Skip to content

settings_scale

just-Addict edited this page Jun 16, 2016 · 8 revisions

##SETTINGS

###scaleY

The seekbar-with-gimmicks has two main methods to scale the waveform to your preferred height. They are replay gain scaling and minmaxrms scaling, from here on referred to as magnitude scaling. Replay gain scaling will of course only do something if the track contains replay gain tags. Magnitude scaling will always work regardless. Both methods will use the scaleY setting to determine the amount of scaling to be applied.

The default value is set to 0.95 but if you are going to use the time scale it is adviseable to set it not higher than 0.8 so that a peak heavy track doesn't make it harder to see the ticks on the scale. They do automatically take on the inverse of whatever color is shown near top and bottom borders, but alternating colors from many successive peaks reaching into the scale area may cause confusion.

Below images show the effect of this setting:

scaled Scaled to 0.95

not_scaled No scaling applied

Although it is not advised, you can supply a scaling factor larger than 1.0 which will make the above waveform look like this: over_scaled Over scaling to 1.35

Situations where you might want to over scale likely won't come by that often, but in cases where a track only contains a few high peaks and these peaks don't show in the final waveform*, over scaling could be used to still make the waveform fit the window (for whatever reason you see fit).

*Due to how the component works, when a track is analyzed it is invariably going to be scaled down in terms of samples. With a maximum of 2048 samples per track this may sometimes result in a single high peak getting averaged down due to neighbouring low values just before and after that peak in the same sample. If that single peak does make it into one of those 2048 samples, it still can be averaged out due to the low peak neighbouring samples and/or size of the component window. Another major factor is track length, as this means that per sample a longer period audio data is going to be averaged into a single sample. The final result is that a track will still be scaled according to the highest measured peak when the track was analyzed, but may visually end up filling only say 3/4 or less of the full height of the channel segment/component window.

When you think this is the case, tools such as Audacity which are also able to display a waveform, will allow you to zoom in and verify the existence of these peaks. Another thing you could do is open a second wavebar window and use this effect in scrolling mode in the second window, together with the zoom highlight directive. If the track does contain only a relative few high peaks and if these did make it into the samples then it will turn up in this view as the scroll mode creates a view that stretches the waveform to a size several times larger than the length of the component window. It also will cause the component to use the highest detail level so that every sample will get used to draw the waveform. However, since the sample resolution is still relative low, there will not be enough information to draw the waveform in a higher resolution, so much of the data between sample points will be interpolated data which makes scroll mode look like a regular waveform at a lower detailLevel setting

Besides the scale setting there are also a couple of scale directives that will change the way scaling affects the waveform.

Clone this wiki locally