Somny

I made this model to use my Sony SRS-XB100 as a portable omnidirectional speaker for room acoustic measurements using Room EQ Wizard or custom routines.

The extended base includes several features:

• Slots for gluing in the 5/8 to 3/8 microphone stand adapter, allowing the speaker to be mounted on a regular microphone stand.
• Slots for M3 heat inserts to secure the two parts together.
• Hollow legs for optional reinforcement with glued-in steel wire.

The directivity patterns for both diffuser variants have not been tested yet, as these are still early prototypes. The cone is expected to create a narrower directivity pattern 90° off the speaker axis, while the parabolic design is intended for wider vertical coverage.

The vetrical coverage of the inverse parabolic variant was measured under free-field conditions and presents the following vertical polar pattern calculated according to ISO 3382-1.

In the context of the standard mentioned, this variant can be used for room acoustic measurements.

Both parts were originally printed from PETG with organic supports.

To properly print the hollow sections of the diffuser, support blockers should be placed through the legs. The diffusers will likely print better when oriented upside down.

Any suggestions are very welcome!

Polar patterns

Inverse parabolic variant (diffuser_parabolic_inv)

The vetrical coverage of the inverse parabolic variant was measured under free-field conditions and presents the following vertical polar pattern calculated according to ISO 3382-1.

In the context of the standard mentioned, this variant can be used for room acoustic measurements between 250 Hz and 4 kHz. At 125 Hz, the requirement is not fulfilled, but not drastically.

Reverberation measurements using REW

The following graph compares the differences in T20 and T30 measured using

• Somny with Room EQ Wizard (REW) software (with the diffuser_parabolic_inv)
• B&K Type 4295 with Dirac 6 software

The comparison is done for 8 microphone-source configurations, the DBX DriveRack RTA-M microphone was used for all the measurements.

These results show that, overall, the measurement setups don't give significantly different results, but the per-band-analysis reveals some significant differences. In the end, the user needs to assess whether the difference is acceptable or not.