Plate Solving with ImageSolver and Color Calibration with SPCC

Plate solving matches the stars in your photograph to a catalog. ImageSolver adds an astrometric solution: a mapping between image pixels and sky coordinates. It identifies where the image is pointed without stretching or color-calibrating its pixels.

SpectrophotometricColorCalibration (SPCC) then measures catalog-matched stars and calculates a color balance using their spectral information, your imaging response, and a chosen white reference. The resulting correction applies across the image, including the nebula or galaxy between the stars.

The sequence is: prepare a linear RGB image → solve it → configure SPCC → apply once → inspect and save. This lesson teaches those two operations, rather than prescribing the order of every processing tool.

Use a calibrated, aligned, stacked broadband color image with stars still present. For a one-shot-color camera, start with a properly debayered RGB integration. For a monochrome camera, combine the aligned R, G, and B masters into RGB first. Keep a separate luminance master aside; SPCC’s color step operates on RGB.

Work on a duplicate and save the untouched integration. Remove unwanted stacking borders before this exercise, then solve the image at that geometry. If you later crop, rotate, or resample it, confirm its solution is still valid before another catalog-based operation.

Linear means no permanent stretch has been applied. STF changes only the display, so you can use it to see a dark linear image. A stretched TIFF or JPEG is not a replacement for the original linear master. Run this exercise before star removal, saturation edits, or adding Hα into broadband RGB.

This first lesson excludes dual-band captures and mapped SHO/HOO images. Their filter responses and color goals require a separate narrowband workflow; ordinary broadband settings cannot recover missing broadband color information.

Have Gaia DR3 registered for local ImageSolver use and Gaia DR3/SP registered for SPCC. They are separate selections in Gaia Preferences. If this is not done, use the linked installation guide under Continue learning before proceeding.

Gather the approximate image-center coordinates (right ascension and declination), the observation date/time in UTC, and the image scale. ImageSolver can often read these from image metadata, but missing or incorrect values need correction.

Image scale can be entered as arcseconds per pixel, or calculated from focal length and pixel size. Include reducers or Barlows in the effective focal length. Binning, resampling, and drizzle change the output pixel scale, so use values describing this image rather than blindly copying the camera’s native pixel size.

For example, 206.265 × 3.76 µm ÷ 500 mm gives approximately 1.55 arcseconds per pixel for an unbinned, unresampled image. A 2× drizzle output has roughly half that scale, about 0.78 arcseconds per pixel. These are examples, not settings to copy for another telescope.

Activate the full RGB image you want to solve. Menu wording can vary between PixInsight releases; use the installed tooltips for controls that differ.

Look for a completed astrometric report with center coordinates, resolution, rotation, and saved solution information. A warning is not automatically a failure; read the final result and any actual error message.

Compare the reported field center and scale with your acquisition plan. If available, inspect a coordinate readout or annotated view for recognizable objects. An image that looks unchanged is normal: plate solving adds coordinate information.

Save the solved image as a new XISF file, for example Target_RGB_solved.xisf. If WBPP already supplied a valid solution at the current image geometry, you can reuse it instead of solving again simply because this lesson includes that step.

Open Process → ColorCalibration → SpectrophotometricColorCalibration, or find it under All Processes. Reset a reused instance before configuring it so that another image’s settings are not carried over unnoticed.

Select the response curves for the equipment that collected these data. For monochrome RGB, match the sensor QE curve and the separate red, green, and blue filters. For a color camera, use the documented combination representing its Bayer responses and any additional filter; a bare sensor QE curve alone does not describe the complete color response. Follow the installed curve descriptions and process help to avoid counting the same response twice.

If an exact camera or filter entry is missing, check official updates and FilterManager documentation. Use a documented compatible response when available and record any approximation; choosing a familiar brand name is not a substitute for matching the response.

Keep narrowband mode off for this broadband exercise. Select Gaia DR3/SP in Catalog Search and start with automatic limit magnitude. Leave Apply color calibration enabled if your version exposes that checkbox.

A white reference defines what counts as neutral; it is not a star you must click in the image. Use the reset instance’s default for a first exercise and record its name. Different references represent different color conventions, so changing the reference to chase a preferred hue changes the meaning of the calibration.

Background neutralization is an optional part of SPCC. It estimates the background color; it does not remove a spatial gradient or turn real colored nebulosity into empty sky.

If there is an identifiable patch of background, create a preview over it. Avoid bright stars, halos, nebula, galaxy light, dust, and stacking edges. Enable Background Neutralization and its Region of Interest option, then use From Preview to select that preview. The region limits background estimation; it does not restrict the whole color calibration to that small rectangle.

Do not force a background preview if the frame is filled with emission or dust. Leave background neutralization off for this first color-calibration test and handle the background with an image-appropriate workflow. Keep rejection limits at defaults initially; SPCC’s lower and upper limits are statistical sigma values, not ordinary pixel-value thresholds.

If your workflow uses MultiscaleGradientCorrection, follow its separate plate-solving and SPFC prerequisites. This lesson’s SPCC step is not a replacement for SPFC, and it does not require MARS just to run ImageSolver and SPCC.

Enable Generate graphs if available. Apply SPCC to the solved linear RGB image using Apply to active view, or drag its New Instance triangle onto the image. Wait for completion and inspect the console for matched stars, fitted samples, and a completed calibration.

Review the reported fit graphs. Look for a usable spread of samples following the fit rather than very few points or a cloud with no relationship. There is no universal good-star count or scatter threshold for every field; a successful run still needs a visual check and sensible camera/filter settings.

Reset STF and recompute a linked RGB STF. An old display stretch can make a sound calibration appear to have a new color cast. Compare both images with consistent display settings, then inspect stars and faint structures at 1:1. Color calibration need not make every star white or produce strong saturation.

Save Target_RGB_SPCC.xisf and retain the solved pre-SPCC version. If you change settings, return to that checkpoint for a fresh comparison rather than repeatedly applying different calibrations to the same already-corrected pixels. Save the process instance and note the white reference, response curves, background choice, and PixInsight version.

ImageSolver cannot align the reference field: recheck the image center, units, effective focal length, and output pixel scale. Then inspect the stars and confirm local catalog coverage. Change one cause at a time before adjusting advanced matching settings.

No astrometric solution: make sure you solved the same image you are now passing to SPCC. A filename containing “solved” is not proof that its current pixels retain valid coordinates.

Missing Gaia database: open Gaia Preferences and check the relevant entry. DR3 is the local solving catalog in this exercise; SPCC needs DR3/SP. A configured entry does not help if its files were moved or an external drive is disconnected.

Insufficient photometric samples: verify linear data, intact unsaturated stars, correct astrometry, suitable response curves, and spectral-catalog coverage. Do not try to fix a stretched image just by raising the catalog magnitude limit.

Unexpected color: recompute linked STF, check the camera/filter response and white reference, and review any background region. For dual-band or mapped narrowband data, stop and use a dedicated workflow rather than treating broadband calibration as universal.

Before asking for help, keep the full error text and record the image’s processing state, camera/filter setup, image scale, selected catalog, and software version. Forum replies linked here include community advice; the installed help and current official documentation take precedence for version-specific controls.

Your RGB image is still linear, contains stars, and has a verified solution at its current geometry. SPCC used DR3/SP and response settings appropriate to the acquisition. You inspected its console and graphs, refreshed the display stretch, and saved a separate calibrated checkpoint.

You can now continue with the remaining processing sequence for your data. Keep the calibration record so you can explain what was done when comparing images or returning to the project later.