[Housteau]

I have a dedicated 2 channel listening room.  I had it specially built for that purpose being modeled after a friends room that I know to work extremely well.  The only differences I made were a foot here and there to bring it to more optimum acoustic dimensions.  It ended up at 15.2' x 21.5' x 10'.  The 10' height is an average of the 12' cathedral peak.

The room had basic room treatments, you know the average stuff like tube traps and panels for first reflections.  My 1/3 octave analysis showed a very good curve.  Since Martin Logans are dipoles getting a good solid and stable image is more difficult that with point source speakers.  That is just the way it is.  However, even realizing that I have always had some difficulty in creating the kind of image and focus I thought the room should get.

So, I have been spending time on different acoustic forums, this being one, trying to educate myself more on the subject.  The first thing I learned was that a 1/3 octave analysis is basically useless.  The sample points are too widely spaced and average out many peaks and dips as if there were none, when in fact there could have been very large narrow ones that just canceled each other out.  This was the case in my room.  I downloaded a set of test tones from Ethan Winers site: 

Test Tones

It includes tones from 10 Hz up to 300 Hz in 1 Hz increments.  Boy was I shocked at how sawtoothed my real room response was.

The second thing I learned was that a cathedral ceiling can be very good for several reasons, but also bad for some things such as imaging and focus.  Now, what to do about what I found out?  I continued my research and decided to build my own acoustic control devices, bass traps and panels for reflections with help from Bryan Pape.

 http://www.sensiblesoundsolutions.com/. 

I still have a few more to build and install, but so far the improvements can be both measured and heard.  I decided to stop my project, just for a while to take a breather to enjoy my progress thus far and make a few graphs to share my experiences.

Here is a graph of my room before my new treatments were added made with my CLS1 / RS1b hybrid system:

     

Now this second graph was made with my new treatments in place using the CLS1 / RS1b hybrid:

   

Before I took the new measurements I sat down for a listen.  I would rather trust my ears anyway .  Sure enough I could tell the difference right away.  Everything was much clearer, bass tighter and the imaging was improved.

However there is something very odd going on in the 250 Hz region.  It appeared to possibly be a function of the speaker and not the room, or at least not totally the room.  So, I did a second graph using the RS1b system complete.  Here it is:

       

As expected the zone below 150 Hz is basically the same as the previous graph since the crossover to the bass towers is in that area.  The exception is a more pronounced dip centered around 90 Hz, but that may have been a mistake made by me.

What I found most interesting is the different nature of this system above the crossover region.  Both systems sound much better now and I am not quite finished yet. 

The front left corner.
 
 

The back wall.
 

The cathedral peak towards the front wall.
 

The prints hanging on the walls are framed in glass, but I have them mounted far above the listening position and tilted upwards.

[fajimr]

nice room... is that a 'new' type of treatment on the back wall? 
the last person who didn't like your speakers?  criticised your taste in music? 

[JoshK]

Couple of questions.... where are you measuring from?  Are you using digital xo's?  If so, are you using time delay to time align the towers to the panels?

Have you tried measuring the bass towers near field to determine their own response?

[Housteau]

The sculpture was made by an acquaintance of mine years ago as he was finishing up Graduate school:

http://www.patrickmillersculpture.com/portfolio.htm

His early days were focused on sciencefiction themes.  This one of mine is actually a self sculpture inspired from Star Wars when Hans Solo was frozen in carbonite.  His art exhibit at the time had nine of them all around the room looking in.

"where are you measuring from?"

I took the measurements from my listening seat at ear level.  The logarithmic chart I used gets conjested early on and so passed 80 Hz I just plotted a point every 10 Hz to show all the trends.

"Are you using digital xo's?"

No.  I am using a modified Infinity RS1b crossover for both the CLS hybrid and the complete RS1b system.  I switch the upper range units in and out when I choose with just minor adjustments.  The bass tower drivers are 8" and actually work well with the CLS1s.  Some folks have tried towers with the larger 10 -12" variety, such as what came with the Infinity Betas, but they are too slow.

"If so, are you using time delay to time align the towers to the panels?"

Infinity set the time alignment of the RS systems by providing recommended footprints of speaker placement to follow.  As long as both sections of each channel are within that zone from each other, the system is aligned.  Now that works with the Infinity ribbon midrange / treble section.  The CLSes are a bit different and quicker due to the nature of electrostatics.  Still, I have pretty much stayed within the original system guidelines with good results. 

"Have you tried measuring the bass towers near field to determine their own response?"

No.  I have just gone by the published specs, but you do raise a good point.  With older system such as I have, abnormalities can show up as circuits and drivers age.  That would have an effect on my measured responses.

Dave 

[bpape]

Many years ago in a previous life, the store I was managing had the RS-1b's.  We took in a pair of Ohm F's on trade one day.  After hours, we pulled the ribbon towers of the RS-1's and swapped in the Ohm's - WOW. 

Those RS-1b's are really quick and go down pretty low.  The xover has a lot of control and as Dave said, if you follow the physical alignment parameters, you can mate alot of things with them.  We even had some Quad's one day up on stands with the towers on the bottom.  Surprisingly, they kept up with the Quads pretty well.

Bryan

[fajimr]

The sculpture was made by an acquaintance of mine years ago as he was finishing up Graduate school:

very cool- I do like it by the way.  your friend has some amazing bronze sculptures!

[Housteau]

I have already posted this update on the original forum I had started this with:
http://www.musicplayer.com/cgi-bin/ultimatebb.cgi?/ubb/forum/f/26.html
But, since I have been here as well and others may not be aware of the other forum, I thought I should do the update here as well.  There may be some here that could benefit from what I have to share.

Well, I have been back at it.  I took a little while to just sit and enjoy things for a bit, but in these last few days I have been able to finish up the acoustic treatments for my room.  To make things a bit easier I'll include the graph of my room before I started my new treatments.
 
  

I was really surprised on how bad this graph looked, but this was the first time I had measured one Hz at a time.

Now, this next graph is with all my room treatments complete.  No other changes have been made except for some new cables here and there.  The system is the same as in the previous graph, my CLS / Infinity hybrid.

    

As you can see these room treatments have helped to improve upon several of my troubled areas.

The new treatments I added since the last time include the 4th and last corner of the room.

 

This corner was a bit tricky because it had the AC intake right there.  So, I built a small platform and stacked the triagular 2" 703 pieces on top just above it.  It is a very quiet AC system with the oversized vents and other things specific to its installation.

The other section I newly treated was the diagonal intersection of the front walls to my cathedral peak.

     

These were also tricky.  I made them with stacked triangular sections of the 2" 703, just like the room corners, but half the size.  Spray adhesive held them together and a reinforcement rod was run through the middle.  The rod ends are wired to the ceiling.  The covers are just 1" x 2" fabric covered frames used for aesthetics.

The only treatments built for high frequency absorption are on the front wall and the cathedral peak.  All the corner units are built reflective of HF.

I believe this is the last of my treatments, at least for now  .  So, now I plan on working on positioning of the speakers and listening seat.  I have a few more options open to me since I am using a 4 piece system.  However, these bass towers cross over fairly high at 150 Hz and need to remain close to the CLSes to maintain proper phase.

My first attempt was only a move of a few inches forward and tilting the bass towers toward the listener.  Three areas got better, the 37Hz, 90hz and that area around 250Hz or so.  That last one surprised be because that frequency is carried by the CLSes and not the Infinity bass towers.  I guess that the close proximity of both speakers affects the dipole nature of the CLSes.

However, the dip at 40Hz got deeper and wider, along with a higher peak at 60Hz, not good.  So, this next graph shows move #2.

 

This was basically the same location, but with the bass towers pointing straight ahead instead of at the listener.  This kept a lot of the previous improvements with a reduction of the ones at 45Hz and 60Hz.  I think this is where I will stay.  At least for now  .

[Housteau]

Just for fun I thought I would show you the difference in what the standard 1/3 octave measurements look like compared to the true response measured one hertz at a time.  Here is my final graph once again:



Now this is what the 1/3 octave graph looks like:



The measurements were taken at a less volume, approximately -1 db, but you can see the difference.  The only problem area that shows up is the one around 60 Hz, because that is one of the target frequencies of the standard 1/3 test.

[Ethan Winer]

> Now this is what the 1/3 octave graph looks like: <

And now you know why I made those test tones you used 1 Hz apart, rather than at 1/3 octave spacing!

--Ethan