X-T2 is out, (offically) Who is interested

[David Sutton]

Purchased an X-T2 last week and  just had a play with the write settings.
With the raw file set to lossless compression, boost mode on (no battery grip - I like the smaller form factor), continuous shooting at 4 frames a second and with a class two card I got bored holding down the shutter after 150 shots. It may have eventually filled the buffer, but it's a bit academic. At three frames/sec I think you would fill the card. That's good enough for what I have in mind.
At present I don't have a raw converter that can read the new files. I've already paid for the new One1 Raw which comes out next month and don't really want to pay for a PhotoNinga upgrade yet, so I may try RawTherapee, which has had good reports once you get over the learning curve. But that applies to them all really.
David

[skirkp]

@Ken Bennett  -- no, not using JPEGs.  I develop from raw files in Capture One (can't use the compressed raw format, but that's OK.  I compress a day's shooting as a single directory once they are processed.)  I did try the 100, 200, 400 DR settings when I was shooting jpegs on the X-Pro2 and didn't see much value there.  Since I don't save JPEGs, I only see the small jpeg that shows in the viewfinder and back screen, and once or twice I have seen the dots there as well -- only in very low light situations.

scott

[Alan Smallbone]

Scott that is odd. Was the image way underexposed and lightened in processing? I saw something similar with an Fuji raw but that was over 4 years ago and have not seen it since. It was a bug in the Capture One raw converter for the X-Pro1 but that was resolved a long time ago. Have you tried a different raw converter? Even a trial version should give you an answer if it is the raw conversion. Many raw converter will read the dynamic range setting in the exif and process it accordingly, which is one reason I always leave it at 100%. Much like the lens profiles are embedded in the EXIF and applied, with LR you cannot disable that, but you can with C1.

Alan

[skirkp]

No, I have gotten the sparkly dots in both outdoor bright settings and indoor mixed lighting, always with a fair range of light levels in the picture.  I'm just shooting the raw frame, not keeping a jpeg or setting parameters like DR.  Highlight tone is -1, shadow tone is +1, to get a little more range out of the viewfinder, which I find rather contrasty (I am used to X-Pro2 and Leica M finders).  I have opened a repair ticket with Fuji.

scott

[rdonson]

Scott,  just curious which metering mode are you using when you get these results?

[flaca]

X-T2 Sharpening again.

I've been using Pete Bridgewood's settings with by X-E2, but I just got an X-T2 and after a bit of fiddling I settled on a different sharpening approach, for one particular photo anyway. The image was a deer, not quite sharp (100-400 with TC handheld...) so it seemed like I needed to push the sharpening to give the apperance of a sharp image if that makes sense. Deer fills centre 2/3 of the frame after cropping a bit.

The following settings look oversharpened when zoomed in, pretty nasty with all sorts of worms etc etc. But when viewed full screen on a 27" monitor these settings give a considerably better result than Pete Bridewood's, Martin Evening's or Ron's in post 167. To my eyes the fur and eye looks sharp(er) with more contrast, without obvious artifacts. I'll print it so presumably the oversharpening will be even less relevant, and I guess (hope!) will produce a better print.

Amount  - 29 - higher than this quickly got nastier
Radius   - 2.4
Detail    - 100 - moving down to 80 made little difference
Masking - 0 - I initially had it around 70 but it didn't seem to make much difference

Other NR settings at 0

After writing all that, I noticed that main difference here seems to be Radius. The fur seems to benefit from a much higher radius than usual. Or maybe because the image wasn't quite sharp?

I've no idea whether this works with other images, e.g. ones that are properly in focus to start with (!), or other non-furry subject matter. And others might have better eyes than me of course. Views and suggestions welcome. Happy to send the image if anyone wants a play.

Cheers,
Scott

[skirkp]

Scott,  just curious which metering mode are you using when you get these results?

Shooting in low light, I got sparkles in the viewfinder and on the image when using center-weighted.  I switched to spot metering for a bit, and didn't have them.  Same thing outside, switching exposure mode from one to another plausible setting ended a session with some sparkles.  It may be that this happens when I have just switched on, but it can stay that way for several shots.  For the most part, I use center-weighted, not multipoint or overall average.

scott

[rdonson]

Scott, thanks for sharing your experiences.

After my post on sharpening settings I've now processed a few hundred more photos. My approach now is to use those settings as a default applied during import into LR. In many cases they are sufficient but in others there is more work required. My observation now is that sharpening is very dependent on the subject and the lighting.  For final looks I resort to 2:1 viewing to judge the sharpening. It's easy for me  to overdo sharpening with the higher resolution of the X-T2.  I find this is especially true if adding various forms of contrast enhancements.

Someday perhaps Adobe will figure out how to actually demosais and sharpen X-Trans files to everyone's satisfaction.

[rdonson]

Shooting in low light, I got sparkles in the viewfinder and on the image when using center-weighted.  I switched to spot metering for a bit, and didn't have them.  Same thing outside, switching exposure mode from one to another plausible setting ended a session with some sparkles.  It may be that this happens when I have just switched on, but it can stay that way for several shots.  For the most part, I use center-weighted, not multipoint or overall average.

scott

This sounds like it might be a bug and something that might be addressed in the firmware release from Fuji that was recently delayed. Just a guess though.

[flaca]

Scott, thanks for sharing your experiences.

After my post on sharpening settings I've now processed a few hundred more photos. My approach now is to use those settings as a default applied during import into LR. In many cases they are sufficient but in others there is more work required. My observation now is that sharpening is very dependent on the subject and the lighting.  For final looks I resort to 2:1 viewing to judge the sharpening. It's easy for me  to overdo sharpening with the higher resolution of the X-T2.  I find this is especially true if adding various forms of contrast enhancements.

Someday perhaps Adobe will figure out how to actually demosais and sharpen X-Trans files to everyone's satisfaction.

Hi Ron,

That sounds good, I''ll probably start with your settings for most images. Thanks for doing the legwork there.

Just out of interest, I had some time to play around this afternoon (very rare). Various people have said that either Photoshop Smart Sharpen (I think) or Nik Sharpener give better results than Lightroom. I tried both on the deer shot I mentioned. Photoshop did indeed give much better results than any of the Lightroom settings I tried, with no worms in sight. Zoomed in to e.g. the deer's face, with a lot of photoshop sharpening applied, it looked like it had only sharpened the contrasty bits of the image and even binned some less contrasty detail - but with no artifacts other than looking a little cartoonish when zoomed close. When viewed at Fit, it looked much sharper and very natural. Very pleasingly, Photoshop didn't sharpen the grassy background too much either, whereas Lightroom did (I know local sharpening would sort that). Whilst I know Lightroom pretty well, I''m a Photoshop novice so the ease with which I obtained the results was surprising. Better to keep everything in Lightroom of course, I am lazy when it comes to workflow, but I can see myself doing this for problem images.

Nik Pre Sharpener didn't produce anywhere near as good results, but that might be my inexperience with the tool or maybe I need to use Output Sharpener as well - which starts getting a bit fussy for me.

Cheers,
Scott

[armand]

Played with more files and I have a couple of observations.

- for 24 MP you need a faster shutter speed; sounds obvious but it's one thing to know it and another when you see it in the field; granted, is not as bad as my Nikon D750 where I have serious issues getting a sharp picture at 1/FL but worse than the X-T1

- the second thing I just realized today when I shot more fall foliage is the Adobe screwed up, again; the profiles are nowhere as close as they were for X-T1. The big difference is for Velvia, on red leaves where the adobe adds a lot of magenta

- I'm not that impressed with the high ISO quality, I don't think is really better than X-T1; at least on many files

[rdonson]

Armand,  thanks for sharing your observations. 

I zoned out on shutter speed needing to be faster with the X-T2.  It explains a LOT for my hand held work and I will now adjust my Auto ISO settings accordingly.  I was wondering why my 18-135 often produced good results but not as high a percentage as I was getting from my X-T1.

A lot of my recent work has been on tripod.  The X-T2 with the 60mm Macro can produce some very nice results as can the 10-24.  I haven't shot much with the 100-400 and the X-T2 though.

With regards to high ISO I've haven't gone above 12,800 with the X-T2.  It's not a stunning improvement over the X-T1 but in many cases for me it is a bit better.  In some of my work the 12,800 from the X-T2 looks like 6,400 with the X-T1.  That is a generalization based on limited a limited number of images and not a scientific test. 

I was in the Blue Ridge and shot a lot of landscape and foliage and processed in Provia rather than Velvia.  The work I've shot that looked good in Velvia was always quite small.  I typically process Provia in Lr and then if I need to work on the color it's off to PS or use Topaz Clarity.

What's up for me when time permits is to see how well Topaz Detail 3 and Topaz Denoise 6 work with the X-T2 files.

[skirkp]

This sounds like it might be a bug and something that might be addressed in the firmware release from Fuji that was recently delayed. Just a guess though.

Feels that way to me, too.  I had sparkle problems in about four of 260 shots in weak light last night at a wedding with lots of dark suits, always just after switching the camera on.  This camera is too good to send back and lose for a month or so, so I am hoping for a firmware fix.

Scott K

PS -- per a suggestion on another forum -- I had set the dust-shaker to be ON at startup as well as ON at shutdown.  I set it OFF for both and made a few test shots just after startup in high DR situations without seeing the noise sparkle problem.  So I'm back to the default, which is dust removal only at shutdown, an d haven't seen this problem recently.  Is anyone else a belt-and-suspenders type with dust removal set on both ends of a session, and if so, do you see my problem?  That might be the answer. 

[Alan Smallbone]

Played with more files and I have a couple of observations.

- for 24 MP you need a faster shutter speed; sounds obvious but it's one thing to know it and another when you see it in the field; granted, is not as bad as my Nikon D750 where I have serious issues getting a sharp picture at 1/FL but worse than the X-T1
.......

Boy I sure rediscovered this last week shooting some cars in a museum with wide open apertures..... The xt2 is a nice improvement with a nice wake up call to improve one's technique.

Alan

[MBehrens]

Can you all elaborate on the faster shutter speeds required by 24MP sensors. Though you all seem to agree this is an obvious need, this is my first 24MP camera and am learning the ins and outs.

I have had the X-T1 for a couple of years and have absolutely loved the results from it and have no problem getting incredibly sharp images. I seldom have to visit the details panel. My first forays with the X-T2 were disappointing, after all of the hype my images all looked much softer than the T1. I took to the garage to set up a controlled shoot between the 2 cameras and found that on a tripod the they are about equal. Wide open the 35mm 1.4 with the 24MP is definitely softer which I attribute to the smaller photocells challenging the lenses ability to resolve the image. Once stopped down a bit things cleared markedly.

I'd really like to get dialed in on this and any input is appreciated.

Thanks,
Morey

[armand]

...
Wide open the 35mm 1.4 with the 24MP is definitely softer which I attribute to the smaller photocells challenging the lenses ability to resolve the image. Once stopped down a bit things cleared markedly.

...

That shouldn't be. Maybe at pixel level but resized at the same size the one from X-T2 should be sharper, even if not by much.

Can you all elaborate on the faster shutter speeds required by 24MP sensors. Though you all seem to agree this is an obvious need, this is my first 24MP camera and am learning the ins and outs.

The X-T1 has a very well behaved shutter and with the 16MP you can get away with a shutter speed of 1/focal length (with non stabilized lenses) or even less and get sharp pictures.
When you increase the MP and add a different shutter which goes to 1/8000 vs 1/4000 in X-T1 that 1/FL is the minimum shutter speed and often you will need faster shutter speeds.
What looked sharp in X-T1 at 1/FL might require 1/1.2-1.5FL in X-T2. Add a mirror as in my Nikon D-750 and 24MP and I will need 1/1.5-2FL to get a reasonably sharp picture. Keep in mind this is at the pixel level, once you downsize most of the difference should go away.

[MBehrens]

Are you saying that the X-T2 has less well behaved shutter? I doubt that very much. And there is no mirror, I understand mirror flap and using lock-up to eliminate it. I'm curious as to what causes this syndrome. Is this a result of smaller and smaller photosites on the sensor? Everything around the sensor is just a little bit, maybe a lot, better. The sensor is the same size. The only thing that is different is the smaller photosites. If we get a 36MP sensor in the X-T3 then we may have to use 1/2FL? I was very happy with 16MP, not sure I'm liking the direction higher resolution is taking me.

[armand]

I don't know for sure but considering it goes to 1/8000 vs 1/4000 it might be less dampened.
That exposure rule applies if you want all 24MP; if you downsize back to 16MP the shutter rule should be the same as X-T1 (if the shutter by itself is not that different).

[Alan Smallbone]

Are you saying that the X-T2 has less well behaved shutter? I doubt that very much. And there is no mirror, I understand mirror flap and using lock-up to eliminate it. I'm curious as to what causes this syndrome. Is this a result of smaller and smaller photosites on the sensor? Everything around the sensor is just a little bit, maybe a lot, better. The sensor is the same size. The only thing that is different is the smaller photosites. If we get a 36MP sensor in the X-T3 then we may have to use 1/2FL? I was very happy with 16MP, not sure I'm liking the direction higher resolution is taking me.

I think the shutter is well behaved. What I think is there is more resolution and basically just need to have better technique. I think the greater detail and the great optics, it just shows up more. I don't see it as a problem, but a reminder to be more careful and use better technique. I really like the XT2 and also the XT1 and all of my previous Fuji cameras. They are a joy to use, make photography a lot of fun and make it very intuitive to take pictures. I think when the interface is less intrusive the image making is more fun and easier. The tool is no longer the problem, the operator is in full control.... 

Alan

[Ken Bennett]

I think the shutter is well behaved. What I think is there is more resolution and basically just need to have better technique. I think the greater detail and the great optics, it just shows up more.

I had the same issue when I got the Canon 5Ds -- shutter speeds that were no problem hand held with the 5D3 suddenly gave me all kinds of obvious camera movement blur. I solved that by treating the 5Ds as my "medium format" system, I shoot it locked down on a tripod for landscapes, etc., or in the studio when I need the resolution.

I have honestly not noticed as significant an issue with moving from the XT1 and X Pro 1 to the XT2 and X Pro 2 sensors.