Typophile - 1000 UPM Limitations - Comments

Thanks for the replies. I

Mike Jarboe — Mon, 28 Apr 2008 14:10:42 -0700

Thanks for the replies. I love Typophile.

The advantage for larger UPM

dezcom — Mon, 28 Apr 2008 12:45:39 -0700

The advantage for larger UPM sizes is the control over curve and point placement as well as bezier handle placement. There is not a resolution change like you see with pixels. The grid made by the UPM controls where all points must fall so that any inbetween position is not possible. Most fonts won’t need the difference but some can be helped—those which need more curve control.

ChrisL

Is the advantage of higher

James Puckett — Mon, 28 Apr 2008 12:21:49 -0700

Is the advantage of higher UPM fonts just getting more finite control over the beziers for type that needs to look “perfect” at huge sizes?

While UPM 1000 is typical

dezcom — Mon, 28 Apr 2008 10:36:53 -0700

While UPM 1000 is typical for PostScript fonts, 2048 is typical for Truetype. If you are outputting to OpenType of either flavor, you can use 2000 if you wish or 2048 as well. There was a thread on this in the past with some posts by Adam Twardoch explaining it in great detail. Try a search on UPM and see if they turn up?
I have just had a problem with a font requiring higher resolution and bumped it up to 2000 with no problem so far. I hope to release the font soon so I will post any issues that I find in testing the higher UPM.

ChrisL

This was discussed several

Ralf Herrmann — Mon, 28 Apr 2008 10:33:47 -0700

This was discussed several times on Typophile. 1000 UPM was the standard value for Type1 fonts. Feel free to use 2000 for an OpenType font.

1000 UPM Limitations

Mike Jarboe — Mon, 28 Apr 2008 10:17:50 -0700

My understanding is that most fonts have the design space (UPM) of 1000 units and I’m curious if anyone knows why this is the standard?

The reason I ask is because I’m beginning to feel the limitations of that space when working with small scale details, most notably when creating complex angles, intersecting shapes and especially when working with circular shapes (at a small scale or while merging or deleting from).

Often times the point(s) of intersection when merging shapes or deleting from them do not land on a perfect grid coordinate, thereby distorting the end shape that I’ve created however slightly.

Does anyone have any thoughts or experience with this topic?

Is there anyone that defaults to a (UPM) of 2000 because of this issue and to work in fine detail?

It seems that a (UPM) of 2000 would be a more flexible space to work in, generally speaking.

Any feedback is greatly appreciated.