Hilde45, the impulse graph provides a single (multi frequency, I think) short sound (the impulse) and records every reputation of that impulse until it completely decays. You don't need to run the scans again, just select the impulse tab in the row of options above the frequency graph. Set the graph limits at -0.002 on the left and .050 on the right, 0db on the top and -65 dB on the bottom. Each vertical line is a reflection of the initial impulse. You would like to have everything from 3m to 20 m on the X axis at -20 dB. Anything above -20 dB will erode imaging because the brain will merge those early reflections with the 0m impulse signal, thus obscuring the spacial information in the recording. Anything longer than 20m will be perceived as room spaciousness.
Next, set the limits to -.002m on the left and 0.5m on the left. I set the bottom limit to about -85 db. This will show you how long it takes for the impulse to decay to the level of room noise, which is really a picture of how spacious your room will sound. This is a matter of personal taste. Some want as close to an anechoic chamber as they can get. In other words, those folks advocate that everything after the original impulse should be at the level of room noise. Most of us like added spaciousness of the room. My room currently shows a gradual decline to the level of room noise (the more or less flat zero slope area at the right part of the graph) at 0.3m. m is time in milliseconds.
Next, set the limits to -.002m on the left and 0.5m on the left. I set the bottom limit to about -85 db. This will show you how long it takes for the impulse to decay to the level of room noise, which is really a picture of how spacious your room will sound. This is a matter of personal taste. Some want as close to an anechoic chamber as they can get. In other words, those folks advocate that everything after the original impulse should be at the level of room noise. Most of us like added spaciousness of the room. My room currently shows a gradual decline to the level of room noise (the more or less flat zero slope area at the right part of the graph) at 0.3m. m is time in milliseconds.