The 4K Fuji X-T2 is here

[gatopardo]

12 hours ago, damoke said:

Video starts at about 2 mins in, shot with ND filter.

Colors are just fantastic. I agree with him that classic Chrome is a good profile for video, with the in camera highlights and shadows tweaks he suggested . Velvias are more for stills.  Thank god he did not grade the footage. Lots of people destroy otherwise great footage because of this grade everything trend. Color and skin tone is one of the most important aspect of an image and will be the first thing a normal viewer will find as wrong, if poorly handled. Detail is far less important for a normal viewer. But XT-2 has plenty of it. 

[Lothar]

2 hours ago, gatopardo said:

Colors are just fantastic. I agree with him that classic Chrome is a good profile for video, with the in camera highlights and shadows tweaks he suggested . Velvias are more for stills.  Thank god he did not grade the footage. Lots of people destroy otherwise great footage because of this grade everything trend. Color and skin tone is one of the most important aspect of an image and will be the first thing a normal viewer will find as wrong, if poorly handled. Detail is far less important for a normal viewer. But XT-2 has plenty of it. 

Well, I agree that Classic Chrome is a profile which you just don't want to touch, because of its gorgeous colours. I made a little video of a trip lately, and I just did some WB and exposure correction (maybe grain). The problem is, that you just can't do much with it in post. There is little latitude. You wont have a lot of dynamic range. There is a big difference between Classic Chrome and Pro Neg Std regarding DR. I agree totally, that skin tone is key. There was a post lately of a video with a girl that had a yellow skin tone. You don't really watch the video, you just wonder if she has a problem with her liver and if she went to hospital. If I could have the Classic Chrome colours with the flexibility of Pro Neg Std., that would be heaven. Unfortunately, there is no Film Simulation LUT for F-Log.

 

4 hours ago, Taranis said:

There must be a difference in their measurement method, as all their results are lower than those of SamuelH:

X-T2 UHD: 21ms vs 29.5ms
A7SII UHD: 25ms vs 30.4ms
A6300 UHD: 34ms vs 39ms

At least it's consistent regarding the ranking. Which means, I don't need to dig into the mathematical basis of both tests to find out which one is off.

 

[kidzrevil]

16 hours ago, damoke said:

Video starts at about 2 mins in, shot with ND filter.

the image quality is astounding. wow ! Funny I don't see the weird frame skipping thing that were in other videos, maybe there is a setting that causes that judder ? 

[gatopardo]

1 hour ago, Lothar said:

Well, I agree that Classic Chrome is a profile which you just don't want to touch, because of its gorgeous colours. I made a little video of a trip lately, and I just did some WB and exposure correction (maybe grain). The problem is, that you just can't do much with it in post. There is little latitude. You wont have a lot of dynamic range. There is a big difference between Classic Chrome and Pro Neg Std regarding DR. I agree totally, that skin tone is key. There was a post lately of a video with a girl that had a yellow skin tone. You don't really watch the video, you just wonder if she has a problem with her liver and if she went to hospital. If I could have the Classic Chrome colours with the flexibility of Pro Neg Std., that would be heaven. Unfortunately, there is no Film Simulation LUT for F-Log.

 

At least it's consistent regarding the ranking. Which means, I don't need to dig into the mathematical basis of both tests to find out which one is off.

 

Can I ask you for 5 sec footage of Classic Chrome and Pro Neg Std in a situation like this photo to illustrate that dynamic range difference? Thanks

[Lothar]

Well, I did such a test already. I was doing such a DR test plus I wanted to stress test the tracker of resolve so I switched off the OIS. I cut out the windows and pushed the shadows of the room. Here are 2 shots of the test and the links to SOOC footage of this test.

Classic Chrome footage: https://www.dropbox.com/s/2olmvdgjtjpr9gh/ClassicChromeDR.MOV?dl=0

Pro Neg Std. footage: https://www.dropbox.com/s/clll4ofg8yjf6jf/ProNegStdDR.MOV?dl=0

 

 

 

19 minutes ago, Lothar said:

Well, I did such a test already. I was doing such a DR test plus I wanted to stress test the tracker of resolve so I switched off the OIS. I cut out the windows and pushed the shadows of the room. Here are 2 shots of the test and the links to SOOC footage of this test.

Classic Chrome footage: https://www.dropbox.com/s/2olmvdgjtjpr9gh/ClassicChromeDR.MOV?dl=0

Pro Neg Std. footage: https://www.dropbox.com/s/clll4ofg8yjf6jf/ProNegStdDR.MOV?dl=0

 

 

 

Sorry, I just have a hard time to grab a still without color correction. So, the jpeg of the Classic Chrome footage is color corrected, too. It is almost SOOC (no cut out) and you can't push it much, because there is no information in the shadow. It's just shadow. ;-)

[gatopardo]

19 hours ago, Lothar said:

Well, I did such a test already. I was doing such a DR test plus I wanted to stress test the tracker of resolve so I switched off the OIS. I cut out the windows and pushed the shadows of the room. Here are 2 shots of the test and the links to SOOC footage of this test.

Classic Chrome footage: https://www.dropbox.com/s/2olmvdgjtjpr9gh/ClassicChromeDR.MOV?dl=0

Pro Neg Std. footage: https://www.dropbox.com/s/clll4ofg8yjf6jf/ProNegStdDR.MOV?dl=0

 

 

 

Sorry, I just have a hard time to grab a still without color correction. So, the jpeg of the Classic Chrome footage is color corrected, too. It is almost SOOC (no cut out) and you can't push it much, because there is no information in the shadow. It's just shadow. ;-)

Thanks for the footage. I've checked  and the difference in latitude is no very big. The Classic Chrome colors are way more nicer IMHO though. Are both with the shadows and highlight dialed to maximize the DR as enabled by the latest firmware? 

[Lothar]

6 minutes ago, gatopardo said:

Thanks for the footage. I've checked  and the difference in latitude is no very big. The Classic Chrome colors are way more nicer IMHO though. Are both with the shadows and highlight dialed to maximize the DR as enabled by the latest firmware? 

Yes, they are. Only the pre-production models had a firmware where this feature was disabled. The production firmware 1.0 does have the possibility to set the highlight and shadow tone attributes. I'm just waiting for the firmware upgrade which enables internal F-Log. ;-)

And yes, the difference is smaller than what I remembered. And I really need to calibrate my monitor. When I grabbed the stills in resolve they came out brighter and I saw more detail. The two items I was looking into were the second pillow and the basketwork of the left flower pot. And by pushing the shadows I saw more artefacts and less detail in CC than in NS. The differences at a more usable level are smaller.

I think that some film simulations are great in some scenes whereas they don't work at all in other scenes. Shooting stills this isn't a problem, because I can choose the film simulation after the fact due to the internal raw converter. But if I shoot video, I have to know it in advance. That's why I like the idea of using a more neutral film simulation and choose the look in Resolve. I am still testing different lighting conditions and different backgrounds. For instance, I also like Provia, but if you are on a green meadow, the green is just too intense and you wish you dialed in CC before.

All in all, if you don't want to spend a lot of time in your preferred color grading tool, CC is the simulation to go with, in my opinion.

[Lothar]

Ok, I admit, you can't get IBIS via firmware upgrade. But if I had bought a Sony A6300 I would be pretty upset now. Regarding the schedule of the product announcements of A6300 and A6500 I would feel ripped off.

But thinking about possibilities for a firmware update for the X-T2, I could imagine the following

• internal F-Log
• histogram and or zebra in video mode
• Kodak 5219 film simulation (ok, maybe a Fujifilm motion picture film sim)
• continuous video AF improvements
• etc. (I am pretty sure, I forgot some other things which can be fixed through firmware evolution)

If I think about the improvements I got for my X-E2, a lot of things are possible.

[Hanriverprod]

a review mostly on its video.

Colors look amazing.

Focus is fast but jumpy.

Needs full articulating touch screen.

Needs ibis.

[Lothar]

6 hours ago, Hanriverprod said:

a review mostly on its video.

Colors look amazing.

Focus is fast but jumpy.

Needs full articulating touch screen.

Needs ibis.

I wonder, is AF really a high prioritized feature as he stated in the video? I know, you could use it for gimbal work, but other than that, I thought it's more a feature for soccer moms. I think even with DP-AF it's not always focusing in the speed you want and it still can't read your mind. The longer the take, the higher the probability, that AF will ruin it. Isn't stopping down the lens still the preferred way to capture moving objects? Because I am still skeptical regarding AF for video it wasn't a feature I was looking at, in the first place. That said, for stills it was an important feature and I am completely satisfied with its performance.

[Hanriverprod]

30 minutes ago, Lothar said:

I wonder, is AF really a high prioritized feature as he stated in the video? I know, you could use it for gimbal work, but other than that, I thought it's more a feature for soccer moms. I think even with DP-AF it's not always focusing in the speed you want and it still can't read your mind. The longer the take, the higher the probability, that AF will ruin it. Isn't stopping down the lens still the preferred way to capture moving objects? Because I am still skeptical regarding AF for video it wasn't a feature I was looking at, in the first place. That said, for stills it was an important feature and I am completely satisfied with its performance.

I used to think that way until I tried some of the new canon bodies, and now to me, autofocus seems usable in many narrative and professional settings. DPAF + touchscreen is pretty amazing and other companies need to be held to that standard.

[Lothar]

After all, I am pretty pleased with the DR of Classic Chrome. This still (grabbed from Resolve) shows a color corrected scene where I pushed the shadows and had a power window on the sky to bring down the highlights. Finally, I added some contrast and saturation to bring back some life. Please keep in mind, this is 8-bit 4.2.0.

As you see it was a cloudy day. This image is more colourful than as saw it with my eyes.

[Hanriverprod]

He tests:

Samyang 12mm f2
Fuji XF 35mm f2
Fuji XF 90mm f2
Fuji XF 18-55mm f2.8-4 OIS
Fuji XF 16mm f1.4
Fuji XF 27mm f2.8

[Hanriverprod]

 

[Hanriverprod]

He finds flog to be similar to (6300) s-log 2 in dr but less than slog 3. Also slog2 has less noise and slog3 a bit cleaner.

He finds flog to have big improvement on dr over internal profiles but also more noise and less sharp.

Slog profiles improve dr over sony's internal profiles but not as much as flog improves dr over fuji's internal profiles.

He recorded externally for x-t2 to Atomos Ninja Assassin 4K.

[TheRenaissanceMan]

2 hours ago, Hanriverprod said:

He finds flog to be similar to (6300) s-log 2 in dr but less than slog 3. Also slog2 has less noise and slog3 a bit cleaner.

He finds flog to have big improvement on dr over internal profiles but also more noise and less sharp.

I.E, less NR and less sharpening. Perfect. 

And less flat than SLOG 3 is good. That means probably none of the same banding problems.

I'm fine with slightly less DR/more noise than the A6300, as that's still solid. I'm willing to trade it for color science, ergonomics, and better native lenses.

[Lothar]

45 minutes ago, TheRenaissanceMan said:

I.E, less NR and less sharpening. Perfect. 

And less flat than SLOG 3 is good. That means probably none of the same banding problems.

I'm fine with slightly less DR/more noise than the A6300, as that's still solid. I'm willing to trade it for color science, ergonomics, and better native lenses.

Yes, I guess they bypass NR and Sharpening when going to HDMI. I hope they can do it if they offer internal F-Log. Who knows how they wired the h.264-Engine into the line. But it looks like it would make sense to hire a ninja. ;-)

[Lothar]

8 hours ago, Lothar said:

Yes, I guess they bypass NR and Sharpening when going to HDMI. I hope they can do it if they offer internal F-Log. Who knows how they wired the h.264-Engine into the line. But it looks like it would make sense to hire a ninja. ;-)

Getting up early and posting while still sleeping is not good. Of course, they don't bypass NR if cinema5D recognized NR at higher ISOs, but I agree that it's fine if the image is a bit noisier. The cinema5D review and this video makes me thinking about getting a Ninja, although it'll make the combo bigger and less portable.

[TheRenaissanceMan]

The relevant question is whether there's less NR and sharpening over the HDMI, or when using F-LOG (like Canon turns off sharpening in C-LOG). If it's the latter, we could potentially get that more natural image from the internal recording. It'd still be nice to record externally for the beefier file to grade, but having the option to use it stealthily would come in handy.

[Hanriverprod]

Panasonic needs to catch up.