Fuji X-E2 ISO test.

[pixmedic]

ive been asked about how I like the fuji, and how its low light performance is.
i knew it was good up to 1600 with zero noise reduction, but i had not had it above that yet.
so...time to find out. here it is. ISO 3200 and ISO 6400 with ZERO noise reduction or any other editing except a quick crop in LR and export. its basically SOOC so you can see how it looks. I tried to make the exposure more or less "correct" when I shot these, hence the different shutter speeds.
shot with fuji X-E2 and 18-55 f2.8-4 lens.

EXIF data is intact.
cropped ONLY and exported in LR. NO other adjustments made.
18mm, f/3.2, 1/60th, ISO 3200

18mm, f/3.2, 1/125th, ISO6400
same as above, cropped only.

 

[xenskhe]

I find the color washing out somewhat at the higher isos not surprisingly. The standard color setting is quite subdued anyways. This is maybe the first or one of the few 'enthusiast' digital cameras i've had were the color can often use a boost rather than cut. Is this NR at zero or -2? I'm using RAF nearly all the time, the raw converter being so straightforward and quick to work with.

I did try some 800 ISO b+w jpegs the other afternoon; pleased to find that with shadows and highlights at +2, the results were something like the Lumix 'dynamic b+w' which is those cameras' strongpoint imo.

 

[pixmedic]

NR at 0. wasnt much color there to begin with. would have to try on some bunnies maybe.
i use LR as the raw converter. seems to do well enough.

 

[xenskhe]

NR at 0. wasnt much color there to begin with. would have to try on some bunnies maybe.
i use LR as the raw converter. seems to do well enough.

These dont look washed out to me. In Silkypix maybe I'm applying too much chroma NR for the camera - the software interface still looks like it did 10 years ago; in terms of degrees of scale (the sliders), it's easy to apply too much NR to what is a much more modern sensor.

A new experience for me using these x cameras is that i hardly ever check the histogram or review the pictures as i take them. Either with matrix or c-w average, the exposures are good. It's a quick camera to use with manual lenses and EVF on the street. I press the control wheel to activate the focus peaking before raising the camera to my eye, quick focus and shoot, camera back onto my chest, walk away without any reviewing. Like a film camera.

 

[Gary A.]

Looks good to me ... a bit dark but that is an exposure/processing thing not a high ISO problem.

 

[jaomul]

Isn't there some talk that Fuji iso numbers equal that of higher numbers on other brands? I know it's supposed to be a "standard" but I believe the exposure differs on the Fuji ones.

I think these look good for their settings

 

[pixmedic]

Isn't there some talk that Fuji iso numbers equal that of higher numbers on other brands? I know it's supposed to be a "standard" but I believe the exposure differs on the Fuji ones.

I think these look good for their settings

I don't know if fujis ISO numbers are any different in relation to other cameras.
All I can say is if the numbers are at least close, I'm pleased with what 3200 and 6400 would clean up like and 1600 was good with no NR at all.

Sent from my SM-N900P using Tapatalk

 

[Derrel]

If you Google the string "are Fuji's ISO numbers fudged?" you will turn up a number of links, like this pretty popular one, which BTW we brought up a year or so ago on TPF: Does Fuji Cheat with its Sensors?

If one looks at the DxO lab testing results pretty carefully, the marked ISO values for many,many cameras are not actually dead-on and in lock-step with the ISO standards, but are "off" a little bit. Some people speculate that the numbers on Canon and Nikon and other brands of cameras diverge from the scientifically accurate numbers so as to make their cameras appear to perform better under review conditions in tests that will be posted to the world wide web and to the bigger blogs and review sites, as in, "Look how amazing the SonNiCan 760D-Alpha performs at ISO 6,400!" when the 6,400 setting is actually more like ISO 4,000.

On-line tests and on-line web and blog articles are the new marketing grounds for cameras these days--the old photography magazines of the past eras are no longer of much influence...people see a lot of web articles, where a 3x3 inch square with an ISO number in 14-point type under it is how they judge Camera A against Camera B, in many cases. So deliberately "fudged" ISO numbers are actually not just a Fuji thing, but more like an industry-wide thing.

 

[Derrel]

The article has been amended since I first read it!!! They have an addendum at the end now! It states:

"Update: Thanks to some awesome people like Iliah Borg, we now know the reason why Fuji RAW files appear darker. Turns out that Fuji has a special tag (0x9650) in its RAW files that highlights the necessary midpoint compensation for RAW files to interpret and make necessary changes. Below are the values for the Fuji X-T1:

ISO 200 / 0.72EV
ISO 400 / 0.72EV
ISO 800 / 0.72EV
ISO 1600 / 0.72EV
ISO 3200 / 1.38EV
ISO 6400 / 2.38EV

So keep the above in mind when looking at Fuji RAW files and comparing them to other cameras. If you are using a RAW converter from Adobe (and potentially other RAW converters), make sure to look at the above table for adjustments needed to make images appear as they should. Big thanks to Iliah Borg and the LibRaw team for discovering the Fuji tags and letting us know!"

 

[pixmedic]

The article has been amended since I first read it!!! They have an addendum at the end now! It states:

"Update: Thanks to some awesome people like Iliah Borg, we now know the reason why Fuji RAW files appear darker. Turns out that Fuji has a special tag (0x9650) in its RAW files that highlights the necessary midpoint compensation for RAW files to interpret and make necessary changes. Below are the values for the Fuji X-T1:

ISO 200 / 0.72EV
ISO 400 / 0.72EV
ISO 800 / 0.72EV
ISO 1600 / 0.72EV
ISO 3200 / 1.38EV
ISO 6400 / 2.38EV

So keep the above in mind when looking at Fuji RAW files and comparing them to other cameras. If you are using a RAW converter from Adobe (and potentially other RAW converters), make sure to look at the above table for adjustments needed to make images appear as they should. Big thanks to Iliah Borg and the LibRaw team for discovering the Fuji tags and letting us know!"

Apologies, and you can go ahead and call me dumb. ....but could you please translate that into English for me?
I'm not sure I'm understanding it right... or at all.

Are the numbers beside the ISO the difference between fuji ISO and other cameras iso's? Or is it saying fuji automatically will need that much adjustment? Or fuji is already making that adjustment? I'm totally lost.
What does it mean when I'm comparing ISO 3200 on my fuji to ISO 3200 on another camera?

Sent from my SM-N900P using Tapatalk

 

[Derrel]

The way I read it is that the tone curve is supposed to be adjusted by those values to correct the image for its basic brightness. As for comparing ISO values from one camera against another, that's what this article was mostly about, and the answer really lies in brightening the Fuji RAW files in post, in software. As you can see, the Fuji images as-metered or as-shot come out looking under-exposed, compared to the Sony images. And the Sony XT-1 seems to do better, make better pictures, when the images are brightened up in **software, later** than by exposing more in-camera. The XT-1 he says has lower chroma noise than the Sony he was testing against (keep in mind the Sony A7 II was shooting 11-bit, highly-cooked raw files back then...the real weakness of the early A7..not 14- not 12-bit raws, but 11-bit maximum!)

Some sensors do great when their dark areas are "lifted" in software, some sensor reveal a lot of chroma noise when their images are "lifted" in software: Canon 5D SR for example...plenty of color noise in the shadows, Nikon D750, astounding shadow recovery possibilities.

As I read it, the x9650 tag Fuji has embedded into the raw files would be the average mid-point gamma correction needed to get the right brightness to bring the images up to about the ISO standard that one would expect. Fuji apparently feels that its sensor does pretty well with a pretty high-degree of in-camera "lifting" of the brightness in software: under-expose in the field, brighten the raws in software.

We've entered the era now with sensors where different brands of sensors (Sony vs Canon especially) and different types (Foveon, Bayer, X-trans) that what we used to know about how to expose is no longer really true. The old expose to the right mandate, the ETTR is Gospel, that idea is now not always the best idea; we've gotten to the point now where deliberately UNDER-exposing, and then brightening the image upward, with modern software, is actually a better way to expose. And he says he sees that in the Fuji XT-1's sensor--better to under-expose, then brighten the raw up, in software, later.

 

[pixmedic]

But, does that mean if I shoot at ISO 6400 that the article is saying I need to add 2+ stops in post? That doesn't sound right

Sent from my SM-N900P using Tapatalk