Jump to content

redonwhite

Members
  • Content count

    13
  • Joined

  • Last visited

Everything posted by redonwhite

  1. Hey guys, I was pointed here by MEB from another thread about this effect. Adobe Illustrator really has this figured out, see below for comparison. I think Affinity should move in this direction too, even simple work becomes challenging if you have to deal with these hairline fractures. Every drawing needs a labor-intensive extra step to be production-ready (my current workaround is duplicating offending objects a few times, works well unless there is transparency involved), and once you add all the extra objects you lose ease of editing. Below is what it looks like in Ai Of course, I have no idea how your anti-aliasing rasterization works, but my guess is you have to make some kind of a special case for edge overlap? Or maybe do rasterization on whole image instead of layer-per-layer basis? Or run some extra post-processing to make sure total alpha = 1 for every pixel? Hope this gets some love in future versions!
  2. There is a bug that is a show-stopper for me right now. Affinity Designer seems incapable of rendering adjacent shapes correctly. If I have two polygons with straight edges and I align their nodes precisely using snapping (or, as in the attached file, by duplicating one polygon and moving only other nodes elsewhere), resulting edges will not meet properly and some background will be seen through what seems like a pixel gap between the edges. This is serious as even simplest vector work actually becomes impossible in many cases! Attached photo and screenshot show what I mean and compare with the exact same case in Adobe Illustrator. This bug seems to persist under any combination of any values of the following settings: pixel view, retina pixel view, vector view, view quality, use precise clipping, renderer (nVidia drivers, WARP drivers), force pixel alignment, move by whole pixels. This bug persists in exported images. I would consider this high priority. clip test.afdesign
  3. Thanks MEB, I'll move my nagging to that thread :)
  4. Of course the triangles are white in my example just for illustration purposes. I think the Ai team found a solution for this. Maybe they are identifying overlapping edges and treating them as some sort of a special case. Also, I know there are other threads here where people complain about AD's aliasing overlap for shapes that are on top of each other and share an edge (so both on the same side of the edge). Possibly both issues have the same solution.
  5. I noticed that when I choose "Only snap to visible objects" and I alt-click a layer/group/object to hide everything else, everything else is still a snapping candidate, i.e. it seems only objects with the visibility checkbox unchecked are actually considered "not visible". How it should be: objects are "invisible" when they are actually invisible in the interface, regardless of the method used to hide them.
  6. Hi MEB, thank you for your work on AD :) Yeah I thought this was the case, just think that it could be more intuitive. Thanks. Since we are talking about the "isolation" mode, I think it's not very useful right now compared to Illustrator's or say Blender's since you can't even drag-select, as me and many others have pointed out here: https://forum.affinity.serif.com/index.php?/topic/1640-ad-isolation-mode/
  7. If I have a combined path object with a hole in it (eg. a donut made of two separate circles by Geometry > Combine), no object will be selectable through the hole in the framing object, either through regular or ctrl-selection. This will only manifest if the target object is below the framing object in the layer window. In the attached file, I can select the green dot, but not the red one. selectability_test.afdesign
  8. redonwhite

    Snapping feedback

    Bumping an old thread a bit, I think the issue here relates to what I just posted about in this thread: https://forum.affinity.serif.com/index.php?/topic/29406-less-confusing-naming-for-snapping-options/
  9. Hi guys. This had me confused for a longer while and I think could be named better. In the snapping options, AD currently has "Snap to shape key points" and "Snap to object geometry". I was looking for an option to snap to nodes, which I think is the most common use case. I thought it would be under "Snap to shape key points", since I wasn't sure what you call a point/node/vertex/key point/control point in AD lingo (I mean, every program seems to have its own naming). I was very confused to see that it resulted in snapping to curves or their extensions; this is exactly what I thought the second option does (geometry). Turns out "key points" are actually extended guides from any objects (which in my document translates to "random lines at crazy angles appearing everywhere"), and "object geometry" means curves and nodes with preference given to nodes. May I suggest renaming it like so: Snap to object geometry > Snap to nodes, edges and intesections Snap to shape key points > Snap to points of importance and extended edges or whatever this is actually supposed to do, I'm not super clear on this despite reading help and trying to use it in my document. Or maybe the snapping options need to be broken down a little further? Cheers
  10. redonwhite

    [AD] Isolation Mode

    Hello! I'd like to cast one hundred trillion votes for isolation mode please! Let me clarify: as it stands now, AD has no isolation mode. Hiding other layers is not what isolation mode is, that's just "hide others". From the user's perspective, isolation mode is a divide et impera approach to editing complex artwork. The essence of isolation mode as the user expects it to be (not just how it is in Illustrator, many other programs have variations on this, Blender being one) is that the user's actions all work as before, but only apply to isolated objects. In particular that necessarily includes marquee selections, the most often used tool! Hiding other objects is OK but not necessary; Blender does this, Illustrator has the white-out which is an elegant solution in my opinion. Note that I might actually need to see other objects, for example to have them as snapping guides! Please implement this guys, AD deserves to be perfect.
  11. Bumping this a bit. I appreciate this can be toggled in preferences, however: I think you guys should select a default mode (proportional or non-proportional) and stick to it. Othertwise, scaling properly requires previous knowledge of whether the object I'm scaling is a vector, a raster, a group etc. This is usually not obvious and requires trial-and-error, which disrupts my flow. So maybe you could please change this? Cheers Edit for clarification: I think good defaults are extremely important, otherwise the interface becomes a customization tarpit like Blender's :)
  12. Hi! New to Affinity so apologies if this is already possible (if so please tell me how!). Searched the interface and forums but came up with nothing. I think it would be nice if slices created from layers, artboards or shapes could be optionally made to inlcude adjustment layers and generally objects not grouped under the object that defines the slice. Right now custom slices (the ones that I draw directly on the artboard in the Export persona) show the effects of adjustment layers, but layer-based slices don't if the adjustment layer is not under the layer that defines the slice. Here's my real-life example: I have an AD file with sprites I made for a game, about 100 of them, which I export through slices. I use shape-based slices because they adapt to my shapes automatically and I don't need to fix them when I resize or move the sprites (which is a lot). Late in development I noticed the colors are off on all the sprites compared to the rest of the game, so I added just two or three adjustment layers on top of everything and it looks much better now. This is where trouble starts: the adjustment layers have no effect on the images exported by my slices. The slices only export the base shapes, as if the adjustment layers on top of them didn't exist. The slices only "see" the adjustment layers if the adjustments are grouped under the layer from which the slice is generated, but this setup is not practical as I would need to add 200 or 300 adjustment layers, and modifying the colors again would be a nightmare. In fact non-layer object visibility could be a per-slice export setting just like file format and compression, this way we could have two slice export formats - one with and one without adjustments, background objects etc. I suppose that implementing this would be failry easy as the functionality is clearly there, just needs to be exposed to the user. Cheers and keep up the good work on this program!
×