Interrupting Impulse Response

[Ciscokid]

Hi, I was reading some of the other impulse threads trying to figure out my graph. I am gathering that when measuring the impulse response we are looking at reflection points and time it takes to reach the mic. Correct? If so, what are we looking for a huge spike? Should I be tring to achieve a smooth as possible pattern close the the zero line?

One thing I noticed is when I moved my speakers back closer to the wall my spike at around 5ms droped. Another thing that is on the graphs I posted below, there is a difference of 15ms from the start of the spikes in the right speaker as apposed to the left speaker. What does this mean?


There is no acoustic treatments as of yet. I am planning on putting up some traps in the corners and first reflection point panels. The room is 14 feet wide and is open to the kitchen. So I have no back wall unless you are counting the back of the kitchen as my back wall.Which is around 55 or so feet away. I have also put the filtered respones of all speakers. Looks likes its finally time to order that BFD.

[Wayne A. Pflughaupt]

No replies to this? Are Frankie and I the only ones who want to know something about impulse response?

Regards,
Wayne

[DrWho]

Could you post the impulse responses with dB FS? It'd be easier to read. It's the Energy Time Curves that you wanna plot in %FS and will have more meaning as it pertains to acoustical treatment.

Basically, the impulse response is showing you the magnitude of sound pressure level as it changes with time. Everything you see to the left of the impulse peak (which sould be centered at time=0) is going to be distortion. Basically the program sends a sweep to the speakers to ensure that they play every frequency and then squishes the response back together - pretending as if every note played at the same time. The result of this is that higher frequencies present when the lower frequencies are playing will get shifted to the left - arriving before that frequency should be arriving. So basically everything to the left can be ignored.

To the right, we're going to see all sorts of things. We'll see the arrival-time differences between drivers, any undamped resonances in the drivers/cabinet, reflections throughout the room, and even resonances of the room.

[Ciscokid]

Hi DrWho

Here are the Graphs with dB FS for impulse. I am not sure I am using the correct scale, so I hope this is ok. I also posted the Time energy graphs for left and right speaker. The Left Speaker is in blue and the right is in red.

[DrWho]

Do you hear clipping or distortion upon playback of the frequency sweeps? That behavior to the left of time = 0 is not normal at all. In fact, your ETC for the left speaker looks exceptionally bad...is there anything nearby that might be obstructing its sound? I almost wonder if you didn't bump the mic or anything while taking the measurement. You could always deconvolve your impulse response back into the measured data and listen for any abnormal sounds upon playback that might be contributing to the wierd data.

If you could post your REW file I could do the deconvolving thing and listen for whatever craziness might be going on.

[Ciscokid]

DrWho wrote:

. You could always deconvolve your impulse response back into the measured data and listen for any abnormal sounds upon playback that might be contributing to the wierd data.

If you could post your REW file I could do the deconvolving thing and listen for whatever craziness might be going on.

Not quite sure I follow what you mean by deconvolve .Is the REW file you are requesting the file with .mdat extention? If so the forum won't let me upload it do to the size.

I got back kind of late from my meeting so tomorrow evening I will run some tests to make sure I am gathering good data. I am thinking of swaping speakers side to side and retest. Here is a photo of my room, as you can see the only thing kind of it front of the speaker is the couch and end table.

[Ciscokid]

Ok got some time to run some tests. First off I recalibrated my mic and sound card file. I reran sweeps for both speakers. The left and right speakers still have the same impulse graph. I tried swapping speakers. That did not change anything.

If you look in the photo I posted,I moved the left speaker closer to the center channel and that moved my impulse starting point closer to zero. So I am not sure what that means if anything. I posted both energy graphs for the left speaker. The first one is in its normal location. The second one I placed the speaker just to the left of the center channel. The 3rd is just the impulse with it close to the center channel.


I did not hear any clipping during any of the sweeps I ran. I am still curious what you mean by the deconvolving. Please let me know what file you need. Thanks for all your help.