Some argue that for accurate stereo imaging, basically only the differences between the left and right speakers matters and not the absolute flatness of the SPL function.
I think this hypothesis is testable. This test distinguishes three situations:
A) Both channels play without disturbance
B) Both channels are disturbed identically
C) The channels are disturbed differently
In these three situations, the sound (music) is evaluated for stereo separation. This can be blind tested.
If situations A and B have approximately the same separation and C clearly less, than the hypothesis is true.
If situation A is clearly better than both B and C the hypothesis is not true.
It is also possible that the results are not conclusive.
The distortions can perhaps consist of a some dips in the SPL (perhaps 2 to 5 dips, -6 dB, Q=2 between 500 Hz and 5 kHz).
If the distortion have to be identical, the dips must be the same left and right and put on the same frequencies. If the channels are disturbed differently, the dips remain the same but are put on different frequencies.
Necessary for the tests are a good speaker system and a dsp between sound source and power amplifier. The dsp should be set up in a way that it is easy to switch between the three situations.
I think this hypothesis is testable. This test distinguishes three situations:
A) Both channels play without disturbance
B) Both channels are disturbed identically
C) The channels are disturbed differently
In these three situations, the sound (music) is evaluated for stereo separation. This can be blind tested.
If situations A and B have approximately the same separation and C clearly less, than the hypothesis is true.
If situation A is clearly better than both B and C the hypothesis is not true.
It is also possible that the results are not conclusive.
The distortions can perhaps consist of a some dips in the SPL (perhaps 2 to 5 dips, -6 dB, Q=2 between 500 Hz and 5 kHz).
If the distortion have to be identical, the dips must be the same left and right and put on the same frequencies. If the channels are disturbed differently, the dips remain the same but are put on different frequencies.
Necessary for the tests are a good speaker system and a dsp between sound source and power amplifier. The dsp should be set up in a way that it is easy to switch between the three situations.
Yes, the L+R is mono.
Frequency response, if identical in both channels, is not much of a factor for stereo.
Frequency response, if identical in both channels, is not much of a factor for stereo.
Are you attempting to test this in a domestically sized room? Different types of speakers with varying degree of baffle width, placement from walls ( or a flush baffle wall mount ) and directivity will all behave dramatically different in the same room.
I kind of did this with my phase shuffler and it mostly just changed the tonal balance of the center. FR I have not tried.
https://www.diyaudio.com/community/threads/fixing-the-stereo-phantom-center.277519/
https://www.diyaudio.com/community/threads/fixing-the-stereo-phantom-center.277519/
Am I correct that you are basically saying that the hypothesis is correct (in principle only differences matter) ? Do you perhaps have further information that supports this view ?
A related topic - I sometimes see photos of listening rooms where 2-channel audio is being fed to 3 channels, Left, Right, and center.
I have often wondered how the 2-channel stereo signal is split into 3 channels. Is it as simple at (L), (L+R), (R) ? This seems like the most straightforward way to do it... But I could also imagine a split of (L-R), (L+R), (R-L). This seems like it might not work well with some recordings... ? Thoughts?
j.
I have often wondered how the 2-channel stereo signal is split into 3 channels. Is it as simple at (L), (L+R), (R) ? This seems like the most straightforward way to do it... But I could also imagine a split of (L-R), (L+R), (R-L). This seems like it might not work well with some recordings... ? Thoughts?
j.
Search for Gerzon matrix for example, there are many variations how to do it. @mark100 has experimented with them few years back if I remember.
Here, see this http://elias.altervista.org/html/3_speaker_matrix.html
Here, see this http://elias.altervista.org/html/3_speaker_matrix.html
Many years ago, Paul Klipsch used a passive resistance based line level box to derive 3ch stereo. It is detailed in one of the Dope From Hope papers if I remember correctly.
These days, most home theater AV receivers offer 5 channel stereo options. Might be able to pick up one inexpensively at a pawn shop.
I worked through the Gerzon articles and patent recently and think his approach is this (ignore the subs and generation of the sides/surrounds). I have been trying to decide whether to try and implement this in software v buying a Meridian processor to get Trifield (the one implementation of his patents).
Yes, Elias donated a lot of nice work.
I'm still playing with the details he explains about Gerson's and others efforts.
Nice !
Have you played much with where to set the lin phase hp and low pass filters? I still need to explore that.
Hi OP,
IMHO, the issue of imaging is one of Perceived Imaging. It's not so much whether a 2 or 3D image is able to be recovered accurately or not it's whether we feel it's right or not.
Among the things that come into play in this perception are many things. Some argue that you don't want a flat ideal FR but a little dip around 2.4 kHz. Some commercial speakers took advantage of this and many still do at least off-axis.
In addition the room acoustics play a large part in the perceived imaging. A totally dead room isn't as good as one with diffusion in the right places.
If you want ideal L and R signals try headphones.
IMHO, the issue of imaging is one of Perceived Imaging. It's not so much whether a 2 or 3D image is able to be recovered accurately or not it's whether we feel it's right or not.
Among the things that come into play in this perception are many things. Some argue that you don't want a flat ideal FR but a little dip around 2.4 kHz. Some commercial speakers took advantage of this and many still do at least off-axis.
In addition the room acoustics play a large part in the perceived imaging. A totally dead room isn't as good as one with diffusion in the right places.
If you want ideal L and R signals try headphones.
Thanks.
And I wish 🙂. My listening room (wife would probably call a family room) is stripped down to the studs on the ceiling and front wall for acoustical treatment and I don't even have a center channel. Planning DIY LCR.
Though I did just take delivery of 3x Bliesma T34Bs, so hopefully will get there.
But yes, playing with the LP/HP frequency will be interesting. Gerzon actually mentioned linear phase filters in this role as desirable, though certainly difficult at the time/when he was alive.
Bill
Cool Bill, very interested in how your experiments go.
And Congrats on getting the Bliesma's
And Congrats on getting the Bliesma's