1
\$\begingroup\$

I took two photos, one with flash exposure compensation set to -2EV:

enter image description here

and the other with +2EV:

enter image description here

I find a minor, almost unnoticeable difference between these two photos. But there's a difference of 4EV in the flash exposure compensation, which is a huge difference. I expected the foreground to be significantly brighter with higher flash exposure compensation. Why is that not the case?

Compare these with a photo without flash:

enter image description here

There's a significant difference between the photos with and without flash, but negligible difference between flash exposure compensation dialed all the way down and up. It's almost as if flash exposure compensation is ignored by the camera.

This is on the Sony NEX-5R, with the bundled flash attachment, with the same flash mode (slow sync), exposure compensation (0EV), aperture (F3.2), ISO (200) and other settings. These photos were shot in aperture priority mode.

\$\endgroup\$
6
  • \$\begingroup\$ What were your shutter speeds for the 2 shots? \$\endgroup\$ Jun 15, 2014 at 5:41
  • \$\begingroup\$ 5 seconds in all cases. \$\endgroup\$ Jun 15, 2014 at 5:42
  • \$\begingroup\$ Did you definitely have the flash in TTL mode? \$\endgroup\$ Jun 15, 2014 at 5:50
  • \$\begingroup\$ Sorry, I don't understand what TTL is, despite reading on Wikipedia and several questions on photo.SE. In case it matters, I'm in manual focus mode (and slow sync flash mode, as I mentioned in the question). The NEX-5R doesn't have a menu option for TTL as far as I can see. Am I correct in understanding that as I increase the flash compensation, the foreground should get brighter, but the overall picture should have the same exposure level (since I'm not modifying the exposure compensation)? \$\endgroup\$ Jun 15, 2014 at 6:05
  • 1
    \$\begingroup\$ I'm not talking about the overall exposure, which is fine. I'm talking just about the effect of the flash. \$\endgroup\$ Jun 15, 2014 at 8:58

4 Answers 4

2
\$\begingroup\$

The brightness of the light from your flash falls off with the square of the distance to the subject. For example, if you are using only flash for illumination and something 10 feet away is exposed properly, then something 20 feet away will be 2 f-stops (4x) darker. Something only 100 feet away will be 100 times (almost 7 f-stops) darker.

Conversely, the power of a flash to properly illuminate objects goes up with the square of the distance to them. Camera flashes will have limited power due to needing to be portable, running off of batteries, and not costing a fortune. As a result, camera flashes are only useful for nearby objects, ususally up to a few 10s of feet. For example, using the flash in your point and shoot camera from the stands of a sports stadium to take a picture of something on the field is totally silly. Just because people do it doesn't mean it's a good idea.

Your case is like the sports stadium. Most everything in the picture is so far away that the flash has no effect on it. Regardless of the flash brightness compensation you used, all the light for these pictures was ambient. The flash had nothing to do with any of them, which is why changing its brightness a few f-stops either way didn't matter.

You can see that the flash had no effect by looking at the foreground at the bottom of the picture. If the flash had any visible effect, you'd find it there. The foreground would look brighter than distant objects. This is not the case, so the light from the flash was already so dim as to not matter at the distance of the foreground.

The reason the flash pictures look different from the last picture taken without a flash is how your camera handles color balance. Flash is "white" light, fairly close to sunlight in color. However, most artificial lights are considerably more yellow and orange. When using a flash, the camera assumed the scene would be lit by the flash and applied color compenstation accordingly. However, your scene was in reality lit by yellow artificial light, so you can see the obvious yellow cast. For the bottom picture, your camera apparently has a way to measure the ambient light color, and it adjusted the color balance accordingly. As a result, the colors look more correct relative to how your eyes perceived them at the time. Your eyes and brain have rather sophisticated color compenstation built in, so what you perceived was already adjusted to the ambient light color.

This also points out another problem, which is that you are not working with the raw data in post processing. No compenstation is applied to the raw data, so all the pictures would have the same color cast in raw. Cameras often put additional information, like the measured ambient color, into raw files, which some software may use as the default color compensation. If you are using raw, then perhaps you are seeing the different color balance due to the software defaults. In that case you can go and set them all to something of your chosing. One way to do this is to find something in the picture you know should be white, then tell the software to set the color balance so that it comes out white (or some shade of neutral gray). The building at the right would be something to start with, for example.

Added:

To prove the points above, here are your second (+2 EV flash) and third (ambient) pictures post-processed to the same color balance and exposure. I used a patch of flat wall on one of the buildings at right as the reference for white and as the common exposure point for both pictures.

Now you can actually see a little effect from the flash in the first picture. This was previously swamped by the dark exposure and the yellow color cast. There are two important things to notice:

  1. The effect of the flash only extends a small distance into the scene. It is pretty much gone after 10 railroad ties into the picture. Since most of the scene is at much further distance, the flash has no effect on the other objects.

    The square law is a powerful effect you can't escape. Even if your flash was 4 times more powerful, its effect would only extend twice as far into the scene, which still wouldn't make any difference for anything but a little further along the railroad track.

  2. The flash light looks bluish. This is because I compenstated both pictures for the ambient light, which has a yellow cast. Since yellow is corrected to look white, white will look like the inverse of yellow, which is blue.

\$\endgroup\$
4
  • \$\begingroup\$ Thanks; accepted. Nit: This answer mixes up the answers to two separate questions I'd asked (this one, about exposure, and another about white balance). Ideally, we should keep the answers separate, but in any case, thanks for the answer and for addressing all the factors. I will keep these points in mind. The flash was indeed not powerful enough. I was able to work around that by using a shorter shutter speed (by increasing the ISO or opening the aperture fully). Then the foreground is lit brighter by the flash. And, yes, Lightroom honors the white balance value stored in the RAW metadata. \$\endgroup\$ Jun 15, 2014 at 16:05
  • \$\begingroup\$ Is the following reasoning correct for why opening the aperture or increasing the ISO increases the effect of the flash: in either case, the camera captures more of the (limited) light of the flash. Regarding why the white balance is off, in an ideal world, the camera would examine the photo immediately after taking it, realize that the white balance is off, and then correct it. After all, it's just setting a field in the RAW header, and does not involve taking another photo or updating all the pixels in the photo. Do you have any suggestion regarding my comment to Matt Grum's answer? \$\endgroup\$ Jun 15, 2014 at 16:20
  • 1
    \$\begingroup\$ @Kartick: I answered about color balance here because it seemed relevant and I didn't know about your other question at the time. Higher ISO or larger aperture captures more of the flash light. However, it also captures more of the ambient. The exposure time is how flash versus ambient are balanced. If the exposure time is the X-sync speed of your camera then, it will still capture the same flash light, but much less of the ambient. You will mostly see a dark picture with only near objects illuminated by the flash. Try it. \$\endgroup\$ Jun 15, 2014 at 23:46
  • \$\begingroup\$ No problem, I appreciate your answers. \$\endgroup\$ Jun 16, 2014 at 1:14
2
\$\begingroup\$

There's 2 things to consider.

Firstly the overall contribution of flash compared to ambient light. If the flash is putting out a small amount of light compared to the total ambient light over the course of the exposure then the flash itself and any changes to the flash will go unnoticed.

Secondly flash exposure compensation only biases the flash metering it doesn't soley determine power output, further more if -2EV FEC tells the camera to put out 10 units of light, and +2 FEC tells the camera to put out 160 units of light, and the flash is only capable of putting out 8 units of light, then 8 units of light is what you will get, regardless of the FEC setting.

\$\endgroup\$
4
  • \$\begingroup\$ If one burst of the flash is not enough, is there a way for me to tell the camera to repeatedly fire the flash during the long exposure? \$\endgroup\$ Jun 15, 2014 at 10:50
  • 1
    \$\begingroup\$ @KartickVaddadi Probably not, it's a fairly limited use case, get yourself an external flash and you can manually pop it as many times as you like during the exposure! \$\endgroup\$
    – Matt Grum
    Jun 15, 2014 at 11:07
  • \$\begingroup\$ Next time I'll try using my iPhone to supplement the camera's flash. Though I guess the results will be totally random and hard to reproduce consistently, since the camera won't be able to adjust the iPhone's flash intensity or the number of times it fires, or the color temperature of the dual-tone flash, or the fact that the LED turns on continuously for a second or two and then flashes. \$\endgroup\$ Jun 15, 2014 at 16:13
  • 1
    \$\begingroup\$ Don't use the iPhone as a flash: use it as a constant light source to 'paint' the foreground with light. It will take some trial and error but you can get good results. Of course, a proper flashlight is better. \$\endgroup\$ Jun 15, 2014 at 16:50
2
\$\begingroup\$

your 3 images are al expected. Your flash is very little impact. You see impact on the metal pole to the right and the ground in front of you (look at the lower left). Your flash tells your cameras it near daylight 5500K WB) so it sets that WB. In both images with flash. The ground in the bottom and the right metal pole has a nice color with that WB but the rest which is untouched by your flash is yellowish. With a cold WB tends to get with a warmer setting. The -2EV has a shorter span of good WB in teh ground. the reach is shorter.

The last image uses a better WB for the entire frame, because it is more consistent. The lower part has the same type of light and the pole as better integrated in the frame.

Even if I try to mimick the WB in teh last iamge to match the 2 flash photos better, you see more uniform response, and the ground closest to you is still yellow, and hte pole to teh right is more integrated in the shot (doesnt stand out unnaturally), while in the flash images your closer ground gets brighter and more bluish and the pole shine up.

enter image description here

\$\endgroup\$
4
  • \$\begingroup\$ +1. Perhaps another reason for the pole at the right to be better integrated in the photos without flash is that, when there's a flash, the metal reflects the light, causing it to stand out. \$\endgroup\$ Jun 15, 2014 at 15:58
  • \$\begingroup\$ yes, that's what I mean with with integration vs not \$\endgroup\$ Jun 15, 2014 at 16:07
  • \$\begingroup\$ I meant that there are two different aspects to the fact that it reflects light: its color changes (as you explained), and it's brighter. \$\endgroup\$ Jun 15, 2014 at 16:14
  • \$\begingroup\$ its actually the specular reflection from the flash that I think is the most significant factor to it seeming unnatural. \$\endgroup\$ Jun 15, 2014 at 17:08
1
\$\begingroup\$

With an exposure time of 5 seconds, with the amount of light being gathered, the relatively far distance of almost all subjects, and the relative weakness of the flash used, I would expect there to be very little difference.

The difference between photos 1 & 2, and photo 3 is the white balance.

\$\endgroup\$
3
  • \$\begingroup\$ I will try again with the widest lens I have (to address the issue of things being too far). \$\endgroup\$ Jun 15, 2014 at 10:42
  • \$\begingroup\$ This won't help, as the problem is that your flash isn't powerful enough. Using a wider lens won't make the city in the distance any closer. \$\endgroup\$
    – crunch
    Jun 15, 2014 at 15:12
  • \$\begingroup\$ Sure, but a wider lens will help get more of the railway tracks in the foreground in the frame. Since these are closer to the camera than the ones in the photos I posted above, they should be lit better by the flash. Is that logic correct? I know that the city in the distance won't be lit by the flash no matter what; this question is not about that. It's about the tracks in the foreground. \$\endgroup\$ Jun 15, 2014 at 16:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

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