 ms to about 30 ms can give you back that sense of spaciousness without being detrimental to imaging or phase behaviour. Still it is not perceived as echo or a separate sound source.
This trick is often called a Haas kicker.
While Toole embraces early reflections to get that sense of spaciousness I tend to side up with Dr. Geddes to avoid them as much as possible.
Introduce them later than the mentioned 15 ms (still may be called early) and they make me "believe" I'm in a bigger room.
Even Linkwitz played with that: http://www.linkwitzlab.com/surround_system.htm
My implementation is similar, though I do this within JRiver. Look in this post: http://www.diyaudio.com/forums/full-range/273971-group-delay-questions-analysis-50.html#post4581047 for another example of a Haas kicker but done in a passive way. This is a result of an excellent room with speakers that have second order filters and ported mains and subs. Still the hand off of the signal is exceptional. And much better room behaviour that I will ever get in my living room.
You can clearly see the Haas kicker at ~24 ms. It is there on purpose.
In my measurement it was turned off as I "fake" that signal, much like Linkwitz described.
Last edited:
LineArray, to add something useful to this thread, why don't you post an APL_TDA plot of those wondrous marvels you keep telling us about.
APL_TDA is only a measurement suite... no need to fear it. The DSP suite is separate. It will display the time domain behaviour of your speakers.
Plus it would be entirely on topic for a change...
Using the demo version and a screen grab would suffice. So far all you do is talk and ask others for proof. Why not be nice and play along? You'd only be measuring, not manipulating your signal.
Just to be clear on this. I did not use the OP's software suite to "manipulate" my speakers. I merely measured them with his measuring suite.
I'm also responsible for talking member jim1961 into making his plot as I was dying to see what an excellent room would look like and it helped his quest. Win-win if you ask me. I'd love to see more people actually try instead of condemn this software. It's only a measuring suite bringing insight to what is happening in your room with your speakers.
Another spin-off thread was started in the Multi Way area of this forum with at least one other member playing along: http://www.diyaudio.com/forums/multi-way/285628-apl-tda-acoustic-loudspeaker-analyzing-software.html
(but that's probably how you found this thread?)
Last edited: