Elysia Tools
Navigation
Workflow Playbook
Prepare a consistent image set by correcting orientation, removing unwanted margins, applying one size policy, compressing files, and checking representative outputs before delivery.
Hubs
Start with the receiving layout, not with a compression slider. Record the target pixel box, whether aspect ratio must be preserved, whether transparency is required, the allowed output format, and a size or quality threshold. Choose a representative sample that includes portrait and landscape images, transparent graphics, tight borders, and different levels of detail.
Use image-auto-orient when EXIF metadata may change how a source is displayed. Compare image-trim, image-trim-whitespace, and image-trim-tolerance on the sample: the correct choice depends on whether the margin is truly blank or merely close to the background. Once the rule is defensible, apply it with batch-image-trim and check for subject clipping.
Use batch-resize to implement the declared fit, scale, or crop policy. Then use batch-compress with a quality or size target that reflects the actual delivery constraint. Inspect representative outputs after compression for blur, ringing, banding, transparency changes, and inconsistent framing. Record rejected files and reconcile the final count and filenames with the input set before delivery.
Workflow playbook
Select a representative sample, correct each source image with EXIF-aware orientation when needed, and compare general trimming with whitespace or tolerance-based trimming before choosing one rule for the batch.
Run the selected border policy through batch-image-trim. Review a few varied subjects and stop if the rule removes meaningful content or leaves a visible margin that needs a different tolerance.
Use batch-resize with the documented target dimensions and aspect-ratio behavior. Preserve the subject when possible, and only fill or crop a fixed box when the receiving layout requires it.
Compress the standardized set with batch-compress, then inspect representative files against the agreed size, quality, orientation, framing, and transparency criteria before releasing the archive.
Run image-auto-orient on each source when camera metadata may affect geometry. If orientation is already baked into the pixels or metadata is unreliable, inspect a sample first so the same image is not rotated twice.
Use image-trim-whitespace for clear blank margins, image-trim-tolerance for borders that are close to the background but not identical, and image-trim for a general single-image edge decision. Use batch-image-trim to apply the chosen policy across the set.
Use batch-resize with a fit or scale policy when subjects must not be cropped. Choose a crop-to-box policy only when the downstream slot requires identical pixel dimensions and the representative samples confirm acceptable framing.
Set a measurable size or quality threshold before using batch-compress, then compare representative outputs for file size, sharpness, halos, banding, and transparency. There is no single quality setting that suits every image set.