Binaural rendering from Ambisonics / recreating listening tests in AMBIQUAL/BINAQUAL papers

[listener17]

Hi!

Thanks for the interesting work!

How can I use BINAMIX repository to recreate the binaurally rendered signals from ambisonics (https://github.com/QxLabIreland/Ambiqual/tree/main/validation/audiofiles), as used in the listening test of the AMBIQUAL/BINAQUAL paper?

In other words:
Section 4.4 in the BINAQUAL paper describes a codec compression dataset (“..512, 384, 256, 128, 96, 64, and 32 kbps to produce a range of conditions. They were then rendered binaurally for presentation.”).

My understanding is:
MUSHRA scores are here: https://github.com/QxLabIreland/Ambiqual/tree/main/validation
But, where is the corresponding binaural content used in the MUSHRA test?

[dspanah]

Hi,

The dataset for the BINAQUAL paper is available at the following link:
https://zenodo.org/records/15431990

The codec compression subset of this dataset includes the binaural content used in the MUSHRA test.