So, what's the purpose of going to the anechoic chamber, and how does it typically go? I can do relatively good measurements semi-anechoically in a dedicated room I use for this, as well as outside. but it's good to verify these measurements in the anechoic chamber that I know is pretty accurate down to 40-50hz. Especially the 100-500hz range is difficult semi-anechoically.
Also, if I am to create CTA-2034-data for graphs + spinorama.org, I have to do that at Seas, since I unfortunately do not have 100,000USD for a Klippel NFS at this point in time. It's still not perfect even in the chamber, especially vertical measurements, but at least pretty good.
Doing this at Seas means taking 70(72) individual, manual measurements from different angles, which takes maybe 90 minutes, not including initial setup. So I was at Seas for about 3 hours yesterday. The screen is small (making small deviations look even smaller) and my head is mostly focused on the mechanics of it all, so there's usually not a lot of detecting issues and active fine tuning of these in the moment. What normally happens is that it looks pretty good when I'm there, and then I find a number of small issues when I look at the measurements back home, and I'm forced to go back again - which is probably the case this time around as well.
Anywhooo, let's analyze what we have then.
On-axis - overall this is very tidy, and as we already know this coaxial driver is very competent, very little diffraction even on-axis:
Reference angle (10deg) - I don't recommend pointing the speakers directly at you with any of our systems, so the reference angle is 10 degrees. Here we see a number of small issues.
Note: The overall response is +/-2.5dB from around 43hz to 18.5khz, so we're nitpicking here.
- The bass extension (-6dB at 29hz as reported by Spinorama.org) is a bit lower than actual, since the chamber reports lower bass than what we actually have, compared both to groundplane measurements and the Klippel NFS. So this is a measurement anomaly.
- There's a bit of a peak at 75hz, this is simply the sub being a bit hot in an attempt to compensate for the reduced low-end reporting of the chamber. Again a measurement anomaly
- Small dip at 200hz and small peak at 350hz. Will attempt to fix this.
- 1-3khz slightly lower in energy. This is a sensitive area where listening will determine the end response (not only the measurements), but might test to increase it a little bit.
- Elevated from 5-10khz. Not really a problem as such, but looking at off-axis graphs, there's more energy at 10khz compared to on-axis, meaning we in total get an elevated level around 10khz in the estimated in-room response. Will probably try to dampen this a bit.
If we look at 30 degrees it looks overall very tidy, but we see the elevation towards 10khz mentioned above:
Estimated in-room response + Sound power DI, we can still see the small deviations at 200+350hz, and also the elevated 10khz. So is probably possible to improve, but otherwise pretty even and nice.
