Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

OSM text quality much worse than Google's #3813

Closed
jidanni opened this issue Jun 23, 2019 · 9 comments
Closed

OSM text quality much worse than Google's #3813

jidanni opened this issue Jun 23, 2019 · 9 comments
Labels

Comments

@jidanni
Copy link

jidanni commented Jun 23, 2019

OSM text quality dangerously low.

https://www.openstreetmap.org/#map=19/24.15109/120.69722

pp

@sommerluk
Copy link
Collaborator

Hm, I don’t know how you scale the images, but in general this style is deployed as raster pictures, which obviously do not look good when upscaled by a non-integer factor like 2,22 as you are using.

The original screenshot from OSM without scaling:
Openstreetmap

@polarbearing
Copy link
Contributor

First, I see no 'danger' here. It would help not to use yellow press jargon in reporting issues.

The google example is written in a grey tone, while OSM uses black. The anti-aliasing algorithms might be a bit different, or just the fonts having different line width; but as @sommerluk said, it has to be judged at the 100% scaling, which looks clear to me.

@HolgerJeromin
Copy link
Contributor

HolgerJeromin commented Jun 23, 2019

Another possibility for differences. Google uses vector tiles which can render better with anti aliasing or even retina displays.

And please adjust your demanding tone (in many github issues from you) a bit. We are not paid staff here.

@matkoniecz
Copy link
Contributor

OSM text quality

Is it also problematic when viewing images at 100%?

@Adamant36
Copy link
Contributor

I have a 1920x1080 and things look a tad fuzzy and off for me. Including fences which have always looked sorta jagged and pixelated. Even while viewing images at 100%. I don't have the same issues on my 4K TV though. There lines, text, and also colors look much more vivid and less pixelated. It's something I've been wondering about for a while now. More then likely it's a monitor issue or something to do with the specific type of monitor I'm using. I figured 1920x1080 would be a good enough resolution to not cause fuzziness. Apparently it's not. So, any actual issues with text quality if there are any could just be the monitor it's being viewed on.

@matthijsmelissen
Copy link
Collaborator

This is likely caused by the fact that we do not provide 2x (retina) tiles on openstreetmap.org. That would also explain why the problem only occurs on some screens.

You can find a deployment of this style with 2x tiles here: https://osm.rrze.fau.de/testhd.html Is there still a quality difference between these tiles and Google?

@jidanni
Copy link
Author

jidanni commented Jul 9, 2019

Problem 1 is that older users will have to hit CTRL++... to read street names often anyway.

Problem 2 is even on https://osm.rrze.fau.de/testhd.html , each time one zooms the mouse wheel, the names momentarily become readable, but then are re-rendered smaller.

If I zoom https://osm.rrze.fau.de/testhd.html finally the names become comfortably readable, but there is no URL to compare the same place on OSM.org, nor can I go to Taiwan to try how placenames look there. P.S., I'm using HDMI-1 connected primary 1920x1080+0+0 and GitHub at 125% zoom while I type this.

@jidanni
Copy link
Author

jidanni commented Jul 9, 2019

And I cannot hit CTRL+++ on https://osm.rrze.fau.de/testhd.html because it just hits the zoom button on the map.
OK comparing https://www.openstreetmap.org/#map=19/49.57475/11.02909
we indeed see https://osm.rrze.fau.de/testhd.html is better!:
27581-2

@matkoniecz
Copy link
Contributor

There is no example of scaling broken for untransformed images. And this style with 2x tiles appears to work fine.

So problem is not in style itself.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
None yet
Development

No branches or pull requests

7 participants