Template:Painted variants/Tutorial/Automation

From Team Fortress Wiki
< Template:Painted variants‎ | Tutorial
Revision as of 18:22, 28 December 2011 by I-ghost (talk | contribs) (updated to version 0004)
Jump to: navigation, search


To aid in the generation of Painted Images, a tool-set has been created to reduce the amount of tedious tasks required to generate Painted Images.

Tools

automateCaptureImages.ahk

This script automates the editing of the VMT for each paint, screenshotting and pasting each image into Photoshop.

automateCaptureImages_slow.ahk

Same as above with increased pauses - may be necessary for taking the images of attached items.

renameImagesEasyUpload.bat

This tool converts the meaningless names exported from Photoshop into useful names for manually uploading to the Wiki. The first section of the name relates to that images position in the Painted Variants template.

renameImagesWikiNames.bat

This tool requires additional parameters (renameImagesWikiNames <"hat name"> ["style"]). It converts the Photoshop exported files into the file names as they would appear on the Wiki. This is useful if you're using the script below to upload the files; if not: renameImagesEasyUpload is easier.

upload.py

This tool will upload all images in the /png subdirectory. Use renameImagesWikiNames.bat before using this. This uses the modified wikitools found here; If you already have this installed in your site-packages, delete the included wikitools as appropriate.

Requirements

  • Python
  • AutoHotkey
  • Photoshop
  • Technical experience manually generating Painted Images
  • Familiarity with the wiki's practises

Process

This is roughly written, intended for users familiar with the process.

  • Run TF2 and then exit it.
  • Extract the relevant VMT files from the content GCF, place them in the appropriate directory.
  • Paste this into the VMT somewhere, preserving the indentation in the VMT: "$color2" "{216 190 216}".
  • Open HLMV and the appropriate model to create a registry entry.
  • Close HLMV and open regedit. Set the model's rotation as noted on Template:Painted variants/Tutorial/Rotations.
  • Re-open HLMV and the model. Set the FOV as noted on the rotations page.
  • Open Photoshop, create a document the same size as your HLMV window - ensure this document is unsaved and named "Untitled". Also ensure there is only one layer; usually "Background" or "Layer 1".
  • Run automateCaptureImages.ahk. Don't interrupt the process.
  • In photoshop, crop and remove the background of the images as usual.
  • QUALITY CONTROL - Check the images, make sure there are no duplicate images - if so this suggests the automation process is too fast for your computer... and it's already pretty slow, so just ragequit. :)
  • In photoshop, File - Scripts - Export layers to files.
    • Destination to wherever appropriate. If you're using the upload script, make it the /png subdirectory of these tools.
    • File name prefix should be blank.
    • PNG24
    • Transparency and Trim Layers ticked.
  • Run the export and wait until it finishes.
  • If you're using the upload script, run renameImagesWikiNames.bat in /png. If not, copy renameImagesEasyUpload.bat into the export directory and run it.
  • If using the upload script, simply run it (upload.py) and then clean the /png directory of all pngs after verifying the images have uploaded successfully (leave the .bat in there). If not, upload the images to the appropriate places - their number prefix is their position from the top left of Template:Painted variants.
  • QUALITY CONTROL - Check the uploaded images against the history of the images they replaced - if they replaced any. Otherwise check them against another hat's colours - make sure the colours look similar.

Feedback

As the toolset is very new and very experimental, please provide feedback on the tools on the article's talk page.