Not a member yet? Register now and get started.

lock and key

Sign in to your account.

Account Login

Forgot your password?

Serifims and Arch-angles

Sep 19 2008 Posted by in multimedia | Comments Off on Serifims and Arch-angles

[…] the rule-of-thumb for fonts was to use sans-serif, lest the serifs disappear in the laces, so to speak. Now that we are entering a world of de-interlaced, HD video, does the rule still apply?

Ah, tricky one – a “yes, but”/”no, if.”

It wasn’t that we couldn‘t use serifs, it’s just that their details are too fine at the point sizes most editors choose, resulting in twitter from single-pixel-height lines because the edges only appeared in every other field 30 times per second. Progressive displays like computer monitors, DLP, and LCD televisions are unaffected, and these constitute the majority of HDTVs as well, though tube displays (like mine) remain interlaced.

Much of the HD broadcasted is interlaced. Both 1080i and 720p formats can have about the same perceived quality, but use two different methods to get there. With 1080i, only half the vertical lines of a 1920 x 1080 frame are displayed during each of 60 fields per second, forming a complete image 30 times per second as we’re used to in standard-def (SD). With 720p, all 1280 x 720 of its pixels are displayed in their entirety 60 frames per second.

So, getting to your question, a 1080i signal displays 540 lines at a time. Standard-def NTSC, with its 480 visible lines, displays 240. Assuming comparable-sized TVs, in HD, there are more lines crammed closer to one another. But they’re still interlaced, and in theory still flicker on superfine details. However, a progressive display receiving a 1080i signal will deinterlace it first using hardware, so these artifacts are practically invisible. Though it doesn’t mean they’ll look good.

Text-wise, keep in mind that a 24-point font may appear the same from one graphic to the next regardless of the image dimensions, but most HDTV screens physically are not four-to-six times bigger than their SD counterparts (otherwise there’d be no quality improvement). So if 24 was your minimum point size in SD, you should at least double that in HD to keep the perception equal. It’s still a good practice to avoid fine graphical edges, but serifs are okay, and you don’t have to blur them anymore, just keep them at at a respectable size.

Also remember that all HDTV formats use square pixels just like computer monitors, not the rectangular (.9 ratio) ones you had to account for in 601 and DV. And sRGB is the closest Photoshop profile to HDTV. There’s much confusion about whether to use the 0-255 luminance range or 16-235. Blu-Ray definitely clamps the signal down to 16-235, as do most HDTV inputs, even HDMI. I would match that of your footage/project and deal with it in post depending on your output format.