Fujifilm X-Pro1 details leaked!!!

[rover]

Oh, I forgot to thank you for your cooperation.

 

[rover]

http://shopping.yahoo.com/articles/...-digital-camera-makers-struggle-to-keep-pace/

 

[Spyro]

Well I think this camera is cheap LOL
In the segment mirrorless-system-with-OVF iτ's about 1/4 the price of the competition, so...

I also think of it it a good sign that most people don't consider it a professional camera, because the last thing I wanted is most people's idea of a professional camera hanging around my neck 🙂

 

[dfoo]

How does the AF work with the X100? Can you tell what the camera is focused on? With an SLR you can generally tell more or less due to DOF. One of the things that really irritated me with the Hexar AF is that from looking through the VF you couldn't easily tell (without looking at the distance scale).

 

[emraphoto]

to this day the x100's af reputation puzzles me.

 

[JohnL]

"The viewfinder is located on the top left corner, which can make it a little awkward for left-eyed shooters like myself to handle"

Better than having it in the middle which makes it awkward whichever eye you want to use!

 

[Paul T.]

How does the AF work with the X100? Can you tell what the camera is focused on? With an SLR you can generally tell more or less due to DOF. One of the things that really irritated me with the Hexar AF is that from looking through the VF you couldn't easily tell (without looking at the distance scale).

The Hexar is a good example in a way, because it is possible to mis-focus. But then, the distance scale is in the VF, and it's easily good enough to indicate if you're focused on the wrong element - you just eventually learn to keep an eye on it. I only mis-focus when I'm out of practice.

I doubt the X100, or any other camera with contrast detect AF, will ever measure up to the Hexar. But learning to live with its quirks, especially learning what it's focused on, it's likely it's good enough.

The internet X100 AF speculation might well be over the top - but if it persuades Fuji to go one better, that can only be a good thing.

 

[JohnL]

Fuji has admitted it has poor AF compared to D-SLRs. A $2000+ camera should not have inferior autofocus or anything else. Sorry.

Damn, I guess all my M- Leica cameras are a super rip off considering that they have no AF of any kind 🙂

No, you are OK! All the M-Leicas are qualified to be $2000+ on both counts. They don't have inferior autofocus or anything else.😀

 

[Focal Plane Circus]

Here is more information to fuel this discussion (while it remains open) from a Spanish review of the X-Pro1 system: http://www.dslrmagazine.com/digital...peciales/fujifilm-x-pro1-primer-analisis.html

The article's lens comparisons, in particular, should kindle more debate. There is probably even more incendiary material in the Spanish text for those who can read it. The rest of us will have to bicker over the photos and charts.

Update: Google translation of the Spanish review: http://translate.google.com/transla.../fujifilm-x-pro1-primer-analisis.html&act=url

On other topics:

Now, if I want a digital camera with a viewfinder that allows me to shoot right-eyed ala a RF, I have exactly two options right now: a NEX-7 or (soon) a XPro-1.

Correction: at least 3 option right now: (soon?) a NEX-7 or (soon) an X-Pro1 or (later) a future X-Pro [see below]

As far as RF goes, contrast-detection is not as fast as phase-detection across the full spectrum of uses. Maybe it someday will be, but not right now.

Near the end of this interview, Kayce Baker says that a future X-Pro camera will definitely have phase-detection: http://www.imaging-resource.com/news/2012/01/12/imaging-resource-interview-kayce-baker-fujifilm

We will now return you to our regular flame-thrower thread.

 

[Mcary]

When shooting with the X100 I use a single user selected and illuminated focus point just like I did when shoot with the 5D. Personally I can't see any difference in accuracy or speed even when shooting wide open at F-2. Also just like with the 5D sometimes I need to move the focus point around a bit in order to find an area with enough contrast to get a lock.
So if the X-1Pro has even a slight upgrade in AF speed and accuracy it'll be an improvement to what I'm use too.

 

[splitimageview]

I going to wait until actual user reports start to come in before getting too excited about potential AF problems.

Why waste a golden opportunity! 🙂

 

[noimmunity]

Here is more information to fuel this discussion (while it remains open) from a Spanish review of the X-Pro1 system: http://www.dslrmagazine.com/digital...peciales/fujifilm-x-pro1-primer-analisis.html

The article's lens comparisons, in particular, should kindle more debate. There is probably even more incendiary material in the Spanish text for those who can read it. The rest of us will have to bicker over the photos and charts.

The chart that maps resolution and S/N ratio against the Canon 5D and the Nikon D700 purports to show that the Fuji easily outdoes both in terms of resolution, and only loses to the D700 for S/N starting at ISO1600 (while it remains ahead of the 5D).

If true, that would be very impressive. Finally, we have some concrete indication of what Fuji's claim that their new X-Tran Sensor is better than previous full frame can mean.

Then there are the MTFs, which also appear to favor the Fuji lenses against Canikon counterparts...

The review does notice, as we all have, that Fuji has positively encouraged apples to oranges comparisons...

The review also mentions that manual focus should be improved over the X100. Cool.

Thanks for the link...

 

[willie_901]

The X-Pro1 Photosite Arrangement Is Not Random

The X-Pro1 Photosite Arrangement Is Not Random

The aperiodic photosite array in the X-Pro 1 is not random, nor are the photosites randomized. I'm not even sure the pattern is actually aperiodic, but my math skills are too pathetic to comment further about that.

Here's Fuji's graphic of the array design.

http://www.fujifilm.com/products/digital_cameras/x/fujifilm_x_pro1/features/

Here's an interesting analysis of the array.

http://kevinpurcell.posterous.com/fujifilm-x-trans-color-filter-array-layout-ha

 

[Focal Plane Circus]

The aperiodic photosite array in the X-Pro 1 is not random, nor are the photosites randomized.

Here's an interesting analysis of the array.

http://kevinpurcell.posterous.com/fujifilm-x-trans-color-filter-array-layout-ha

Interesting indeed. I had noticed the repeating 9x9 pixel blocks and that the 3x3 blocks within every 9x9 block are identical, except that every other block is turned 90 degrees, giving the whole sensor a checkerboard pattern. The rest of the analysis is news to me.

I can now see an implication of the 2x2 RGGB blocks that "can sample color info with about half the resolution of the array" (another checkerboard): the X-Pro1's sensor is potentially an EFR sensor.

The author of the analysis is also pretty interesting. His Twitter page tells more about him:

https://twitter.com/#!/kevinpurcell

Thanks for the links! 🙂

 

[Doug]

"Random" is not the right word to describe the regular repeating arrangement in this sensor, but I can see where one would struggle for a better word to distinguish it. Maybe "diversity", as each row and column contains all three colors, unlike the Bayer array. I also don't see how the XP1 sensor would have been inspired by film grain; perhaps that thought arose from the misuse of "random". Certainly an interesting development, and interesting to see how the results may show improvement.

 

[ferider]

If anybody cares, for correct terminology:

Nothing is random in computer science, except noise.

When "random-ness" of random number generators is characterized, the generator's periodicity is used. The longer, the more "random". Every random number generator is periodical.

The new Fuji sensor has a longer periodicity (is more "random") than the Bayer array. The arrangement of pixels has higher "entropy".

Roland.

 

[rxmd]

If anybody cares, for correct terminology:

Nothing is random in computer science, except noise.

When "random-ness" of random number generators is characterized, the generator's periodicity is used. The longer, the more "random". Every random number generator is periodical.

I guess that's true for computational random number generation, but if you really need aperiodical random numbers you can always get them from some stochastic physical process (such as from noise in electronic components or from radioactive decay).

Not that this has much to do with camera sensors, anyway 🙂

 

[willie_901]

Even noise has structure. There is white noise, red noise, etc. Modeling noise is important when you have data with pathetic signal-to-noise ratios. Radio astronomers are very interested in modeling noise.

It is difficult to find physical phenomenon that is truely random. Radioactive decay may be one example.

On the other hand there are a lot a ways to generate/observe pseudo-random noise and with few exceptions the results are sufficiently random.

This stuff is related to digital photography because the photon counts are effected by errors (noise) from numerous sources.

 

[astrosecret]

Film grain is random... i noticed that the fuji algorithm wasn't random as well. Honestly they're excellent at marketing (which is all the 'random film grain-like sensor' is) but seem to rush things into production. I hope the x pro 1 is different. Though the x100 has its place in my heart, Fuji is all show. The sensor itself, which is wonderful, is by sony. and the physical design of the camera is all fuji, that is to say they are responsible for difficult AF, poor menu layout, poor battery performance, poor manual focus, sluggish startup/read/write speeds, poor OVF/AF coupling, etc. But the size/sensor keep it up there as one of my favorite cameras despite the poor mechanics.

 

[ferider]

Not that this has much to do with camera sensors, anyway 🙂

Just trying to answer Doug's search for a word (he used "diversity"). The pixel placement in the X-Pro1 sensor has higher entropy when compared to a Bayer sensor.