Give me a full frame trichromatic sensor already!

Started Jan 18, 2018 | Discussions thread
Jack Hogan Veteran Member • Posts: 7,310

fPrime wrote:

Iliah Borg wrote:

If SMI on trichromatic is lower that on a regular Bayer, following Jack's calculations, how does that fit into your argument?

Jack has acknowledged his source data was imperfect. His SMI numbers are completely bogus...

Pardon me for begging to differ

I stand by my estimates of SMI from the given, nearly identical captures in those natural D65ish conditions: the Standard Back turns out to be relatively more accurate with an SMI of around 87 vs about 81 for the Trichromatic. FWIW 87 is excellent and 81 is noticeably less so, based on my experience. A couple of other clarifying comments:

1) I don't care which is better or worse, I am in it for the intellectual curiosity.

2) I am pretty comfortable with the results produced by my procedure (matrices and metrics). If you have any questions about any part of it I welcome them, we may both learn something new in the process.

3) When I say things like "these figures say nothing about how well ‘accuracy’ generalizes outside of the very limited ColorChecker tone set" I do it to inform the reader who may not have thought about that - but that doesn't mean that the numbers are bogus. On the contrary, for the limited sRGB tone set represented in the captured ColorChecker (not exhaustive but fairly typical of ones found frequently in real life) they are indeed valid and correct.

If you would like me to estimate SMI for the two backs in different conditions just point me to two nearly identical raw captures of a target with published reference specs and I'll be happy to run it on them. I asked Usman Dawood at Fstoppers for his raw files but he wasn't forthcoming.


Post (hide subjects) Posted by
(unknown member)
(unknown member)
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow