kimcarsons

Indeed. The more time I spend with the X-T2 the more I've come to realize that the improvements it offers over the X-Pro1 for still image quality/performance are marginal. This goes for the autofocus performance too. Yeah, it's much faster under a very specific set of ideal circumstances, but the rest of the time it falls back to a contrast detection cycle which is exactly the same speed (and accuracy) as the X-Pro1. Honestly, I think I could only recommend the X-T2 over the other cameras if you really need to shoot video (esp 4K) and need to do it with Fuji's lenses for some reason, or you have a similarly strong desire to shoot long/fast bursts. It would be nice if the X-Pro1 had a finer grained grid of contrast AF points like the X-Pro2 and X-T2 do though (possible via firmware update but unlikely).

I find the most usable mode to be zone AF with the smallest zone shape. It's much quicker to move a box that size around. Unfortunately, unlike what the marketing material would have us believe, the X-Pro2 and X-T2 still seem to prefer to focus on the background if it's anywhere inside the focus box/point (except, of course, when you would like it to [e.g. focusing on an nearer arm/shoulder instead of a face when both are inside the box]).

Yeah. It was a big disappointment for me when I first experienced it on the X-Pro2. I assumed it would be force-sensitive like the Thinkpad trackpoint/nipple. If you're limited to individual clicks or a slow repeat rate, then the joystick isn't really any better/faster (except that you're not overloading functions on the D-PAD buttons). I use it, but I don't find it to be any better than e.g. using the D-PAD on the X-Pro1.

ISO 100 is an expanded ISO. It's just like taking an image at ISO 200 (over exposed by 1EV), then adjusting the exposure by -1EV in your RAW processing software (only the camera does it for you)---meaning you'll lower the noise floor, but you'll also probably have some blown out/unrecoverable highlights. Does that answer the question?

Are you sure you really need the wide aperture? The X-T2 video looks pretty good at ISO 6400. Also, an LED video light is pretty handy and not too big to pack. I have this one, it's pretty good and compact if you don't use the high capacity batteries: https://www.amazon.com/YONGNUO-YN300-Air-Temperature-3200K-5500K/dp/B0157PE1ZC/ref=sr_1_1?ie=UTF8&qid=1480700196&sr=8-1&keywords=yongnuo+yn300+air

The best lenses for video on Fuji are the 18-55 and the 18-135 (both are silent focusing). Primes are pretty useless for video because of Fuji's lack of IBIS---you'd need a big shoulder rig, heavy tripod, steadicam or gimbal to get a stable shot. All of the zooms except for the 16-55 have OIS, which makes handheld shooting possible.

Yeah. Use an external microphone or even separate audio recorder. Anyone serious about sound quality is going to do that anyway. That being said, many of the Fuji lenses focus very quietly, especially those with the LM designation.

UPDATE: These problems still exist with FW 1.10.

UPDATE: This problem still exists with FW 1.10.

UPDATE: This problem still exists with FW 1.10.

Didn't fix any of the bugs that bug me though...

The Fujinon 35mm F2 basically has no optical distortion correction, leaving it with extreme barrel distortion. I suggest, rather than take my word for it, you do the experiment yourself. If you process the RAW image software distortion correction disabled in your RAW processing software (or use software which doesn't have the a profile for this lens [e.g. rawtherapee]), you'll see it on pretty much any subject, even chaotic/organic scenes like grass (the corners are also very soft on this lens), but especially any scene with straight lines in it. Fuji cameras do not allow one to disable the software correction in the in-camera JPEG processing. The main consequences of this kind of lens/camera design are that: 1) The field of view of the lens (optically) is greater greater than the field of view of the digitally corrected output. 2) The actual resolution of the final image is lower than it would be with an optically corrected lens, because the image is being digitally stretched (and those soft corners discarded.) 3) Pixels are not sqaure/rectangular or of uniform dimensions across the frame (you can get some weird artifacts in the corners from the stretching of the correction.) 4) You could never use such a lens on a film camera (this is one of the ways mirrorless manufacturers get away with this kind of design without much public outcry.) And of course, 5) The lenses are smaller, lighter, and cheaper to produce.

I'll take your word for it. My basis of comparison is a 50" 4K (UHD) screen, looking at center frame crops with my face about 1.5' from the screen (same subject, framing, lens). Yeah, I can see a very subtle difference in sharpness in this scenario. But if I was comparing two images of different subjects, I couldn't tell you which came from the X-Pro1 and which came from the X-T2. They're extremely close in IQ. If we were talking 16MP vs 36MP, it would probably be a different story... Even with the same subject, fit to screen (i.e. zoomed out), I can't tell the difference (4K is only 8MP after all)...

My personal theory? It's Fuji's marketing department's way of ensuring that something you need/want is missing from each model, so that you end up buying both.

You're unlikely to notice a significant difference between 16 and 24 megapixels with any of Fuji's lenses (zooming in to 100%). And you're certain not to notice any difference at normal display and print sizes. This isn't to say that the lenses aren't sharp enough for 24MP---they are. It's just that the 8MP difference is not very significant. The new cameras are faster overall and have better autofocus performance and much better video quality. But for stills, and especially for slower paced shooting, rest assured you're not missing much by getting one of the 16MP bodies. I wouldn't pay the full retail price for the X-T1, though.

I agree. I was happy with the IQ and resolution of the X-Pro1. A 16MP X-Trans III sensor would have had better low light performance (perhaps 1/2 to 1 stop), less rolling shutter, faster burst rate, faster EVF, higher video frame rates, no crop in 4K video mode, etc. That extra 8MP really isn't doing anything for me. It's also annoying that there's no 16MP JPEG mode, (M is 12MP).

My point though is that there is lots of in-camera enhancement for JPEG, especially with Fujifilm. That is, there is no "natural" color profile, just the film simulations (presumably the difference between blue and purple could matter quite a bit in a criminal investigation). Some Fuji lenses (and other bands of mirrorless camera lenses) are not optically corrected for distortion, so the in camera processing is doing some heavy lifting there (again, one would imagine that in a forensic setting a ruler is being included in the frame and perhaps physical measurements are being made from the image, which distortion/distortion correction may invalidate). I don't think there's ever been a Nikon (or probably any DSLR) lens with as much optical distortion as the Fujifilm 35mm F2. So this question may be one hasn't had its day in court yet, so to speak.

Out of curiosity, how does your employer feel about RAW vs JPEG and especially the requirement of heavy software distortion correction for lenses like the 35mm F2?

That seems like an awful lot of gear. If you're doing all tripod work, why do you need the fast primes?

Well that certainly looks better than anything I ever got out of the X-Pro1 above ISO 1600... However, I think it's largely because of the lighting and the subject. I will say that this image does exhibit the main problem with X-Trans (esp high ISO) and that is the much reduced chroma detail. Notice that her teeth and the whites of her eyes are the same color as her skin. I sincerely doubt the real scene shared that quality. If this were white-balanced to correct for the color cast of the light, it would be more obvious. Children make a poor example for the 'waxy-skin' issue anyway because they have perfect, smooth skin naturally. It's on older people with splotchy skin, blemishes, burst capillaries, bloodshot eyes, etc. that it becomes noticeable (i.e. that those features are missing from the images.) It doesn't have to be skin either. It just has to be something with fine (particularly red or blue) color detail against a different ground color. You'll never get as much color differentiation/detail from X-Trans as a Bayer sensor of the same resolution. It's just a fact: The red/blue photosites simply aren't there in the same proportion and frequency. And it's also a fact that this applies equally to all generations of X-Trans.

That would cost more and be harder to use than a Pentax or Olympus with pixel-shift built-in, so what would be the point?

I'm not missing it, in fact I pointed it out: the NR setting only affects the luminance noise reduction, it doesn't affect the chrominance NR. I have the X-Pro1 (X-Trans I) X70 (X-Trans II) and X-T2 (X Trans III) so I do have some basis for comparison. Personally, I find the X-Pro1's output to be rather unlovely above ISO 1600, but then I'm not going for the gritty B&W street look either. Admittedly, I shoot RAW + JPEG and rarely like the JPEG output from any of the cameras (NR-wise) for high ISO shots better than what I can get from processing the RAW files (I can preserve more detail processing them myself.) I always look at the JPEG as a baseline, though. I'm afraid that all of us spouting our subjective opinions on the matter isn't doing much to clear things up. Has anyone competent done a direct comparison between the generations of X-Trans that shows the 'waxy skin tones' issue and how X-Trans I is immune from it?

Is this any different than using align_image_stack on an image series and taking the mean?

Perhaps you should re-read the OP.

You have to take the DP review comparison tool with several grains of salt. For one thing, the JPEG images are not camera JPEGs, but JPEGs generated by ACR. The subject of this thread is SOOC JPEGs...