If you wish to keep your interface, I recommend that you download the v1.8 script (still attached to the first post), attach to your post and change the instructions to reflect the fact that you need to use v1.8 and no later.
I've changed a few parts of the script to accommodate the Python interface (because it's my first attempt at programming in Python, I had to rely more on the bash script for passing status between scripts).
Okay. I will change my first post according to your recommendation.
Attached is the new 'presets' directory with decimal values (integer values are indeed not necessary) and with the 'Classical' setting corrected. I've made a quick check, it looks like the others a correct...
I assume this is because you adjusted for the preamp - it seems that the mbeq_1197 LADSPA plugin simply doesn't compensate for clipping artifacts.
Perhaps there's a better plugin we can use (I've tested tap_eq and it seems promising, but it's limited to 8 bands), but in the meantime we shouldn't correct the preamp on the VLC presets.
Attached: 'presets' without preamp correction. Changed interpolation method (gives better result on the sides) and corrected an error on the "reggae" setting. I hope that this time it's good...
I'm not so sure our hear can really make the difference between a 15-bands and an 8-bands equalizer. It just takes more time to adjust 15-bands...:razz: So, if the tap plugin is more convenient... why not?
Also, there is a quiet interesting idea for hear-disabled people. It would consist in being able to equalize separately left and right channels. Because, in some cases you can hear correctly from one hear and need some help to hear the trebles or basses from the other one.
I checked your latest presets (without preamp correction), and they still cause clipping; when I manually reduced all the bands by 3.0, clipping vastly reduced (and in some presets, disappeared).
When using VLC's equalizer (with the PulseAudio equalizer disabled), no clipping occurs even when you move the bands to the maximum/minimum positions, so the accuracy of the values themselves are not the problem.
I propose that we take your latest presets, reduce all values by 3.0 or more, test the results, and adjust any remaining presets that cause distortion.
Presets, huh. Been looking to tinker with my voice for some VOIP-enabled games.
Going to give this a try.
edit: how can I select what stream do I want to apply this to?