Hi Elias,
Thanks for sharing these
2 quesions:
1. How is the width?
2. Can you try making it a flooder (just put it on floor, flip the front driver to the top)?
The prototype box width is 22 cm.
I did quickly try aiming the box at the ceiling located at the ear height, but I'm not sure what design feature it could add ?
- Elias
Of course it is not perfect yet.. My first proto is made of cardboard and 10€ fullrange elements
- Elias
I have a feeling you will make it perfect...
Card board box coming out this weekend to do an exact replica..... But with chr70 drivers. I can compare to my Gerzon setup... In fact... I will use the same box for both methods. To much fun eh...
This thread is growing really fast!
Er.... That question followed your post about listening experiences (mostly in the center imaging issue), so, about the width, I meant the lateral dimension of sound stage, not the box (so it's a how, not a what)
About upfiring, this is what I thought:
When the signal is fully panned to one side, say, full left, there'd be a half of sound emitted from the center driver which is pointing at the listener. That would dominate the perception of source location (which is at the center now, of course). But it shouldn't be. The sound should appear at the far left.
In a brief experiment of (stereo) flooders I've done, I found there's much less 'hole in the middle' drawback of ordinary setup, and a much better continuity of the full 'panaroma'. Simply put, the speakers disappear.
So I thought, maybe upfiring SSS can also benefit from this -- less direct source localization, so it disappears when the sounds coming from far left or far right
No?
Er.... That question followed your post about listening experiences (mostly in the center imaging issue), so, about the width, I meant the lateral dimension of sound stage, not the box (so it's a how, not a what)
About upfiring, this is what I thought:
When the signal is fully panned to one side, say, full left, there'd be a half of sound emitted from the center driver which is pointing at the listener. That would dominate the perception of source location (which is at the center now, of course). But it shouldn't be. The sound should appear at the far left.
In a brief experiment of (stereo) flooders I've done, I found there's much less 'hole in the middle' drawback of ordinary setup, and a much better continuity of the full 'panaroma'. Simply put, the speakers disappear.
So I thought, maybe upfiring SSS can also benefit from this -- less direct source localization, so it disappears when the sounds coming from far left or far right
No?
We still do not know what is the audible advantage (not theoretical - audible) advantage of SSS cardboard over pillowed cardboard, what did sound wrong/worse in case of the latter?
so far I am not at all convinced that this matrixing can provide any advantage apart from purely psychological - that is looking more serious/sophisticated/cool than a pillow in front of the speaker
OTOH it's possible that it creates problems (Markus points)
so far I am not at all convinced that this matrixing can provide any advantage apart from purely psychological - that is looking more serious/sophisticated/cool than a pillow in front of the speaker
OTOH it's possible that it creates problems (Markus points)
The 'stereo to 3-ch matrix' on my 3 separate speakers system works pretty good, though. It sort of making up the weak spot of ordinary stereo. The center channel ensures excellent center image, while the other 2 provide a very wide stage.
sure - but I am not comparing apples and oranges, my questions/reservations concern strictly and only the "SSS cardboard" vs "pillowed cardboard"
Last edited:
So it seems usually the extreme case will be L = 1, R = 0. And they will be in same phase.
But, what is the frequency range panning is performed ? Can we assume some low freq limit for amplitude panning is commonly used studio practises, say > 100Hz ? > 200 Hz ? > 400 Hz ?
- Elias
Most recordings are made of monophonic sounds that are level panned over the whole frequency range: The signal of a single channel is simply split into two channels (left and right). The level in each of the resulting two channels is determined by the panpot.
The only time err time comes into play is when working with stereo mic'd tracks (or effects). But even then level panpots are commonly used on these two stereo tracks.
Last edited:
so - was there any practical - audible - problem with pillowed cardboard or is it just "theoretical" problem or a problem of silly looks?
I've listed some of the problems earlier in this thread.
Additional problem with The Cardboard was overly spacious sound making an unnatural bathroom effect noticeable especially on studio recordings, but not so much on natural ambiance recordings. So, it did not perform with all the recordings.
Now, with the SSS this is partly solved. SSS sounds better with studio recordings, and with natural ambiance recordings not worse than The Cardboard.
If you need more reassurances, it may be best to try DIY ! My humble suggestion is you put your three way flooder project on hold and start building SSS
- Elias
where?
interesting, You have never mentioned such problems with it before
interesting, I wonder how much better and - especially - why?
If you need more reassurances, it may be best to try DIY ! My humble suggestion is you put your three way flooder project on hold and start building SSS
- Elias
perhaps I will... as soon as You give a DIY try to a flooder in Beveridge placement
I asked You first, many months ago - so it is Your turn and only after then mine
To address the potential problem of tonal balance of the negative vector dipole effect if left uncompensated, now there is a solution !
The signals for the three elements of the SSS are as before:
Lo = L - xR
Ro = R - xL
Co = xL + xR
where parameter x is between:
x = 0...1
Rearranging the terms:
L - xR = (1 - x)L + x(L - R)
R - xL = (1 - x)R + x(R - L)
xL + xR = x(L + R)
And utilising definition of MS stereo:
M = L + R
S = L - R
The three element signals of the SSS can be written as:
Lo = (1 - x)L + xS
Ro = (1 - x)R - xS
Co = xM
Heureka !
Because L and R remain in same phase in typical mixing practises, the only term causing negative vector is the S signal. S signal forms a sideways dipole in the SSS. This can be compensated easily ! I've drawn one possible configuration below:
Stereo signal is fed into MS matrix to form M and S signals. Then dipole correction is applied to S signal, and compensatory phase correction is applied to M signal. Then using inverse MS matrix signal is transformed back to L and R format which is fed into SSS speaker.
This is expected to correct any possible issues in frequency dependent tonal balance depending on input signals
- Elias
The signals for the three elements of the SSS are as before:
Lo = L - xR
Ro = R - xL
Co = xL + xR
where parameter x is between:
x = 0...1
Rearranging the terms:
L - xR = (1 - x)L + x(L - R)
R - xL = (1 - x)R + x(R - L)
xL + xR = x(L + R)
And utilising definition of MS stereo:
M = L + R
S = L - R
The three element signals of the SSS can be written as:
Lo = (1 - x)L + xS
Ro = (1 - x)R - xS
Co = xM
Heureka !
Because L and R remain in same phase in typical mixing practises, the only term causing negative vector is the S signal. S signal forms a sideways dipole in the SSS. This can be compensated easily ! I've drawn one possible configuration below:
Stereo signal is fed into MS matrix to form M and S signals. Then dipole correction is applied to S signal, and compensatory phase correction is applied to M signal. Then using inverse MS matrix signal is transformed back to L and R format which is fed into SSS speaker.
An externally hosted image should be here but it was not working when we last tested it.
This is expected to correct any possible issues in frequency dependent tonal balance depending on input signals
- Elias
I'm in the understanding that in most recordings bass is mono, but then it must be because bass instruments are panned in the center and not because of frequency dependent panning. That is fine for SSS !
Ortoperspekta is a three speaker concept where speakers are surrounding the listener, not a single speaker concept.
SALORA orthoperspecta
An externally hosted image should be here but it was not working when we last tested it.
- Elias
Er.... That question followed your post about listening experiences (mostly in the center imaging issue), so, about the width, I meant the lateral dimension of sound stage, not the box (so it's a how, not a what)
I see
Well, now that you know the box dimension you can start building a prototype too More about the imaging width below:
This is a very good point ! I have encountered some hints of this happening too.
The book solution to this could be implementation of Gerzon psychoacoustic shelving filters for center and side channels ! Gerzon aka Ambisonics and Trifield filters !
In his papers he suggest a transition happening around 5kHz. According to his theory there should be different spatial energy distribution below and above the transition freq.
I would start by introducing a low pass shelving filter for the center channel, and a high pass shelving filter for the side channels. Corner freq 5kHz. Maybe start with 3dB shelving.
In a brief experiment of (stereo) flooders I've done, I found there's much less 'hole in the middle' drawback of ordinary setup, and a much better continuity of the full 'panaroma'. Simply put, the speakers disappear.
So I thought, maybe upfiring SSS can also benefit from this -- less direct source localization, so it disappears when the sounds coming from far left or far right
No?
Yes, floodering makes the speakers to disappear. But pretty much same thing happends to the imaging. This is based on my limited experience with flooders.
While ceiling firing can provide benefit in some freq range (e.g. tweeters) it may not be a good idea to use it over whole audio freq band in my opinion.
- Elias
Yes I have, these comment apply to The Cardboard:
"On other aspects of The Cardboard I must state that it does not sound good with non natural music. Take some tracks of house music, terrible. Try some tunes of Hed Kandi disco, beyond bearable limit..."
Much better
Actually, so much better it justifies the potentially more complex design.This is because (I think) with SSS the central direct sound is explicitly defined, full spectrum. With The Cardboard, and especially with the pillow trick, direct sound is absent. As said direct sound blocking method can work very well only if recording is made in natural ambience, but studio records do not perform well at all but are overly spacious.
perhaps I will... as soon as You give a DIY try to a flooder in Beveridge placement
I asked You first, many months ago - so it is Your turn and only after then mine
I have done it already, many months ago
The results are proudly presented in the oustanding thread: see post #144http://www.diyaudio.com/forums/multi-way/197396-unconventional-techniques-achieving-oustanding-stereo-imaging.html
Now it's your turn
- Elias
where is the bathroom effect mentioned? terrible can mean anything
what causal link can there be that having explicitly defined center can cure bathroom-like aspects of sound?
I don't get it
I have done it already, many months ago
no mention of Beveridge placement in that post, besides if in Your experience
then You clearly did something wrong with the setup and Your experience is very limited indeed
check out Radugazon's flooder tests
Now it's your turn
- Elias
oh stop it Elias
I was dying my stereolits in 2005-07 and You would have never discovered the thing without my numerous posts on this forum, there were times when You were very.. uhm... sceptical and I was very patient but oh please don't thank me... oh ok!, anyway, You're welcome
Last edited: