advertisement


Speaker/Room Measurement Witchcraftery

I've still not heard any explanation as to why the LS50 cannot convey Freddie Hubbard in full reach or Miles with harmon mute, and certainly nothing in the measurements explains it. However I believe Dave has heard much better results from trumpet recordings since.

I thought they were great on Miles, I heard Someday My Prince Will Come, which is one of my system tests, superb on sax too with Sonny Rollins The Bridge. Again it needs stating this was with superb amplification, many multiples more expensive than one would normally partner a speaker of that price. I’m a big fan of having some valves in the picture with a small speaker, and the pre in this case was a high-end Croft or Conrad Johnson (can’t remember which as the owner changed from the former to the latter, but I can’t remember when). They cast a huge wide soundstage with remarkable presence and heft.

I have also heard the LS50 up against a ProAc D1, the 1SC replacement (the 1SC possibly being my favourite ported small stand-mount) in the same system and they were no worse to my mind. I could happily have lived with either speaker (the owner preferred the 50s). I'd have picked the D1 as I hate the looks of the Kefs! Stands were Target R4s, and one should never underestimate how much heft and solidity they can bring to a small speaker.
 
Very tempted to get some LS50's to play around with. The actives I heard a couple of years back were superb.

I'm just wondering if that cancellation at 45hz could possibly be due to the LS50's being out of phase with the subs below the ports tuning freq (52hz). I'm not sure what the bungs will be doing. Have you measured the 50's on their own or measured with the subs inverted to see if the 45hz cancellation improves? (If it does it will only shift the problem above 52hz)

Ideally you'd have a high pass filter on the 50's around 100hz to take out the ports. A lack of time alignment between your 50's and subs will also cause anomalies in the FR, not to mention bass presentation. You can use REW to do a loopback test and determine any time delay between the 50's and the subs then re-position accordingly.


 
the subs cross at 80hz I think. even if the sub were out of phase, it cannot explain the 40hz dip.
Well it depends on the slopes and wherever there is an overlap of output you'll get a cancellation if they are out of phase..interesting that the R50 plot an their own doesnt show the deep suckout at 40hz.
 
The ZRBs are very carefully aligned with the LS50s in both the time and frequency domain.
Ports are blocked. My measurements confirm that the foam bung provided with the LS50s works just as well as the cork that plugs the fart hole in my CAOW1s, which as already stated, have almost identical port tuning and rolloff characteristics.

It's worth remembering that most music rolls off below 40 or 50Hz (not so with HT) so maybe my 30Hz boost nicely counteracts this? It certainly sounds "right".
 
Well it depends on the slopes and wherever there is an overlap of output you'll get a cancellation if they are out of phase..interesting that the R50 plot an their own doesnt show the deep suckout at 40hz.
true. s man, have you made sure the sub and mains are in phase?
 
The ZRBs are very carefully aligned with the LS50s in both the time and frequency domain.

When you said you were moving the subs around I just assumed there wasn't any time alignment.

Ports are blocked. My measurements confirm that the foam bung provided with the LS50s works just as well as the cork that plugs the fart hole in my CAOW1s, which as already stated, have almost identical port tuning and rolloff characteristics.

Yes, I see the measurements of 50's vs your CAOW's now. If the ports are in fact blocked, I'm surprised they don't sound constipated!
 
Just been looking at some measurements I took the other week after installing a REL sub with my ported Eatons. With the port stuffed I get less of my normal peak at 40Hz, but a massive troughs at 45Hz and 65Hz, with nearly as much of a dip in between. Then a big peak at 70Hz. With open port it is much more even between 45Hz and 70Hz. And yet... I ended up with the ports stuffed. So could it be that ports are useful in making measurements look nicer, but not so good at making music sound better?
 
I thought they were great on Miles, I heard Someday My Prince Will Come, which is one of my system tests, superb on sax too with Sonny Rollins The Bridge.

I thought they were pretty good with sax (albeit with a slight upper mid nasal emphasis, IIRC). With trumpet, I can imagine them doing ok with Someday My Prince or KoB. I heard them with Round about Midnight recently and they were ok with that extremely close mic'd sound too, but not great. The bigger challenge, I think, is Lift to the Scaffold. Then you also have a different challenge with Freddie Hubbard on 'Maiden Voyage' - I believe this is harder to get it right. I've heard the LS50 with good amps too - my Unico Pre/DM are very powerful and extended - but I'm still to hear them get the reach, bite and attack of Freddie Hubbard's solos.
 
Just been looking at some measurements I took the other week after installing a REL sub with my ported Eatons. With the port stuffed I get less of my normal peak at 40Hz, but a massive troughs at 45Hz and 65Hz, with nearly as much of a dip in between. Then a big peak at 70Hz. With open port it is much more even between 45Hz and 70Hz. And yet... I ended up with the ports stuffed. So could it be that ports are useful in making measurements look nicer, but not so good at making music sound better?

Additional bass point sources can be beneficial in helping to destruct modes, however with subs, ported speakers are a PITA because of the out of phase behaviour below their tuning frequency. If the subs cross above the port tuning frequency, they're out of phase below and in phase above, then vice versa when you invert. The only way to integrate them is to low pass them below the port tuning frequency using a low pass filter that mirrors the main speakers natural roll off. Better to high pass the mains at a frequency that gets the ports out of the equation provided your low pass sub filter is symmetrical.

FR measurements are very misleading when integrating subs because of the sub/room interaction. You're never quite sure what's down to the sub exciting the room or bad sub integration with main speakers. The mains/subs can be out of phase and sound crap however the room masks this in the FR measurements. You need to get the sub integration right first then deal with the room. DSP is a boon in this respect but its still a challenge getting it right. Its probably more difficult than a speaker designer integrating a mid and a bass driver in the same box.
 
Additional bass point sources can be beneficial in helping to destruct modes, however with subs, ported speakers are a PITA because of the out of phase behaviour below their tuning frequency. If the subs cross above the port tuning frequency, they're out of phase below and in phase above, then vice versa when you invert. The only way to integrate them is to low pass them below the port tuning frequency using a low pass filter that mirrors the main speakers natural roll off. Better to high pass the mains at a frequency that gets the ports out of the equation provided your low pass sub filter is symmetrical.

FR measurements are very misleading when integrating subs because of the sub/room interaction. You're never quite sure what's down to the sub exciting the room or bad sub integration with main speakers. The mains/subs can be out of phase and sound crap however the room masks this in the FR measurements. You need to get the sub integration right first then deal with the room. DSP is a boon in this respect but its still a challenge getting it right. Its probably more difficult than a speaker designer integrating a mid and a bass driver in the same box.
simple, you measure the subs separately, then the mains separately, then measure both subs and mains with the subs at 180 degree and then 0 phase
 
See post #48.
If my ZRBs were out of phase I doubt I could generate a set of room responses that are so close to the theorical figures - the difference is mostly due to the surface absorbtion coefficients.

I wasn't entirely joking when I said "My conclusion is that my speakers are so damned accurate at LF that they can be used to measure the room!"
 
ZRB nearfield measurements....

Frequency response with active filter in place:

ZRB_nearfield.jpg


Group delay:

ZRB_GD_nearfield.jpg
 
Final effort (since this thread seems to have run its course).
2nd attempt to emulate JA's measurement method. This time with the port open and no ZRBs:

LS50_full_range_JA_method_port_open.jpg


And, for comparison, the original with the ZRBs:

LS50_L_R_15_average_6th.jpg


It's hard to account for the differences above 200Hz although the 1st attempt was measuring L+ R together and the 2nd was with L and R measured separately.

My conclusions:
- Overall I think It's quite hard to get meaningful measurements and they can be tricky to interpret.
- REW is very impressive, especially for the asking price!
- The ear is the final arbiter.

Final thought:
Why does wizardry have a positive connotation and witchcraft have a negative one?
 
See post #48.
If my ZRBs were out of phase I doubt I could generate a set of room responses that are so close to the theorical figures - the difference is mostly due to the surface absorbtion coefficients.

I wasn't entirely joking when I said "My conclusion is that my speakers are so damned accurate at LF that they can be used to measure the room!"

Im just surprised you wouldn't flip the polarity of either the Kefs or the subs and run a quick measurement even just from curiosity-but it's your time you are playing with I suppose.
I'm intrigued what absorption coefficient differences can account for a 30dB suckout at 45hz-that surface could be a room treatment wonder material!
 
The difference up above 200Hz is likely just a tiny difference in mic position. I’ve found measurements largely useless for this reason to be honest! I’d try turning the sub down a few db.
 
It is a pity you didn’t measure the Kiis when they were there it would have been useful to have a known reference.
Keith
 
I have put the measuring kit away now and haven't got the energy to try a ZRB out of phase measurement.
Since I was running the LS50s au naturale I thought I would have a listen to them... quite nice speakers with an easy going sound and waffly bass (as usual from ported speakers). For the last hour or so I reverted to the LS50s with ZRBs... I call it a "condom off" experience.
 


advertisement


Back
Top