Is your feature request related to a problem? Please describe.
Currently, the manual white balance values must be iteratively and manually tuned based on visual appearance. This is not very rigorous, and hard to ensure good calibration between machines.
Describe the solution you'd like
A more rigorous way to tune white balance, used by Fabien Lombard, is to capture an image with no flowcell/ibidi chamber, and then use software to measure the mean RGB value of the image. Then that value can be used to calculate the white balance gains needed to achieve some desired mean RGB value. Fabien requested (in the 2023-08-31 software taskforce meeting) that the software provide an interface to automate the adjustment of the manual white balance gains (e.g. after the user presses a single button).
This feature request was made by Fabien.
Describe alternatives you've considered
Another option which would be somewhat less automated and less integrated into the software, but which could be an intermediate prototype provided as an experimental feature in one release and then integrated more fully in a subsequent release, could be to provide a standalone tool where the user pastes in an image with no flowcell/ibidi chamber, and then the tool proposes a pair of manual white balance gains to set. Then the user can paste in an image acquired with those gains, and either the software tells the user that the gains are good (if only one iteration of trial-and-error for gain adjustment is required) or it provides them the next pair of gains to try (if white balance adjustment must be an iterative process).
If providing the frontend functionality to paste in an image is too difficult for the prototype (because it'll require choosing and adding a javascript library), we could just have the prototype ask the user to upload an image. It should be very easy for the user to right-click the live camera preview stream and either save the current frame or copy the current frame to their clipboard.
Is your feature request related to a problem? Please describe.
Currently, the manual white balance values must be iteratively and manually tuned based on visual appearance. This is not very rigorous, and hard to ensure good calibration between machines.
Describe the solution you'd like
A more rigorous way to tune white balance, used by Fabien Lombard, is to capture an image with no flowcell/ibidi chamber, and then use software to measure the mean RGB value of the image. Then that value can be used to calculate the white balance gains needed to achieve some desired mean RGB value. Fabien requested (in the 2023-08-31 software taskforce meeting) that the software provide an interface to automate the adjustment of the manual white balance gains (e.g. after the user presses a single button).
This feature request was made by Fabien.
Describe alternatives you've considered
Another option which would be somewhat less automated and less integrated into the software, but which could be an intermediate prototype provided as an experimental feature in one release and then integrated more fully in a subsequent release, could be to provide a standalone tool where the user pastes in an image with no flowcell/ibidi chamber, and then the tool proposes a pair of manual white balance gains to set. Then the user can paste in an image acquired with those gains, and either the software tells the user that the gains are good (if only one iteration of trial-and-error for gain adjustment is required) or it provides them the next pair of gains to try (if white balance adjustment must be an iterative process).
If providing the frontend functionality to paste in an image is too difficult for the prototype (because it'll require choosing and adding a javascript library), we could just have the prototype ask the user to upload an image. It should be very easy for the user to right-click the live camera preview stream and either save the current frame or copy the current frame to their clipboard.