Choose a change and preserve the source
A background, a label and a color effect serve different goals. Decide what your copy needs: placing a transparent logo on a color, identifying a sample or preparing a grayscale version. Keep the original so trials do not accumulate transformations that are hard to undo or remember.
Each operation generates a new image. For multiple steps, download the output and load it into the next tool; content does not transfer automatically. Preserve a suitable intermediate copy during trials to avoid repeated lossy exports. Review the finished variant at its intended size and against its actual background, rather than relying solely on an enlarged tool preview.
Backgrounds and borders change different properties
Image background composites transparent and semitransparent pixels over the selected color. Output becomes opaque even in PNG or WebP. It does not remove a photographed backdrop: an already opaque white wall stays white. Try a known transparent graphic to understand which parts change before working with your final source.
A border extends the canvas outward. A 400 × 200 px source with a 16 px border becomes 432 × 232 px. Verify that the new measurements still satisfy your destination. Color palettes and conversion supply codes for these adjustments; inspect the actual combination, since a code alone does not establish whether it works visually with your content.
Place text and check readability
Text on image accepts one line of up to 100 characters. Adjust position, color, opacity and size relative to the shorter side. Labels fit the available width, so a long phrase can appear small despite a large requested size. Choose a short message that does not obscure data or the subject.
The download contains pixels with no editable text layer. Keep the wording and options for later revisions. A visible mark identifies a sample or version but does not prevent copying or cropping. Check contrast in light and dark areas. Rotating after adding text also rotates the label, making operation order part of your design decision.
Understand grayscale, inversion and pixelation
Grayscale preserves dimensions and converts colors to gray tones. Inversion replaces each RGB channel with 255 minus its value, turning pure red into cyan. PNG and WebP can preserve alpha in these operations; JPG uses white where transparency existed. Open the output and verify the format selected for the copy.
Pixelation reduces the image to a grid and enlarges it without smoothing. Larger blocks strengthen the effect while final dimensions stay unchanged. The entire image is affected, with no region selector. Do not assume this is sufficient to remove identifiable information: inspect the result and use an appropriate procedure when actual data removal is your purpose.
A border that expands the canvas
Load the 400 × 200 px sample in Add image border. Use 16 px thickness, black and PNG output. Generate, download and open the copy. The border is added outside rather than consuming part of the original graphic.
Border: 16 px Color: #000000
Output measures 432 × 232 px because each dimension adds two 16 px borders. Original contents retain their dimensions inside the expanded canvas.
Verify dimensions in Image information. Try Image background with the transparent source: transparent areas are filled, while already opaque regions keep their colors.