Yes, gain is about analogue value’s. So a higher iso value set with the dial on your camera causes internaly the camera to increase gain on the photocell readout circuit.
After this the virtual image gets mapped and digitalized into a rawfile.
FRV is a great for it’s sellingprice.
Camera histogram and dxo histogram are based on RGB values not r,g,b,g values. Aka tiff and jpeg not raw.
Maybe i can find my test for you.
It is and you compensate it with iso.
You could say this is only a way of wording things.
But this has consequences on what happens with your images. Sensors are a bit complicated.
Look at those ones :
And enjoy.
It’s more. @OXiDant is using a so called RAW histogram. Watching the first video the signal is amplified before A/D conversion. So the sensel values are already iso corrected. He is saying no.
George
But did you look at where the EV0 is situated? 3 Stops to the left relative to the highest values.If you just change the iso the shape, the rough form, will not change. And if in both images the EV0 is situated 3 stops to the left relative to the highest value you get a similar histogram. I can’t play with it since I don’t have it and dowloaded a trial in the past.
George
No always; it depends on iso and kind of sensor.
there is no (analog) signal after ADC … after ADC there will be only digital data 
… so any “signal” amplification can happen in analog domain
Must happen I think.
I don’t think that analogue signal is stored somewhere.
George
as usual a wrong generalization… in some cases, for example - old digital MF backs, “ISO” ( any nominal ISO used ) was nothing else but a tag value written in a raw file to give a hint to a raw converter to perform a data correction …
in a proper generalization a nominal ISO value dialed in or selected automatically can result in any combination of analog gain before ADC + digital multiplication after ADC ( so that data written in raw files will be already adjusted ) + tag values hinting to raw converters to do multiplication during raw conversion - and how things are vary depending on a specific camera model
there can be NO analog amplification at all … it all depends on a specific model with a specific sensor in it
At base iso.
George
I will carefully and slowly watch those two videos up above - I suspect they have a lot of good, useful information.
My D780 is still in (M)anual mode, as I need to learn how to properly adjust the controls for AutoISO which I haven’t done yet. I got to my brother’s home yesterday, and while he took an early-morning walk with his dog, I brought along my camera looking for things that seemed interesting, and worthy of a photo. In all, I captured 7 images.
That is how I captured an image.
I didn’t have a tripod, and maybe I should have put the camera in spot metering mode, set the ISO to a reasonable value, perhaps 200, and metered for the brightest part of the image, then set the exposure compensation to +2. I will try this next time.
(I also didn’t (yet) set the camera to auto ISO, as I haven’t practiced with it yet. That can come after I do the things I’m comfortable with, both shooting as I described.)
This has turned into an interesting discussion, but I think my basics still apply:
No i am saying iso isn’t a factor in exposure.
Iso value dial on your camera is setting certain gain value’s in the electronic readcircuits. (thus analogue) this optimising (the underexposed) sensor readout so your image is as clean as possible. Thus yes analogue signal improvement not in post nor after the ADC.
The other thing i have said is that the isovalue number is written in the rawfile so the rawconverter knows the expected lightness of the preview image. So the rawfile stores the same “exposure” in every shot but each isostop higher more gain has bin used to improve the image before it becomes a rawfile.
Test set camera on raw plus jpeg.
Use tripod, wait for darker times.
Start with base iso 3 or 4 stops underexposed lock shutter time and aperture.
Then increase iso value until your camera tels you the image is “propper exposed”.
Load rawfiles and oocjpegs in a propper viewer.
You will see that raising in post the underexpost raw image’s they look like the "propper exposed " oocjpeg. Same noisyness.
One thing speaks for using higher iso then base iso: no camera is trully isoinvariant.(which would implicate that for raw you don’t need the iso-dail.)
So using higer isosettings on your camera means probably a more adjusted camera electronics for lower light.
My opinion - when Aperture, Shutter, and ISO are set correctly I get a good image in both my film and digital cameras, assuming I am using the appropriate ASA film.
If I then open up my lens much wider, my image appears over-epposed.
If I then close my aperture much smaller, my image appears under-exposed.
If I raise my shutter speed a lot, my image appears under-exposed.
If I lower my shutter speed a lot, my image appears over-exposed.
If I raise my ISO (or put in film of a higher ASA), my images appear over exposed.
If I lower my ISO (or put in film of a lower ASA), my images appear under exposed.
For me, it’s a triangle - adjust any one of them, and I must compensate with at least one of the other two.
It is a factor in the desired exposure. The camera’s light meter wants a cerain amount light on the sensor. At base iso that’s gained with the aperture and shutter speed. Raising the iso lowering the result of these 2 settings. So it is a factor in exposure. That anlogue signal is multiplied depending of the iso used and that signal is send to the A/D converter. There you get your sensel values.
That’s because using the exposure compensation in post the same result has as raising the iso. Exposure compensation is one of the few tools that uses the raw data: it’s linear. The converter doesn’t work with the iso values nor do they work with shutter speed or diafragma.
George
Having considered most of what is written above, here is my resulting question.
I set aperture based on my desired depth of field, and other factors.
I set my shutter speed, or try to, so my hand movements don’t move the camera, blurring the image. I also try to set the shutter based on what I’m photographing - do I want everything sharp, or do I want something (such as a propeller !!!) blurred.
I am often moving around, with lighting changing all the time, so either I (or my camera) needs to control the exposure. From what @Joanna writes, I’m supposed to do that, but very often I don’t have time to do so.
This is where I hope to use auto-ISO, as that won’t disturb the other two components of my exposure control, aperture and shutter.
In the Nikon D780, in the “Photo Shooting Menu” there is a specific selection for this, “ISO Sensitivity Settings”.
Under that there are three settings, along with a switch to turn auto-ISO on or off.
The main control, turning auto-ISO on or off is easy:
Then there are three selections I can make:
I currently have maximum sensitivity set to 16,000 - is this reasonable?
I currently have minimum shutter speed set to 1/500 - again, is this reasonable?
With Photolab, and the D780, I could change the maximum sensitivity up to 51,000 ISO, and higher. If I set it to 16,000 I don’t think it will ever get that high.
My thoughts are that if I’m wandering around, and suddenly see an image I want to capture, the above settings will get me an acceptable image, as long as I focus.
But if I’m taking the time to capture a specific image, I’m pretty sure I would want to set something manually - but auto-ISO is likely to get an acceptable image. Of course, if the camera is on a tripod, or I am setting up to capture a specific image, I think I would leave auto-ISO turned off.
Yes that is pretty reasonable, all in all.
Most modern cameras allow you a lot of automated and semi automated controls as well as manual ones. If you can trust the camera to make the decision for you, it can save you time and help you focus on the image, composition, lighting etc. With experience you will learn how your particular camera works and what you can rely on and what you cannot rely on and need extra attention.
Manual mode is great when you are not in a hurry and you a shooting a scene where conditions are stable. Semi auto mode, such as Auto ISO, Aperture priority mode, Shutter Priories mode, or Program mode
Program mode on a DSLR camera is a semi-manual mode where the camera sets the exposure for you, choosing the correct aperture and shutter speed based on the amount of light available. It is similar to shooting on auto, but with the option to manually adjust the settings if needed. The camera algorithmically picks from a list of predetermined settings based on the amount of light measured when it meters. Program mode unlocks other functions that give you more creative control over your images. It is ideal for point and shoot moments when you just need to quickly snap a picture.
I tend to use Program mode in street photography or just travel stuff, but if I’m photographing something like architecture or landscape I want manual mode. If I was shooting fast action or going for particular creative effect I would probably choose one of the semi automated modes that is appropriate for the effect I’m going for.
Indeed what you wrote.(i think you understand this better/more then me.)
When i was trying to understand what ISO and ISO value in combination with noise really is, this partly analogue and digital optimisation inside the camera just before it stores the rawfile makes it even more difficult to grasp. I got a lot of help from people who where able to calculate sensors sensitivity specs, who lectured me about this pushed me in testing concepts with small tests.
To be clear as far as i understand it.
Official. The signal is not the latent image. Every photowell/cel creates 1 signal. Which is 1 readout and 1 charge level. =>SNR is based on that. Noise is something what only is visible on lower signal, or in long exposure moments.
The ADC does shift 16 bits which contain 12bits of data of 1 photocell and 4 bits as grid location. I think. This could be already 32bits in high res camera’s. To deep to be really interesting to know.
What is interesting is:
The sensor is powered with a gain, this gain is in base iso optimised to capture maximum of charge caused by foton’s. None til filled is the DR of the sensor.
As always the botom part will be very worse in SNR. There is always electrical noise. I think the Captest is to see that noise level.
Any gain enlargement to “improve signal” will enlarge alway’s also the electrical noise(i thimk it is called readnoise.).
But that noise can be filtered by comparing 2 well together and detect that noise signature. (internal optimisation software connected to the isovalue dail value’s.)
What also is “amplified” is shotnoise and other noise types. => dxo prime is dealing with that hopefully. Me was told that “amplified” is the wrong word because the ADC into finally a rawfile is far more complex to be called that.
To keep things a bit simple, a good exposure is measured in lux/s to a certain level (something with 18% grey?)
A long exposure of 2 min wil not cause the same exposure quality as a 30 sec one. (shotnoise and thermal noise of the heating up of the sensor.) therefore the iso dail is a usefull item. To avoid to much longexposure noise.
But the exposure triangle is a relic of the film era.
You read the “film speed” and in the electronic era of dslr and other camera’s you set the asa/iso of the box in the pasm-wheel so it knows which auto shuttertime it needs to calculate. My canon ql featured a lightmeatering and a asa dail, the Aperture and shutter time you needed to do yourself. The light meatering was set by that ASA dail.
So if you forgot to change that after a faster film al your images where underexposed on the film emulsion.
When film got digital file the asa/iso was converted to the now known ISO. A standard for converting a signal readout in to a certain lightnes to impose a correct exposure at the cost of extra visible noise,
Once you grasp the concept of what exposure really is and what is post exposure adjusting and the probable pre exposure adjusting can be, then it gets easier to understand why certain things happen in a imagefile.
And the in post, rawconverter, even extra instructions for further improvement is only connected to the label xxxx ISOvalue. The only way the rawconverter can be instructed to create a certain lightness/brightnes is or in the digital modified rawfile or by the metadata inside the rawfile.or both.
Sorry for the rant/ramble but it’s sometime very interesting to see the misconception that isovalue is part of the exposure. Lenssize and sensor size does. Lux/s per mm2 and bigger photocells does help in capturing more light.
Again, it’s not “part of the exposure” but part of the “wanted exposure”.
George
I think what people mean by ISO being “part of the exposure”, they really are referring to so called exposure triangle.
The Exposure Triangle:
I’m sorry but all this rambling on is doing nothing to make anything understandable.
I have used the exposure triangle of aperture, speed and ISO for nearly 60 years.
Before digital we bought film at a certain ISO and exposure was calculated based on that ISO.
Nothing has changed, despite all the preceding techno-babble.
All anyone needs to know is that it takes three settings on the camera to set the exposure - nothing more, nothing less.
Can certain folks stop trying to confuse the heck out of those who are struggling?
