Jump to content

Imaginary

Members
 View Profile  See their activity

  • Content count

    58

  • Joined

  • Last visited

Community Reputation

15 Neutral

See reputation activity

About Imaginary

  • Rank
    Member

Recent Profile Visitors

569 profile views
  1. Imaginary started following [Both Apps] Most recently used brushes, PLEASE REDUCE STARTUP TIME, Stock panel: Keep search term when switching archive and and 7 others
  2. Imaginary
    My perception when changing to 1.7 was different. It was a huge decrease in startup time.
  3. Imaginary
    Sure Walt, that's a way. But not that much less cumbersome... But of course there are many issues with Photo which have a higher priority.
  4. Imaginary
    Please keep the entered search term when switching between the different stock archives. At the moment the search term is cleared and has to be reentered whenever you switch to another archive which makes browsing really cumbersome.
  5. Imaginary
    +1
  6.    Imaginary reacted to a post in a topic: Masking when there is no mask?
  7. Imaginary

    Masking when there is no mask?

    Imaginary replied to dmstraker's topic in Photo Bugs found on Windows

    +1
  8.    Frozen Death Knight reacted to a post in a topic: [AP] Paint brush color picker should use color picker's source setting
  9. Imaginary
    Thank you, James, for those detailed explanations of the technical background. It starts making sense now. Nevertheless I think that colour decontamination should be an option and have an intensity slider like the other parameters. And I also know that documentation is just an imposition for software developers, but the the "Refine selection" part really lacks a lot of information about what is going on when particular options are chosen. So it really needs a thorough revision. And if only to avoid false impressions...
  10. Imaginary
    It is not a matterof fairness. I was not searching for issues, I just wanted to correct the mask after refining and stumbled across this. If anything, the software is not acting in a fair manner, because of a completely counterintuitive behaviour... Besides I cannot see the point in having pixel layer and mask separated when both only work together... If the mask only works with this exact pixellayer changes, it is just useless to have the separation. Well, then why isn't colour decontamination an option in the refine dialogue? And why isn't it mentioned anywhere? Indeed, I tried to improve the mask after refining.. Well, using output to mask doesn't give you the exact same result as previewed in the refine dialogue as I already pointed out about two years ago: [AP] Refine selection preview issue
  11. Imaginary
    Yet another issue with selection refinement. Refine selection with output to new layer is completely broken. Steps to reproduce: 1. Use the selection brush to make a selection 2. Refine selection witht the Matte brush 3. Choose output to "New layer with mask" and apply 4. Switch off the mask, the new pixel layer is completely destroyed. It shows the traces of the Matte brush Sadly this is just one more issue with refining selections. Come on Serif, you should put a lot more effort in improving this function. PS is much more reliable here. AP 1.7.3.481 Windows 10
  12.    Tom C reacted to a post in a topic: Brush modes when painting in masks…
  13.    IPv6 reacted to a post in a topic: Brush modes when painting in masks…
  14. Imaginary
    I came across a strange behaviour concerning image layers vs. pixel layers. In the attached file there are two groups each with the same layers: Well, actually there is a little difference between the groups: in the lower group the bokeh image is an image layer, in the upper group the image layer was rasterized to an pixel layer. Both bokeh layers have the same blend ranges: The mask layers in both groups are completely white. When the mask is switched off, in both groups the resulting images are completely the same. But when the mask is switched on, although as an empty mask it should not have any effect, in the image layer group the resulting image differs from the "unmasked" version. This looks like a bug to me. AP behaves as expected when blend ranges are reset. It looks as if the "white color" of the mask layer has some effect because of the setting in the source layer ranges. Well, only in the image layer group. Strange.... AP 1.7.3.481 Windows 10 GirlWithBokeh_Mask.afphoto
  15.    GRH reacted to a post in a topic: [Both Apps] Most recently used brushes
  16. Imaginary
    1. Select the "The Mask" layer mask you want to improve 2. Go to Channels and right click "The Mask Alpha" -> Create Grayscale Layer 3. A new pixel layer is created which can be edited with the paint brush in overlay mode 4. Make sure the new pixel layer is not a child layer, move it outside if necessary. 5. Choose "Rasterize to Mask" from the context menu of this pixel layer. 6. Move the new mask where you need it. This is a very cumbersome way and there is no live preview possible either. Masks should be treated like pixel layers when it comes to manipulating. You need a live preview. The greyscale layer step should not be necessary. It is kind of frustrating to see that in two years since I am an Affinity user nothing was improved concerning masks.
  17. Imaginary
    +1
  18.    CLC reacted to a post in a topic: [AP] Refine selection foreground/background issue
  19.    IPv6 reacted to a post in a topic: [AP] Paint brush color picker should use color picker's source setting
  20. Imaginary
    No one who has ever stumbled across this?
  21. Imaginary
    Well, actually it is not the poor primary selection that bothers me too much here since this is something highly dependent of contrast. Low contrast, poor selection, fair enough. I mentioned this in my original post. What is really annoying is the fact that once you are in refining mode, you cannot improve a poor selection by precisely determining particular foreground/background areas. It is impossible because AP tries to intelligently decide where you might want the fg/bg border to be. Little hint: it is exactly where it was painted! So any selection "improved" by fg/bg brushing will inevitably be worse than before because fg/bg bleeeding is created that cannot be corrected without new bleeding. Just take the image from the original post an try it yourself. It is sad that there was no improvement whatsoever in "Refine selection", the behaviour has not changed at all, even this bug still exists.
  22. Imaginary
    When sampling a color with the paint brush tool by pressing the ALT key, always the global color as it can be seen on the screen is sampled. If there are brightening or darkening layers above the current layer, the sampled color is brightened or darkened once more, which ends up in a completely wrong color. The actual color picker tool has a source setting "Current layer" which samples the correct color. So either the paint brush tool should consider the color picker tool's source setting when sampling a color or it needs an own setting for sampling colors.  There is a dark blue circle on the middle layer used as sampling source. The top layer has a brightening effect. The color for the lower dot on the right was sampled with the actual color picker tool with source set to "Current layer", the color is correct. The color for the upper dot was sampled with the paint brush tool's inbuilt color picker by pressing ALT: the color chosen is already brighter than it actually is and is then brightened once more. 
  23. Imaginary
    I'd like a brush category "Recently used" on top of the list that automatically shows the ten (number could be configurable) most recently used brushes. Would make it much easier to switch between brushes of different categories. Alternatively (or additionally ) a category "Favourites" would be nice, to easily find frequently used brushes. 
×

Important Information

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.

  I accept