TextMesh Pro Font Render Mode.

If you have not already done so, I suggest you watch the Font Asset Creation video which covers the important aspect of font asset creation as well as these rendering options.

I also recommend you watch the Working with Material Preset video as these are also very important concepts to understand with TextMesh Pro.

P.S. Distance Field 16 is the default / recommended choice.

 

SDFAA should be fine for pretty much all use cases with perhaps the exception of when you will be rendering large text such as titles that exceed 90 point size on screen with large outline or shadow.

 

Every mode has it PROS and CONS and use.

Having said that, I now use SDFAA 98% of the time so that makes it my default go to

 

This is because the wrong shader is used. Make sure that once of the Signed Distance Field shaders is assigned to the material used by this font asset.

I suspect that you might have first create a font asset using one of the bitmap modes and then switched to one of the SDF modes after. As such, the material was first assigned a bitmap shader.

 

The SDFAA is still the goto for dynamic font assets. For static font assets and mostly large text (72 point size or larger) that is heavily styled like titles, SDF16 is usually the goto.

In terms of the visual artifacts, these are usually the result of a Sampling Point Size to Padding Ratio that is too small. You should aim for a ratio of about 10% which will give you a decent effective range on material property while allow you to maintain a good sampling point size. There are several posts detailing the Sampling Point Size to Padding ratio and how it affects / works with fallback font assets, etc.

Since bold adds dilation which is affect by this sampling point size to padding ratio, this is why you would see this when using bold if your ratio is small.

When inspecting your font asset, what is the sampling point size and padding values?