Skip to content

Added Z-A mega sprites#236

Open
Jessseee wants to merge 1 commit into
PokeAPI:masterfrom
Jessseee:add-z-a-megas
Open

Added Z-A mega sprites#236
Jessseee wants to merge 1 commit into
PokeAPI:masterfrom
Jessseee:add-z-a-megas

Conversation

@Jessseee

@Jessseee Jessseee commented Jun 16, 2026

Copy link
Copy Markdown

Not trying to take away from the other pull-request, but there hasn't been activity on that PR in a bit. So, I added the A-Z mega sprites by @kyledovey (shared here) to the sprites/pokemon/ folder. Let me know if (and which) sprites should be duplicated to other directories in the repo. There were also new sprites for Mega Floette, Mega Victreebel and Mega Dragonite, but given that they already had sprites I left them out of this PR.

As requested by @anhthang on the other PR,

  • I renamed the files to match the id numbers from PokeAPI
  • Made the images square (96x96 pixels) using FFMPEG:
    ffmpeg -y -i "input.png" -vf "scale=96:96:force_original_aspect_ratio=decrease:flags=neighbor,pad=96:96:(ow-iw)/2:(oh-ih)/2:color=0x00000000" "output.png"

Also @kyledovey, if you have any more Z-A sprites (back, shiny, etc) that need adding just send them over and I'll include them in this PR.

@Naramsim

Copy link
Copy Markdown
Member

Hey, we had the PRs from @FallenDeity bringing them in.

#235 #233 #234

Is this PR for the same images?

@Naramsim

Copy link
Copy Markdown
Member

We could run the same downscaling algorithm @FallenDeity used.

@kyledovey

Copy link
Copy Markdown

Thanks Jessseee, appreciate it!
image
I think the scaling might be a bit off, it looks like theyre all being made to fit 96x96, even the smaller ones, so the pixels are getting a bit distorted. I haven't looked into how what the standard format is, but likely it wants empty space around them? or trimmed with no empty space? not sure. I think FallenDeity would know better

@FallenDeity

FallenDeity commented Jun 16, 2026

Copy link
Copy Markdown
Contributor

Hey, we had the PRs from @FallenDeity bringing them in.

#235 #233 #234

Is this PR for the same images?

I havent done megas yet so this should be good only thing is I think 1-2 of them should be removed for rework for example zygarde mega is too tall compression ruins the image

@Jessseee here is the ones I had prepared locally for all front and back + shiny variants za_megas.zip the zip file has original sprites scraped from kyles site and a nested normalized/ dir containing the resized sprites (both og and resized are renamed)

you can use them and add them to this pr, I took them from @kyledovey 's site https://pokengine.org/collections/10qmfbx6/Nintendo?megas&shiny&page=11 and processed them for all variants feel free to give them a look

the algo i used goes as follows

  • if smaller than 96x96 add transparent pixels (standard u can see in older smaller sprites)
  • if larger attempt cropping the transparent pixels and worst case use nearest neighbor to downsize while maintaining aspect ratio

ps: make sure to regenerate the sprite_audit_report.csv and update kyle's credit section in readme to include za megas

these were the resized ones in all 4 variants

10278
10281
10287
10288
10290
10291
10292
10294
10299
10301
10304
10305
10309
10311
10312
10313
10325

Present in front + front_shiny, but not back + back_shiny

10279

Present in back + back_shiny, but not front + front_shiny

10308
Pokemon ID View Original Size New Size Scale
10301 Front 131×145 87×96 0.66
10301 Back 97×146 64×96 0.66
10312 Back 124×96 96×74 0.77
10312 Front 121×103 96×82 0.79
10299 Front 115×96 96×80 0.83

i think zyagrde suffered the worst in rescaling, followed by darkrai you can compare them based on original and normalized/ in the zip when making the pr, i think the other than zygarde rest seem acceptable here is my resize script if you wanna experiment https://pastebin.com/kWHqCAK7

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants