Fuji XPro/Xtrans a mistake?

Started May 16, 2012 | Discussions thread
mjl699 Regular Member • Posts: 309
Re: Fuji XPro/Xtrans a mistake?

One thing that I think is going on, which is not being discussed here, is that you cannot restore what is blurred away with a Moire filter. What I am saying is (my opinion only):

1. If you ignore colour when you are processing, and just take pixels into account, the ultimate resolving power of a sensor with a Moire filter is smaller than one without with the same size and pixel count. So the X-Pro1's resolution should be better than an equivalent sensor with a Moire filter.

2. If you have a Moire filter, edges are always blurred. Even when they are in focus. That is the purpose of the Moire filter - to smear light of one colour across more than one pixel. If they were not blurred across more than one pixel, then Moire would be the result whenever colour patterns with "resonant" frequencies (in XY space) were incident on the sensor. So if you imagine the edge between two squares placed side by side, one black and one white, then you can see that some of the light from the white square must fall into the black square as a result of the Moire filter. So where the X-Trans sees an edge (there will be at most 1 grey pixel if the sharp edge falls onto the centre of a pixel and some white light is incident), a Bayer sensor must see black, then grey (2 or 3 pixels, I don't know, but more than the X-Trans) then white. No edge. You can also see that the Moire filter has a colour as well as a geometric effect (a thicker strip of grey has appeared from nowhere). What this means is that to render an edge for any sensor with a Moire filter you have to do some sharpening to get an edge that looks "in focus". This has an effect on the processing and on the final "look" of the image, and in my opinion an artificially sharp look is one of the main ways poor quality JPEGs from digital cameras under-perform versus film. In trying to find an edge, they have to "create" one that is not on the sensor. In doing so they can really mess up the look of the fine detail. Photograph hair on a 16 MP sensor with a Moire filter and then do it with the X-Pro1 - you'll see what I mean. Ever wondered why grainy film can look better than razor sharp JPEGs? I think this is why. Even wondered what makes the X-Pro1's rendering so damnably good? I think this is why. Straight from the camera at 100%, the X-Pro's edges look more natural, more like the film look.

3. The density of colour pixels in the X-Pro1 is slightly less than that of a Bayer filter (I think it is (R-G-B) 8-20-8 for the X-Pro1 and 9-18-9 for a Bayer sensor. Given that this is an 11% difference in the density of R and B and also that the spacing of R and B is uneven, its seems quite possible to me without having done any kind of mathematical analysis that the colour accuracy of a X-Trans sensor might be lower than the colour accuracy of an equivalent Bayer sensor. Someone more knowledgeable than me needs to comment.

4. The Nikon D800 has a much higher pixel count than the X-Pro1. So much higher in fact that for every 1 X-Pro1 sensor pixel, there are 2.25 Nikon Sensor pixels. That's amazing to me. This means for every 1 pixel you pass over going across or down the X-Pro1 sensor, you will pass over 1.5 pixels in the Nikon. This is a massive numerical advantage and will eliminate to some extent the benefit of the missing Moire filter in the X-Pro1. It stands to reason that the Nikon D800E will be the best except in the presence of Moire . But then the Leica M9 (18 MP, no Moire filter) has been about a while, must suffer from Moire, and you still see amazing photos from that without Moire. So the D800E could be truly awesome, and maybe the best of the best in the ultimate image game.

5. Some of the DPR samples for the X-Pro in the comparison tool seem to be out of focus. My images appear better in RAW at 100% than theirs. I am staring at images reflected in the pupils of my subjects and looking at hair, etc, so I don't think it’s just that I am looking at images with larger structures and less fine detail. Be interested in seeing a DPR comment on that and also in seeing if correcting that leads to a better comparison between the X-Pro1 and other cameras (e.g. versus Canon EOS mk III and Nikon D800).

X-Pro1. Great images, good handling, fun, etc etc. Well done Fuji for the best images I have ever been able to take. Keepers fall out of the camera. My only real complaint is focussing is a little slow on moving toddlers indoors in poor light - you have to take a few shots to get one in focus. But everything is liveable for the results, including Silkypix, which I find as annoying as everyone else!

 mjl699's gear list:mjl699's gear list
Canon EOS 400D Panasonic Lumix DMC-GX1 Fujifilm X-Pro1 Canon EF 50mm f/1.4 USM Canon EF 28-135mm F3.5-5.6 IS USM +5 more
Post (hide subjects) Posted by
(unknown member)
(unknown member)
(unknown member)
(unknown member)
(unknown member)
(unknown member)
(unknown member)
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow