Watching the video, I noticed, that I tried it in a different way:
Load a file with the original encoding.
Set File>Options>Mpeg to write ISO-8859-1
Return to the files and select them.
Now press Ctrl-S to save the new encoding.
I can see that it has been done as I watch a column that shows the encoding.
The special characters (I have German umlauts) are still displayed.
BUT: when I now change the code page for e.g. TITLE to e.g. traditional Chinese, I get a ? for the umlaut as apparently this character has no match in the chinese character set.
Another experiment:
Take a file with UTF encoding.
Apply an action to convert the code page.
You will notice that any special character is now translated to that code page
Braitsa (Fantasiestück) becomes Braitsa (Fantasiestьck)
(the encoding stays at UTF)
If I now modify the settings for saving to ISO then the title becomes
Braitsa (Fantasiest?ck)
So, the code page translation always is relative to the OS code page, I think.
If no matching character is found in the current OS environment then you get a "?".
UTF-8 and ISO have only 8 bits to address a character which leads to 256 variations. This is not enough to cater for all the different languages.
UTF-16 provides 32767 possibilities. This is enough to provide characters for any language in the world.