As discussed on Matrix, there is a bit of confusion when it comes to audio handling in kdenlive.
- Generally speaking, “Volume” and “Gain” is not the same thing.
“Volume” adjusts level in the recorded range while “Gain” amplifies / boosts the Signal, therefore can add Noise etc. Usually it is clearly distinguished what is what, i.e. in Mediacomposer, Volume is a prominently placed mixer slider, while Gain has to be applied as an extra effect. This is probably to discourage inexperienced editors to unintenionally use Gain.
Currently, in kdenlive, there is no differentiation bwtween gain and volume and its not clear what the effect does: it is called “Volume” effect, it’s slider is labeled “Gain”, and it seems as if it does both, depending on where the user moves the slider. Only there is no UI feedback about it.
Apparently all audio level related effect go back to this melt effect:
Setting the Max_Gain variable to 0 would make this a pure volume adjustment. But generally speaking, the features of the MLT effect are not well represented in the UI dialog.
IMHO, all this functionality could be transparently united in one effect with a better UX:
-
A checkbox which enables / disables gain, with a fixed Max amount which could be defined in the settings.
-
When gain is enabled, the slider range goes to a higher dB level then 0, defined by the max_gain setting. The “normal” slider should use a “relative” project wide definable scale with a definable zero level (in Europe / EBU space normally -18db but depends on the region) and a visible headroom part, to make the user easily understand that technically, 0db is the most they can adjust a clip to. The “real” level of the clip could be shown as well, important is the relative volume.
-
A “normalize” button with a target volume number field and an accompanying switch, which toggles between RMS and Peak target volume. Clicking the normalize button auto adjusts the clip to the desired volume. If desired volume is not reachable without gain, it asks to enable gain.
-
Visible feedback (i.e. level is displayed with a red background, when adjustment is in the “gain” region
-
Information field which displays RMS and peak volume for the selected clip
In a lot of use cases, this would make the “embedded” Volume Effect the only audio tool that needs to be accessed and adjusted by a user.
I would personally vote for the embedded effects to be enabled as a default in the settings AND being enabled in the effects stack, to minimize Interaction. Maybe they could be “invisible” disabled if no parameter in them has been changed to keep them out of the processing pipeline if not needed.
The seperation of the different MLT effects features in distinct effects (Volume, Gain, Normalize etc) could be removed and the amount of audio effects reduced for better UX.