short sweeps vs long, accuracy of FR

[atledreier]

I have been tinkering with REW again, and came across something a bit weird, I think.

Looking at this graph:



The green curve is a 1M sweep (4 sweeps)
The purple curve is a 128k sweep.

To my eyes the purple is more accurate, with deeper valleys. The green one SHOULD be more accurate, as there is much more data there to analyze. If that is the case, I'm very very happy with my curve!

Still, even at 1M I can't get anything useful out of the impulse response. I see you guys can tell speaker distances to an inch from looking at impulse responses, I don't see that in mine. What should I be looking for? And with what settings? I want to analyze my room treatments and imaging, along with subwoofer timing. That's maybe for another thread though.

EDIT: Figured out the impulse thing. I used 200Hz sweeps, which is way too low bandwidth for a proper impulse.Went to 20k and all the detail popped out. So I guess I'll go reflection hunting.

[DrWho]

How loud are things when you're doing your measurement? If you're noticing that much variation, then chances are you're experiencing effects from the noise floor in your room.

Also, what are you using for a mic and how do you have it mounted?

[atledreier]

I did the measurements at close to reference level. I use an analog RS meter on a tripod located in my sweetspot.