Hi
My lens and camera combination is Canon 85mm f/1.8 and 5D markII.
I don't want to use tripod or flash indoors.
I have heard that there is possible to change the exposure by 2 steps if you shoot in raw.
I don't think I want to use higher ISO or shorter shutterspeed as they are pushed to its limits.
Then I think that I can underexposure every picture I take indoors and change the exposure back with imagining software, but I guess that I will loose some image quality. Could this be a better option than using very high ISO settings?
Regards
-Bjørn
[quote name='Bjørn Tore' timestamp='1298925328' post='6404']
Could this be a better option than using very high ISO settings?
[/quote]
No. But do a simple test to cinvince yourself ...
take the same image twice ... once with say 6400iso and normal exposure ... and once again
with 1600iso and same settings for exposure and aperture (therefore 2 stops underexposed)
and pushed in raw-conversion by 2 stops.
I bet, the 6400iso image will look better.
Just a thought. ... Rainer
[quote name='Bjørn Tore' timestamp='1298925328' post='6404']
Hi
My lens and camera combination is Canon 85mm f/1.8 and 5D markII.
I don't want to use tripod or flash indoors.
I have heard that there is possible to change the exposure by 2 steps if you shoot in raw.
I don't think I want to use higher ISO or shorter shutterspeed as they are pushed to its limits.
Then I think that I can underexposure every picture I take indoors and change the exposure back with imagining software, but I guess that I will loose some image quality. Could this be a better option than using very high ISO settings?
Regards
-Bjørn
[/quote]
Both are basically the same!
Sensors do NOT change sensitivity with changing ISO settings. The camera just applies more amplification of the received light.
When you underexpose and then amplify the exposure on the computer, you are doing on the computer what the camera's electronics otherwise would have done.
As such, it will not make a big difference, whether you let the camera do the amplification, or your computer.
As written above: You won't see an improvement, because you can't trick the laws of physic.
Color and luminance noise will have the same levels *and* you have a higher risk of posterisation in critial parts. Underexposure will reduce the bit depth your sensor chip is theoreticly able to deal with. By compensating the underexposure during postprocessing you "stretch" the bits and *may* cause visible colour gaps. It is not an issue in most cases.
If you have troubles with visible noise artifacts/reduced details using built-in noise reduction in Camera RAW/Lightroom or DPP (or other RAW converters) I strongly suggest to take a look towards tools like Noiseware, Noise Ninja, NeatImage and Topaz DeNoise. Dealing with high noise levels is always a compromise between noise reduction and loss of detail. DPP tries to deal noise with two rulers. Noiseware has some 3 dozens, but you won't have to use them and just rely on default settings most of the time.
Ciao, Walter
Reading the original post, I started to wonder what you are photographing.
The night-time/indoor work I do covers theatres, concert halls (classical), dance stages and press conferences etc. So far, I've managed to get by with various primes (f1.4 upwards) and maximum 3200 iso (usually 400-800) occasionally I'll use a monopod or tripod, and when necessary (and possible) a flash.
I can't think of any indoor situations where you would be expected to get good photos without high iso, or a flash, or good lighting, or a tripod. You'll have to compromise on one of these.
So my question is: are you being realistic by restricting yourself to this extent? Maybe you should let us know what you want to photograph, then maybe someone has a good solution.
No, underexposing by 2 stops and pushing in post will make a visibily worse quality image.
The main reason for this is that you have now lessened the (already low) dynamic range of your image. This will be very noticable in the shadow areas of your image which I am guessing there are going to be a lot of since you are running at f/1.8 and querying about high ISO.
There will also be a lot more shadow noise/banding in the underexposed shot as digital cameras introduce a lot more noise in the darker areas than light.
I would be interested in seeing the test described above by Rainer (with no Noise Reduction) which I expect would back up my 2 reasons above.
Allan
Is it possible that underexposed shot + exposure slide in ACR vs. high ISO setting usage can have same consequences in resulting image? I really ask this because I'd like to enhance my knowledge in that respect. Firstly, I think it highly depends on how dark is the capture RAW image (which is to be corrected in ACR for example). I fully agree that the sensors don't have changing sensitivity but the amplification of the signal does the job if the ISO is increased. But is is the same when it comes to manipulate the raw data in ACR?
If you underexpose by using higher shutter speeds at -lets say- base ISO, you'll get less light photons hitting the sensor, and apart from brightness issues you'll have limited saturation, tonal and dynamic range. OTOH, if you shoot at appropriate shutter speeds at higher ISO setting, -depending on the high ISO performance of the camera- you'll have color/luminance noise, limited sharpness, and comparatively more tonal and dynamic range (if the captured image at the first case was very dark). So, latter has sharpness and noise problems in addition, but comparatively better in terms of tonal/dynamic range. Isn't that right or am I missing something?
Serkan
[quote name='PuxaVida' timestamp='1299060435' post='6425']
Is it possible that underexposed shot + exposure slide in ACR vs. high ISO setting usage can have same consequences in resulting image?
[/quote]
No, albeit the effect is quite similar.
When you expose correctly, the following happens:
1) Each sensor-cell produces an analog output-voltage.
2) This voltage is amplified by an analog amplifier (which is controlled
by the ISO setting you selected).
3) The amplified voltage is digitised by an A/D circuit.
When you underexpose, the following happens:
1 )Same as above.
2) The voltage is amplified less, due to the lower ISO setting. It will therefore
not reach the same maximum level that is reached above.
3) The amplified voltage is digitised by an A/D circuit.
4) The digitised value is enlarged
It might seem, that there is no difference between doing an analog amplification and
a digital amplification ... but there is ... it is hidden in step 3.
With a properly amplified value as output of step 2, the A/D step can make use the
full bitdepth ...while the output value of the underexposed image will never reach
the possible maximum, and therefore the A/D step will always produce "low"
values (due to the underexposure). But even if you enlarge these values later, the
point is, that you started with a reduced number of values ... and the number of
distinct different values does not increase.
Example ... a properly amplified signal produces digital values between
0 and 1023 after digitization (10bit) ... an underexposed signal (by 2 stops)
will only produce values between 0 and 255 ... and even if you multiply
by 4 (and therefore increase the range) to 0 to 1023 ... you will not use
all possible values. ... 0*4 = 0 ... 1*4=4 ... the 1 and the 2 and the 3 will not
be used. ... You will see this effect as posterization especially in the dark areas.
Thanks for the explanation... And I assume the results in terms of noise levels in analog amplification vs. digital bit-shifting will vary from camera to camera. I haven't tried it on my camera but I suppose underexposing and then pushing up in RAW converter produces less noise than ISO boosted images... But the results regarding the other IQ criterias like color saturation / sharpness / dynamic & tonal range is another issue...
Serkan
[quote name='PuxaVida' timestamp='1299072803' post='6438']I haven't tried it on my camera but I suppose underexposing and then pushing up in RAW converter produces less noise than ISO boosted images...[/quote]
Test it and post results. I tried ...
Again: You can't trick the laws of physic.
Ciao, Walter