Draft: DO NOT MERGE Fhg/check gain problems

This is just for communicating the findings on checking unexpected excessive application of the limiter, pointing to potential problems in the decoding/rendering. The code and scripts in the branch were modified to put out some limiter statistics and find suspicious test points. The test were run using the following scripts: ci_linux_template.json smoke_test_limiter.sh by running

./smoke_test_limiter.sh FhG/check-gain-problems

The following results were created with suspicous values (the thresholds for marking a value as suspicious were set relatively low).

limiter__er.html

limiter__ext_orientation.html

limiter__ext_orientation_jbm.html

limiter__fec.html

limiter__headrot.html

limiter__hrft_fb.html

limiter__hrft_swb.html

limiter__hrtf_wb.html

limiter__jbm.html

limiter__non_ism_default.html

Generally apart from the problems that are relate to #987 (closed) and !1382 (merged) the only other identified cases are also MC and there mostly output to Ambisonics. I checked a few of those manually and this looks not like a real problem, but more because of the nature of the long test item used in the test.