[mrlittlejeans]

When you are measuring the distance from the back walls, where on the panel are you measuring from?  The point closest to the wall, the mid point of the panel, the tweeter or the point farthest from the wall?  This assumes you have the speakers toed in, if not, it is a moot question.

I picked up a pair of used 3.6's and had them sent to Magnepan to be rebuilt.  Got them Friday and played with placement for a while.  I found that my soundstage was best at around 8' between the speakers as measured from the inner edges.  The outside edges were ~2' from the side walls and the inner edges were about 52" from the bottom of some built in bookshelves on the front wall.  Of course, the shelves themselves extend out about 18-24" from the back wall and there isn't anything in the shelves so depending on how you measure, I could be as far as 76" from the front wall.

Attached is a pic of the room.  Please ignore the mess.  This is a rental as my house, along with my 3.7's, was caught in the forest fires here in Colorado almost two months ago so I'm rebuilding the system.  I picked up these 3.6's as I was never completely happy with the 3.7's and wanted to go back to 3.6's to make sure I wasn't confusing nostalgia with reality as I always thought my old 3.6's sounded better than my 3.7's.

I'm noticing a bit of huskiness on female vocals that I don't recall hearing before.  This may be room dependent or due to the fact that I just hooked everything up Saturday and the equipment (particulary the Audible Illusions L3) needs to break in more.

[andyr]

I suggest you need to define whether you have your 3.6s ribbons-in or ribbons-out. 

Soundstage is very different.


Regards,

Andy

[mrlittlejeans]

They are currently tweeters out as shown in the pic (which may be hard to see as the speakers are so dark).  Mayhaps I'll switch them up tonight and try tweeters in.

[MGbert]

They are currently tweeters out as shown in the pic (which may be hard to see as the speakers are so dark).  Mayhaps I'll switch them up tonight and try tweeters in.

mrlittlejeans:

While you're experimenting, could you check out a theory of mine?  Take a mirror, put it along the side wall where you can see the reflection of the 1.6 from the listening position.  If it is not edge on (ie, the null is pointed at the first reflection point) could you rotate the 1.6's so the null edge is visible in the mirror and see if there is an improvement in clarity of the phantom center?  I believe it has to do with reduction of comb filtering effects.  It seems to have that effect in my admittedly tiny space, so it would be interesting to see the effect it has in a larger space.  Trust me, yours is huge compared to mine...     Have fun!

MGbert

[josh358]

mrlittlejeans:

While you're experimenting, could you check out a theory of mine?  Take a mirror, put it along the side wall where you can see the reflection of the 1.6 from the listening position.  If it is not edge on (ie, the null is pointed at the first reflection point) could you rotate the 1.6's so the null edge is visible in the mirror and see if there is an improvement in clarity of the phantom center?  I believe it has to do with reduction of comb filtering effects.  It seems to have that effect in my admittedly tiny space, so it would be interesting to see the effect it has in a larger space.  Trust me, yours is huge compared to mine...     Have fun!

MGbert

I've tried that experiment. When I null the speakers, I hear an improvement in midrange clarity, which is entirely in keeping with theory. However, if the total path length of the first sidewall reflection is >=10 feet, it's actually desirable to have some sidewall reflections because the ambiance is generally sonically desirable -- it will tend to expand the image laterally. Diffusion can improve the effect although it may be too close to the listener for a QRD diffuser, in which case a different diffuser design has to be used.

I haven't listened for/noticed a difference in center image stability, though, so I'll be curious to see the results, if any.

[josh358]

When you are measuring the distance from the back walls, where on the panel are you measuring from?  The point closest to the wall, the mid point of the panel, the tweeter or the point farthest from the wall?  This assumes you have the speakers toed in, if not, it is a moot question.

I don't know that there's any answer to your query. What counts is the path length from the acoustical center of the driver to the first reflection point on the front wall to your ears and this will be different for each driver. Also, the shelves will affect different frequencies differently. They'll be transparent to lower frequencies which will just diffract around them. Not so to higher frequencies. What they do will depend on what's on them, too.

For purpose of description, I'd probably just measure from the middle to the wall.

I'd be concerned that the shelves would interfere with imaging and perhaps cause cavity resonances. This could be eliminated by putting doors on them. Or you could use an online QRD diffuser calculator and use books and CD's to make a giant diffuser . . .

[BobM]

IMO you have quite a lot of toe in. I would try firing them a bit more straight ahead and maybe sitting back a little farther. From the picture it looks like you have the room for that.

[mrlittlejeans]

Thanks for the responses.  I did a little playing around last night and came to the conclusion that I think the shelves are causing my midrange muddiness and imaging issues.  I brought down another pair of speakers and they exhibited the same problem.  I probably won't fill them as this is only a temporary accomodation until our house is rebuilt and I don't have anything to put in them.  Maybe I'll go buy some sheet rock and just cut it to the right lengths and put it in front of the shelves so I have a solid rear surface.

[josh358]

Thanks for the responses.  I did a little playing around last night and came to the conclusion that I think the shelves are causing my midrange muddiness and imaging issues.  I brought down another pair of speakers and they exhibited the same problem.  I probably won't fill them as this is only a temporary accomodation until our house is rebuilt and I don't have anything to put in them.  Maybe I'll go buy some sheet rock and just cut it to the right lengths and put it in front of the shelves so I have a solid rear surface.

I think that's a good strategy, if you have the time. You might have to move the speakers a bit further in the room afterwards, usually 5' is the point at which imaging starts to get good (10 ms delayed reflection). You could also mount a pair of QRD diffusers at the first reflection points -- GIK makes some foam ones that might be good for a temporary setup, cheap and easy to hang.

[SteveFord]

You've also got an open hallway on the right to contend with.
What's lurking behind the couch?  Could you flip the speakers and couch around?

[MGbert]

I've tried that experiment. When I null the speakers, I hear an improvement in midrange clarity, which is entirely in keeping with theory. However, if the total path length of the first sidewall reflection is >=10 feet, it's actually desirable to have some sidewall reflections because the ambiance is generally sonically desirable -- it will tend to expand the image laterally. Diffusion can improve the effect although it may be too close to the listener for a QRD diffuser, in which case a different diffuser design has to be used.

I haven't listened for/noticed a difference in center image stability, though, so I'll be curious to see the results, if any.

@josh358:

Shhh!  I wanted him to find out on his own - you spoiled it.   

Besides, from his pic and writeup, they are 2 feet from the side walls, just like mine.   

@mrlittlejeans:  when you get paneling, place a section temporarily behind the right speaker, blocking the passageway.  The difference between left and right rear wave return time is not doing you any favors.  And if you have paneling left over, mebbe you could try mini-FRTs?

http://www.audiocircle.com/index.php?topic=118512.0

MGbert