Image Effects

Most of these effects are the same operations browsers use when a web page styles an image with CSS filters. The W3C Filter Effects specification defines grayscale, sepia, invert, brightness, contrast, saturate and blur as standard filter functions, which is why the results look familiar from other apps. The difference here is that the change is saved into the file itself, so the edited picture looks the same in every viewer, document and upload form, not only on a page that applies the filter.

Strength runs from 1% to 100% and starts at 50%. It applies to every effect with a scale: sepia, brighten, darken, contrast, saturation, blur, sharpen, pixelate and face blur. Black & white and Invert colors have no strength because they are all-or-nothing. The preview updates as you move the slider, so judge the result on the first photo before you run the batch. Every photo in a batch gets the same effect at the same strength, which keeps a set of product shots or event photos consistent.

Small values do more than you might expect. For brightness, contrast and saturation, a strength between 10% and 30% is often enough to fix a dull or flat photo without making it look edited. Higher values are for deliberate looks, such as a punchy, high-contrast poster or a faded, washed-out style.

Black and white deserves a note of its own, because it is the effect people use most. A simple conversion can make a colorful photo look flat, since red and green of the same brightness become the same gray. If a black-and-white version lacks punch, run it a second time with More contrast at a low strength; that separates the tones again and gives the classic look. Sepia works the same way and also benefits from a touch of contrast afterwards.

Blur faces is different from the other effects: it detects faces in the picture and changes only those areas, leaving the rest untouched. It is meant for photos of people who did not agree to be shown, such as passers-by in a street photo, children at a school event, or bystanders in a picture used as evidence or in a report.

Hide faces with chooses the method. Blur softens each face so it can no longer be recognized at normal viewing size, while the photo still looks natural. Pixelate replaces each face with large blocks, the look most people associate with anonymized images. Solid block covers each face with a black shape, which removes it completely. Mask shape chooses between an Ellipse, which follows the shape of a head, and a Rectangle, which also covers hair and ears. For blur and pixelate, set Strength high when the aim is privacy; at low strength the person may still be recognizable.

Automatic detection is a starting point, not a guarantee. Faces that are very small, turned away, partly covered or in deep shadow may be missed. Look at the preview closely, and for photos where every face must be hidden, check the result at full size before you share it.

Faces are rarely the only thing that identifies a person. Names on badges, number plates, house numbers and phone or laptop screens can give away as much, so look for them too and crop or cover them as needed. Leave Strip metadata on as well: the GPS position stored by a phone camera can reveal where a photo was taken even when every face is hidden.

Corrections and styles use the same effects with different intentions. To correct a dim indoor photo, start with Brighten at a low strength; if it then looks flat, run it again with More contrast. To correct a picture taken in harsh midday light, Darken brings back some detail in bright areas. Sharpen helps a picture that was slightly soft from a phone camera or from resizing, and works best at low strength; high values add bright halos around edges and make noise more visible.

Styles change the mood instead. Black & white removes distracting colors and puts the attention on shapes, light and expressions, which suits portraits and street photos. Sepia adds the warm brown tone of an old print. More saturation makes food, flowers and travel photos more vivid; push it too far and skin tones turn orange. Blur at a soft strength gives a dreamy look or turns a photo into a background for text, and Pixelate at a high strength turns any picture into a retro, blocky graphic.

Order matters when you chain effects across runs. Brightness and contrast change the tones that black and white is then built from, so adjust them first if a gray version looks muddy. Sharpening should come last, after any resizing, because making a picture smaller afterwards removes much of the crispness you added, and making it larger exaggerates the halos. Blur and pixelate are best applied once to the final picture, since running them twice compounds the effect quickly.

Some jobs need a different tool. A black-and-white PDF document is a job for PDF to Grayscale, which keeps the pages as a PDF. For cutting away part of a photo before applying an effect, the Crop Image tool comes first.

Output format can stay on Same as input or change to JPG, PNG or WebP. PNG keeps transparency, so a logo with a transparent background stays transparent after the effect. Quality applies to JPG and WebP and starts at 85, which keeps photos looking like the original apart from the effect itself.

Effects are applied to the pixels, so the result is a normal image file with no layers or undo history. Keep the original photo if you might want a different strength later, and run the effect again from the original rather than from an already edited copy. Each lossy save at the quality setting adds a little loss, and starting fresh avoids stacking it.

Auto orient turns phone photos upright before the effect is applied, so face detection and the preview work on the picture the way you see it. Strip metadata is on by default and removes GPS position, camera data and comments, which matters most when the point of the edit is privacy. Once the photos look right, the image converter page can change their format, and Merge Images can put a before and after side by side.