Area Text: What determines baseline position?

Question

When creating an Area Textbox object (using the text tool, click drag a box, type inside it), what determines the distance from the top of the box to the text baseline?  I notice with lower case letters that they don't always touch the top of the area box.
...Matt

Answer 1

...do TTF fonts have a setting in them that dictates the em box size compared to the text size...
No.
..., or is it some calculation of TTF font properties?...
No.
Illustrator is showing me the em box is 87.5pt when my text is 100pt.
No it isn't. If you've set the type size to 100 points, then you are telling Illustrator that the em square measures 100 points.
...Does the TTF dictate 87.5...
No.
...or is it somehow calculated?
Assuming you again have the First Baseline option for the Area Type set to Em Box Height, as you did in post 9, and if the value you are reading (87.5) is the distance you are measuring from the top of the text frame to the first baseline, then it is not the measure of the em square, it is the proportion of the designed location of the baseline as measured from the top of the em square, when scaled to the type size you have chosen (100 points), expressed in the unit of measure you are using in your document (points).
Who (or what) chooses 87.5?
You asked this same question in post 9 and it was directly answered in post 10: The designer does.
Said another way: You indirectly specified the actual measure of 87.5 points by setting the type size to the actual measure of 100 points, because that's what the distance between the top of the em square and the baseline scales to when you scale the em square to 100 points. But the value, 87.5, is nowhere in the font itself, just as the value of 100 points is not in the font itself.
In normal use, you don't know what the value of the em square stored in the font is, and you don't need to. In other words, you don't know how many em units the em square is divided into by the designer for the designer's purpose. In TrueType it is often 2048, because that's a nice round binary value. But it doesn't have to be. You can set Trajan to 100 points all day long; you still don't know if its em square is 1000, 1200, 1500, 2048, or anything else. You've just told your computer to treat that em square--whatever it may be--as if it actually measured 100 points.
As for font file format, you can launch Fontographer or FontLab and design a font on an em square of 1024 units if you want, and then export it in a PostScript Type 1 formated file, or in a TrueType formatted file, or in an OpenType formatted file.
A significant difference in TrueType is that it doesn't use cubic (degree 3) Bezier curves like Illustrator does and PostScript does, and OpenType can (OpenType can contain either kind); it uses quadratic (degree 2) Bezier curves, which are far more cumbersome and tedious to draw with, because they require more anchorPoints to achieve the same shape. (Ever tried doing any serious vector-based drawing in the drawing module of a common "works" or "office" software suite?) For that reason,  TrueType fonts are quite commonly drawn using PostScript Type 1 Bezier curves and then auto-converted to quadratic curves. One can open an existing PostScript Type 1 font in Fontographer or FontLab and export it to TrueType without knowingly changing any of the curves; but they will all be changed. Each segment will have one handle, not two.
Something needs to be said regarding the often trotted-out analogy to pre-digital (actually even pre-photomechanical) cast metal type: It is often overstated and therefore misleading when trying to understand the virtual, not material, measure system of digital type design.
Yes, many of the terms are throwbacks to metal cast type. But the physical limitations those terms imply do not apply. Many who like to expound on the terms don't seem to understand this, and lead those whom they are advising astray. (Occurrances of this in this very thread.)
Descenders and ascenders are not  limited to their corresponding lines in a digital font. Neither is x-height. Capheight (of the actual glyphs) is not even limited to within the bounds of the em square. In fact, even a glyph's sidebearings can be set to zero, because there is no metal to actually bear.
The same can be said of the user-defined actual measures. There is nothing preventing you from setting the leading (line spacing) to a measure that is less than the font's em square (type size). Hot metal leading was bars added between the slugs of lines of type. In digital type, it is the distance between corresponding positions in adjacent lines (i.e.; baseline to baseline).
So bringing up how the terms are borrowed from hot metal type is often more misleading than informative. It's better to simply think of all these "lines" associated with digital type design as analagous to the ruler guides you would use in Illustrator or any other drawing program. You set them up as convenient references and snap-to objects; but you are not actually constrained by them as you draw paths.
Look at most any normal typeface and you'll see that rounded glyphs extend beyond the baseline, x-height, and cap height; that the height of all caps is not the same, nor are descenders and ascenders. To see downright wild "boundary violations" (if that's what they really were), look at any highly decorative italic or script with elaborate swashes (Bickham, for just one example).
JET