Prepare and Export PixInsight Images for Web and Print

A web image needs manageable pixel dimensions and file size. A print file needs the right crop, enough pixels for the chosen physical size, and the color profile and format requested by the printer. Start both from a finished master instead of trying to make one small JPEG serve every purpose.

This lesson begins after stretching, color work, star restoration, and other finishing. It applies to broadband RGB, LRGB, and mapped-color SHO/HOO images. Export color management preserves the intended display appearance; it does not turn a narrowband palette into natural color.

Save Target_finished_master.xisf at full resolution before resizing or reducing bit depth. Keep your linear source and useful processing branches too; the finished master is a display-ready checkpoint, not a replacement for them.

Disable STF and confirm that the image still has the intended brightness. STF changes the screen display without permanently stretching the pixels. If the image becomes almost black, return to the stretching lesson and make a proper nonlinear finishing copy before export.

Inspect the complete image and representative areas at 1:1: bright star cores, faint dust, smooth sky, sharp edges, and corners. Check for clipping, banding, halos, noise, crop remnants, and accidental edits. Remove unintended masks before further adjustments on an export copy.

Use a reasonably calibrated and profiled display for color decisions. Avoid judging print brightness on an extremely bright screen or while a night-light color filter is active. Different displays and viewing conditions can still produce differences.

An ICC profile describes how the image’s RGB numbers represent colors. Assigning a profile labels existing numbers; converting between profiles changes the numbers to preserve appearance as far as the destination gamut allows. Do not assign sRGB over an Adobe RGB image as a shortcut for conversion.

In PixInsight, locate ICCProfileTransformation through All Processes. Inspect the source profile associated with the image and use the installed process help for the current controls. If the source is untagged, establish the intended profile from the image’s history and color-management setup before converting. Do not guess a wide-gamut profile to make colors look stronger.

AssignICCProfile is appropriate only when you know the correct interpretation of untagged or incorrectly tagged data. Your monitor’s device profile belongs in display color management, not as the delivery image’s RGB profile.

RGBWorkingSpace and ICC color management are different systems in PixInsight. Changing RGBWorkingSpace is not an export conversion to sRGB. SPCC also serves a different purpose: broadband color calibration does not replace an embedded export profile.

Duplicate the master and rename the view Target_web. Choose dimensions based on the destination’s current upload requirements. For practice, a 2400-pixel long edge is a reasonable comparison size, not a rule for every website. Some destinations create their own smaller display versions.

Open Resample through All Processes and use its pixel-dimension controls. Preserve the aspect ratio: for a 6000 × 4000 image, a 2400-pixel long edge produces 2400 × 1600 pixels. Verify both output dimensions before applying once to the duplicate. Do not force an arbitrary width and height that distort the image.

Use the process’s normal interpolation settings as a starting point and inspect small stars and edges afterward. Resampling changes pixel data. If the size is wrong, make another copy from the master rather than repeatedly shrinking and enlarging the same image.

Judge the resized result at 1:1 and at its expected display size. Optional output sharpening should be subtle and tested after resizing; skip it for a first export if the image already looks crisp. More sharpening is not automatically better for small stars or faint nebulosity.

sRGB is a practical web delivery space. If the copy already has the correct sRGB profile, it needs no repeated conversion. Converting a wide-gamut image can change very saturated colors that the smaller destination cannot represent.

Make Target_print from the full-resolution master, not from the resized web copy. First obtain the print service’s requirements: physical size, aspect ratio, minimum or preferred pixels per inch, accepted formats and bit depths, color profile, borders or bleed, and whether automatic image corrections can be disabled.

PPI means image pixels per printed inch. It differs from the printer’s ink-dot DPI. Required pixels = print inches × requested PPI. For example, a 12 × 8 inch print at 300 PPI uses 3600 × 2400 pixels. A 6000 × 4000 image corresponds to 20 × 13.33 inches at 300 PPI before any crop.

300 PPI is a common planning value, not a universal minimum. Larger prints viewed from farther away may use less; follow the provider’s guidance. Changing a resolution tag alone creates no additional detail. Upsampling creates more pixels, but cannot recover detail that was never captured.

Match the composition to the print aspect ratio. A 3:2 image does not fill an 8 × 10 inch rectangle without cropping, borders, or distortion. Choose a deliberate crop or ask for borders; never squeeze the image to fit. Recalculate available pixels after cropping.

Use the lab’s requested delivery space. Many photographic services accept sRGB, while others support Adobe RGB or another specified workflow. Wider gamut is useful only when the complete delivery chain supports it. Keep the print branch independent of the web conversion.

If the lab supplies a printer-and-paper ICC profile, read whether it is for soft proofing or for actual file conversion. Do not automatically convert the delivery file to that device profile. Soft proofing simulates the output on screen; it is not itself a required change to image pixels. Use a color-managed proofing application if needed and request a small test print for critical work.

Do not convert an astrophotograph to CMYK merely because it will be printed. Follow the specific provider’s instructions. If printing yourself from another application, coordinate its color handling with the printer driver so the same conversion is not applied twice.

When accepted, a 16-bit unsigned-integer RGB TIFF with an embedded requested profile is a useful print handoff. Use File → Save As on the print copy, select TIFF, and check the format dialog’s sample type, profile embedding, and compression settings. Choose no compression or supported lossless compression. Avoid floating-point TIFF unless the receiving service explicitly supports it.

Some labs instead require 8-bit JPEG or TIFF; supply that specification rather than assuming 16-bit is mandatory. Keep processing at the master’s precision until the delivery stage, reopen the final file, and verify it before upload.

For web, open the export in a color-managed browser and inspect the uploaded version too. The service may resize or recompress it. Compare at a similar displayed size with STF off in PixInsight; a fitted preview and a 1:1 image can look different even with the same pixels.

For print, check the ordering preview for orientation, crop, borders, and the selected physical size. Disable optional automatic enhancement if you intend to preserve the prepared appearance and the service permits it. A proof or small test print is a better basis for adjustment than making every print file arbitrarily brighter.

Evaluate a test print under suitable lighting. Paper and other print surfaces reflect light rather than emitting it like a screen. Save any print-specific brightness, color, or sharpening adjustment as a separate version for that size and material.

Keep filenames descriptive, for example Target_print_12x8_sRGB_16bit.tif. Record the actual dimensions, profile, format, bit depth, crop, and provider specification; a filename alone is not proof that the file matches them.

The export is almost black: check whether the apparent stretch was only STF. Return to the finishing master and establish its real nonlinear state before trying to brighten a JPEG.

Colors differ strongly outside PixInsight: check the source profile, conversion versus assignment, embedded destination profile, viewer color management, and display settings. Do not change global color settings just to compensate for one mismatched viewer.

The sky has blocks or bands: compare the high-precision master with the resized copy and saved file. JPEG compression and reduced precision can reveal problems; try higher quality or lossless output when accepted, and avoid pushing gradients after 8-bit conversion.

The print is cropped or soft: compare the ordered aspect ratio, crop, delivered pixel dimensions, and effective PPI. Changing the DPI label does not fix a lack of pixels.

Before delivery, confirm: correct version, permanent stretch, intended crop and dimensions, correct embedded ICC profile, accepted format and bit depth, no obvious artifacts, and a successful reopen of the exact saved file. Keep the master untouched for future exports.

This is an original beginner workflow. PixInsight’s installed process and file-format help is the reference for current controls. The Adobe references explain general output concepts; their Photoshop menu instructions are not PixInsight commands. Your chosen print provider’s current requirements take precedence over the example settings here.