My client wants to add three new weights to and existing 12 font family. I remember struggling with the whole family naming and style linking process for the original twelve a couple years ago and these additions have me worried…
The original set contained roman and italic Black, Bold, Semibold, Regular, Book and Light weights. They want to add Extralight, Thin and Hairline.
The numerical weight value for the existing Light is 300. If I use the FL defaults, the next lighter weight values are Extralight - 200, Thin - 100, and then Ultralight back goes back up to 200.
I don’t understand this progression. Would I get into trouble if I went with Extralight - 250, Thin - 200, Hairline - 150 instead?
And another concern — I am starting to draw the Thin glyphs and I am realizing I don’t quite have the fidelity I would like for these really thin stokes. I am thinking about changing the UPM size to 2000…
But is is smart to do that when all the rest of the family is set the standard 1000 UPM?
I have a lot of drawing to do but I would feel better if I could figure out some of these technical issues in the mean time and could really use some advice!
Comments
(FontLab)
…by 20 or so. Then interpolate this with the Regular to make the other thin and light weights.
If it doesn’t seem like you are getting enough precision at UPM 1000, add an extra BCP to problem joints—the slight offsetting will never be noticed.
The whole linking thing is still stressing me out. I have such a difficult time with that whole process. I have printed out Lücke's treatise and I study Twardoch's charts. But still...
Maybe it is just me but, dang, that system seems crazy and convoluted. Makes my head spin!
My client want the whole family to link up and display properly in the browser list from Thin to Black.
I am still in the drawing stage. I will start testing soon but things that seem fine on my mac don't always work on other and older systems.
You can also hire people to master the fonts for you.
It says I CAN put in my own custom values. I am worried about wether or not I SHOULD.
I may consider hiring somebody with more experience dealing with this sort of thing in the end but I would like to give it a shot myself first.
If you're hoping to see a piece of Adobe software knocked off its perch, look for Sketch (or other competitors, including Adobe's new XD) to replace Photoshop as the de facto web design tool.
I do recommend not going below 250, else you will have problems in a significant portion of Windows GDI apps. If you and your client don't care about them, no worries. But my general experience is, such a decision is likely to bite you eventually.
Then again, CSS only accepts multiples of 100 for specifying weightclass for the web, despite my protests, because the W3C wants to make it impossible for us to make a font family that works everywhere. (OK, that's my slightly biased/frustrated interpretation of the discussions. Obviously I was not happy with the outcome.)
The confusion existed because the initial CSS implementations must have read the examples and default values and interpreted them as hard law.
Technically, it doesn't. But what happens is, if a font has a weight value of less than 250, the GDI renderer would artificially embolden the font, supposedly to ensure that the font doesn't get so thin that it is difficult to see or print. In practice, this causes very thin fonts (below 250) to be displayed bolder than less thin fonts (250 or higher). To avoid this issue, the "best practice" is to avoid values lower than 250.
I guess if you don't care about what happens with GDI, then nothing's stopping you from using weight values below 250.
What!? We can make font families that work anywhere thanks to the support and threats of the w3c to get woff done.:)
The issue here, I think, is responsiveness of several kinds. Our company's progress in this area required inventing our own replacements for optical size, weight and width.
Typography today, needs relative specifications, (rem) as well and as absolute ones, 12 pt. ...it's got relativity, on the web, just for size. But to have relativity work in weight and width, one needs to leave the standards. Our Adios to those standards came when we had to make a 21st century font catalog online.
What do you think the standards owners are thinking when they push variations, while continuing to put on this parametric shit storm show for 20 Years? I mean, there is no variations name table and a bunch of other stuff, in the proposal. If they are be thinking of instances named with a series of numbers from the axis, or picking os2 values out of a hat?, or none, I feel sorry for their users.
I think the vartzars might think if one limits the scope of variations to what helps compression, then one can keep the parametric identities of the resultant styles as dumb as today, but they ain't done it yet, so their nuts are probably cold. Maybe when folks get closer to actually using variations and the os2 table together, they'll warm to reform?