-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
How smooth can in room response realistically get?
- Thread starter Mawclaw
- Start date
I never said that decay and absorption are same things !!!
I said "decay is representative of the room absorption and directivity of the loudspeaker".
Loudspeaker directivity is also a factor : in same room, decay will be different with ie an omni, cardioid or horn speaker.
When sound wave hits the walls/floor/ceiling some of it gets reflected, some absorbed and some pass through, so your neighbours can also enjoy your music.
Ratio between those three effects varies with frequency (LF more easilly passes through than HF) and also depends of what your walls are made of. With LF the heavier they are (made from denser material) the more reflections you're getting. As frequency raises it more becomes dependent on the material of the surface of the walls.
Decay is mostly about the part that gets reflected. Surely, speakers with wider spread will cause more reflections than narrow beaming ones, but decay is still dictated by room and not by speaker.
Pio2001
Senior Member
I am not sure about all this. But here are my decay curves according to REW, measured with both speakers playing, from the listening position :
jlo told me that it was rather unusual to have these curve rising with frequency, instead of decreasing. This is however consistent with the absence of curtains, carpet or wooden floor in the room.
I suppose that if we like to have a speaker that has a flat frequency response (as measured in free field), and if this speaker is playing in two different rooms, one with a decreasing decay curve, and another with a rising decay curve (like the one above), then its frequency response, as measured from the listening position, this time, might have more treble in the second room.
Experts seem to disagree about this. Or, I would rather say that they agree that we don't know the answer.
In this video, Paul Hales gives an interesting point of view about it, from 46:44 to 50:30 in the interview :
He shows the example of a horn speaker that has a very flat frequency response measured anechoically, and then shows the frequency response of this speaker in a large room. We can see that the curves are decreasing : -8 dB from 50 to 500 Hz, and then the curves are flat above this, because the horn has a narrow directivity and what is measured above 500 Hz is not very different from the anechoic response itself.
He says that with such measurements, an automatic EQ device would flatten the low frequencies (or rise the high frequencies) about 4 dB.
Then he explains that in a large room, since we focus our attention on the direct sound, and because the automatic EQ system don't "know" that the speaker has a narrow directivity, this correction would not sound right, and we would end up manually changing the target curve.
Well he explains this better than me.
Pio2001
Senior Member
It works well here. From 200 to 10 kHz, if I set 0.2 as DI at 200 Hz (I can't read the actual value), it tells me
Critical distance at F1 : 0.7 m
Critical distance at F2 : 1.6 m
Target at F2 compared to F1 : -5.2 dB
And if I set 1 instead of 0.2, it is -4.6 dB.
These values are in good agreement with my preferred curve.
As you Pio and Scoot said, the big question is : how should be the target when the decay or the DI is non-monotonic (ie inversion of the slope as per your room) ?
Should the target keep monotonic or follow the inversions ?
In my software, I keep it monotonic. I think it avoids big mistakes but it may not be ideal for every setup.
Last edited:
Last time we discussed your corrected FR you said that you lowered 20-80Hz range by 3-4dB from standard Harman preferred curve (10dB slope over 20Hz-20kHz range) because your speakers couldn't handle LF well.
Pio2001
Senior Member
Yes, but here, the calculator takes only two points into consideration : 200 Hz and 10000 Hz. I didn't do anything fancy in this frequency range.
Frankly I don't see the point in that calculator as room EQ anyhow doesn't correct anything above 400Hz, so the natural slope, dictated by room absorption, speakers linearity and distance from LP to speakers, will remain untouched by room EQ.
Unless of course you start EQ-ing the speaker, but that is completely different story.
