I would almost never go as close as 1 m. I'm usually at 3 or there abouts. I don't obsess over the actual value as long as its not too close, and 1 m is far too close. LF are always a measurement problem, but I don't believe that they have any meaning in a free field. LF are dominated by the room and the sourcxe is all but irrelavent. Its the HF where gating works fine that I am concerned with and indoor gated measurements work just fine.
Lynn Olson(excellent posts as usual), Gainphile, Dr. Geddes, thanks again everyone! This is a great board. The knowledge and experience here is just awesome.
OK, it sounds like my living room may actually be too small to get useful measurements. I guess I'll have to try. If it doesn't workout, it's out in the street I go. Why does space have to cost so much in the Bay Area? Don't answer that. I was hoping to build my lazy susan and get a new mic stand and boom tomorrow, but the wife's Honey Do list is ever growing. In just 2 days time while I've been away earning a living she's come up with about 8 hrs worth of things she'd like me to do tomorrow if I work efficiently. Ugh. At least there will be more to do the next day!😱 Maybe I'll come up some way to fit this all in. Happy wife, happy life--I think. I'll get measurements ASAP.
Thanks again,
Dan
OK, it sounds like my living room may actually be too small to get useful measurements. I guess I'll have to try. If it doesn't workout, it's out in the street I go. Why does space have to cost so much in the Bay Area? Don't answer that. I was hoping to build my lazy susan and get a new mic stand and boom tomorrow, but the wife's Honey Do list is ever growing. In just 2 days time while I've been away earning a living she's come up with about 8 hrs worth of things she'd like me to do tomorrow if I work efficiently. Ugh. At least there will be more to do the next day!😱 Maybe I'll come up some way to fit this all in. Happy wife, happy life--I think. I'll get measurements ASAP.
Thanks again,
Dan
Thanks for the good posts from everyone.
Let us suppose the typical distance between the listener and the speaker is 3 metres and the tweeter is 1 metre high from the floor. The floor bounce travels 0.6 metre longer so it comes in 2ms. Our hearing would take the bounce as part of the original sound. Ceiling bounce would usually come within 4ms.
Consider placing acoustic absorption materials to reduce the floor bounce when doing measurements, the dip caused by the floor bounce would reduce by 3dB - 10dB (in practice, I placed up to 0.5 metre high wool batts but have not seen attenuation more than 6dB). However, when we listen to the speakers we don't have the acoustic absorption materials on the floor. While the speakers can be made to have flat response in free field, the floor bounce will create a large null.
If the speaker and listener's positions are fixed, would it make sense to include the floor bounce (and ceiling bounce) in the measurements, based on which the XO/EQ is to be designed? Indeed, would it make sense to include all early reflections within the 10ms window in the measurements?
Yes, time domain issues are important. But those issues can only be addressed with the speaker cabinet design, room treatment and speaker placement. Time domain issues can not be corrected with EQ.
If our hearing mixes sound within the first 10ms, should we include the floor and ceiling bounces into our measurements, provided that we are not designing a commercial loudspeaker but one we use at our own home with fixed positions of the speakers and listeners?
Regards,
Bill
Let us suppose the typical distance between the listener and the speaker is 3 metres and the tweeter is 1 metre high from the floor. The floor bounce travels 0.6 metre longer so it comes in 2ms. Our hearing would take the bounce as part of the original sound. Ceiling bounce would usually come within 4ms.
Consider placing acoustic absorption materials to reduce the floor bounce when doing measurements, the dip caused by the floor bounce would reduce by 3dB - 10dB (in practice, I placed up to 0.5 metre high wool batts but have not seen attenuation more than 6dB). However, when we listen to the speakers we don't have the acoustic absorption materials on the floor. While the speakers can be made to have flat response in free field, the floor bounce will create a large null.
If the speaker and listener's positions are fixed, would it make sense to include the floor bounce (and ceiling bounce) in the measurements, based on which the XO/EQ is to be designed? Indeed, would it make sense to include all early reflections within the 10ms window in the measurements?
Yes, time domain issues are important. But those issues can only be addressed with the speaker cabinet design, room treatment and speaker placement. Time domain issues can not be corrected with EQ.
If our hearing mixes sound within the first 10ms, should we include the floor and ceiling bounces into our measurements, provided that we are not designing a commercial loudspeaker but one we use at our own home with fixed positions of the speakers and listeners?
Regards,
Bill
That is a very good question but there really is no well-grounded and scientifically proven answer!
Best, Markus
From my perspective this is never a good idea. Its not like you can EQ out the reflections - only at a single point would this be possible. The only "good" way to do this is to get rid of these VER - especially the floor and ceiling. The first 10 ms should be as reflkection free as possible. Then EQ the direct field of a CD speaker and you have the best solution.
If you read some of the other threads it's like the more educated and experienced you are the more likely you are to be wrong. 🙄
I was refeering to some other threads where the people were all chest pounding about who's right, with the single common thread being that I was wrong. I don't know how I survived being wrong as much as some people think that I am. I certainly know I could never have graduated if that were the case.
Do I still owe you a book chapter? I forgot about that. Remind me when I get back to the US.
OK, your speakers with CD may be able to reduce (but not eliminate) the null at higher frequencies only. Non CD speakers won't work as there is no way to get rid of the floor and ceiling bounces which come in within the first a few ms.
So you will listen to some big null(s) and peak(s) if you don't include the floor and ceiling bounces into the measurements.
Yes I am talking about single point. I have only one chair in my dedicated listening room of 5.1m x 6.5m. My speakers are not for sale. They are for one person to listen to in that room and are not to be used in any other rooms.
But you have two ears.
Do what you like, you asked and I gave you my opinion. Get rid of the reflections, don't EQ them.
So did the mixing/mastering engineer when he prepared the music you're listening to. How did he compensate for those interferences? Did he EQ his speakers or did he EQ the sound signal?
Best, Markus
I know quite a few people with PhD's, and there is a consensus amongst them that by far the most important quality you need is to be able to be obsessed by something to the point of bloody mindedness.
You need some brain power as well of course, but the former property is far more important.
rcw.
You need some brain power as well of course, but the former property is far more important.
rcw.
I made a jig for measuring polars and did some graphs with REW of this build. 1/6 octave smoothing for a clearer picture--or maybe I did 1/3. I can't remember now. Why didn't I write it down? Notice the difference between L and R. It just shows how much baffle width has to do with dispersion. Sorry Dr. Geddes, I still haven't done the HolmImpulse. I have PC phobia since getting this Mac. I know it's a better tool. I can't do real time gated measurements, or overlay graphs. Do these still look OK? My left and right speakers are mirror images of each other so I did them both and turned them in the same direction.
On axis left speaker:
On axis right speaker:
11.25 degrees off axis left speaker:
11.25 degrees off axis right speaker:
22.50 degrees off axis, left speaker:
22.50 degrees off axis, right speaker:
33.75 degrees off axis, left speaker:
33.75 degrees off axis, right channel:
45 degrees off axis, left speaker:
45 degrees off axis, right speaker:
Thanks again!
Dan
On axis left speaker:
An externally hosted image should be here but it was not working when we last tested it.
On axis right speaker:
An externally hosted image should be here but it was not working when we last tested it.
11.25 degrees off axis left speaker:
An externally hosted image should be here but it was not working when we last tested it.
11.25 degrees off axis right speaker:
An externally hosted image should be here but it was not working when we last tested it.
22.50 degrees off axis, left speaker:
An externally hosted image should be here but it was not working when we last tested it.
22.50 degrees off axis, right speaker:
An externally hosted image should be here but it was not working when we last tested it.
33.75 degrees off axis, left speaker:
An externally hosted image should be here but it was not working when we last tested it.
33.75 degrees off axis, right channel:
An externally hosted image should be here but it was not working when we last tested it.
45 degrees off axis, left speaker:
An externally hosted image should be here but it was not working when we last tested it.
45 degrees off axis, right speaker:
An externally hosted image should be here but it was not working when we last tested it.
Thanks again!
Dan
I think that this is quite true when one is working on the degree, but you need to broaden yourself if you are going to be succesful working in industry. Industry doesn't have much use for obsessions - they want people who are flexible and can readily adapt to new projects. I didn't work in the precise area of my 'obsession" for decades. Only got back to it when I left industry. Now that I work for myself I can be obsessive again.🙂
So these are not gated measurements? They contain room reflcetions?
Thats a very course polar resolution.
The mids seems to be independent of angle and the tweeter is beaming like crazy. No apparent crossover problems however. Or isn't there a crossover?
Yea, they contain room reflections, but I did all the things you told me to do except I did them at 1 meter--like you said not to. There was no other way to get a smooth graph off axis. At 2 meters the on axis and 11.25 degrees off axis look essentially the same, but further off axis they just get crazy. My room is long but narrow. So room reflections shouldn't be too bad I think, but still worse off axis I bet. The further they get off axis, the more energy that firing straight into the wall which is only about 4-5 ft from the speaker. My room is just too small for much more distance and a clear looking graph. I'm going to do it again with Holm at my next convenience. Real time gating sounds like a great idea. I just wish I had a bit more space, my house is small.
Course resolution? Do you mean I should have applied more smoothing?
There is a crossover somewhere near 250Hz. That is approximate.
Do you think I should add a tweeter? The response from all three positions across the couch looks pretty good--a little too much of the 15k peak in the center, but the sides are flat other than the reflections and room mode issues. Maybe I could just toe them in more?
Thanks you very much!
Dan
Course resolution? Do you mean I should have applied more smoothing?
There is a crossover somewhere near 250Hz. That is approximate.
Do you think I should add a tweeter? The response from all three positions across the couch looks pretty good--a little too much of the 15k peak in the center, but the sides are flat other than the reflections and room mode issues. Maybe I could just toe them in more?
Thanks you very much!
Dan
Oh yea, that tweeter is supposed to be a 4" full range speaker and I forgot to use my calibration file for the mic. So I actually have a little more bass and 1 db less treble in the top octave. The bass starts dropping off on the mic at 100Hz, but it is slow. Down 11 db by 5Hz.
Dan
Dan
Last edited:
- Status
- Not open for further replies.