Chuck Gerlach
I have used Room Optimizer™ on three different systems/rooms and got similar (positive) results in all three cases. By placing the speakers according to the program results, the frequency response was indeed the flattest (after running the SigTech impulse response test and converting to frequency response).
In each of the three cases, however, it had the speakers closer to the wall behind the speakers (front wall?) than anyone had them without the program. As a result of this placement, soundstage depth was considerably reduced even though the bass was better. And no one was willing to give up the depth! What I found interesting was that the distance from the sidewalls as computed by your program was less than 1" different than each of the three systems were before your program provided the data! I am most impressed. I have tried the other programs that were on the market and they were completely worthless (and incorrect!).
RPG
In the program binder we remark about this trade off between stereo image and bass response. It is important to point this out, since people may not be aware that there is often a trade off. One of the best arguments for subwoofers is the fact that they give you the opportunity to separate the bass transducer from the imaging part of the playback system. One thing to also note is that there is no single best solution for the locations of listener and loudspeakers. We take advantage of this by allowing a search range for listener and loudspeakers. When a client’s preference or physical constraints demand that the loudspeakers be beyond a certain distance from the front wall, just set the minimum x search limit to this value and the program will not look for solutions closer to the wall than this value. The program will constrain its search between the minimum x and maximum x limits imposed and determine the optimum solution within this range.
Chuck Gerlach
When I constrain the ratio distance (by checking the box) I get significantly different results than if I don’t check it. I don’t think I understand exactly what this function does.
RPG
This question indirectly refers to your comments above. One attempt to provide good imaging as well as good bass response is to apply a stereo angle constraint on the angle subtended by the two lines connecting the listener and the left and right speakers. This angle is constrained by placing limits on the perpendicular distance between the listener and the line connecting the left and right speakers. The conventional 60° angle forming an equilateral triangle is obtained with this distance set at 0.88. Thus a user can look for an optimum solution between 0.88 and some upper limit (program defaults to 1.33).
You will get different answers depending on the search limits, but remember that each solution is an optimum solution in this multi dimensional space. What you are searching for is the best solution within the constraints the user imposes on the optimization. The idea of using this stereo angle constraint was to try to simultaneously get good bass response and good imaging by adhering to triangular listening geometries that have the been known to provide good imaging.
Chuck Gerlach
I also get different results when I compare using extreme starting positions and not using extreme starting positions. What is going on with this function?
RPG
It is important to remember that this is a multi dimensional error space and that all listener and loudspeaker solutions are acceptable. They can be evaluated by their standard deviation. The lowest standard deviation is the most acceptable solution. All optimization searches are started from random positions. However, we have found in multi dimensional optimization that starting searches from random positions near the extremes of the error space allows the search to examine more potential solutions. This is a parameter that was used in program development and will probably be removed from user control in later releases.
Chuck Gerlach
I also get different results if I increase the number of solutions from 4 to 6. Which is the most accurate? And if the 6 is better than the 4, would 8 be better than 6?
RPG
Infinity is the best, but unfortunately most of us don’t have the time. Remember that each solution starts from a completely random position and the more of these searches the better. The intelligent search engine hopefully examines most potential solutions, but there is no substitute for doing this as many times as you have the patience for. As you do a multiple solution search you will notice in the error view that some solutions are lower than others are. The lower the better!
If you select confirm solutions, there will be a second confirmation search from a position slightly perturbed from the found solution to see if the error stays the same or gets lower. This check is a standard practice in multi dimensional analysis to insure, as best as is possible, that the solution is not a "local" or good intermediate solution, but a "global" or best solution.
Chuck Gerlach
Thanks for your help. And if you ever "get south" I’d love for you to hear my system. Between absorbers, 24 Skylines and SigTech, it is pretty amazing.
RPG
I am looking forward to it.
Home:
Products:
Room Optimizer™
Testimonials and FAQs
|
|
|
Copyright
2000, RPG Diffusor Systems, Inc. |