Hi folks,
I bought a cheap SPL meter recently, and thought I'd attempt a rough n ready FR plot. So, I made a CD full of sine waves from 22hz to 20khz, in varying steps - not quite logarithmic but useful.
The bass is smoother than expected, but what the heck is going on at 900hz?! Is this severe comb filtering from my rectangular speaker baffles? What can I do about it?
I might try my hand at a sharply tuned panel filter to reduce the peak at 200hz.
Low bass and treble can be ignored as the SPL meter won't be nearly accurate for these, and subjectively they're ok. The meter is +/- 3db accurate for the rest, which is sufficient to highlight my problems.
I bought a cheap SPL meter recently, and thought I'd attempt a rough n ready FR plot. So, I made a CD full of sine waves from 22hz to 20khz, in varying steps - not quite logarithmic but useful.
The bass is smoother than expected, but what the heck is going on at 900hz?! Is this severe comb filtering from my rectangular speaker baffles? What can I do about it?
I might try my hand at a sharply tuned panel filter to reduce the peak at 200hz.
Low bass and treble can be ignored as the SPL meter won't be nearly accurate for these, and subjectively they're ok. The meter is +/- 3db accurate for the rest, which is sufficient to highlight my problems.
Attachments
This curve looks to me like bad roomacoustics, maybe you should
try improving your listeningroom.
specially the downpeaks at 70hz and 115hz might have a massive
impact on bassquality.
I once tried to measure my speakers with the pc (Hobbybox) and
a good micro, but had similar results.
I was able to get a useful measurement by covering the speakers,
a chair with the micro and other stuff with loads of blankets.
Then i had a "linear" curve from 200hz up.
But lower frequencies can't be damped like this...
Damned, seems you can't measure your box simply at home...
Michael
try improving your listeningroom.
specially the downpeaks at 70hz and 115hz might have a massive
impact on bassquality.
I once tried to measure my speakers with the pc (Hobbybox) and
a good micro, but had similar results.
I was able to get a useful measurement by covering the speakers,
a chair with the micro and other stuff with loads of blankets.
Then i had a "linear" curve from 200hz up.
But lower frequencies can't be damped like this...
Damned, seems you can't measure your box simply at home...
Michael
I'm sorry, it looks like I missed out one key point - this measurement is from as near lsitening head position as practically possible. It is not a nearfield response!MikeB said:
I am not so interested in measuring my speakers' response at this stage, just finding out what I am currently hearing.
This position is just in front of the rear wall and is about 3 metres away from the speakers.
Another thing - that dip at 115 is drawn wrong - it 'peaks' at 120hz and is on the 50db line ie. bottom of range of SPL meter! I just missed drawing the line to it, probably because it's so whoppingly bad.
The subjective bass quality is, as you'd expect, variable according to the material being played!
Some tracks sound very impressive and accurate, with little overhang, and good impact. There is generally always good extension. However, some tracks sound dull with weak bass and not enough punch - the dips may be showing themselves up here.
I measured expecting to find some big peaks to try and tame, but instead am worried about my dips
(actually, there are no audibly terrible peaks)
Can I passively boost my dips? Perhaps some resonator panel or enclosure that takes in bass and sends it out at 70hz or 115hz?
This would be nuts wouldn't it?
Some tracks sound very impressive and accurate, with little overhang, and good impact. There is generally always good extension. However, some tracks sound dull with weak bass and not enough punch - the dips may be showing themselves up here.
I measured expecting to find some big peaks to try and tame, but instead am worried about my dips
(actually, there are no audibly terrible peaks)Can I passively boost my dips? Perhaps some resonator panel or enclosure that takes in bass and sends it out at 70hz or 115hz?
This would be nuts wouldn't it?
That's a known effect to me...
Maybe you should play the material with the weak bass, and walk
through the room, checking if the sound changes. If these dips
are coming from roomacoustics, you will have hot spots. This means
at some point these frequencies are nulled out, at other points they
are boosted. This means, that the dips become at some other point
in the room peaks.
This anwers the question if a passive boost to these frequencies
are nuts...
Maybe you should play the material with the weak bass, and walk
through the room, checking if the sound changes. If these dips
are coming from roomacoustics, you will have hot spots. This means
at some point these frequencies are nulled out, at other points they
are boosted. This means, that the dips become at some other point
in the room peaks.
This anwers the question if a passive boost to these frequencies
are nuts...
Well, with so little info to work with, my guess is that if the stand's sides are ~7.5" apart, then it's a standing wave issue combining with the standing waves across the baffle face. There will also be one in the long dim, maybe the 400Hz one?
Regardless, I recommend you rethink how you have these spaced up. Adding a front/back and filling them with kitty litter, sterilized sand, or Portland cement to mass load them would be my choice. Mass loading the top is a good idea also. Heavy potted plants work well and damp/diffuse the > -3dB BW somewhat.
GM
Regardless, I recommend you rethink how you have these spaced up. Adding a front/back and filling them with kitty litter, sterilized sand, or Portland cement to mass load them would be my choice. Mass loading the top is a good idea also. Heavy potted plants work well and damp/diffuse the > -3dB BW somewhat.
GM
You need to use warble tones or centre frequency filtered pink
noise to get a good idea, a set of single tones will never give
you a good idea of whats going on, and ideally you need to
combine measurements from a number of listening positions.
http://www.interstudio.co.uk/sndchk.htm
I use this (with my ears), the warble sweep tones are great for
quick aural checks, the 1/3 octave pink noise good for measurements.
Can't find the thing at the moment.
sreten.
noise to get a good idea, a set of single tones will never give
you a good idea of whats going on, and ideally you need to
combine measurements from a number of listening positions.
http://www.interstudio.co.uk/sndchk.htm
I use this (with my ears), the warble sweep tones are great for
quick aural checks, the 1/3 octave pink noise good for measurements.
Can't find the thing at the moment.
sreten.That depends heavily on what is causing this dip. If a suckout is caused by standing waves for instance it doesn't help much EQing it out. If you feed a suckout of that type with more power, you just use more power to cancel itself out ! What helps in this case is moving the speakers and/or improving the room acoustics.
Regards
Charles
GM said:
I've never thought about the 'stand' dimensions producing standing waves before. But this is exactly why I posted - to find things I'd not thought of! I guess you looked at my crappy old website to see the speakers. I will post a pic of my current setup when I get home. BTW the side-walls inside my speakers are totally undamped, as when they were things were a wee-bit dull. Now, this is the same dimension as the stands of course - could be a major factor!!!
sreten said:You need to use warble tones or centre frequency filtered pink
noise to get a good idea, a set of single tones will never give
you a good idea of whats going on, and ideally you need to
combine measurements from a number of listening positions.
http://www.interstudio.co.uk/sndchk.htm
I use this (with my ears), the warble sweep tones are great for
quick aural checks, the 1/3 octave pink noise good for measurements.
Can't find the thing at the moment.
I always appreciate your experience-guided advice Streten, but why do you always tell me to do something complicated?
lol
Streten,
I actually have Sound Check 2! I'm not sure how to get anything useful from the pink noise and warble tones though...
The sine waves seem to make good sense, and now that I have a CD of them I can properly hunt down the troublesome rattles! Some I won't be able to fix though, such as the entire plaster dividing wall behind me wobbling at 34hz...
I actually have Sound Check 2! I'm not sure how to get anything useful from the pink noise and warble tones though...
The sine waves seem to make good sense, and now that I have a CD of them I can properly hunt down the troublesome rattles! Some I won't be able to fix though, such as the entire plaster dividing wall behind me wobbling at 34hz...
Huh ?
Use the 32 pink noise bands with your sound level meter,
and plot the frequency response at your listening position,
(and several others over an arc if you have the time).
Using my ears the sweep tones are easy for spotting
problem areas and the 1/3 octave bands great for
identifying the frequencies of the problem areas,
if you can't hear discontinuities with the 1/3 octave
bands, then simply put they are not there.
Tracks 34 and 35 are the quickest tests, but only really
bad problems are immediately apparent over these.
sreten.
Use the 32 pink noise bands with your sound level meter,
and plot the frequency response at your listening position,
(and several others over an arc if you have the time).
Using my ears the sweep tones are easy for spotting
problem areas and the 1/3 octave bands great for
identifying the frequencies of the problem areas,
if you can't hear discontinuities with the 1/3 octave
bands, then simply put they are not there.
Tracks 34 and 35 are the quickest tests, but only really
bad problems are immediately apparent over these.
sreten.
Ah, ok. Will try the 1/3 octave bands next. Thanks.
It looks like I have found the culprits of two peaks already - my internal width is 19cm - which is half a wavelength of 890hz - this is right where my biggest peak is. There is one much less important one at around 370hz, which possibly comes from the height of the stand section.
http://www.mcsquared.com/wavelength.htm
(actually, GM pointed this out, and I'm grateful to him)
Now, how much is coming from the open stand section, and how much from inside the working enclosures?
I have some 18mm ply (just made a cdp platform with some) and some 6mm MDF, and some 18mm MDF. So I can get to work on closing the stand section - just closing it on the front should work AFAICT. And if I attach plywood it should alter/improve the resonance of the speakers, presently being just MDF.
My question now to anyone reading is: how can I alter the 900hz/speaker width resonance inside the speakers, without reducing the liveliness of the presentation?
How about attaching a thin piece of wood to one side at an angle, so as to spread the resonance evenly?
Thanks for reading and suggesting!
It looks like I have found the culprits of two peaks already - my internal width is 19cm - which is half a wavelength of 890hz - this is right where my biggest peak is. There is one much less important one at around 370hz, which possibly comes from the height of the stand section.
http://www.mcsquared.com/wavelength.htm
(actually, GM pointed this out, and I'm grateful to him)
Now, how much is coming from the open stand section, and how much from inside the working enclosures?
I have some 18mm ply (just made a cdp platform with some) and some 6mm MDF, and some 18mm MDF. So I can get to work on closing the stand section - just closing it on the front should work AFAICT. And if I attach plywood it should alter/improve the resonance of the speakers, presently being just MDF.
My question now to anyone reading is: how can I alter the 900hz/speaker width resonance inside the speakers, without reducing the liveliness of the presentation?
How about attaching a thin piece of wood to one side at an angle, so as to spread the resonance evenly?
Thanks for reading and suggesting!
Simon,
I suggest you move your microphone a bit while measuring. If the microphone is fixed, it is normal in a room that you get ripples. But if you can move the mic around a imaginary sphere of approx 1 or 2 foot diameter, then, some ripples will disapear... magic
This method is better for measuring in room.
F
I suggest you move your microphone a bit while measuring. If the microphone is fixed, it is normal in a room that you get ripples. But if you can move the mic around a imaginary sphere of approx 1 or 2 foot diameter, then, some ripples will disapear... magic
This method is better for measuring in room.
F
gary f said:
Thanks for the tip - this is something I've considered, and will try to somehow use when carefully re-measuring this range.
I've just sawed a coupla bits of pine to form baffles for the stand sections of the speakers. I also put a cushion in the cavity of each one.
I held my meter steady and tried 900hz at the same level as before and compared it to the frequencies above and below... and the peak seems to have gone down by 1-2db!
I find my bass response a tad thick, could a highish level at 140hz-230hz cause this? Or maybe it's the peak at 85hz.
When I listen I hear the thickness in the mid-bass region - 45-65hz is what I keep thinking it sounds like.
I can't easily relate the character of sound I hear to the measured response!
When I listen I hear the thickness in the mid-bass region - 45-65hz is what I keep thinking it sounds like.
I can't easily relate the character of sound I hear to the measured response!
Sorry for my english... it's not so good
repeat your measurement in an anechoic room or at least in a wide open space and you shoul obtain a linear reponse but any peaks or gaps in the frecvency can be corected by equalizers
try http://www.linkwitzlab.com/ ... you can find a lot of good ideas here about how to minimize thye standing waves in your room...
unfortunatelly i don't have the resources to build such things but I have tried using other drivers and i'm pleased... they sound much better but their spl lack dissapoints the bass freaks but you can alweays use a conventional sub
repeat your measurement in an anechoic room or at least in a wide open space and you shoul obtain a linear reponse but any peaks or gaps in the frecvency can be corected by equalizers
try http://www.linkwitzlab.com/ ... you can find a lot of good ideas here about how to minimize thye standing waves in your room...
unfortunatelly i don't have the resources to build such things but I have tried using other drivers and i'm pleased... they sound much better but their spl lack dissapoints the bass freaks but you can alweays use a conventional sub
Thanks for the advice Sharky, but I'm not looking to replace my speakers right now. I've not finished getting the best from them yet. I'm also pretty much broke.[Ro]Sharky said:
I'm not looking to use EQ either, until I'm rich (haha) and can afford one of those fancy DPS boxes, and to tri-amp ie only eq the bass section.
I've worked out the cause of all the serious peaks and dips now and am considering how to improve things. I will try to report back with an improved curve soon!
My dip at 125hz is, I believe, a null of the F3 of length. This should be more like 135hz, but as the rear wall is plaster-board, the effective room length is longer I believe. That's the only explanation for the dip I can think of. Just need to try and not sit too far back to avoid this dip! I've always preferred it closer to the speakers/out from the rear wall.
The dip at 70hz is from being near the middle height and width wise, so at the F1 null for those two - which are nearly the same distance
I don't think there's anything I can do to fix this - the room is too small for manouvre.- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.