You might have heard of the term "native ISO" before and wondered what it refers to. This excellent video discusses the topic and what it means for your cameras.
Coming to you from ProAV TV, this great video dives into the topic of native ISO. Back in the film days, ISO referred directly to the film's sensitivity. In other words, when you put film with a different ISO in your camera, you were directly changing the sensitivity of the medium. On the other hand, in the digital era, ISO is a bit more complicated. A digital sensor generally has only one sensitivity level (dual native ISO cameras are another topic), for which the term "native ISO" designates the ISO speed at which a full well gives a maximum signal. In other words, you can think of native ISO roughly as the highest ISO at which a post-sensor gain is not applied to the signal. Whereas you can actually change the ISO in a film camera by changing the film itself, changing the ISO in a digital camera roughly equates to changing the gain applied to the signal coming from the sensor. Check out the video above for the full rundown.
If you're passionate about taking your photography to the next level but aren't sure where to dive in, check out the Well-Rounded Photographer tutorial where you can learn eight different genres of photography in one place. If you purchase it now, or any of our other tutorials, you can save a 15% by using "ARTICLE" at checkout.
An interesting video and answers a question which I had been mulling over in my mind as I could not reason how a sensor could replicate ISO. Now I know it is signal amplification via positive or negative gain from the native ISO. So I decided to try and find the native ISO for my D7200. A Google search brought up a review of the camera from 2015 on FStoppers where the reviewer said the native ISO is 25600. This seems unlikely to me and I suspect the term has been used incorrectly. Here's the link.
https://fstoppers.com/gear/nikon-d7200-quadruples-low-light-capabilities...
It always seemed wrong to my mind that Digital Cameras perpetuated the concept of ISO as part of the exposure triangle as I couldn't see how the sensor itself could change sensitivity and by and large it doesn't.
Grain in film images was a function of the material used to record the image although to some it became an artistic effect.
I suspect many equate noise to grain but they are not the same if I understand the video correctly.
Ideally a digital image will be noise free and then processed to achieve the final artistic picture.
I have never understood the desire to equate digital imagery with film I suspect it is marketing to help people understand traditional concepts. Fuji I think have even created filters to mimic old film types. I don't see why anybody wants to do that. We should be exploring new ways of presenting images not hankering for the past.
Enough. I still don't know what the Native ISO of the D7200 is for sure. 🙂
Edit: 5 hours later. I still have not been able to find out the native ISO of the D7200 but what is worse I now have no idea whether it's a Toshiba or Sony chip that is fitted. There is no definitive answer so far as I can see. Plenty of opinions based upon someone's Uncle's next-door-neighbour's pen-friend's cat's suggested rumour source but nothing easily and simply found. Will ask Nikon directly.
Edit 2. 10.38 BST 07Apr2020 I have had a reply from Nikon.
"Thank you for contacting us and for your interest in our products. Information about the native ISO from our cameras is confidential as well as the manufacturer of the sensor. We can only confirm the base ISO as it is the lowest number before Lo, for the D7200 this means 100. If you have other questions, feel free to contact us a again."
So I've asked them why it is confidential. As I cannot see how that information would hurt the company any way but especially as the D7200 is an obsolete camera now.
Or should I treat the answer as a subtle coded message...............?
Knowing the native 'ISO' would allow camera users to maximise their image quality.
It wasn't hard to find directly from the Nikon website.
Now for using the term ISO, it's only beacause the ISO term in film was for the sensibility to light of the film. Now in digital camera it's the sensibility to light of the sensor. It is not achieved the same way but it have the same role in the exposure triangle. That's why they call it ISO.
Your link doesn't mention native ISO at all. Richard's question remains unanswered.
As Joe has mentioned that is the range of selectable ISOs listed in the camera which can be chosen. It could be that 100 is the native ISO and everything above that an amplification but the man in the video suggested that the signal can both be amplified and 'throttled' which implies that the native might be for example 240. It is that figure which I am interested in.
Point of interest.....film cameras do not suffer from "Native ISO"
That was because each film had a 'fixed' ISO viz their native ISO. With digital you amplify the signal to vary ISO - with a film camera you change the film. With digital it is the sensor that has a native ISO essentially 'the camera' as you cannot change a sensor (afaik).
Thank you sir. I do understand. Have been shooting film for about 60 years now. Digital is (was) a bit new to me and had to acquaint myself with its verbage. Thank you.
Why the downvote? It's both true and humorous...
Wasn't me. 🙂
If i did that, it must have been a mistake. Sorry sir
I can't find where the downvote is. Please point it out so I can change it.
Timothy, it wasn't you, you cannot 'vote' on your own posts (for obvious reasons) as the buttons are greyed out. It might be somebody hit the wrong button or made a mistake or just failed to see the humour in your comment.
My first SLR was a Pentax ME Super. I saw a lovely one the other day and was sorely tempted but the cost of film and developing etc is so high now that it really was not a practical proposition.
🙂
Yes thank you. I was replying to Jon Kellett. Thank you sir. I love Pentax. Had a K1000 to train for camera repair decades ago. Ah...the memories. Enjoy yoyr day and lock up, if you're isolating yourself.
Sorry for the confusion, I was commenting about the downvote you got from Joe Dredd - Totally unfair of him.
You know... I miss the excitement of seeing the prints for the first time. I love the smell of the prints, of the negs, of the photo lab. I just don't miss the cost and low res.
Yes i understand what you mean. I mostly shoot slides for agencies and myself, but do shoot print film as well. I like waiting for 'Christmas' to arrive after I've turned in the slides for developing. Have a safe day.
I'm a bit lost. I thought the native ISO was a combo of the sensor sensitivity and the parameters of the amplifier in camera. The sensor is what it is but the engineers design a amplifier that runs for arguments sake from ISO 100 to ISO 6400 and anything above or below that ISO rating is a digital manipulation in camera (not analog). From the video, it seems that there may be only two ISO ratings that are the true native ISO. Any one care to correct me if possible.
That is essentally what the man said as I understood it. Each sensor has a native ISO. Some sensors have 2 (more?) which increases the range of ISOs used where digital noise will be minimised. And he amplification starts from a second level a fresh rather than being exponential.
That's what he says in the video, which isn't really true at all. Old cameras in particular (but also some newer ones) have multiple amplification levels, just look at the "sawtooth shaped" read noise curves of cameras like the Canon 6D, Nikon D5 and most sensors before the "isoless" sensor which Sony started manufacturing since 2013/2014 if i remember correctly which have a straight line.
Now Sony moved to the dual gain setup, a lower gain which allows you to use all the DR of the sensor effectively and a high gain (usually real value of 640 ISO) with extremely low read noise when it's needed.
There's on-sensor amplification and signal processing, then there's second stage amplification and processing - The second stage is your selectable ISO. My understanding is that on dual-ISO sensors there's two first stage amplification and signal processing systems.
It's all analog at this point and the old pre-amp and power amp practices still apply, amplifying in the pre-amp provides a lower noise level than amplifying later.
I am under the impression that while CCD sensors are analogue, CMOS are not. Not sure how that effects things.
They're both produced using CMOS tech, both are analogue at the coal-face, but have completely different theories of operation.
In a CCD an image is projected through a lens onto (in effect) a capacitor array (the photoactive region), causing each capacitor to accumulate an electric charge proportional to the light intensity at that location. Once the array has been exposed to the image, a control circuit causes each capacitor to transfer its contents to its neighbor. The last capacitor in the array dumps its charge into a charge amplifier, which converts the charge into a voltage. Conversion to digital is after this point.
CMOS sensors contain rows of photodiodes coupled with individual amplifiers to amplify the electric signal from the photodiodes. There's also an output preamp before the ADC. Canon have a good guide here: https://global.canon/en/technology/s_labo/light/003/05.html
So both are analogue at the very first stages and both then convert to digital after amplification and signal processing.
Thanks John, I've bookmarked that link to read later..🙂
Not entirely true. Many sensors, (CCD & CMOS), do absolutely no analogue amplification.
Drives me crazy! Nothing has changed in photography. All it is, is some who know not, teaching others what they do not know.
For one, it is not ISO; it is sensitivity. Film had a fixed, (or native) sensitivity value, (measured in ISO), and sensors also have a fixed (or native) sensitivity.
One cannot change this sensitivity, but one can change the exposure index, (EI), on the light-meter, both with film, and with a digital sensor. Just as with film, changing the EI does not let it more light, but lets in less light. This is later compensated by overdeveloping. (…And dual-range sensors also only have one “native” Sv. The dual nature has to do with how the signal is read, and not the saturation of the medium).
With film, one develops the film with chemicals, and overdevelops by using a combination of higher temperatures, stronger concentrations, and/or longer times. With digital sensors, one develops by converting the analogue signal into a digital representation, and overdevelops by using a combination of reading technique, (dual-range), analogue amplification (or gain, hardly done these days), and/or digital amplification, (the most common way these days), whether by in-camera processing, or in desktop software processing.
In both cases, the result is the same; reduced DR, and increased shot noise. There is nothing new under the sun. If one understood these concepts from the film era, they would not be hard to understand at all.
…And the “triangle” of TV/Av/Sv is not an “exposure of the medium” triangle; it is a “how to use your light-meter” triangle. An exposure triangle has Tv/Av/Lv variables.
Sv is not an exposure variable. EI is an exposure variable, and raising it results in a reduced exposure, & increased development, and with that, reduced DR & increased shot noise.
DR reduction is not due to increased sensitivity, but due to the medium not being fully saturated. Shot noise is not due to increased sensitivity, but due to fewer photons.
[EDIT]
Also, it is NOT I-S-O. It is not an abbreviation. It is not an acronym. It is a word, pronounced either as, “eye-so,” or as “isso.” Either pronunciation is correct, the second one often used in Latin-based languages, such as Spanish, French, etc.
[/EDIT]
Thanks for that further explanation Karim you have clarified my point that ISO should not be a term applied to digital imagery. What would really be better would be a 'sensitivity' control that is fully variable and with live view and EVF showing the effects in real time this could be a real boon to exposure and do away with the need for 'exposure compensation' as presently presented. (I am assuming that expcomp varies the 'ISO.')
With traditional photography we were accustomed to capturing images with set values for sensitivity but with digital this needn't be so although I doubt the ability to vary it so finely would be noticeable in the final result. However, cameras could become simpler if 'ISO' and expcomp were combined into one control which the operator could use to change the 'sensitivity.'
By perpetuating film terminology I personally feel manufacturers have muddied people's understanding. New Tech should use appropriate descriptions
And still I don't know the native 'ISO' (or sensitivity value 1) of my D7200. (Whether I need to or not is irrelevant, I just want to.)🙂
My point is that ISO, is what is used both to measure exposure index, and sensitivity, but, whether the medium is film or sensor, there is only one sensitivity, and an infinite (theoretically) number of possible exposure indices, and they are all measured in ISO.
To answer your question, the “native ISO,” a.k.a., the sensitivity, of a sensor, is typically the lowest Sv available outside any “extended” setting. That is to say, if your camera has an EI range listed as, “ISO 100-6400 normal, 64-12,800 extended,” then the Sv is ISO 100/21°.
…And, yes. Photography educators ought to stop calling it, “the ISO.” It is akin to an HVAC professional, calling the thermostat, “the degree Fahrenheit.”
First damn time I see a video on the topic that is actually correct!
Eeeh not that much, he says that higer gain means higer noise, but that's not true. Higer gain actually means LOWER noise in the final image given a certain exposure time, that's why you shoot with High ISO in the dark instead of using ISO 100 and boosting it in post production... if ISO 100 really had less noise nobody would be using anything beside it.
Mathias Elmeskog , I cannot say that he was actually correct, (he made a few mistakes), but more correct than most.
Paolo Bugnone , you are incorrect, (as is the guest). Noise is not caused by gain. It is caused by two things; few photons, and few electrons. amplification only makes the already existing noise be more apparent, but it was there the entire time.
Noise is the ratio of the read value, to the actual value. Say for example, from a certain area of the scene, hitting 1000 pixels, an average of 2,000 photons are incident on each pixel in a ¹/125 second period, but each pixel actually measures somewhere between 1,992, to 2,008 photons. The noise ratio is very low. Reduce the aperture by about five stops, such that an average of 64 photons are incident on each of 1,000 pixels in a ¹/125 second period, but each pixel actually measures between 54 to 72 photons. Now the noise ratio is high, (although the standard deviation is the same; ±8 photons). When this image is over-developed, (that is, gain is added), the noise becomes obvious.
This is called “shot noise,” (re: Schottky), and is a result of the Poisson distribution, (re: Poisson). Gain cannot mitigate this at all.
But what analogue gain does is add more noise, as electron amplification means adding electrons from a separate source, with its own Poisson distribution issues, or more shot noise. Digital gain adds nothing more.
Also, most modern DSCs do not do analogue gain, anyway. Most of them do digital amplification.
One wants to start with as much signal as possible, and do as little analogue gain as possible, to reduce shot noise as much as possible.
Karim, question would the noise be greater when the ratio is wider because the pixels are less saturated? Which is what I think your last paragraph is saying.
So in practice the exposure should be as close to maximum saturation as possible to minimise noise?
Which would presumably mean the sensitivity should be as close to the design sensitivity of the sensor as possible. (If we were allowed to know what that is?)
But when creating a correct exposure we know this may not always be possible due to other constraints (TV, AV etc)
Thanks
Correct. One ought to use TV to adjust motion blur, Av to adjust DoF, and use Lv, in as for as they can, the get the highlights of the subject as close to saturation as possible.
If that is not possible, then use digital amplification to increase the brightness. Most modern cameras use digital amplification for full stop EI adjustments. If your camera is not one of those, then set it to the lowest non-extension EI value, and amplify in software, (which is always digital amplification).
Preferably, use software which uses a 32-bit floating-point pixel pipeline. At least one major popular proprietary software vendor still uses an 8-bit integer pixel pipeline.
I know what shot noise is, and that gain doens't change shot noise.
You are however ignoring read noise. Gain is there to combat read noise by boosting the original signal over the floor of read noise of the ADC, I repeat... if higer gain meant more noise why would anybody shoot past ISO 100 when you could just recover it in post?
I heard the question the first time. I honestly do not know the answer. I shoot at ISO 100 because, if it ever comes to that —and in my line of photography, it hardly ever comes to that— I can, and have, recovered in post.
I have two answers to the question; ¹ it is easier, (not necessarily better), to have the camera do the noise reduction and overdevelopment, that doing it oneself with desktop software, and ² they do not know better.
…And analogue amplification does not necessarily get rid of read noise. It mitigates the issue to some extent, (as usual, nothing comes from nothing, and it adds more not noise), but there are other ways of doing that which have proved far better, which is why fewer and fewer systems are using analogue amplification.
Hi everyone. Hope you are all well. Yesterday I installed the new firmware 1.3 in my Canon R5. As I observed, the C-Log3 has much more light (in daylight). Some guy told me in his comment that it is "because in C-Log3 the ISO automatically rises to 800. That is the native ISO." Do you think so? In my case. I did not see any bad influence, noise in the image. The image is not affected in terms of sharpness. I suppose it will allow me to use a more wide-open aperture. What do you think about all of it? Thanks
C-Log3, or any Log, does not change the sensitivity of the camera. (Nothing you do can actually change the sensitivity). What does change is the camera engines interpretation of middle grey, and the relationship between black, (no light, zero charge), white, (maximum saturation, full well capacity), and middle grey.
A mathematical formula is devised which put all three of those points on the same curve. This means that middle grey is more left of centre of a normal histogram, meaning that more of you image values will be above middle grey, thus making your image look brighter.
Since middle grey is being re-evaluated, (or re-interpreted, really), it may be equivalent of adjusting one's exposure index, and thus one's development. One can then possibly say that the exposure index is now ISO 800/30°. Nevertheless, since sensitivity is based entirely on how much light it takes to saturate the sensor, be it silicon or halides, the sensitivity has not changed at all.
What has changed is the response curve. This is not new to digital photography. Chemical emulsions all had their unique response curves.