8
\$\begingroup\$

I were using Fujifilm x-t20 and I took two photos: one with 'Auto' switch on and the other with 'Auto' switch off.

I compare the two and I cannot find noticeable difference.

automatically exposed image

and the settings

screenshot of auto settings

Here is in manual mode

manually exposed image

and the settings

screenshot of manual exposure settings

As a beginner I am not very confident in my choices of the settings. That's why I tried to compare what the camera would pick in auto mode.

My question are:

  1. To me the settings are quite different: ISO 200 vs 400 and 1/400 vs 1/60, but the outcomes look almost identical. Do these values (ISO and Shutter Speed), at least in a well-lit situation, have a large range of tolerance? If so, what is the range?

  2. Which set of settings are better objectively?

\$\endgroup\$
4
  • \$\begingroup\$ Did you use change all settings manually or leave any in auto? Did you use the camera meter or an external meter? What metering mode did you use? What was the rationale for your setting choices? \$\endgroup\$
    – xiota
    Commented Oct 22, 2019 at 21:30
  • \$\begingroup\$ First you might answer yourself why you picked those settings. \$\endgroup\$
    – Gerhardh
    Commented Oct 23, 2019 at 10:21
  • 1
    \$\begingroup\$ Something to keep in mind here--modern cameras tend to be quite smart about such choices, the main reason to override the computer's choices is because you know something it doesn't. Examples: There's snow everywhere, I had to dial in two stops of correction to get decent shots. I cranked the aperture down all the way because I wanted a starburst around the bright light. My objective is that branch, not the sky behind it. The target was too small for even spot metering, I took my reading off the rest of the plant. \$\endgroup\$ Commented Oct 23, 2019 at 11:56
  • 2
    \$\begingroup\$ @LorenPechtel Yes. I'd add: When you know something or when you have an opinion about a particularly slow or fast shutter speed or depth of field. \$\endgroup\$
    – mattdm
    Commented Oct 23, 2019 at 12:59

2 Answers 2

22
\$\begingroup\$

It's hard to really tell from the small versions here — which is a lesson in itself, because at 1280x850, which is a perfectly fine online viewing size, the differences really don't matter that much.

However, in this case, I think Auto probably did make some better choices.

Shutter Speed

You picked ¹⁄₆₀th of a second. This is fine, but probably slightly susceptible to blur from camera motion shake or subject movement due to wind. Auto mode picked ¹⁄₄₀₀th, which will better freeze things.

ISO

You picked 200 and the camera picked 400. Like most modern largish-sensor cameras, there's not much difference between 200, 400, or 800 on this camera in terms of noise, but it's possible the auto mode enabled the expanded dynamic range option (where the camera actually shoots at a lower ISO and then brings up the shadows in internal post-processing). Because of the way that works, this is only available at higher ISOs than the base 200. (The difference isn't super-visible in the result, but I can imagine a little more detail in the shadows — but that might also be because the auto image is a little brighter… see below.)

Aperture

This is really the big one. You picked f/18, and the camera picked f/8. All camera-lens systems are subject to diffraction, which means that while the narrower aperture gives you more depth of field, there's less sharpness overall. See What is a "diffraction limit"? for more. In general, it's better to not go beyond f/8 except in special circumstances.

Overall Exposure Value

Although it's subtle, the auto-exposure image is a little bit brighter overall. You can particularly see that in the sky. This seems to be less than a full "stop" of difference — each stop is a doubling or halving (See What is one "stop"?), and because of the way human perception works, generally differences of less than one stop aren't a big deal (and as a rule of thumb, plus or minus a stop from a middle value are reasonable valid exposure choices without really feeling over- or under-exposed).

So… between f/8 and f/18, there are 2⅓ stops — going in that direction, darker, so we'll call it negative 2⅓. (See Why are f-stops not linear? for how this is calculated.)

And between ¹⁄₄₀₀th and ¹⁄₆₀th is just about the same: 2⅔ doublings (which is 2⅔) — but in the direction of more exposure.

And finally, 200 to 400 is a full stop — leaving a ⅔-stop discrepancy, which seems close enough to be in the ballpark of the visual brightness difference in the result. As Michael C notes, the difference in metering is explained by the slight (but significant) change in framing, with more sky in one shot than the other.

(See What is the relationship between ISO, aperture, and shutter speed? for more on this "stop math").

\$\endgroup\$
7
  • \$\begingroup\$ I haven't shot with any of the Fuji "X" series, but I would imagine the ISO shown in the EXIF info reflects that actual analog amplification used rather than the "effective ISO" after digital processing. But the ISO 400 shot is not "brighter" than it should be compared to the ISO 200 shot, it is "darker" than one would expect for an image exposed 2/3 stops brighter than the ISO 200 image. \$\endgroup\$
    – Michael C
    Commented Oct 22, 2019 at 13:12
  • \$\begingroup\$ After opening both images in separate tabs and switching back and forth, I think the ISO 400 image is about 2/3 stop brighter than the ISO 200 images. Look at the sand in the shadow in the center foreground. \$\endgroup\$
    – Michael C
    Commented Oct 22, 2019 at 13:15
  • \$\begingroup\$ The difference in metering is fully explainable by the difference in framing. There's more sky in the shot exposed 2/3 stop darker (which really makes a difference with corner vignetting in the upper left corner), and less sky in the shot exposed 2/3 stops brighter. \$\endgroup\$
    – Michael C
    Commented Oct 22, 2019 at 13:26
  • \$\begingroup\$ @MichaelC Yep, I'll add a note on that too. \$\endgroup\$
    – mattdm
    Commented Oct 22, 2019 at 13:32
  • 1
    \$\begingroup\$ For those worried about noise when the camera shoots at a lower ISO then pushes the shadows - don't be... The sensors on Fuji X cameras are really close to iso-invariant, which is what lets them implement this. (The "best" iso-invariance is actually on the older X-T1, with the newer X-T2 and X-T3 only slightly behind.) \$\endgroup\$ Commented Oct 23, 2019 at 5:13
6
\$\begingroup\$

The XT-20 is an APS-C sensor size camera. f/18 is a quite narrow aperture for that that will limit the achievable sharpness due to diffraction. At the same point of time, you have a wide range of distances on the image and bringing all of them into sharp focus is a challenge even at the wide angle of 18mm (about 24mm full-frame equivalent). Basically it depends on your artistic vision (and your focus point) whether you want a differentiation in sharpness. To be honest, I find it non-obvious what the artistic intent of that photograph was, so it's tricky to subscribe where to go.

ISO200 is less susceptible to image noise, but the sensor you have here is still pretty solid at ISO400 and would still have leeway for sharpening or exposure processing. Also this is not a dark scene.

1/400 s exposure time is probably playing it a bit oversafe for 24mm EFL at least for a static scene. Assuming unwindy conditions, your choice of parameters leading to 1/60 s seems to still be pretty much ok at this focal length and with image stabilisation on, assuming you have somewhat steady hands.

So the main difference I expect to be with sharpness: f/8 is bound to be sharper in areas that are in-focus but have quite a narrower in-focus region. f/18 will have more spread out sharpness but a bit lower total sharpness. However, due to the very low noise, the diffraction unsharpness can be sharpened up gracefully without amplifying noise to distraction.

There is one additional difference between those choices that you don't see in this photograph: if there is a really bright light source (sun or strong reflection of it), the "sunburst" diffraction pattern around it will be significantly different for f/8 and f/18.

\$\endgroup\$

Not the answer you're looking for? Browse other questions tagged or ask your own question.