Methodox Wiki
Product documentation
Home

Image basics

An image file, an editable document and the picture on your screen are related but different. The document can contain many layers and instructions. The canvas shows their combined result. An exported JPEG stores a finished raster picture of that result.

Pixels, dimensions and zoom

A raster image is a grid of samples called pixels. A 4000 × 3000 image has 12 million pixels. Its width and height describe the data, regardless of how large its window happens to be.

Zoom changes your view, not the file. At 100% you can inspect individual image pixels without the reduction used by Fit view. Use Ctrl+1 for actual size and Ctrl+0 to fit the picture. High zoom reveals pixels; it does not add detail. Check both the whole composition and its edges at actual size before exporting.

Aspect ratio is width divided by height. A 4000 × 3000 image is 4:3. If you need 16:9, choose whether to crop away some content or add canvas space. Stretching width and height independently distorts objects. Lock the ratio when resizing a photograph unless distortion is intentional.

Change What happens Example
Zoom Only the view changes Fit a large photo into a small window
Image ▸ Resize... Pixels are resampled Make a 1200 px wide delivery copy
Image ▸ Canvas size... The frame changes without scaling the picture Add room for a caption
Crop The chosen region becomes the frame Remove distracting edges

Resampling estimates new pixel values. Enlargement cannot restore detail that was never captured. See Transforming for interpolation choices.

Pixels per inch (PPI) relates image pixels to a physical print size. A 3000 px wide image placed at 300 PPI is 10 inches wide; at 150 PPI it is 20 inches wide. Changing only that resolution metadata does not create pixels. Resizing with resampling does.

The editor labels some resolution controls DPI. In this context they specify image pixel density. A printer's hardware dots per inch describes ink or toner dots and is a different measurement: several printer dots may reproduce one image pixel. Choose print dimensions in the print preview and assess whether your original contains enough detail for the viewing distance.

Transparency and the checkerboard

Alongside red, green and blue, a pixel may carry alpha, its coverage or opacity. An alpha of zero is transparent; full alpha is opaque. Partial alpha makes smooth edges possible.

The checkerboard is the editor's way of showing transparency. It is not part of the picture. If you see a checkerboard in a JPEG itself, it has been baked into the pixels; JPEG has no alpha. Use PNG for a cut-out with transparent edges and test it against both light and dark backgrounds.

Several controls can reduce coverage at once:

Control Scope Typical use
Selection Where an editing operation is allowed Paint inside a chosen shape
Attached layer mask Where one layer appears Hide a background reversibly
Clipping Coverage supplied by a lower layer Put a texture inside lettering
Layer opacity The whole layer Fade an effect uniformly
Pixel alpha Coverage already stored in the pixels A transparent edge

For an otherwise opaque pixel, 50% layer opacity × 50% mask coverage gives 25% coverage before other compositing factors. Changing opacity cannot restore pixels that were erased. A white mask cannot reveal image data that is already transparent.

See Masks for painting and selection-to-mask steps.

Bit depth, palettes and gamut

Bit depth per channel describes numerical precision. An 8-bit channel has 256 possible values; a 16-bit integer channel has 65,536. Higher precision can reduce rounding and banding during heavy edits. A file's bit depth does not guarantee that every tool processes at that precision. Writing an 8-bit result into a float file cannot restore discarded values or highlight detail.

An indexed palette is different: each pixel chooses a colour from a table. An 8-bit indexed image has at most 256 palette entries, not 256 values independently available in each RGB channel. It can suit flat graphics but may need dithering to approximate photographic gradients.

Gamut describes the range of colours a space or device can represent. A wider gamut is not more bit depth. Spreading the same number of values across a larger range can make steps more visible. See Colour management before choosing a working space simply because its name sounds more advanced.

Lossy, lossless and editable

Lossless compression preserves the pixel data given to the encoder. Lossy compression approximates it to save space. JPEG quality controls that tradeoff; PNG compression level mainly trades encoding time against size.

A lossless file is not necessarily a lossless editing document. PNG may preserve its exported pixels exactly while losing your separate text, masks and adjustment layers. Converting full colour to a small palette can also change colours before lossless compression begins. Save an editable .mie master and create delivery copies from it.

Raster, vector and metadata

Raster layers contain pixels. Vector shapes and text describe geometry and lettering that the editor redraws at the required size. Rasterizing converts those descriptions into pixels and removes their special editability; duplicating first lets you keep an editable original.

Metadata describes the image rather than drawing it: camera settings, captions, copyright, dates and sometimes location. An ICC profile describes how to interpret its colour values. Metadata and profiles can be retained or dropped independently of how the pixels are compressed. See Opening and saving for format choices.

Next