kcat
commented
9 months ago Unless there's some bugs in the diffuse-field equalization and surface-weighted diffuse-field average algorithms, the quality of them will be dependent on how closely a given HRTF matches your head, and the frequency response of your headphones (diffuse-field equalization effectively works by taking out what's common in all the responses, leaving just the differences between the responses, which tends to work better across the range of frequency responses headphones can have).
As for the SCUT KEMAR HRTF itself, unfortunately it can't be included with OpenAL Soft. At least, according to the SOFA file of it that I have, the license is CC 3.0 BY-SA-NC, the NC being Non-Commercial which is incompatible with the LGPL that allows commercial use. You can certainly make, use, and distribute .mhr files created from it in accordance with its license, but it can't be distributed built-in to OpenAL Soft itself without violating one of the licenses. Unless there's a different version of the dataset with a different, more compatible license.
TheBestKmanEver
ThreeDeeJay
i did some playing around with 'makemhr' with the SCUT KEMAR HRTF (Near Field with multiple recorded distances) and found that i have better locational sound when i disabled the diffuse field EQ and the surface weighted settings for making the MHR, it is kinda weird sounding at first but i do recommend running 'alhrtf' to test to see if you like it after making the mhr
also i noticed a small issue with the the scut hrtf on the hrtf database on airtable, the radius is set at 0.105 rather than the default of 0.09 (oops this was my fault actually lol)