'Towards an attention-based letterfitting model' ... electric bubbles! https://www.aldusleaf.org/2019-03-17-letterfitting-attention-model.html …
-
-
One of the first things I look at with with any autospacing method is whether it manages to set the spacing on the left side of M and N in contrasted serif types slightly tighter than that of H etc.. It's a subtlety, but subtlety is kinda the name of the game. +
@SebastianLoschpic.twitter.com/3Se8CIXFG0
-
It seems possible that a model using a concept of even grey between blurred glyphs might pick up this. Although it might also over-compensate for the thin verticals, especially in really high contrast types.
End of conversation
New conversation -
-
-
Still smells over-simple. For one thing, any method that doesn't do "managed collision" cannot be optimal.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Yes, and this suggests that ‘halo overlap’ should be accumulated in some non-linear fashion – the parallel edges of “VA” should count as a stronger overlap than an approaching serif or point, such as “ly”.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.