You don’t lower shutterspeed your camera is raising shutter speed to keep lightnes of the oocjpeg the same but infact your lowering exposure when you rais ISO value when lightconditions arn’t change. That is when you don’t manual set Aperture and shutterspeed that is.
If you set shutterspeed and aperture fixed and raise ISO value your only brighten the OOCjpeg. And a raise slightly sensitivity of the sensor readout i suspect.
Nope, this so called dynamic range is about mapping of the charged wells, photocels.
When you use base iso then you have the max of the sensors capability. Yes. But every step you take with iso value the sensor doesn’t change it’s spec’s.
Extended iso is stretching the boundery’s of a sensor. (hanging over the balustrade so to speak.)
Let’s think.
Why do i raise ISO?
Because my preview image is too dark. Underexposed.
If shutterspeed and aperture gives a certain amount of lux/s which is say nothing till 50% of the maximum charge (whitepoint) so it’s effectively 50% underexposed.
This won’t change if i just change ISO value. What you infact do is use smaller steps of mapping in order to map the smaller differences in the photocel charges due the fact of less charge differences in the grid. Black keeps black but white will be greyisch in your rawfile. Aka when you optimising the readout mapping the maximum dynamicrange declines due the fact there’s less difference between minimum and maximum capture per each photocel.
George – you may be just fooling yourself here because the camera’s histogram is jpeg-based.
OXiDant suggested in an earlier post that you perform a simple experiment where you will see the results in a RAW-based program (FastRawViewer). By doing so you will prove to yourself that ISO is not part of exposure for RAW files.
If you prefer the armchair approach, there is a lengthy, excruciating discussion of this topic in DPRevived (see link below) where some real experts, e.g., Iliah Borg (FRV) and Jim Kasson among others, weigh in on the matter. Also, some nutcases, I’m afraid, but the information is in there.
Whether any of this is important for your day-to-day photography is another matter entirely.
For a constant lit scene at fixed aperture, if you double the shutter speed you will halve the exposure of a RAW file. Your ISO setting will not change the exposure of the RAW file.
My camera has an light meter. It meters the light and calculates the exposure settings, a,s and iso. When one changes , some of the others must change. Iso doesn’t change the amount of light but the needed amount of light. And that we are doing with the ligh meter, in or out the camera.
Yes, it is output based, like the jpg,tiff etc… You can’t compare it with the sensor. You can’t use it for calculating with stops. All these programs that uses RAW histograms and are playing with exposure stops are wrong in my eyes.
Look here Exposure value - Wikipedia
Your “output based” frame of reference is revealing. There, the “exposure triangle” construct, originating as it does in film technology, has obvious utility for rendered images. We’re in agreement.
But I’m assuming that most of us on the forum work with RAW files and are trying to get the best out of our cameras and processing software. So, it is a bit surprising to hear you go on to decry use of RAW histograms, for example, as somehow “wrong”. Why do you say that? You may not be one to “look under the hood”, so to speak, but others do want to peek now and then. Just might learn something new, useful or not.
The exposure triangle is input side based. The lightmeter calculates the amount of light to be used based on 1)the physical amount of light through the aperture and shutter speed and 2) the magnification factor: the iso. That process can’t be undone, only corrected.
The a/d converter converts these values into binary values, usually 12 or 14 bits. This is what you call your RAW data.There’s not much you can do with it. You can’t calculate with stops by example despite it is done. Every relation with the sensor is gone. The only thing you can see is the room at the low and high side of the histogram. Just like with what you call the jpg-histogram.
In the conversion to a RGB-raster pixels are created. The values are changed so that they fit in a color space. Somewhere in this line also a gamut correction is added. On the way to the monitor another correction is done: to the monitor’s color space.
I don’t see any use of a RAW-histogram.
And in this context iso is influencing the light.
In digital cameras, ISO is not a measure of sensitivity but of amplification of the analogue signal before it gets digitised.
If you allow the camera to choose the ISO automatically, it takes away a measure of control.
When I shoot, I spot meter, zoomed in, and then reframe. By setting auto-ISO, as soon as I reframe, the spot measure point finds a different subject and changes the exposure.
The only time I can see auto-ISO being of any possible use is if you want to let the camera decide the exposure and use something like matrix or centre-weighted metering.
The question is - how much control do you want over the exposure your camera records?
That’s what I’m saying too.
I must say the since I got the Z6II I use auto iso often. The camera doesn’t make much noise and the in-camera image stabilization let me use lower shutter speed.
I always shoot in the A-mode.
When I need a fast shutter speed I use the S-mode.
High- and low-key I never do/did.
With bright highlights and deep shadows I use the blinkies and histogram and correct the exposure according them.
The histogram I use always.
With the D750 I used the ae back button. Using that I could aim the camera on the light parts, lock the exposure, than aim on a subject, lock the focus, and after that I can composite the framing. I don’t do that always,.
The exposure for this image was carefully spot measured on one of the spotlights at +2EV, to make use of the available dynamic range at 400 ISO and required a 10 stop ND filter to allow such a low ISO and f/16 to create the stars from the lights.
Now that is what I call fiddly, trying to manipulate your thumb on one button whilst turning a dial whilst half-pressing the shutter.
True, iso algorithm doesn’t change the physics of a sensor. It optimises the use of the sensors physics by calibrating/adjusting the gain of the Analogue circuit in order to get a propper readout which need to be digitised. In the metadata is the isosetting storred so the rawconverter knows which lightnes the image in RGB should be.
Halfpress or backbutton Exposure lock will prevent that.
When i want some extra control over exposure as in which part of the scene needs to be exposed correctly i can use exposure compensation dail or use the backbutton to lock exposure when i use the pinpoint modes.
Most of the time i am in a wide box mode, lazy thinking rawfiles are flexible enough to adjust the wrongly guesses of the camera.
My aproach towards @mikemyers was about if you want some automated comfort wile clicking around, auto iso is the least harmless because it isn’t part of the actual exposure it only influence it by changing for instance shutterspeed if you allow it. (A-mode.) Then it spiraled down in “does iso is a part of exposure or not.”
Most of the time i enjoy the fruits of carefully build automated m43 camera.
Aim, decide frame, set aperture, check shutterspeed, adjust EC if i want. Focus point on item of interest and click. Job done.
Oke shot of detailed scenery i wiggle some more to let iso not be a blurring factor.
Turn off auto iso and see if i can get away with iso 400 800 ish. If it was raising the iso too far in i-iso that is.
Edit:
In the link the guy say’s
Citaat For some of my street photography work (when both myself and the subject(s) are moving) I’ve hit a sweet spot where I’ve nailed it that 1/500th to stop the motion coupled with f5.6 or f8 for sufficient dof is mandatory.
For this scenario I’ll use auto ISO and let it float. I’ll maybe use + exposure compensation for ETTR purposes. It depends.
That was my initial advise.
ISO matters only if it’s ruining your image you wanted to show.
Restrict auto ISO just before the drop off of quality which you can decide in several site’s. My m43 sensor is 1600iso before prime denoise , 3200iso using dxo’s prime denoise and deepprime does even allow 6400iso in some cases.
I have it restricted to 3200iso
I doubt that. Iso does magnify the signal in some way in the camera. So the raw sensel values are produced from that magnified signal. It doesn’t happen in the raw converter. The raw converter is creating the RGB-pixels and modify them with a gamma correction and to a certain color space.
I aim the camera to the light parts in the environment. I don’t use the dial. Would be useless with the D750 since I don’t have a real time histogram. I may try it with the Z6.
Here are some night shots from the rail station in Liege, Belgium.
Ok
Let’s cut it in pieces.
1 isosettings in a camera does a few things.
1 - it sets the gain according to the in the factory decided optimum settings , each isovalue has its own gainvalue.
This gain is decided and set to keep as clean as possible , noisewise, readout from the sensors pixels/photocells on diverent levels of (under)exposure.(lux/s value’s)
2 - the same isovalue is used to create a preview in your camera LCD (the real live preview is also using/factoring in, shuttertime and aperture values to show in front on your LCD the oocjpeg outcome.)
3 it is used to process the OOC jpeg.
So in order to give the rawfile’s data a marker which give away which lightnes the camera would give the jpeg you need to have a isovalue written inside the metadata. (rawfiles has r,g,b,g channel values which are converted by demosiacing in to RGB “pixels” in the rawdeveloper.) those r,g,b,g channel values are just done by the mapping of the ADC. Rougly speaking value iso is then like km, m, cm.mm to give the number it’s realword value. (like i send you this: Please make for me a box with the mesurements 167x145x10.) if you don’t know the value of 1 in size you can’t make it as i want. So i have to send you please make it 167cmx145cmx10cm.
I said rougly. I wonder if you alter metadata and change the isovalue the rawdeveloper previewed a different image in lightness.)
I still doubt that the digital sensel values are not iso corrected. To me the magnification of the analogue signal is done before the A/D conversion.
Your conclusion in your test post 1858
Maybe this is wrong. When shooting at base iso of 100 the camera produces a histogram of a certain shape. When using iso 200 and the same a and s the exposure is multiplied by 2 but will have the same form unless clipping occures.
Now look at your FRV histogram. It gives you a EV0 exactly 3 stops below the maximum value. So the histograms will look the same. I don’t have FRV or Rawdigger.
