Jump to content

IPv6

Members
  • Posts

    355
  • Joined

  • Last visited

Everything posted by IPv6

  1. May be robust scripting then? This is really huge task which may well worth a year(s) of development... But... why there is no scripting? why??? (i`m patiently waiting anyway)
  2. > new features Back to serious things. Really hoping this will be scripting (with deep document-altering possibilities) and not some 3D-bla-bla-stuff. Lack of scripting is a real major workflow limiter for now just 5 cents
  3. Hi @NathanC I recorded the video - at this time links was not restored (as i mentioned) and even document size was broken (definitely something shady going on here) after switching to history item before resize. Same happens with CtrlZ sometimes Look at attached file (saved with history) Screen Recording 2022-06-30 at 15.15.52.mov linkLostBug.afphoto
  4. Two bugs found: 1) If we have document with linked pixel layers (pixel content link) and then "resize document" - then layers are not linked anymore after resize (pixel content link lost) 2) Undo resize sometimes DO NOT RESTORE the links (pixel content link) back. Not happen everytime, but i experienced it, so you press Ctrl-Z and get broken document after that without noticing
  5. BTW, may recomend vtracer for manual converting colored images (png) to svg https://github.com/visioncortex/vtracer It`s quite good, can handle colored fills (so it is better than potrace), it`s opensource
  6. Indeed, but there is a proposal that already works for Firefox, so... would be really cool to allow inport-export this feature via this attribute, at least (for example) https://www.w3.org/Graphics/SVG/WG/wiki/Proposals/Variable_width_stroke https://developer.mozilla.org/ru/docs/Web/SVG/Attribute/stroke-width This is really important use-case for curves (separated curved lines without fill, but with width and opacity), but currently AD gives no possibility to prepare such things in other applications for import into AD
  7. > it's pity we can't make our own True I have version 9, it is ok. Preview on dowscaled image is fast, full processing takes 20-30 seconds for 5000*2000 image on average PC, for some quite heavy lua-based processing. Work on Mac and Win really well, you can setup controls for your own comfort and overall it is usefull. Yes, you can have access to only one layer - but you can load arbitrary images as inputs for filter (for me this is enough). There is also ability to use real curves as inputs - which is super-useful, since you can shape parameters as needed without additional hassle and tons of extra parameters. I don`t know another product with such extended per-pixel scripting support, if you know - would be curious to know too. It is a shame that full-fledged per-pixel scripting is not a common functionaly in every pixel-based DCC. There is similar stuff in nuke/natron but they are not so user-friendly and reusable, after all
  8. i recommend you to buy FilterForge and spend time on Lua scripting. It is not so fast, but covers literally every need for combining images with custom logic, with antialiasing and stuff. It is working as plugin (with some custom efforts - can be googled) Affinity photo is cruelly useless for that kind of tasks // "displace adjustment" is really a total garbage, "Perspective Live filter" is a pain to use for simple ofsset and so on
  9. No // procedural filter can use only pixels from the same layer, not the others Which is a real pitty Instead of "beneath layer" it would be cool to have "Layer" kind of input, so sampling this imput wopuld be just another method call inside formula. That way more then one layer could be added to procedural logic
  10. Are there any way to import-export curves as LINES with variative width? All usual formats are imported/exported as filled areas, completely killing the "lineness"
  11. Bookmarks would be cool, agreed! >Many operations are performed on the currently selected layer. With the ability to independently display (and select) different layers, it will be quite complicated and quite confusing - the user will perform operations on other layers than intended. Imho it is not a problem, several outliners are meant to be synced (selecting layer in one selects it in all). This exactly the way to elect the layer you want FAST, without scrolling and opening/closing groups. There is also many operations that do not require layer to be selected - like ctrl-click on layer to select content/mask - etc
  12. Currently there is only one "Layers" outliner possible, that shows whole tree. This is inconvenient when you have some frequently used "masks" layers - you always have to scroll the tree. Would be great if some group or whole tree can be detached into another "Layers" panel, so you can access different layers inside hierarchy without scrolling
  13. Stiil a problem in [1.10.2.1178] With this bug it is quite easy to "lost a layer" while linking them - when there is a lot of groups, etc
  14. Still, they are applied - in RGB32 mode // its an unfortunate decision made by Serif
  15. bump... Lack of smoothing in Liquify require excessive manual work sometimes //
  16. Yes, live warp is amust... Especially because live warp has a potential to be much faster than live liquify (which is slow as for now // unfortunately)
  17. Nice request! There is a workaround for SOME cases - spare channels actions are recorded properly. So you can record "Spare channel creation" on one layer and "Load spare channel into ..." on another. This gives kind of inplace copy-paste, properly recorded in macro. Mostly useful for masks
  18. Yes, this would be cool... And same request for multi-selection support in Smudge brush and in Mix brush - would be great if this tools get used to pixels from the same layer tool is applied. Like recording macro with smudge/mix brush on single layer and replay on others btw, smudge brush can not be recorded in macro - long-standing bug - but anyway, to illustrate an idea
  19. Please, add support for layers modes on Mask layers, So "ADD" will actually add alphas, and "Substruct" will actually substruct alpha values. This will greatly extend what is possible with masks and in many cases simplify layer structure for some effects Currently user can change blend modes for alpha-only layers, but mode is silently ignored... which is feel "broken" anyway
  20. Thanks for pointing on other thread, attaching crash dump too 9cb5c406-247e-490b-88ab-b0cb3a6eaecc.dmp
  21. For me crashes, latest release, no openCl, reproduced again and recorded (this time crash from second scale, not from the first, you can see on video) crash.mp4 . Just scale with control, without control scale works ok
  22. I have a file (see attach) with pattern layer. Every time i trying to scale this layer - AP crashes as soon as mouse button is released. Steps: - choose tiles patter layer - activate arrow tool, blue dots will arrive on layer (for scale/rotate) - press and hold control (to keep center) and drag scale dot a bit - release mouse button or control -> AP will crash // water_caustic1.afphoto
  23. Still a problem in (1.10.0.248) Pixel data pasted well, but when the layer itself is pasted -> crash crashinfo.txt
  24. If the pixel layer with adjustment is covered by other layers - any attempt to merge adjustment into layer makes it transparent, effectively erasing it content. Look at video for reproduction. Metal computing off, fresh document Screen Recording 2021-06-05 at 17.00.32.mov LayersLostBug.afphoto
×
×
  • Create New...

Important Information

Please note there is currently a delay in replying to some post. See pinned thread in the Questions forum. These are the Terms of Use you will be asked to agree to if you join the forum. | Privacy Policy | Guidelines | We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.