3 dB is what one expects when adding two incoherent sources while 6 dB is expected for completely coherent sources. At LFs one wants the LF sources to be incoherent so I guess it's a good thing.
So if you are using a 4 sub set-up to help smooth the response you don't see any increase in sensitivity over the bandwidth?? There has to be some?? If you look at that it's 6db for 1/4 WL and 3db for 1/2 for MC. So for 1/2 same as a non coherent source?
Rob 🙂
Have you ever measured the freq resp of a smarthpone?
I have, and they are not even close.
Some are even equalized plus there is some additional compression etc going on.
So they are most definitely not accurate enough and only work well as relative measurements devices in the speaking range.
(as long as the compression doesn't kick in).
Not to mention the different types of microphones they use, which differs per brand, per model.
As well as the shape of the phone and position.
Even calibrating them for a certain SPL is a tricky business.
Anyway, extremely important point is that there is no knowledge how they perform exactly.
Even much less in a general sense (your mileage may vary A LOT!!)
So we simply can't rely on it.
If your answer is that they are "accurate enough", please show data that shows that this in true for many phones in general.
Even holding a calibrated omni microphone in the wrong way, especially in proximity of things (your body for example) will easily add a couple of dB's.
Last edited:
And for those following along...here's a calculator for summing incoherent sources..http://www.sengpielaudio.com/calculator-leveladding.htm
(With links for coherent summations.)
My wife, who is an audiologist, uses here Iphone meter all the time and has compared it to her pro setups with decent results. We are not arguing court cases here and tracing to the NBS is not necessary.
I didn't say that there was no increase in sensitivity.
You are still mixing up free field conditions with a closed space. At 1/2 lambda (out -of-phase) in a room, there will be nulls at many places about the room, i.e. -infinity dB gain.
How do you define what you are asking when the sum of subs in a room is all over the map in frequency and space? Sure 3 dB is a decent rule-of-thumb as an average, but the variations with frequency and location are enormous. What does "sensitivity" even mean in this context.
Hello Earl
Yeah sensitivity wasn't the best choice of words. I also recognize that any speaker placement in a room will have issues at certain frequencies due to placement room geometry and wavelength. That said I just did a sanity check and low and behold I got approximately the same numbers I did when I set up years ago.
I am also making the assumption that the listening and speaker positions have been optimized for the smoothest overall response.
These subs are a mono pair, different locations with EQ and level matching approximately 1/4 wavelength.
This is a 30Hz sinewave meter on slow and A weighting. A solid 5 with some variation both ways. Single and pair.
I understand that as the frequency changes SPL will vary and can't be uniform across the whole bandwidth. That is just like any other speaker pair.
I also agree that bandwidth limited noise would be a better test signal.
Rob 🙂
Attachments
This has absolutely zero to do with "court cases". Kinda dramatic thing to say.
But just one anecdotal personal sample is not enough to say that every smartphone is just fine estimating sound pressure levels.
"Decent results" also still doesn't provide any real numbers btw.
Some phones might be reasonable, but like I said before, your milage may vary a lot.
It's just bad engineering and science to just assume these kinds things.
It's just poor practical advice to use as a general statement for people who are unknown to all of this.
I am personally always very careful claiming such statements, because I know that the average reader, reads it as the truth.
I have and tested several phones, I wasn't surprised how high the error was.
Best cases were good, within a dB or so. Worst cases where off by more than 3dB.
Some having quite the error in the frequency response, meaning that in some cases you could over estimate certain things way to much.
There was also quite the discrepancy in how to hold the phone, as well as how dirty, old and dusty the microphone was.
Conclusion still is that your milage may vary an awful lot.
There are way to many variables, incl practical ones, to be able to rely on these devices.
If you wanna have an extremely rough estimate if something isn't 30dB(A) or 100dB(A), sure.
But just can't quantify any numbers to it.
Even with cheap microphones for measuring, there is quite some discrepancy.
There are plenty of articles online that show how bad this is.
The more you know, definitely, I did question the accuracy. I was more worried about scaling but non the less I compared results with my u-mik and found it surprisingly precise. This is some time ago, and what I remember is feeling like, wow.... so the spl meter is precise and the theater really did hit those super loud peaks?
I would definitely consider the quality, I mean this thing lives in my pocket, I do construction, a material, a liquid, leaving its signature on the diaphragm is a possibility.
That was then, Now I have a newer iphone, so scientifically I would want to at least re evaluate the situation. I am currently waiting on a cord for my U-Mik.
Rob,
Using "A" weighting, 30Hz reads -40 below actual level.
The reading of 80.3 dBA (slow) could be 30Hz at 120.3dB SPL, 2nd harmonic (60Hz) at ~105dB SPL, 3rd harmonic (90Hz) at ~60dB SPL, or any combination of fundamental and harmonic distortion adding up to 80.3 dBA.
At any rate, we can't tell which frequencies are adding to the +5dB gain in those two different locations from a dB meter.
Thanks hey if you mess up might as well go all the way! I confused A and C will give it a try using C. As soon as my furnace shuts off! Cold here!!
LOL Rob 🙂
Last edited:
Hello weltersys
Thought about the balance of your post about harmonics and did an FFT comparison using Clio Pocket. Just loaded the new 3,0 software so I had to give it a spin. Getting the same 5 Db and looks clean to me WRT harmonics not being an issue.
Rob 🙂
Thought about the balance of your post about harmonics and did an FFT comparison using Clio Pocket. Just loaded the new 3,0 software so I had to give it a spin. Getting the same 5 Db and looks clean to me WRT harmonics not being an issue.
Rob 🙂
Attachments
Yes, the third harmonic at -41dB below the fundamental (about 1% HD) is no issue (how about at 120dB..), and the gain of 5dB at that location and frequency when doubling cabinets is near the ideal of 6dB.
Other frequencies probably won't be so close to the ideal 😉
Other frequencies probably won't be so close to the ideal 😉
About 10 years ago an old Holden hearse was sold here. The ad stated that this catafalque was used to deliver bread from a bakery😎.
For some reasons I didn't buy this vehicle.
Last edited:
Time to dampen the slot...
I've got 120"x120" of cotton batting. I understand that density plays a roll in the efficacy of the sound absorption . The batting in multiple layers should get the job done don't you think? Less mess than rockwool, less messy than poly/cotton/wool fill, supposedly more effective than polyester batting due to tighter structure.
I've got 120"x120" of cotton batting. I understand that density plays a roll in the efficacy of the sound absorption . The batting in multiple layers should get the job done don't you think? Less mess than rockwool, less messy than poly/cotton/wool fill, supposedly more effective than polyester batting due to tighter structure.
Density plays a role, but thickness is usually the dominate factor. To work at LFs it has to be thick.