08 May 2015
Stereo sound Imaging solution
In 2 channels stereo always may be observed a bit no convenience of getting proper localization of sounds. Perfect image requires an ideal positioning of listener along the vertical line of two channels phase coherence. ANY side movement to left or right effect in annoying changes of timbre caused by phase disturbances. . That ideal position by audiophiles is called „sweet spot”. In long period of listening time staying in that line is troublesome and additionally listener tends periodically to confirm right position. Wideness of base is also not constant with small even movement of head.
Author experiences it the most with standard stereo recordings.
Next trouble rose with visit of author’s fellow who is used at home to movie theater. He paid attention and complained to flutes „traveling” between speakers. Actually it was nothing surprising to author, who such exactly traveling considers the proof of perfect point to point operation of Standard Stereo system. It is just like that, and responsible are mistakes of phases which author presented earlier. The essence of trouble is that two speakers produce sharp picture for one central line but listener has two ears and has to choose weather to position symmetrical both in a bit out of phase or one in good phase and other in wrong phased fields. As presented on drawing the size of „sweet spot” does not cover both ears instantly and that is evaluated by author by experience with pink noise. The same test with reach sounding recordings unveil this problem in a way of „shaking sopranos” with horizontal shift. .
Standard stereo images
Positions of listener marked left, center amd right In Green marked „sweet spot” with good phase coherence , in red bad phase coherence . Grey are ears.
Because distances b,a,c which we want them to be equal cannot be equal, so phase coherence of channels sums is not existing. As we see only one ear, else in left or right position may experience matched phases. This of course decreases loudness of affected frequencies in second ear and for such frequencies image is shifting. That shift is jumping accordingly to notes and listener’s head movement. Author underlines, that it is the most noticeable in pure (sinus) tone instruments like example flute. With other it is just trembling sopranos. .
No any support could be found, since years for author problem seemed unsolved. Last time he evaluated a bit unheard (for him also) solution of two tweeters for each speaker column. Ha ha … primary it sounds weird – by popular imagination .to get sharp picture we should fix sound sources at defined vertical line of column with overall size for sopranos as small as possible- analogy with one projector for left screen and one for right seems to be clear. Small size because for sopranos we need to take also care to prevent surfaces around active surface. The reason of it is to diminish disturbing reflections mixing with source. It seems not popular knowledge that timbre depends a lot on surrounding around tweeter.
But two projectors for each will blur picture. Is it true for sure?
May be, but let’s see High End class type MAGNEPAN speakers , ribbon type – where „tweeters” are not positioned in center line of column but are fully wide about 10″ . They do not form reflecting surfaces due to their no reflective structure. Author, while listening to them, had got just opposite impression of better than usual imaging and much less troubles to get in sweet point. This is by all accounted to perfection of construction but are we not wrong?
We make fault thinking about audio picture as analogy to projected movie. In audio we do not get direct „painting” – all painting is made in our imagination and our signal is only acoustic pressure coming from not clear for listener space point. (In case of doubt it is simple test by covering one ear and then trying to localize speaker by other) . That is why size of speaker has nothing to do with sharpness of stereo imaging. So let’s be not worry about this horizontal dimension.
What are we getting for this two tweeters? First note – they must be positioned one by one horizontally. Vertical positioning is not discussed here. Then for best result one of them should be in line with other drivers to avoid more confuse. . Following is drawing with two horizontal tweeters:
Base of operation for bi-tweeter stereo system
As shown, for each ear we can align tweeters in a way to deliver by chosen pair of left and right column phase coherent sum of left and right channel, even in case when other pair signal (shown in orange) is not aligned. All most valid distances are equal = a and our „sweet point” has grown horizontally few times This way we get compensation of damage and all imaged instruments are almost „fixed” independent from head movement.
Author by quick switching on-off additional tweeter in constructed by him columns tested the result and observed more than 50% wider image and much more stability of timbre of sound and that was for author decisive to resign from widely known one tweeter concept. Anybody may try it and in case of three way columns it may be even better idea to double also mid-range.In authors columns mid-ranges are not used and serial connected in the same phase two tweeters cover frequencies above 4,8 kHz.
Stereo sound improvement. One of two tweeter’s stereo units tested by author.
UPDATE: 2015 -May 22. – added few notes for better understanding