Article

Some aficionados of Sigma cameras like to claim its Foveon sensor images possess a mysterious image quality missing from competing technologies. The effect, unfortunately, isn’t easy to describe because it is is intangible and hard to quantify – but you will know it when you see it, they say. In some ways, this echos the claims sometimes made about the special “look” of Leica lenses, their special “draw”.

Fans refer to the “effect” variously as “a 3D effect”, “verisimiltude”, “micro-contrast” and just plain “beautiful”, whilst critics dismiss it as fantasy.

Whether you are a believer or a critic, determining whether the effect is real or imagined is not easy – it frustratingly refuses to submit to simple analysis. There are a number of possible explanations for this difficulty. On the one hand, there is a hint of elitism apparent in the claims of some believers: non-believers have been accused of being “trolls” and of various kinds of mental, perceptual or moral deficiencies. On the other hand, even I (a heart-on-the-sleeve sceptic on this question) have seen tantalising glimpses from occasional Foveon images of a suggestion of “depth” to the rendition that is very pleasing.

It would be easy (and trite) to dismiss the fuss as cultish “fanboyism” but the conviction with which its adherents press the claim does seem to transcend the usual brand based tribalism. There is fervent conviction at work. I think the question of existence or non existence of the “effect” can’t simply be explained away as internet forum nonsense; it deserves investigation. So, is it real? Or just the imagination in overdrive?

I’ve given a lot of thought about how to find out and having dismissed a number of unlikely schemes, I settled for just about the simplest. I have decided to shoot pairs of photographs of a range of everyday subjects with different lenses and settings and under different lighting conditions. I will make these available unlabelled for anyone to inspect. Voting links will be provided so observers can vote for their favourite from each pair or submit a no preference opinion. Over time this will build up a database of preferences that could be analysed to yield a definitive answer to the question of whether there is a Foveon effect.If the answer proves to be a “Yes”, then we can move on to part 2 – isolating the cause and working out how to exploit it to good effect in our images.

Below you will find the beginnings of a library of Foveon vs Bayer CFA sensor images.Click on the thumbnails to view the full size composite (large file)To contribute to the database, email your preference clicking an appropriate voting link.

Note: This project stalled pretty early on and developments in both Bayer sensors and the Merrill version of the Foveon sensor have left the quest behind. I did get quite a lot of emails from people making their guesses who somehow found this page.  Of entirely academic interest now, but using the two pairs of images originally supplied, it was clear from the results and comments that people couldn’t really tell the images apart consistently but they tended to prefer the Bayer images.  I don’t think this means a great deal as the images used were not ideal. Until a proper test can be performed, the question remains as open today as it did when Foveon first appeared.

ID	Images
0001		Vote for left image	Vote for right image	Vote No Preference
0002		Good light image