Measuring my speakers

[tvrgeek]

You wouldn't earn to swim if you did not know there was deep water.

[DavidL]

I also have the 4 volume anthology Loudspeakers published by the AES. f you think Joe D-Apolitto's book is difficult, then try reading through that, although I have learned a lot from those too. Nothing doing well comes easily but takes lots of study and dedication. I think suggesting to a beginner Joe D-Apolitto's book is the right thing to do just to show them what they don't know yet, but if they apply themselves to learning, then it will be worthwhile.

[Barleywater]

Originally Posted by tvrgeek

You really need to change to MLS type measurements. Sweep is great for below 200, but above that, to separate the room from the speaker, you need a gated pulse.

Great results are possible with both MLS and sweep methods. Both are capable of measurement from virtually DC to Nyquist. However: MLS is more sensitive to noise in environment. Swept sine has better S/N than MLS. MLS brings out worst in slewing behavior of drivers and electronics. And: REW has perfectly useful gating features too.

[wintermute]

yes I used to solely use MLS signals but decided to try swept sine (so I could get distortion measurements) and actually found that I got better measurements with swept sine than with MLS. This was using holm impulse.

Pink noise measurements in REW (using the RTA) are good at the listening position as well, especially if you play around with the averaging. see this post by Markus for some ideas What is the ideal directivity pattern for stereo speakers?

Tony.

[tvrgeek]

If you are not using a gated measurement, you are measuring speakers in a room. To measure speakers, use the gated measurements.

When you say "better" does that mean smoother, prettier, or closer to what you think you hear?

[Pano]

Originally Posted by wintermute

yes I used to solely use MLS signals but decided to try swept sine (so I could get distortion measurements) and actually found that I got better measurements with swept sine than with MLS. This was using holm impulse.

Strange, because I seem to get identical results. Wonder what the difference is?

[wintermute]

From memory I found (at least where I was measuring) that I got less noise with the swept sine compared to MLS. I think that they both gave the same FR but I think that I could move the gating lower with the swept sine than I could with the MLS without the higher frequencies looking like they needed a shave I'll have to be a bit more scientific the next time I'm doing some measurements.

Tony.

[Pano]

Yeah, let me know what you find. I'll look at it too.

[tvrgeek]

Originally Posted by Pano

Strange, because I seem to get identical results. Wonder what the difference is?

You have a big dead room? At 1/3 octave smoothing, or nearfield, they are close. At 1M, 1/24 the differences are large. Below about 200 Hz, I use sweep.

[tvrgeek]

Originally Posted by wintermute

From memory I found (at least where I was measuring) that I got less noise with the swept sine compared to MLS. I think that they both gave the same FR but I think that I could move the gating lower with the swept sine than I could with the MLS without the higher frequencies looking like they needed a shave I'll have to be a bit more scientific the next time I'm doing some measurements.

Tony.

Are you sure it is noise, not diffraction, reflection, basket reflections and other nasties?