Hi,
I shoot very short exposures so I stack thousands of frames. I’ve created scripts to handle this by stacking them in blocks of 10 minutes of exposures each.
One thing I do to keep intermediate disk storage space usage minimized, as it stacks several thousand lights, is I use RICE compression on the files (ex: the 32-bit calibrated and registered lights). This works wonderfully and I’ve checked the noise and FWHM of the output compared to no compression and the RICE compression essentially adds little to no additional noise.
One mystery is I’ve played with the different RICE quantization levels including levels 4, 8, 16, 32, and 64 and the file sizes are smallest with 4 and increase and are the largest with 64. But the documentation states that a larger number is “more quantization” and more loss so I would expect file sizes to shrink with higher levels of quantization. Quant setting 64 produces about 2x the file size as Quantization setting 4.
Quote: If it is too large, one undersamples the pixel values resulting in a loss of information in the image. If it is too small, however, it preserves too much of the noise (or even amplifies the noise) in the pixel values, resulting in poor compression.
Can you please solve this mystery for me? It appears that level 4 is actually the highest level of quantization from the file size perspective.
This is SIRIL 1.4.0b2
Regards,
Steven