Thanks!

There’s a big difference here. The channels I synthesize actually exist, and based on their data, I derive a formula for converting between photometric systems. If a channel doesn’t exist, there’s no reference result from which to derive the formulas. Any estimate of a nonexistent channel is rough.

Example coefficients obtained using the least squares method based on a sample of one million pixels from all OPAL maps: F467M(F395N, F502N) = 0.18 + 0.80 a - 0.58 b - 0.65 a² - 0.37 ab + 2.16 b² + 5.87 a³ - 9.09 a²b + 4.84 ab² - 1.73 b³ The root mean square deviation σ = 0.01206. [3/3]

For example, OPAL Cycle 22 maps of Jupiter (Jan. 2015). Rotation A has all the necessary filters for high-quality spectral reconstruction: F395N, F467M, F502N, F547M, F658N. Rotation B missed F467M and F547M. These were synthesized from F395N, F502N, and F658N. [2/3]

I'm working on a true color processing of the outer planets maps, projected by the HST team. The selection of filters varied from year to year, resulting in slightly inconsistent spectral reconstructions, so I tried to recreate the data for the missing filters using synthetic maps. It worked. [1/3]

True color processing of Mars: ten years apart, Hubble and Psyche captured roughly the same region. You can see Sinus Sabaeus, with Arabia Terra above and the South Polar cap below.

Map of Mars in true colors: images from Mars Express were projected by Michael et al. (2024) and post-processed with TrueColorTools. TCT used the Tikhonov regularization to reconstruct a spectral cube from four High Resolution Stereo Camera (HRSC) channels and simulated human perception of it. [1/3]

I present three maps of Titan: one in visible "true" colors (albedo scaled and gamma corrected), one in enhanced colors (linear albedo scale with a factor of five), and one of the lakes. In this thread, I will briefly describe how the impenetrable haze was bypassed.