Home |
Search |
Today's Posts |
#1
|
|||
|
|||
Room Response Curves
I'm seeking some applied acoustic advice.
I am using RPG's Room Optimizer software to place a pair of full-range speakers. From room dimensions and user's constraints on positioning, the software suggests optimal speaker and listener positions. Room Optimizer tries to flatten both the "speaker boundary interference response" and the "modal response." In the examples of graphs I have seen (including my room), the former gets much flatter than the latter. Questions: 1. What do those two responses mean in practical terms? (For example, should they be added together to produce the perceived response?) 2. In finding optimal positions, R.O. weights the two types of response equally, but that can be changed by the user. Given that the room already has substantial acoustic treatments (bass traps and absorptive/reflective panels), would one type of response be more important than the other in determining the sound I perceive in the room? 3. I am considering adding digital room correction (Tact 2.2X). If I added that, would the answer to the preceding question change? And a general comment: R.O. gives results based on exact listener and speaker positions to the nearest 0.1 inch (0.25 mm). Since no room is exact, no positions are exact, and nothing is a point source, wouldn't it be more useful to give area-averaged results over, say, 3 or 4 inches? Any guidance that might help me avoid moving two very heavy speakers around, denting my pine floors with each move, would be welcome! Mike Prager North Carolina, USA |
#2
|
|||
|
|||
Mike Prager wrote:
I'm seeking some applied acoustic advice. I am using RPG's Room Optimizer software to place a pair of full-range speakers. From room dimensions and user's constraints on positioning, the software suggests optimal speaker and listener positions. What I have learned in a long lifetime of experimenting (and dicking around) is that speakers are positioned for IMAGING, not for frequency response. You position them according to all of the reflection patterns, room surfaces, listener position, and a dash of experimentation, but you do this for imaging, ignoring frequencies for the moment. When you're happy with that, then you EQUALIZE for frequency response. This little lesson seems reasonable to me, since imaging depends a lot on geometry, and freq response depends more on the room and EQ. I mean, what are you going to do if, after you use one of these computer programs for speaker placement, if the freq response is smoother, but the imaging sucks? OK, let's talk about imaging for a moment. How do you know - that is, what are you looking for in good imaging? Stereo is not just a clump of sound in the left, or in the center, or in the right speaker. There should be a smooth transition all the way across the stage. Like if you play a solo piano, you can tell where every key is, or if there is a jazz big band, you can place most of the instruments. Depth imaging is a little tougher, and a loftier goal, but what you should be getting is imaging that pools OUTSIDE of the actual speakers, in a region behind the plane of the speakers, from left of the left speaker to right of the right speaker. Sounds seem to be coming from virtual points in space, fairly pinpointedly, if recorded that way, anywhere in that region. If all of this holds as you walk around, even better, but that part is hard to obtain in smaller rooms. Now - a good starting point for everything would be to have the main speakers (L, R, and C) positioned so that the L and R are 1/4 of the room width from front and side walls. Then place subwoofers in the front corners and never move them (!). If the main speakers are not of a designed-in radiation pattern, then experiment with toe-in, to get an even soundstage as you walk across the room. Gary Eickmeier |
#3
|
|||
|
|||
Gary Eickmeier wrote:
What I have learned in a long lifetime of experimenting (and dicking around) is that speakers are positioned for IMAGING, not for frequency response. [...] When you're happy with that, then you EQUALIZE for frequency response. This little lesson seems reasonable to me, since imaging depends a lot on geometry, and freq response depends more on the room and EQ. I mean, what are you going to do if, after you use one of these computer programs for speaker placement, if the freq response is smoother, but the imaging sucks? [...] Gary, thanks for an interesting perspective. By using a computer program, I managed to get worse imaging AND worse bass response! So I am back to positioning by trial and error, and yes, I am planning to use some equalization to smooth out the response if necessary. Mike Prager North Carolina, USA |
#4
|
|||
|
|||
In article ,
Gary Eickmeier wrote: Mike Prager wrote: I'm seeking some applied acoustic advice. I am using RPG's Room Optimizer software to place a pair of full-range speakers. From room dimensions and user's constraints on positioning, the software suggests optimal speaker and listener positions. What I have learned in a long lifetime of experimenting (and dicking around) is that speakers are positioned for IMAGING, not for frequency response. You position them according to all of the reflection patterns, room surfaces, listener position, and a dash of experimentation, but you do this for imaging, ignoring frequencies for the moment. When you're happy with that, then you EQUALIZE for frequency response. I use Apogee speakers and the RPG program. I found that when I positioned the speakers to smooth out the humps the imaging also improved. Box speakers may be different. |
#5
|
|||
|
|||
Robert Peirce wrote:
I use Apogee speakers and the RPG program. I found that when I positioned the speakers to smooth out the humps the imaging also improved. Box speakers may be different. To be fair to RPG, I should add that my room has features that their program just can't model (and is not claimed to model), and that undoubtedly influences the results. However, when I took a look at CARA (the more complex alternative), I found it too complex for the amount of time I have available for tweaking. For those who put in the time, it is said to give remarkably good results. Mike Prager North Carolina, USA |
#6
|
|||
|
|||
Mike Prager wrote:
To be fair to RPG, I should add that my room has features that their program just can't model (and is not claimed to model), and that undoubtedly influences the results. However, when I took a look at CARA (the more complex alternative), I found it too complex for the amount of time I have available for tweaking. For those who put in the time, it is said to give remarkably good results. But still, these effects (of computer aided room positioning) operate only at the bass frequencies, and what you want to do is position the main speakers for imaging and place the subs in the corners (ref. Tom Nousaine and a few others) and then adjust gains and EQs to taste. The actual effects of positioning the main speakers will be relatively minor in this scenario. So save a little money and trouble. Gary Eickmeier |
#7
|
|||
|
|||
Mike Prager wrote:
I'm seeking some applied acoustic advice. I am using RPG's Room Optimizer software to place a pair of full-range speakers. From room dimensions and user's constraints on positioning, the software suggests optimal speaker and listener positions. Room Optimizer tries to flatten both the "speaker boundary interference response" and the "modal response." In the examples of graphs I have seen (including my room), the former gets much flatter than the latter. Questions: 1. What do those two responses mean in practical terms? (For example, should they be added together to produce the perceived response?) 2. In finding optimal positions, R.O. weights the two types of response equally, but that can be changed by the user. Given that the room already has substantial acoustic treatments (bass traps and absorptive/reflective panels), would one type of response be more important than the other in determining the sound I perceive in the room? 3. I am considering adding digital room correction (Tact 2.2X). If I added that, would the answer to the preceding question change? And a general comment: R.O. gives results based on exact listener and speaker positions to the nearest 0.1 inch (0.25 mm). Since no room is exact, no positions are exact, and nothing is a point source, wouldn't it be more useful to give area-averaged results over, say, 3 or 4 inches? Any guidance that might help me avoid moving two very heavy speakers around, denting my pine floors with each move, would be welcome! Mike Prager North Carolina, USA Mike, hello from North Carolina. There is a lot of useful information on room setup he http://www.linkwitzlab.com (check out "room acoustics" and "toneburst cd", the website gives you enough information to be able to generate the tonebursts yourself in CoolEdit or code them in C, MATLAB or similar) Room-analysis software doesn't take into account the frequency response of the speakers or their radiation patterns. There is no substitute to measring the in-room response, which is neither hard nor expensive. I use a system consisting of Behringer ECM8000 microphone - mixer - computer sound card (Echo Gina) and CoolEdit software. I found this particularly helpful in matching a subwoofer to my monitors. It is also helpful in detecting early reflections (you can actually see the early reflections on the computer screen and know exactly how early they are). As far as frequency response is concerned, one should try to do as good a job as possible without equalization. However, some resonances (peaks) can be tamed with an equalizer. Averaging over the "listening area" is a good idea, since one never sits in exactly the same spot all the time. |
#9
|
|||
|
|||
Mike Prager wrote:
3. I am considering adding digital room correction (Tact 2.2X). If I added that, would the answer to the preceding question change? I would just use classic setup techniques, such as dividing the room up in thirds and avoiding early reflections from the side walls and use the Tact to fine tune the raw setup before correction. Tact recomends that you strip the room off treatment and put the speakers in the corners, but imo this is a marketing gimmick to maximize the difference between a corrected setting and 'bypass.' I think you will find RO fairly useless once you start using the Tact, which is a very versatile piece of work. |
#10
|
|||
|
|||
wrote in message
... Tact recomends that you strip the room off treatment and put the speakers in the corners, but imo this is a marketing gimmick to maximize the difference between a corrected setting and 'bypass.' Moving the speakers closer to the corners can make a huge improvement in the low frequencies in terms of clarity and impact, especially when used with DSP correction. There are speakers that are designed to be placed in the corners; e.g., the Allison One and the Tact W210. Of course, moving the speakers back towards the walls will increase the room gain (i.e., more bass) as the low frequencies' reflections from the walls become more in sync with the direct sound. The Tact RCS will EQ-out the added room gain but retain the improved time behavior, resulting in tighter, clearer bass with greater impact. I believe Tact has an explanation on their website for their recommendation. I also was skeptical until I moved my B&W801's half the distance back towards the front corners and was astounded by the improvement in the bass, even with the same target frequency response curve in my Tact RCS (when it comes to low frequencies, time response is as important as frequency response). If you have an RCS, I recommend that you give it a try. Or if you at least have a tone control, move the speakers back and use the tone control to reduce the added bass. If you don't have corners, moving the speakers back towards the front wall will have a lesser effect. Regards, Tip |
#11
|
|||
|
|||
Tip wrote [and I shortened]:
Moving the speakers closer to the corners can make a huge improvement in the low frequencies in terms of clarity and impact, especially when used with DSP correction. [...] the low frequencies' reflections from the walls become more in sync with the direct sound. The Tact RCS will EQ-out the added room gain but retain the improved time behavior, resulting in tighter, clearer bass with greater impact. As you said, Tact does offer the same explanation on their Web site. Still, it is useful to hear from a disinterested party who has done it. The conventional recommendation to keep the speakers away from the front is to improve imageing, particularly depth of image. What was your experience in moving your speakers towards the corners -- was there a tradeoff between better bass and less soundstage illusion? Mike Prager North Carolina, USA |
#12
|
|||
|
|||
Hi Mike,
"Mike Prager" wrote in message ... Tip wrote [and I shortened]: Moving the speakers closer to the corners can make a huge improvement in the low frequencies in terms of clarity and impact, especially when used with DSP correction. [...] the low frequencies' reflections from the walls become more in sync with the direct sound. The Tact RCS will EQ-out the added room gain but retain the improved time behavior, resulting in tighter, clearer bass with greater impact. As you said, Tact does offer the same explanation on their Web site. Still, it is useful to hear from a disinterested party who has done it. The conventional recommendation to keep the speakers away from the front is to improve imageing, particularly depth of image. What was your experience in moving your speakers towards the corners -- was there a tradeoff between better bass and less soundstage illusion? Yes, there definitely is a tradeoff, and you have to hit a happy median. I guess it depends on what's more important to you - imaging or bass (I'm a bass player ;^). But this is the idea behind the Tact W210 & W410 "corner-woofers": put the woofers where they sound the best (in the corners), the main speakers where they sound the best (away from the walls), and integrate the mains and the subs with the RCS 2.2X, which provides the digital crossovers, level matching, and time alignment in addition to room correction (needed for the corner placement). Diffuser panels on the walls near the speakers may provide a solution, but I decided to go the 2.2X/W410 route instead. Regards, Tip |
Reply |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Forum | |||
on topic: we need a rec.audio.pro.ot newsgroup! | Pro Audio | |||
Transient response of actively filtered speakers | Tech |