how do you judge exposure?

[beachbum]

i realise that the LCD on my DSLR is brighter than my monitor, which i deem it to be accurate. this makes exposure tricky to judge.
the pic will look okay on the DSLR, but when transfered to my monitor, it will look like it is 2 stops under. i trust my monitor cause it is inline with the lab that i am using
The LCD brightness has already been reduced to the minimum level.

Then i tried to rely on reading the histogram on the DSLR to get a more accurate exposure reading, but i realise that the histogram shown for the image on the DSLR looks different from the one shown on photoshop.

now, i have to dial in exposure compensation for all my shots when i shoot. i don't think this is the optimum way of working.

anyone has any experience on how to deal with this?? please advice. Now i can't shoot with confidence.

 

[eos]

I experienced the same problem initially with the false impression the LCD is giving. Then I relied on histograms and though I did not compare if it's the same as that in Photoshop, it pretty much relayed accurate exposure data. After a while, somehow got the hang of how the system behaved (consistently underexpose under certain circumstances) and did not rely as much on histogram, except for important shots.

 

[Prismatic]

Hmmm... why do you rely on the LCD for your exposure? Shouldn't you be reading off the exposure meter???
If you know where to meter for your shots, your shots will come out okay what.

 

[Larry]

Originally posted by beachbum
Then i tried to rely on reading the histogram on the DSLR to get a more accurate exposure reading, but i realise that the histogram shown for the image on the DSLR looks different from the one shown on photoshop.

that's strange... don't see why the histograms should be different if it's the same image. are they vastly different?

btw, isn't there some settings for you to control the brightness on the LCD monitor? i know i used to regularly dial a -1 value on my previous D100's LCD setting.

 

[mpenza]

Histogram on the camera should be accurate right? That would be the most objective way of judging exposure.

 

[liuhao]

haha...i have exactly the same problem as you have.

these days if i see a perfect exposure picture on the LCD of DSLR, i know it's underexposed. if it's overexposed, it would be just right. heehee...

Originally posted by beachbum
i realise that the LCD on my DSLR is brighter than my monitor, which i deem it to be accurate. this makes exposure tricky to judge.
the pic will look okay on the DSLR, but when transfered to my monitor, it will look like it is 2 stops under. i trust my monitor cause it is inline with the lab that i am using
The LCD brightness has already been reduced to the minimum level.

Then i tried to rely on reading the histogram on the DSLR to get a more accurate exposure reading, but i realise that the histogram shown for the image on the DSLR looks different from the one shown on photoshop.

now, i have to dial in exposure compensation for all my shots when i shoot. i don't think this is the optimum way of working.

anyone has any experience on how to deal with this?? please advice. Now i can't shoot with confidence.

 

[ckiang]

1. LCD on digicams are never accurate in terms of evaluating exposure/colour.

2. Histograms are calculated differently by different programs, so don't be surprised if your camera displays a different histogram from Photoshop, for e.g

The histogram is the best way to evaluate exposure. How to use it? Read this article


Regards
CK

 

[liuhao]

Originally posted by ckiang
1. LCD on digicams are never accurate in terms of evaluating exposure/colour.

2. Histograms are calculated differently by different programs, so don't be surprised if your camera displays a different histogram from Photoshop, for e.g

The histogram is the best way to evaluate exposure. How to use it? Read this article


Regards
CK

that's a good article. thanks a lot.

maybe for landscape hisogram is useful. for portrait, i just want the face of subject has right exposure and dont care about overexposure of sky, underexposure of shadow area. but i can hardly judge this from hisgram. still have to look at LCD. and LCD is not accurate.

 

[jasonpgc]

Originally posted by ckiang
1. LCD on digicams are never accurate in terms of evaluating exposure/colour.

2. Histograms are calculated differently by different programs, so don't be surprised if your camera displays a different histogram from Photoshop, for e.g

The histogram is the best way to evaluate exposure. How to use it? Read this article


Regards
CK

For DSLR, the best way is to use the Spot meter or Partial Metering. You should be focusing the attention on the subject, NOT on the bar chart. If you still got a hell of problem, use the auto exposure bracketing feature provided by the DSLR. Sort out the bad ones when you are back home :bsmilie:

 

[Hobbes234]

Originally posted by ckiang
1. LCD on digicams are never accurate in terms of evaluating exposure/colour.

2. Histograms are calculated differently by different programs, so don't be surprised if your camera displays a different histogram from Photoshop, for e.g

The histogram is the best way to evaluate exposure. How to use it? Read this article


Regards
CK

hi CK,

kinda new at this histogram thingy... little trouble understanding the article
so does it mean that if your histogram is a bell shaped curve, you would have better exposures.. (and the ends of the curve don't drop too quickly to 0)


thks

 

[BraveHart]

Judge? Based on experience. After awhile you'll be quite comfortable in predicting the effects of a particular aperature/shutter setting in a given scene - make changes where appropriate. Don't rely too much on technology. If needed...bracket your shots.

I'm a film user so I know nuts about the histogram...but what I do feel is that don't make a science out of your art.

 

[ckiang]

Originally posted by Hobbes234
hi CK,

kinda new at this histogram thingy... little trouble understanding the article
so does it mean that if your histogram is a bell shaped curve, you would have better exposures.. (and the ends of the curve don't drop too quickly to 0)


thks

Perfect one will be bellshaped - that means the image has an even separation of tones. Anything that extends way to the right of left can mean trouble.

Regards
CK

 

[jasonpgc]

Originally posted by ckiang
Perfect one will be bellshaped - that means the image has an even separation of tones. Anything that extends way to the right of left can mean trouble.

Regards
CK

I Disagree.

This depend on what sort of subject you're taking.

If it is a night shot, you would expect to see a U curve, where most of the scene will have extreme shadow for the dark sky and huge spikes on the highlight for areas lit by artifical sources.

If it is a picture is taken in cloudy day under shade, you would see a n curve. Because most of the color falls in the mid tone. The contrast is low, so there is almost no shadow or highlight.

If it is a very flat curve, the picture will have a even distribution of tonal colors. This normal for a picture with 1/3 shadow, 1/3 midtone, 1/3 highlight. This does not means it is the best histogram. It just means that the distribution is even.

However, the histogram does not help you to determine how a particular area or color in your scene should appear in your photo or slide, (Only the Spot, Partial, CW meter does so :bsmilie: ). The histogram just tell you how the tones in the scene are distributed. A bell shape histogram does not means you have a good exposure. If you are shooting something with only highlight and shadow, you are in deep shit if you get a bell shape histogram :devil:

 

[erwinx]

Originally posted by jasonpgc
I Disagree.

This depend on what sort of subject you're taking.

If it is a night shot, you would expect to see a U curve, where most of the scene will have extreme shadow for the dark sky and huge spikes on the highlight for areas lit by artifical sources.

If it is a picture is taken in cloudy day under shade, you would see a n curve. Because most of the color falls in the mid tone. The contrast is low, so there is almost no shadow or highlight.

If it is a very flat curve, the picture will have a even distribution of tonal colors. This normal for a picture with 1/3 shadow, 1/3 midtone, 1/3 highlight. This does not means it is the best histogram. It just means that the distribution is even.

However, the histogram does not help you to determine how a particular area or color in your scene should appear in your photo or slide, (Only the Spot, Partial, CW meter does so :bsmilie: ). The histogram just tell you how the tones in the scene are distributed. A bell shape histogram does not means you have a good exposure. If you are shooting something with only highlight and shadow, you are in deep shit if you get a bell shape histogram :devil:

i agree, and look, a luminous landscape article!

http://www.luminous-landscape.com/tutorials/understanding-series/understanding-histograms.shtml

 

[Zerstorer]

Few images ares perfect bell shaped curves. The important thing to note is that there should not be an abruptly cutoff peak on either side. e.g.the right side, which will indicate blown to white highlights.

Histrograms are useful mainly for judging whether the dynamic range of the scene has been captured. It doesn't necessarily tell you if its the optimal exposure, since it is extremely difficult to correlate the correct peak with that of your main subject, unless it is a smooth low contrast midtone subject.

 

[beachbum]

thanks for all the informative replies.

i shoot using partial metering very often, and it has never failed me back when i used film. However, my experience with digital has not really been pleasent, esp. exposure wise. e.g. like D60 has an underexposure problem.

i guess the best bet would be to rely on the histogram and bracket when in doubt!

cheers

 

[mpenza]

actually for film, you may actually have over or underexposure. It's the lab who compensated for the exposure when printing.

similarly, if you send your digital images for printing, the lab will usually compensate for you and you could get properly exposed prints from badly exposed images.

 

[RCheah]

Thought this may come in useful, an article on "How do you know if your D30 is underexposing?".
(http://www.dlcphotography.net/ChuckTips.htm#anchor7648)

This little test confirmed what I had suspected, that my D30 is underexposing by 2/3 of a stop. Now my exp comp is permanently on +2/3.

As for how to judge, I'd say aga aga lah. Since there is no such thing as perfect exposure? Then based on your own preference I guess.