Expert Advice Wanted Here
- Thread starter jcmusic
- Start date
I see what looks like your crossover points at 80hz, 400hz and 4khz and a microphone that was placed about half way height of the speaker. What were you trying to measure?
david
Those are two nearfield measurments one meter from each speaker, the mic is level with the mid horn and tweeter. I was tryint to measure the speakers without the room effects...
Last edited:
Was there a reason for you picking one meter? If you want room effect then you need to back off and take measurements from seating position at different heights including ear level followed by further measurements around the room, mic position is dependent on what you're trying to figure out and fix. I'm not familiar with your mike, if its very directional you'll even end up with different readings based on the angle you point the mic. Your listening chair will also affect the results, you run different measurements with and without it, specially if it has a high back.
david
If that's at 1m, that's pretty poor response. Even in a live room at 1m you should have fairly accurate data above 3-400hz. I'm assuming not, but was any gating used? If not, try setting the gate to about 300hz.
The 80hz suckout is a room node most likely, not a crossover point. And looks like another room node around 45hz. Based on that I'd guess it's a mid-small sized room in the 13'x18' range. And the left speaker is exhibiting more room interraction than the right. The 400hz suckout is more troublesome because it's likely not attributable to the room, but the speaker itself, and that's a huge dip in a critical range. Then a sharp rise through the midrange into low treble, then lobing effects in the 4-7000hz range, and a sharp falloff on top.
The only good thing I see is no floor bounce, does this speaker have a woofer close to the floor?
The 80hz suckout is a room node most likely, not a crossover point. And looks like another room node around 45hz. Based on that I'd guess it's a mid-small sized room in the 13'x18' range. And the left speaker is exhibiting more room interraction than the right. The 400hz suckout is more troublesome because it's likely not attributable to the room, but the speaker itself, and that's a huge dip in a critical range. Then a sharp rise through the midrange into low treble, then lobing effects in the 4-7000hz range, and a sharp falloff on top.
The only good thing I see is no floor bounce, does this speaker have a woofer close to the floor?
That makes sense, 15" woofer within a 1/4 wave of the floor = no floor bounce cancellation.
To be completely honest, if I were dealing with a speaker that measured like that, I'd rip the crossover out and start from scratch.
If you can get inside the box, try flipping the polarity on the midrange and measure again. That would tell a lot about x-over points and summation problems. Still though, it would need a from scratch crossover design to 'fix' the issues.
Maybe a 2 way system with woofer crossing to the fullrange unit at 400 hz or a 3 way with the mid crossing at 4000 hz , probably a X over designed by " EAR "
JC music I just now read it was klipsch horns with as you say a 2 way( 2 inch full range horn unit) , got that right , 2 inch horn 400 hz - 20 kHz . woofer goes to +- 400 hz or did I miss something
Probably the worst (IMO) price /FR response ratio ive seen in a commercial design was the Kaiser kawero Chiara ,X over stuffed full with duelund stuff but that response auwch
it was in a german mag.
Last edited:
I have to add that in general horndesigns measure quite poorly , probably in a large part due to the used hornunits themselves
There's no reason why poor measurements should coincide with use of a horn or waveguide. We have a couple waveguide based designs in our lineup that measure like rulers, both on and off-axis.
The difference is the professional vs the hobbyist. I come across this all the time within the vintage horn world where people randomly add tweeters or subs to extend the range beyond what the speaker was initially designed for. They sound exactly like what you see in the graph above where each driver almost becomes a standalone unit. The difference in lower frequency response is probably the result of differing corners.
david
Sorry I forgot to add the tweeters, so the speakers are 3 way crossed at 400hz and 6000hz...
Maybe a 2 way system with woofer crossing to the fullrange unit at 400 hz or a 3 way with the mid crossing at 4000 hz , probably a X over designed by " EAR "
JC music I just now read it was klipsch horns with as you say a 2 way( 2 inch full range horn unit) , got that right , 2 inch horn 400 hz - 20 kHz . woofer goes to +- 400 hz or did I miss something
Probably the worst (IMO) price /FR response ratio ive seen in a commercial design was the Kaiser kawero Chiara ,X over stuffed full with duelund stuff but that response auwch
You don't know where he measured the speaker. Therefore, you can't say much about the response. Your advice is nonsense.
Jcmusic,
You need to window out the room effects before you can start to draw conclusions about the FR graph. You can't do that without the impulse response. The impulse response will show the time domain where the room reflections start. You could work back from there in terms of FR. You need to show us where the speaker was in relationship to room boundaries and tell us room size for anyone to draw competent conclusions. Also, it would be nice to see impulse response and group delay too.
Last edited:
Well I may have found the problem i need to measure to be sure but, I am 99% sure I have found what was causing all my trouble!!!
The best you can hope for in a medium size room is good HF measuremts. For example, I had my speakers in the center of a 14'6"x24' room. I had to window out everything below 1khz.
Don't make any drastic decisions based on this measurement without knowing more about what is the speaker vs room.
Real highend speaker manufacturers measure in enormous warehouses with a large crane to get a good full range response or use an anechoic chamber.
Michael.
Don't make any drastic decisions based on this measurement without knowing more about what is the speaker vs room.
Real highend speaker manufacturers measure in enormous warehouses with a large crane to get a good full range response or use an anechoic chamber.
Michael.
Last edited:
Sorry, my advice is NOT nonsense. I've taken tens of thousands of measurements in all different room environments, I know how to interpret the data. If you could read you'd also see I suggested windowing the measurement.
The key is he said the measurement is taken at 1m. That's close enough to know we have pretty decent data down to 3-500hz. Everything will be more jagged than it is with gating, but still not enough to overwhelm the trends shown.
You only know one variable, the distance of the measurement. You also NEED to know the distances of the various boundaries in the room to know how far down the measurement would be accurate. You also don't know what kind of mic he is using. So, yes, your advice for ripping out the crossover is totally bogus.
Unless he's taking the measurement inside a glass box, it doesn't matter what the room looks like! No living space will alter the response to the point of rendering the data useless when taken at 1m. I'm not telling him to model a new crossover design based on those measurements, that would be bad advice. I'm using the measurement to view trends, for which it is totally valid.
I agree with vapor 1 but I find 1 meter to close , I personally measure also a lot at 2 meters or 2,5 which is still well within distance to side or back walls .
LS units do need a certain distance to blend together and form an even response , at 1 meter you might see a dip at the x over point which will be gone at the listenerspot for example 2,5 meters .
Measuring at 1 meter in a acoustic dead room doesn't tell me much about the design
LS units do need a certain distance to blend together and form an even response , at 1 meter you might see a dip at the x over point which will be gone at the listenerspot for example 2,5 meters .
Measuring at 1 meter in a acoustic dead room doesn't tell me much about the design
Avantgarde specifically warns against measuring their speakers at distances under 2.5m. But many reviews show measurements taken at 1m, even with large speakers. Fortunately the measurements shown in Soundstage are taken at 2m and than scaled to 1m distance.
Similar threads
- Sticky
