[CMYK separations/ Halftones for screen printing.]

I need 16bit cmyk mode badly .    Krita  has it nicely for instance  . too bad its slow as hell without proper GPU support

[extremelly beginner-unfriendly software]

Thankfully  chat GPT can instruct you a lot about Affinity soft .    So it's a quick transition  and it can explain you key differences in layer system pretty quick and well.   The bad thing is a lack of features available  in Photoshop like 16 bit cmyk,   and its procedural features are way  behind some other 2d soft . 

 Learning curve is nothing especially hard IMO.       Took me a weekend of random clicks probably coming from Photoshop.     Many things are really more convenient in Affinity like freq separation  , bulding 3d LUT from  random picture pair,   live normal map/ lighting filter,  mesh deform , linked layers, embedded files .   But I still maintain my Photoshop subscription  which went almost twice  recently while I have absolutely zero advantage of their AI direction.    Affinity still lacks basic things.        

[Is there a way to change transparency of "Bleed" erea in Designer?]

1 hour ago, Alfred said:

The items in an artboard’s bleed area aren’t obscured

Thanks for the hint, Alfred.   Looks like artboard works  to some extent . If artboard layer would  stay bellow.    But it doesn't  once you move something

 

3 hours ago, walt.farrell said:

I'm curious why you'd want that.

I need something like pattern preview in PSHop or Krita.    

[Can I have macro know document size somehow?]

6 hours ago, carl123 said:

Arrange > Align Left
Arrange > Align Top

it's not exactly what i need .   i need  a copy be in bleed area outside doc frame.  like when you have first symbol in top left corner and second symbol shifted in X direction to what spreadwidth value is

[Is there a way to change transparency of "Bleed" erea in Designer?]

Or rather have it fully transparent?

[Is there a way to recreate photoshop layer comps in affinity?]

Well. it does  work of a  sort but  making states much more complicated .

 

PS. actually i was too quick to decide.    The newly added to a group objects  keep staying in that group state  but they also stay in other groups  now.  I am totally lost now.    Wonder what that update button really update?    In Photoshop its simple like 2x2 

 

[Can I have macro know document size somehow?]

I want to make a macro that shift symbols to image corners  doing exactly same shift the image size in x/y is .    Does it possible somehow?

[Is there a way to recreate photoshop layer comps in affinity?]

layer states  remember only layers existed  when you create the state initially , and update button doesn't make them include later layers into same "state"   like photoshop does.  

Queryes save only layer with same tag  but you can't have two  tags on same layer  so no way to add same layer in diffrent queryes"    

 So what might be a trick?  I need it exactly how it works in Photoshop   + a way to export them in separate files .

[Scripting]

Well,  at least  paid GPT recently became  increasingly  appreciative and flattery not missing a chance  to tell you how great and pro  your ideas are.   Untill, you turn "Monday" version on and there it tells you what it really think about your brain capacity 😁

[Scripting]

So why I am paying 20 a month  for chatGPT 4.5  "advanced reasoning" .  Feels better than free one .  Or is it just self persuasion and a mind trick ?

Sometimes it sounds  like it has a better idea of how things should work  than  my initial  one. 

 

 

[Scripting]

Well,  with scripts writing  AI is at least doing exactly what you  you need it to do  contrary to image generation AI which always does  something  partly  blurry and never what you wanted exactly.   Just useless fancy pictures in same recognizable styles.  AI material textures  are so ugly an unusable for instance.

[Randomly alternating image spans for textured image brush in Designer.]

So  it could vary .   both in stretch and repeat mode . I mean those red dotted lines in brush dialog.  could we have more spans . At least 5 maybe. And a seed value . So if same seed  we could replace the stroke to another one with same pictures order      With an option to stay perfectly  deformed along the spline or scattered with offset.    I think Serif had something like this before.   

 

[Scripting]

I just hope chat GPT  would be ok with Affinity scripts.     It's what made Blender so  easy to use recently and so much ahead of paid software.   Almost everything Chat writes for Blender works .    It can modify the software  way beyond default capabilities.  I actually half way on modifying Blender to a "perfect" 2d raster/vector hybrid.   Like  Expression once promised.    With cryptomatte support   and all other things we asking  here for years.

   But when I try Chat to write something  for Adobe or 3d max and  it  never works.     Other day I asked chat to write me a simple frequency separation  tool  in Affinity style for Phtoshop + a few extra things like specific kinds of blurs.  It couldn't . I had to do the  action  manually.      But python + pillow +a few other image tools  and Chat GPT does wonders.   It wrote me all possible freq separation types  ever existed found  from some obscure  research publications , some working slow as hell .    A one  separating vertically looking details from horizontal  I am still not sure I understand how it works.     

I very much hope  there will be a mechanism to assess pixel and vector layers data to some quick data save to process them  with GPT made  Python scripts  and bring back .  Maybe by saving super quick  uncompressed tiffs or RAW in temp folder and then read back .    That approach works with Photoshop for me.    

   

[Pattern preview instead of pattern layers.]

A button that would do  locked symbol containers  of every layer  and place them with document size shifts  all around the document frame.  Invisible in the layer stack  pane.     I did such setup manually  and it sort of  works  but even as a batch  it create so messed up  layer stack I constantly break it accidentally  and could hardly figure out a thing.     Please let it be just a one button option for preview and painting.   Photoshop has a nice one bit doesn't work with deformers   and   I don't need it to tile forever , just  8 quads around . it's perfectly enough to get an idea how it would look repeating .   Lets have better one.

[Layer live color adaptation through frequency separation.]

It's a tedious task when you need  to do a collage and need to adapt colors of new layer to average colors of surrounding pixels. Photoshop does auto-blend  but only in RGB  mode and quite frankly I don't like how it does it.   An old and  ancient way is to do it through frequency separation.  You blur background and use it as low frequency  part for top layer.     

Pretty sure it's not that much calculation expensive  procedure  and could work as live filter  doing so on all of RGBA , yes including alpha as an option. 

[16 bit CMYK mode please]

Again, people use Cmyk not only for printing . it's a data storage  format  where all channels always synced whatever geometry distortion/ lens projection you apply . All at once .   I couldn't care less about printing for example, and often about colors and color profiles too when I use it.  Yet I have  been  using it for years.   it's just a convenience  feature. 

Doing only 8-bit pushes me away , toward Photoshop, which lacks certain cool features we have here.     In fact I'd like 32 bit cmyk even more .   Would allow to pack twice more channels, before and after comma in pixel values.        

[Curves/ bones deform as a better version of puppet wrap]

What I wanted to say is both Photoshop puppet warp  as well as APhoto one  are not very convenient and precise , not for facial  animation for example   vs something like 3Dmax or Blender.     Nor for barrel or fish eye distortion compensation either  where you do  gazillion of those control dots and they still do it wrong in some corners  because you have no access to how mesh deformation  influenced / no proper weight distribution .     

 I am just saying we could have perfectly same approach with auto or manual weight control on the mesh by a brush  and curve instead of just dots.  In 3d max it's super easy since you can directly use a spline as a deform-able bone . In Blender it spline IK   on top of bone chain or geometry nodes .

Another thing it needs to be live non destructive  filter.   Going Aphoto or Photoshop after any 3d software  and any deformation feels like useless toys.   Even Affinity mesh deform itself.  Why we cant select several mesh nodes and rotate or scale them around some point.   or why we can't align them in a perfect line or distribute in certain order ?  All this easily doable even in Blender. 

Perhaps it would be easier to just  have a sort of addon for Blender  to quickly bring  background and active layer to from affinity to blender and do all the image deformation there .     A button in affinity photo that export blend file with already embedded images: active layer  and whats behind it  and some ready-made  setup to deform  and then an addone in blender to render the result back to APHoto?   Could be sort of pseudo non-destructive with sidecar .blend files.

 

[Curves/ bones deform as a better version of puppet wrap]

Mesh deform is nice  but not very convenient really.    I deform images in 3d soft using  curves or bone chains   and an automatic weight applied to an image mesh .    But it's too time consuming to setup and render/bake even  with chat GPT made python  scripts. 

Would be nice  to have similar approach in Affinity photo  where we could draw a few vector curves  on top of a pixel layer  and using those curves as bones  with adjustable weight distance  deform the image  in similar manner  mesh deform does.    I assume it would be much easier to do in 2d  rather than in 3d.       Like puppet wrap in Photoshop but much more convenient and easier to control.

Would be nice  tool to make  faces a bit of smiling expression. A silhouettes tweaks etc.                                                                                                                                        

[16 bit CMYK mode please]

People use CMYK not only for printing .  It's used in various scientific  and CG purposes.     Multichannel  format in Photoshop is useless  for example and  people do use cmyk instead .    And having  it only in 8bit with ugly banding   makes it useless for anything but printing.      It exists in Photoshop for a purpose , not just like useless adornment .  And I believe it requires just a minor code touch, not a big project.

 

 

 

[AutoBlending better than in Photoshop please.]

We have that auto-blend option in photoshop . It' inconvenient as hell to really use there  but an Idea is great. Would love it work as live filter . A layer that would blend in into background smoothly  while you move it around.    Or do it Photoshop way please.  It's basically a frequency  separation   beneath .Just to troublesome to do manually  all the time.

 

Patch tool working with mesh or curve deform  would be great too

[16 bit CMYK mode please]

We can have it Photoshop but  in Affinity its only 8 bit .  Please make it supporting 16 bit too

[Can I turn a channel or spare channel into editable gray layer somehow ?]

Thanks guys

 

[Can I turn a channel or spare channel into editable gray layer somehow ?]

I sometimes see grayed out  option in right click menu and couldn't figure out how can I do it?

[Affinty 2.6 - Cropping/Copy/Paste Issues]

From my experience you need to hit "trim and rasterize" , then you can copy only whats in selecting frame.

[Will Afinity Photo ever respect backround color where alpha is black ?]

So a tga file would show it's true background color not black where alpha is 0?     I know I can  save  it with  actual color with alpha 0.001  still less than  1 in 255 8bit scale    bat when i open such file again it's still  perfectly black background there  when  switch off alpha.         It's very annoying  since we use RGBA tgas  to pack unrelated images.    

 

Could we have an option to have it working same way as in Photoshop?