User:Yoshifan52/Sandbox/Recording process

Guides and explanations for the recording processes I follow for Inkipedia uploads, especially those that require group coordination. For those interested in helping or the simply curious

=Recording equipment= I use an Elgato HD60 S [ OBS. my settings For S2 and S3 the Switch's built-in recording function may alternatively be used for most things. note that its resolution differs in handheld mode and recording output reflects that. note that the trim function is quite lossy lossless clipping git link

=Multiplayer recordings=

S3 Fistbump animations videos
Output example / Relevant article

There's a dedicated thread in the Discord where we coordinate recording these, and I maintain a tracker spreadsheet noting which emotes have been recorded thus far. The simplest setup to consistently get fistbumps for recording requires 3 players in a Private Battle; the process takes approx. 2 minutes per fistbump.

Massive thanks to to Xev, Milchik, 6-Frisk, Fancyrat, Sourguppyworm and Superfry for helping coordinate and/or record so far :)

Cosplay guide

 * Player 1 dresses up as VeronIKA:
 * Feminine Inkling with S3 Customization Skin 5.png fifth skin tone, S3 Customization Eye 18.png yellow-orange eye color, S3 Icon Megalobraid.png megalobraid hair, S3 Icon fi slit.png slit eyebrows, S3 Icon Legwear 1.png s1 style shorts
 * Equips the, ,
 * Player 2 dreses up as Hirooooo:
 * Masculine Octoling with S3 Customization Skin 8.png eighth skin tone, S3 Customization Eye 19.png red-blue eye colors, S3 Icon Octolocks.png octolocks hair, S3 Icon mo classic.png classic eyebrows, S3 Icon Legwear 1.png s1 style shorts
 * Equips the, ,
 * Player 3 doesn't need to dress up, as they won't be in any recordings

Private Battle setup

 * The Private Battle room must be opened by Player 2 (the Hiro cosplayer) for the ending screen to line up with Veronika on left and Hiro on right
 * Battle settings: regular Rainmaker on Scorch Gorge
 * Once the room is set up, send out an invite (pool is wikitest)
 * Players 1 and 2 are on team Alpha, Player 3 is on team Bravo
 * Players 1 and 2 equip the same emote per match to record both the left and right variants in one go
 * Players 1 or 2 must record the clip; whoever is recording must enable the first color lock option in settings

Gameplan

 * Player 3 descends anywhere on the map, preferably near the Checkpoint
 * Player 2 inks a bunch of turf
 * Player 1 gets the Rainmaker, travels to Checkpoint with minimal inking
 * Player 2 jumps to Player 1
 * Player 2 splats Player 3
 * Player 1 dunks the Rainmaker
 * Fistbump occurs on results

Notes:

 * The in-match roles of Players 1 and 2 may be fully swapped with no issue
 * Player 3's outfit and weapon do not matter
 * Player 1 must not be splatted
 * Don't use specials
 * Player 3 cannot see the fistbump animation so they cannot be tasked with recording

Upload process

 * Trim video clip to ~5sec (cut out the lead in zooms as well as the fade out squid graphic)
 * Note: it's extremely difficult for this to be lossless
 * Crop to 1000x700; inputting the following numbers into ezgifs crop tool custom fields is ideal: 425-465 84 1000 700
 * Mute the cropped output

=Singleplayer recordings=

Deep Cut (Splatsville)
Requires x100/x333 battle win

Squid Sisters collage process
Output examples ex1-1 ex1-2 ex2 Here are the steps I follow in GIMP:


 * 1) Gather all the images you want to make a collage of in a folder
 * 2) In GIMP, select "File>Open as Layers..." and navigate to the folder containing the images, select all the images, and press "Open"
 * 3) Organize the images into layer folders, and label everything to avoid confusion later. I name my folders "callie" and "marie", and each image is labelled "callie 1-3" or "marie 1-3" in alignment with the guide image I have provided (Callie in Shiver's Team color is "callie 1", Marie in Frye's Team color is "marie 2" / for the single truck day 2 pics a photo of Callie in Shiver's Team color and Marie in Big Man's Team color is "callie 1")
 * 4) Align all the individual layers so the idol's placement is generally consistent from layer to layer; it's impossible to align them fully, so just do your best. The way I do this is by lowering the opacity of each layer to approx. 45%, hiding all but two of them at a time, and gradually moving them to more or less align with each other. I typically use the bottom-most image in the layers box as a "base" image which I intentionally never move. Additionally try to somewhat line up both idols between their separate layer groups somewhat in the cases of their solo (two separate trucks) images.
 * 5) Using the Rectangle Select Tool select a rectangular area that contains a tight area where the idol fits inside the selection box across all layers; how well this works depends on how well you've aligned the idol images cross-layer
 * 6) Adjust the selected area to be a nice, even amount of pixels, round up if you must: the current area of your selection is noted on the bottom of the screen as "Rectangle: X x Y" when you are adjusting it. You can pull the selection to be larger or smaller via large rectangular tabs that show up when you click then mouse over near the selected rectangle's sides, or by holding down alt (to add) or ctrl (to remove). I worked with a 640x752 selection area for that Callie image
 * 7) Note down the width of the rectangular selection, which should still be active. If not, hit "Edit>Undo" to go back to it, or make a new selection using the same process.
 * 8) Select "Layer>Crop to Selection" for each of your layers (or layer groups if you took the optional step; this is quicker). Eventually this will trim your entire canvas, which is the goal.
 * 9) Using the width of your selection noted in step 6, calculate how much wider your image should be with the following formula: [selection width]×[amount of images]=[target width]. Eg. if your selection's width was 300 pixels, and you have 5 images you want to make a collage of, the resulting number would be 300×5=1200. Note: if the calculation result exceeds around 2800, it would be best to vertically stack the images after a certain point.
 * 10) Select "Image>Canvas Size" and input in the "Width:" field the result of the calculation made in the previous step. Make sure the "X:" field under "Offset" is set to 0, then press "Resize".
 * 11) Using the Move Tool move each layer so each version of the idol is side by side, ensuring they're laid out pixel perfect. I zoom in here and there using ctrl+mouse wheel to make sure it's aligned right. It's possible to use the arrow keys to move them perfectly, but it's rather slow.
 * 12) Once you're done aligning everything, select "File>Export As..." and type in your desired filename with the .png format, then press "Export". In the menu that pops up, ensure the "Compression level" field is set to 0, then press "Export".
 * 13) Losslessly optimize your output files here.