@Signhere, for some reason I don't quite understand, you seem to be trying to present a case that well documented issues with earlier versions of the Mobius firmware that have been reported, discussed and demonstrated numerous times on DCT,
RCgroups, YouTube and elsewhere, along with being acknowledged and eventually resolved by the developer are somehow completely non-existent. Just because you can present a video where the problem doesn't manifest doesn't really prove much of anything useful and cherry-picking video that doesn't happen to show well documented firmware problems confuses the issue. For example, driving into the sunset or bright sky at dusk behind a truck with red
& bright yellow stripes won't likely initiate the color shifts being described here and as mentioned previously footage shot at night when the sky is still not quite dark doesn't give a completely fair assessment of night time performance.
One of the vexing things about some of these early color shift issues at the time was that they sometimes only manifested under a unique combination of circumstances. The DSP would overcompensate the WB adjustments only when a particular set of RGB numerical values came into view. Nevertheless, this happened quite often and when it did in these earlier firmware versions they were deemed unacceptable to everyone for obvious reasons.
I think we can all acknowledge that more recent firmware versions have been beset with their own problems and this is perhaps the reason for the focus on these older versions but I vote we march forward rather than backwards. Of course, anyone who wishes to use earlier versions of the firmware should certainly do so if they work well for you and by all means post videos of the results but implying that these problems simply don't exist in these firmware iterations makes little sense.
A quick YouTube search turned up this cyan color shift example. Notice that the color shift occurs shortly after the red CocaCola sign enters the frame on the left side of the road at the 12 second mark where it becomes a large enough percentage of the frame to achieve a tipping point for the shift, continues as more red lettering appears again on the left side of the road and then very
slowly begins to resolve as the car proceeds. THIS is the real and well documented firmware
white balance stabilization issue being discussed here.