I see, for me a wide (front wall dissappears) "window" into a room of spaciousness IS spaciousness
and immersion into a space of sound surrounding me like in a church or a concert hall for me is just surround sound
I believe that surround sound requires some form of rear speakers (hardware) reproducing appropriately processed signals (software)
perhaps as a hardware just one additional rear back-to-back would do
see:
http://www.diyaudio.com/forums/multi-way/10962-stereolith-loudspeakers-question-29.html#post2521656
appropriate software seems to be more problematic
but personally I am not that much interested in this sort of experience of total you are there
I just don't like sounds jumping at me from behind
plus at the real events when I am at the audience in the concert hall the surround sound is essentially disturbing coughs and so on
not much music is coming from behind, human bodies, not to mention upholstered seats, are pretty absorptive
there are exceptions of course as an echoey church for example
best,
graaf
Last edited:
as to Radugazons beautiful test machine I would say that back-to-back section is too narrow to work
I don't have any explanation but my 13 cm wide back-to-back was a total failure, no soundstage, no imaging at all
BTW You are quick at asking questions Markus, much slower at answering them
so let me take this opportunity (for the third time) to draw Your kind attention to couple of my questions posted above that You haven't asnwered yet, in case You didn't notice them:
are they working in one volume or is there any kind of partition inside? from:
http://www.diyaudio.com/forums/multi-way/10962-stereolith-loudspeakers-question-33.html#post2523007
what is exactly pretty amazing about it?
and
Do You mean AVR + an upmixing algorithm?
from:
http://www.diyaudio.com/forums/multi-way/10962-stereolith-loudspeakers-question-40.html#post2525359
and
Could You be a little bit more specific? Which conlusions exactly?
and
What is more to it?
and
which book?
from:
http://www.diyaudio.com/forums/multi-way/10962-stereolith-loudspeakers-question-40.html#post2525362
best!
graaf
Hello,
Yes. Sidefiring has the benefit of the driver on axis is directed to the wall, and the reflection directed to listening position is almost on axis. Better than average.
Actually I did not try it It didn't make any sense to me to put the midrange dipoles against the side wall (perpendicularly) since it will generate maximum of longitudal reflections and minimum of lateral reflections. I can imagine it could be funny though
Yes I'm talking about phantom images not moving nor collapsing when turning my head. This is remarkabe.
Actually, it is written in 'books' since many decades that in conventional stereo triangle the image will shift when turning the head (even only rotate not moving sideways). Turning the head will change ITD and ILD and pinna cues. Everything changes means phantom image collapses.
My cardboard is a step forward as this does not happen with it. Image do not shift nor collapse.
- Elias
Yes. Sidefiring has the benefit of the driver on axis is directed to the wall, and the reflection directed to listening position is almost on axis. Better than average.
No surprise in that. I believe EVERY serious dipole user has tried that at some time.
Actually I did not try it
It didn't make any sense to me to put the midrange dipoles against the side wall (perpendicularly) since it will generate maximum of longitudal reflections and minimum of lateral reflections. I can imagine it could be funny though But I am not sure, what exactly Elias is referring to. I don't believe he is talking about a phantom image following his head movement (like with headphones). He probably is talking about phantom images being not destroyed by turning his head. I feel happy with him that he now has this great experience too - I carelessly had regarded that as a matter of course in a really good system.
Rudolf
Yes I'm talking about phantom images not moving nor collapsing when turning my head. This is remarkabe.
Actually, it is written in 'books' since many decades that in conventional stereo triangle the image will shift when turning the head (even only rotate not moving sideways). Turning the head will change ITD and ILD and pinna cues. Everything changes means phantom image collapses.
My cardboard is a step forward as this does not happen with it. Image do not shift nor collapse.
- Elias
Hi Scott,
Yes, only high freqs are attenuated by the pillow. And that is what is needed! It will eliminate pinna cues from coming from center location. The fact that lower freqs passes the pillow tells that high freq cues dominate. And they must not be inconsistent, thus they must be eliminated.
I have a concept to prove this. Be it proven right or wrong, I'll report back.
- Elias
Yes, only high freqs are attenuated by the pillow. And that is what is needed! It will eliminate pinna cues from coming from center location. The fact that lower freqs passes the pillow tells that high freq cues dominate. And they must not be inconsistent, thus they must be eliminated.
I have a concept to prove this. Be it proven right or wrong, I'll report back.
- Elias
My difficulties are the most common :
it's not working as it was supposed to.
The duty if this pile of wood was multiple :
Then, I've seen the Finland pillow tweak. I tried with truck mudguard, this worked instantly. Ridiculous, but efficient.
Have to say that 2 barriers on axis have no interest. Some frequencies of the direct radiation have to be attenuated. When measuring, it's a big notch centered on 700 Hz.
it's not working as it was supposed to.
The duty if this pile of wood was multiple :
- exploring the CTC (mids section horizontal, second driver wired from the other channel and in phase opp : nothing happens)
- exploring a special topology of filter that works very well on the other proto, unsuccessful again.
- and whatever I did as paradoxical wiring, DCX phase modifications, other placement in the room, I always had a 80 cm wide soundstage.
Then, I've seen the Finland pillow tweak. I tried with truck mudguard, this worked instantly. Ridiculous, but efficient.
Have to say that 2 barriers on axis have no interest. Some frequencies of the direct radiation have to be attenuated. When measuring, it's a big notch centered on 700 Hz.
What do you think about the following theory how the original Stereolith could work?
One channel has a 180° allpass filter. Bass drivers are going on ground, tweeter is bridged between the two channels and 90° phase shift is somewhat below the crossover point, where the human ear is most sensitive.
One channel has a 180° allpass filter. Bass drivers are going on ground, tweeter is bridged between the two channels and 90° phase shift is somewhat below the crossover point, where the human ear is most sensitive.
Experiments in "the old days" had to be rather crude/simple. They were kind of "one note, one direction, one reflection"
I can easily see how that did not reveal enough cues to locate anything at the side of the head. After all there is still this "cone of confusion" at 90° to the side of the head. When I turn my head perpendicular to the stereo axis, sounds at the stereo axis still don't lock to the next loudspeaker, but I can't tell their direction with any precision.My actual system sends more than average and rather "true" copies of the direct sound into the frontal room corners. They all arrive later than 6 ms after the direct signal at the ear. May be this similarity to stereolith helps to explain the robust phantom image. Is there any time limit for pinna cues getting integrated into the localisation of a sound?
Rudolf
Last edited:
The first proto (TQWT with supravox) is now fitted with a quite special passive filter.
To make short, the ILD part remains as it is (left or right), but a part of the ITD related frequencies are sent to the opposite channel. Pink noise gives the localisation on the appropriate side, and the music too, slopes are swallow. No modifications of the FR or of the impulse. On the scheme, the 12 ohms is here to give the illusion of more bass, and the component values are also tailored to regularize the FR.
The results on the supravox are simply amazing, sometimes I am instinctively searching the source with my eyes so much there is a presence. It works also outside, but in the room, with the reflections, I heard a really 180° display on some musics. Just one problem, this kind of "wide range sound" needs a lot of bias from the listener.
I believed this filter could be an universal solution, obviously not, it's impossible to fit on the 3.5". Huge notch, then big overlap required, then no effect...
OTOH, some amps could disagree by blowing something.
Last word : a good way for comparing instantly two systems...with a PC as source, one system, passive, is linked to the analog output of the soundcard, the other system, through it's own DAC is linked to the digital output. Just have to click on the advanced play control panel. This is very revealing and destroys the listener's bias.
As you see below, a low WAF is required too
To make short, the ILD part remains as it is (left or right), but a part of the ITD related frequencies are sent to the opposite channel. Pink noise gives the localisation on the appropriate side, and the music too, slopes are swallow. No modifications of the FR or of the impulse. On the scheme, the 12 ohms is here to give the illusion of more bass, and the component values are also tailored to regularize the FR.
The results on the supravox are simply amazing, sometimes I am instinctively searching the source with my eyes so much there is a presence. It works also outside, but in the room, with the reflections, I heard a really 180° display on some musics. Just one problem, this kind of "wide range sound" needs a lot of bias from the listener.
I believed this filter could be an universal solution, obviously not, it's impossible to fit on the 3.5". Huge notch, then big overlap required, then no effect...
OTOH, some amps could disagree by blowing something.
Last word : a good way for comparing instantly two systems...with a PC as source, one system, passive, is linked to the analog output of the soundcard, the other system, through it's own DAC is linked to the digital output. Just have to click on the advanced play control panel. This is very revealing and destroys the listener's bias.
As you see below, a low WAF is required too
Attachments
Very similar to everyone's experiments, EXCEPT Radugazon's.
i.e. cardboard boxes of different sizes (and packing tape) with several different fullrange drivers (from under 4 inches in diameter to 6.5"s).
Conceptually I already understood that it wasn't the room (..or wasn't primarily the room). So when I took them outside I had a slightly different purpose for doing so (..and at that time I did "test" them without any significant reflections.)
What I was "testing" for outside was the development of depth due to the "front" wall (i.e. the wall behind the speaker). My garage door served this purpose (moving the loudspeaker closer and then further away). There is an effect due to the timing difference between direct sound and the reflection - with greater distances increasing depth (..up to a point where the reflection is to low in spl). Though the effect is modest in most rooms (that are relatively small), it is a good bit more than any perceived expansion due to side-wall reflections. (..front wall reflections have the "good fortune" of working with a monophonic emphasis as opposed to side wall reflections).
Note that these reflections that due produce increased depth are NOT in the intensity bandwidth. In other words they are freq.s primarily below 1 kHz.
Last edited:
I experience depth "without" a front wall. The whole wall is covered in 20cm polyester batting with 20cm air gap.
You experience with what? What's the setup?
let me take this opportunity (for the forth time) to draw Your kind attention to couple of my questions posted above that You haven't asnwered yet, in case You didn't notice them:
are they working in one volume or is there any kind of partition inside? from:
http://www.diyaudio.com/forums/multi-way/10962-stereolith-loudspeakers-question-33.html#post2523007
what is exactly pretty amazing about it?
and
Do You mean AVR + an upmixing algorithm?
from:
http://www.diyaudio.com/forums/multi-way/10962-stereolith-loudspeakers-question-40.html#post2525359
and
Could You be a little bit more specific? Which conlusions exactly?
and
What is more to it?
and
which book?
from:
http://www.diyaudio.com/forums/multi-way/10962-stereolith-loudspeakers-question-40.html#post2525362
best!
graaf
"..the *development* of depth". (emphasis added)
i.e. INCREASED depth.
BTW, "flooder" type radial use with certain ceiling reflections does something similar..
Last edited:
Hello,
I measured the effect of the pillow blocking the direct sound.
Fast Finnish basic course:
pahvi = cardboard
tyyny = pillow
Cardboard box is in the room as would be when listening. Only one side element is playing. Measuring distance 2 m. Omni mic.
Here wavelet CSD without the pillow:
0 ms: direct sound
2.5 ms: diffraction around the cardboard box
4 - 5 ms: floor and ceiling reflections
6 ms: sidewall reflection
7 - 10 ms: floor-sidewall and ceiling-sidewall reflections
Here wavelet CSD with the pillow:
Direct sound and diffraction are attenuated. Otherwise not much change.
From the plot it seems to be that the side hemisphere reflections between 6 - 10 ms are responsible for hiding the speaker, because three reflections are coming from same lateral direction but from different height, the "air" is playing. The pinna cannot catch the speaker location, it disappears.
If there would be only one lateral reflection, situation it would be similar to stereo triangle, and speaker location could be detected. (my hyphothesis)
- Elias
I measured the effect of the pillow blocking the direct sound.
Fast Finnish basic course:
pahvi = cardboard
tyyny = pillow
Cardboard box is in the room as would be when listening. Only one side element is playing. Measuring distance 2 m. Omni mic.
Here wavelet CSD without the pillow:
0 ms: direct sound
2.5 ms: diffraction around the cardboard box
4 - 5 ms: floor and ceiling reflections
6 ms: sidewall reflection
7 - 10 ms: floor-sidewall and ceiling-sidewall reflections
An externally hosted image should be here but it was not working when we last tested it.
Here wavelet CSD with the pillow:
Direct sound and diffraction are attenuated. Otherwise not much change.
An externally hosted image should be here but it was not working when we last tested it.
From the plot it seems to be that the side hemisphere reflections between 6 - 10 ms are responsible for hiding the speaker, because three reflections are coming from same lateral direction but from different height, the "air" is playing. The pinna cannot catch the speaker location, it disappears.
If there would be only one lateral reflection, situation it would be similar to stereo triangle, and speaker location could be detected. (my hyphothesis)
- Elias
I must add that the disappearing of the speaker is possible only if the direct sound is attenuated. It's easy to test by ear too.
- Elias