#90 -5 dB Pregain in Stream Output Format

[Moonbase]

Could we get a "-5 dB pre-gain" option on the stream output ("format" tab)?

Rationale:
Most of us nowadays use ReplayGain 2.0 tagging, thus effectively targeting the program loudness at -18 LUFS. Now EBU R128 actually recommends -23 LUFS and some stations actually experiment with the reduced loudness (= more headroom for clipping-free high-dynamic content).

It’s often (especially in the experimental phase) not feasible to re-gain the whole rotation to -23 LUFS, so I’d just be great if we could use the existing files and simply add a "pre-gain" of -5 dB (LU) for the output of the experimental streams.

[Moonbase]

… so ~~I’d~~ just be great … → … so it would just be great …

[Moonbase]

In python/format.py in class FormatPregain (starting at line #498), change FormatDropdown.__init__(… to:

        FormatDropdown.__init__(self, prev_object, _('Pregain'), "pregain", (
            dict(display_text=_('0 dB'), value="1.0"),
            dict(display_text=_('-0.5 dB'), value="0.944"),
            dict(display_text=_('-1.0 dB'), value="0.891", default=(codec in ("aac", "aacpv2"))),
            dict(display_text=_('-1.5 dB'), value="0.841"),
            dict(display_text=_('-2.0 dB'), value="0.794", default=(codec in ("mp2", "mp3"))),
            dict(display_text=_('-2.5 dB'), value="0.750"),
            dict(display_text=_('-3.0 dB'), value="0.708", default=(codec == "speex")),
            dict(display_text=_('-5.0 dB'), value="0.562")),
            1,
            _("A blanket gain reduction to promote audio quality when using long established lossy audio codecs like mp3 with loud audio material.\n\n"
            "ReplayGain makes this feature generally unnecessary and the correct setting in that case is 0 dB."))

Tested on Linux Mint 19.1 (Ubuntu 18.04-based).

[Stephen Fairchild]

I pushed an update that does pre-gain down to -5.0 dB in half dB steps.