WebGL 2 & C2/C3

0 favourites
From the Asset Store
A cool way for kids to write and practice English Alphabets
  • R0J0hound Yes, that's the one

  • Texture distortion is complicated - Classic had a skewing feature, and everyone expected it to affect the collision mask as well. So then you have to take in to account position, rotation, scale and skew in collision testing, and that starts to turn a simple, elegant and fast collision system in to a real mess.

  • Do you reckon WebGL 2 will come out by the end of this year or so?

    I can't wait to not have to make every sprite power-of-two sized!

  • Try Construct 3

    Develop games in your browser. Powerful, performant & highly capable.

    Try Now Construct 3 users don't see these ads
  • You already should ignore the power-of-two sizing for sprites. In fact that's the worst size for sprites, since they assemble the least efficiently on to spritesheets (considering it uses 1px border on to a power-of-two spritesheet). It only makes a moderate quality difference to non-power-of-two sized tiled backgrounds.

  • Yeah it's a shame it's inefficient, but non-power-of-two sprites aren't smooth or render as well as power-of-two when scaled/zoomed out. Sometimes quality matters!

  • Yeah it's a shame it's inefficient, but non-power-of-two sprites aren't smooth or render as well as power-of-two when scaled/zoomed out. Sometimes quality matters!

    I agree.

  • Yeah it's a shame it's inefficient, but non-power-of-two sprites aren't smooth or render as well as power-of-two when scaled/zoomed out. Sometimes quality matters!

    Only during preview. After export they belong to a power-of-two spritesheet, so mipmapping takes effect (which is what improves the downscaled quality). Though with WebGL 2 it will work in preview mode as well I guess.

  • Only during preview.

    I haven't found that to be true. Maybe I've been doing something incorrect somewhere? If that's actually suppose to be a feature, then I suppose I'll submit it as a bug report.

    Example (capx):

    As you can see, the power-of-two sized sprites move/look smoother and with less pixel jitter. Without being that square size, it seems as if it's doing "point" sampling rather than "linear"!

    Anyway, I hope WebGL 2 won't take too many years, so we can get this one improvement!

  • - ah, you're right. There's a quirk in C2's spritesheeting system where single leftover individual frames are cropped to their original size instead of padded to a power-of-two size, which creates an inconsistency in their display quality. However you can work around it by setting "downscaling quality" to "high", which forces all exported images to be padded to a power-of-two size. But yeah you're right, there are a few odd cases like this which WebGL 2 will fix. Chrome is already experimenting with WebGL 2 - you can experimentally enable it in chrome://flags, but note the C2 engine currently does not attempt to use it even when it's supported. It's a good sign that it should probably be supported some time this year, though.

    FWIW, Construct 3 has a better spritesheeting engine that fixes some of these issues, even in WebGL 1 mode.

Jump to:
Active Users
There are 1 visitors browsing this topic (0 users and 1 guests)