So, under favourable conditions, i.e. f-stop and aperture can be widely adjusted, movement not a factor, etc. etc. is there a "BEST" ISO to use for a particular camera? Or just use the lowest ISO setting for optimal results?
The best ISO to use for overall image quality is the lowest ISO setting, and then expose such that the highlights are clipped just where you find it acceptable.
There's some confusion about sensor signal gain. Some people believe that there is but one "native" sensitivity for a sensor and anything but that is digital manipulation of the data. But, as far as I know, changing the ISO in camera changes the way the sensor interprets the analog signal into digital data. That is, the gain is performed on the analog signal, not the digital one. Another point I believe to be the case is that the "expanded ISO" gain settings (usually below 100 and above 6400, but it varies) as well as some of the ISO settings in between major stops are performed digitally. I don't believe there to be a "native" level of gain.
I'm open to being corrected on either of these issues, of course.
In any case, the best ISO to use for a given image is the lowest one possible that delivers the desired exposure, like anything else. Assuming you don't appreciate the qualities of a grainy high ISO image. If you do, then by all means go higher.
winman3 wrote:
...So, under favourable conditions, i.e. f-stop and aperture can be widely adjusted, movement not a factor, etc. etc. is there a "BEST" ISO to use for a particular camera? Or just use the lowest ISO setting for optimal results?
alundeb wrote:
The best ISO to use for overall image quality is the lowest ISO setting...
JamesBeach wrote:
...the best ISO to use for a given image is the lowest one possible that delivers the desired exposure, like anything else.
Although the intuitive answer -- that the lowest-sensitivity setting possible is always the best, the real world isn't always that simple.
BrianO wrote:
Although the intuitive answer -- that the lowest-sensitivity setting possible is always the best, the real world isn't always that simple.
I am very familiar with those discussions, and they are concerning read noise only. If I want to shoot ISO 160 to get lower read noise than ISO 100, what really happens is that the camera records ISO 200 and then adjusts the signal digitally down to 160. So even if you get lower read noise, you get more photon shot noise, which is still the dominant noise source unless you do special shadow recovery.
I think generally it is considered better to avoid the special "Lo" ISO settings because they tend to sacrifice some dynamic range. Nevertheless, they can still be used to make good photos. There's some technical factors that prevent such settings being classed as official ISO values because they don't strictly comply with the ISO definitions. I don't know the details but I'm sure it has been described by the technical gurus here at FM in the past.
Depends on the camera. I know with my 5D2, it performed better at the 160 ISO multiples instead of the "base" ISOs of 100, 200, etc.
Some people will say that's because the camera is really shooting at ISO 200 then backing it down in the pipeline.
Others say, by looking at DXOMARK sensor data, that the camera isn't really at ISO 200 when you set it at 200, but a bit lower, meaning the camera is actually pushing the shots to 200.
I dunno - What I generally try to avoid is hitting the L or H settings and shooting like I normally would.
However, in real-world A/B comparisons in low-light shooting (and shadow raising), on my particular 5D2, the noise from ISO 160, ISO 320, etc. looked better and had less banding than ISO 100, 200, etc.
cputeq wrote:
Depends on the camera. I know with my 5D2, it performed better at the 160 ISO multiples instead of the "base" ISOs of 100, 200, etc.
Some people will say that's because the camera is really shooting at ISO 200 then backing it down in the pipeline.
Others say, by looking at DXOMARK sensor data, that the camera isn't really at ISO 200 when you set it at 200, but a bit lower, meaning the camera is actually pushing the shots to 200.
I dunno - What I generally try to avoid is hitting the L or H settings and shooting like I normally would.
However, in real-world A/B comparisons in low-light shooting (and shadow raising), on my particular 5D2, the noise from ISO 160, ISO 320, etc. looked better and had less banding than ISO 100, 200, etc.
The ISO calibration from DXOMark is a calibration of the true light sensitivity, that may differ from the manufacturer's specification, but has nothing to do with how the amplification is implemented. All cameras differ somewhat form the spec there.
You cannot really separate the selection of ISO from how you expose with respect to the histogram and clipping. When you select ISO 160 on the 5DII, you get less headroom, and you could have gotten the same good results for pattern noise by exposing to the right with ISO 200. But you would still clip the maximum signal from the sensor compared to ISO 100, so using ISO 160 will not give you the lowest possible photon shot noise even if you expose to the right there.
ISO 160 is a kind of a lazy way to expose +1/3 at ISO 200. If there was a setting for ISO 80, it would have been a nice lazy way to expose +1/3 at ISO 100.
JamesBeach wrote:
There's some confusion about sensor signal gain. Some people believe that there is but one "native" sensitivity for a sensor and anything but that is digital manipulation of the data. But, as far as I know, changing the ISO in camera changes the way the sensor interprets the analog signal into digital data. That is, the gain is performed on the analog signal, not the digital one. Another point I believe to be the case is that the "expanded ISO" gain settings (usually below 100 and above 6400, but it varies) as well as some of the ISO settings in between major stops are performed digitally. I don't believe there to be a "native" level of gain.
I'm open to being corrected on either of these issues, of course.
In any case, the best ISO to use for a given image is the lowest one possible that delivers the desired exposure, like anything else. Assuming you don't appreciate the qualities of a grainy high ISO image. If you do, then by all means go higher. ...Show more →
On most cameras intermediate (between stop) ISOs are not implemented by adjusting the gain but by pulling the exposure. In addition the highest ISO setting is often pushed and not by gain.Which stops are pushed are camera dependent.
GeneO wrote:
On most cameras intermediate (between stop) ISOs are not implemented by adjusting the gain but by pushing the exposure. In addition the highest ISO setting is often pushed and not by gain.
cameron12x wrote:
In other words, shoot at ISO 100 on the 5D2 for the best image quality from that body.
Yes, but this alone does not guarantee best image quality.
The key for maximizing image quality is not primarily the ISO setting, but the exposure. Exposure is a function of aperture and shutter speed. To maximize the exposure, you must operate at the border of highlight clipping, and this will most often require some effort and experience.
alundeb wrote:
Yes, but this alone does not guarantee best image quality.
The key for maximizing image quality is not primarily the ISO setting, but the exposure. Exposure is a function of aperture and shutter speed. To maximize the exposure, you must operate at the border of highlight clipping, and this will most often require some effort and experience.
Absolutely agreed. There are MANY variables which affect image quality.
I just wanted to make sure that I understood correctly that the lowest (non-expanded) ISO on a given DSLR will produce the best image quality for that DSLR (not taking into consideration any other variables).
As an aside, I try to ensure that my exposures are optimally shifted to the right for the best IQ (which I interpret as maximizing DR in that case).
Cameron -- In your hierarchical listing, you left out ISO 50, which I find very useful, and certainly better than ISO 200 for image details and low noise. There might be a slight penalty in high end clipping and compressed DR, but the overall IQ certainly makes it a better choice than ISO 400.
I would rank the IQ of my 1Ds2 and 1D2 as: 100, 200/50 (50 is capable of finer details/lower noise than 200, but suffers slightly reduced DR -- so nearly a tie), 400.
I generally shoot these cameras at ISO 200 for hand held daylight images, depending on overall lighting conditions. Second choice is ISO 100 for best details, noise and DR. But I find the extra stop at 200 to be significant compared to very slight compromise on IQ. On tripod, it would almost always be 100.
alundeb wrote:
Yes, but this alone does not guarantee best image quality.
The key for maximizing image quality is not primarily the ISO setting, but the exposure. Exposure is a function of aperture and shutter speed. To maximize the exposure, you must operate at the border of highlight clipping, and this will most often require some effort and experience.
That is all true, Anders my friend, if a key photography objective was to have the least amount of noise in an image. However, that is often only an auxilliary technical aspect.
I often "expose to the left" where that suits my photographic quality objective better. True enough, here and there I thereby create pronounced noise which has to be dealt with, yet I gain much more in my own aesthetic sense.
For example, I must have seen thousands of images of birds with white colours in their plumage which have been exposed to the right following the internet mantra. Now, as a result, the pictures might exibit some combination of these three adverse appearances:
(1) Highlights are unpleasantly blown in the image, and details are gone where and when it counts.
(2) After a major photoshop intervention, the blown highlights have been "mitigated" in a way that turns them into grey-beige, which is clearly not the target's colour.
(3) The dark/irridescent plumage is left overexposed, I guess in order to "bring out the colour", which again departs from the true target's colouring and renders an unnatural look to the bird.
PetKal wrote:
That is all true, Anders my friend, if a key photography objective was to have the least amount of noise in an image. However, that is often only an auxilliary technical aspect.
I often "expose to the left" where that suits my photographic quality objective better. True enough, here and there I thereby create pronounced noise which has to be dealt with, yet I gain much more in my own aesthetic sense.
For example, I must have seen thousands of images of birds with white colours in their plumage which have been exposed to the right following the internet mantra. Now, as a result, the pictures might exibit some combination of these three adverse appearances:
(1) Highlights are unpleasantly blown in the image, and details are gone where and when it counts.
(2) After a major photoshop intervention, the blown highlights have been "mitigated" in a way that turns them into grey-beige, which is clearly not the target's colour.
(3) The dark/irridescent plumage is left overexposed, I guess in order to "bring out the colour", which again departs from the true target's colouring and renders an unnatural look to the bird....Show more →
+1, Peter, i am with you on this...man i hate digital unnatural IQ look, you get the same results from point and shoot cam
I realize that there may be some instances that you may want to expose to the left for a certain "look," but in general doesn't exposing as far to the right as possible (without blowing out the highlights) give you the most usable DR?
And the greatest flexibility for any desired modifications in post?
I have no problem with this, my friends and fellow discussion partners. There are always some things that don't get mentioned in single posts, and are somewhat taken for granted.
My assumptions were, and some of these may differ based on your workflow:
I always shoot in RAW and adjust for exposure in the conversion program. If you shoot JPG, this opportunity is much more limited, and exposure compensation may result in unpleasant looking images.
I also assumed, that prior to taking advantage of the highest possible exposure, you have already ensured that the highlights you want to preserve are not blown out. Exposing "to the right", I understand, means that you concentrate on getting the right hand part of the histogram, the highlights, correct.
We may also need to repeat that the OP asked about the case where there is enough light, and doesn't have to take into account getting high enough a shutter speed.
