Olympus High Resolution (Pixel Shift) support

Hello all, i am a new poster, and fairly new with Rawtherapee (as well as digital photo processing).
I have a fair experience in computer programming (my past live as i am retired as of now, but only Windows since 1995), photography (as hobby, including digital since 2005) and the owner of a em-5 mark ii camera able to take hires photo.

I am very interested in understanding the way one can extract the best of those HiRes photo, and, reflecting on the subject, ended up in expecting that an end-resolution of about 4 times the sensor resolution photosites is required to extract all the details present in the original photo (plus the x4 to rebuild a clean equivalent bayer sensor capture, but there i may/must be wrong as the bayer image processing down the pipeline may take advantage of the bayer filter layout for processing sharpness and contrast).

I would be very thankfull if one could get me started in this.

Just for clarification on the above message:
It is very strange to compare the non HiRes ORF images to the normal res ORF/ORF images. Olympus normally packs a em-5 mark II 16 mega pixel in a less than 16 MB file, with a strip offset of a little more than 1MB. For an Hi-Res image, it stores the raw output of 64 mega pixel into a exceeding 100MB file, the strip offset being the same.
Examining the Tags, they are no apparent différences in the way the color pixels are managed, other than the order, so they are apparently no différences between the two files types on the basic strip data composition.
For me, i would have expected the hi-res raw file to be around 50 to 60 MB (64/4 = 16, 12.5 * 4 = 50), meaning that another image strip could be present in the raw file.
Has anyone explored this?

Hey

I moved your posts to the RawTherapee category as you originally posted this in the darktable category.

Thanks

Peux-tu mettre à disposition un fichier d’example sur filebin.net par exemple ?

There is one on https://raw.pixls.us/

Here is an interesting image set taken with my e-m5 II : Filebin | 3d9twdqhyjcq2qrh

I have started to look into the binary data, and i believe this is a lost cause because while the single image raw file (with the extension ORI) image data is certainly packed, if not compressed, the multi image raw file isn’t even packed, so this may explain the size increase. This may be due to lack of resources in the camera, but then, this is a pity, because storing the 8 pictures taken would not have cost a byte more of storage and offered plenty of post processing opportunities…