Who uses aperture priority?

[Clix Pix]

Long-time Aperture Priority fan here; it works best for most of the kinds of shooting I do, but yes, of course from time to time I will need to shift into Shutter Priority when the situation requires it. I rarely shoot in Manual these days but haven't forgotten how to do so!

I'm old enough to remember when with regard to film speed we referred to ASA ("A-ESS-A") prior to the advent of ISO (which I have always pronounced EYE-ESS-OH rather than "Eye-So." The latter makes me wince....

 

[Freida]

I was referring to this one I posted


But like I said I’m not really sure which is correct.

wow, that video is super revealing. So basically, I shouldn't sweat the ISO so much when shooting manual as I can fix that in post and not lose any details etc. Thats insane - one less thing to worry about

Really, knowing that ISO is post processing after the image is taken is actually going to simplify the process for me a lot. Really good to know, thank you so much for the video

 

[Apple fanboy]

wow, that video is super revealing. So basically, I shouldn't sweat the ISO so much when shooting manual as I can fix that in post and not lose any details etc. Thats insane - one less thing to worry about

Really, knowing that ISO is post processing after the image is taken is actually going to simplify the process for me a lot. Really good to know, thank you so much for the video

I’d recommend doing your own testing and see what works for you.
Best to use a tripod and cable release for consistency.

 

[Freida]

I’d recommend doing your own testing and see what works for you.
Best to use a tripod and cable release for consistency.

you mean to imply that what he said in the video is not correct? ie. ISO 100 is the same as ISO 25600 when it comes to the image itself and its just postprocessing and therefore it doesn't matter if you take it (RAW) with ISO 100 or ISO 25600 as the image is identical and in lightroom you can just adjust it and have identical result. No?

At least that is what I understood from the video

 

[Apple fanboy]

you mean to imply that what he said in the video is not correct? ie. ISO 100 is the same as ISO 25600 when it comes to the image itself and its just postprocessing and therefore it doesn't matter if you take it (RAW) with ISO 100 or ISO 25600 as the image is identical and in lightroom you can just adjust it and have identical result. No?

At least that is what I understood from the video

Like I said I’d test it. What level of noise is acceptable to me might not be the same to you.

 

[Freida]

Like I said I’d test it. What level of noise is acceptable to me might not be the same to you.

I'm actually not concered about noise but the revelation that ISO is POSTPROCESSING feature.
Ie. basically it doesn't matter what setting on ISO you use, the result is the same.

(sure, you will need to tweak it in post if your ISO is wrong but essentially its the same image). Thats what I meant.
So getting ISO wrong in the first place will not degrade the image quality at all.
Unless I misunderstood something

 

[jerwin]

back in the olden days of digital cameras:

entry level cameras wouldn't let you set ISO in one third increments, thought Auto mode could exploit it.
there was thought to be a quality decrease associated with using fractional ISO.

Let someone from seven years ago explain it:

Most Canon DSLRs only have analogue amplification circuits for the whole stops (100, 200, 400, 800 etc.), when you select one of the intermediate fractional values (520, 640 etc.) uses the closest analogue amplification stop and then uses digital multiplication to give the correct overall sensitivity. E.g. ISO640 is really ISO800 multiplied by 0.8, ISO500 is really ISO400 multiplied by 1.25 etc.

https://photo.stackexchange.com/questions/14141/why-cant-i-set-a-partial-iso-stop
I suspect that videographers requested a more linear ISO response. More convenient than a variable nd filter. Less obvious than changing the aperture or shutter angle.

Astrophotographers are too niche....

 

[Clix Pix]

I'm actually not concered about noise but the revelation that ISO is POSTPROCESSING feature.
Ie. basically it doesn't matter what setting on ISO you use, the result is the same.

(sure, you will need to tweak it in post if your ISO is wrong but essentially its the same image). Thats what I meant.
So getting ISO wrong in the first place will not degrade the image quality at all.
Unless I misunderstood something

I don't know if this will help you or perhaps confuse you more, but....

There is a huge difference between shooting something in using ISO 100 and in shooting the same image using, say ISO 1200. It's all about resolution of the image and how much detail is revealed or obscured and degraded. This information is recorded in the pixels at the time of shooting. Once lost, detail really cannot be recovered in post-processing, even if one has underexposed somewhat at the time of clicking the shutter. "Noise" is also much more evident at higher ISO levels. So, yes, it is important to get in the right ballpark straightaway when determining ISO as one is preparing to shoot the image. For the most detail and less noise, the lower the ISO the better. Shot in-camera at the time of exposure, not something to be dealt with later in post-processing. That said, modern technology has advanced things significantly in digital photography so that what we wouldn't have dreamed of using as an ISO ten or even five years ago actually IS doable now and the image can come out half-decently. That said, though, ideally for the highest quality image with best resolution and showing every detail sharply and clearly, with no or little accompanying "noise" (in digital photography it's referred to as "noise"; in film it was called "grain)," if that helps clarify anything. in essence, t is better to aim for the lowest ISO number possible at the time of adjusting the camera settings and preparing to make the shot under whatever circumstances one is shooting rather than depending on the idea of "fixing" it in post-processing. Getting it right in the camera in the first place can save a lot of time and frustration later on down the road.....!

 

[Freida]

I don't know if this will help you or perhaps confuse you more, but....

There is a huge difference between shooting something in using ISO 100 and in shooting the same image using, say ISO 1200. It's all about resolution of the image and how much detail is revealed or obscured and degraded. This information is recorded in the pixels at the time of shooting. Once lost, detail really cannot be recovered in post-processing, even if one has underexposed somewhat at the time of clicking the shutter. "Noise" is also much more evident at higher ISO levels. So, yes, it is important to get in the right ballpark straightaway when determining ISO as one is preparing to shoot the image. For the most detail and less noise, the lower the ISO the better. Shot in-camera at the time of exposure, not something to be dealt with later in post-processing. That said, modern technology has advanced things significantly in digital photography so that what we wouldn't have dreamed of using as an ISO ten or even five years ago actually IS doable now and the image can come out half-decently. That said, though, ideally for the highest quality image with best resolution and showing every detail sharply and clearly, with no or little accompanying "noise" (in digital photography it's referred to as "noise"; in film it was called "grain)," if that helps clarify anything. in essence, t is better to aim for the lowest ISO number possible at the time of adjusting the camera settings and preparing to make the shot under whatever circumstances one is shooting rather than depending on the idea of "fixing" it in post-processing. Getting it right in the camera in the first place can save a lot of time and frustration later on down the road.....!

Well, I know how ISO works so I understand what you are saying. The topic I was wondering about is the info that in the video about ISO work (its fake). The guy mentioned that ISO is not PHYSICALLY present at the time when the picture is taken. Its simply applied AFTER the picture is taken and its postprocessing thing. So he demonstrated taking identical picture with same settings except ISO and then when taking (i think it was 400 vs 25600 or something like that) and then taking those pictures in lightroom and boost the 400 higher and the images are identical in noise etc.

so technically, what I took from it was that I can shoot everything with ISO set to 100 and in post fix the ones that are too dark and I will get the same result.

Unless I misunderstood the video but thats what he implied and demonstrated there, right? Or am I totally off?

 

[Clix Pix]

Since I have not seen the video in question and was only going by my own many years of experience using cameras, first film and then digital, I cannot answer your questions related to what you apparently gleaned from that video. My advice? Forget that video (not everything on the internet is accurate and correct!). Just get out there and experiment for yourself: shoot the same subject (preferably one which offers some fine detail, such as in feathers or patterns and textures ) with 100 ISO and then 1200 and take that into the computer, open up your editing program and see what happens.....

 

[jerwin]

From DPReview's Invariance comparison tool

fiddle around with the widget yourself, if you like. I went looking for failure.
The EOS R isn't invariant. The Z7 is closer, but the nonpushed 6400 is superior.

If your camera is invariant enough that you can incorporate it into your technique, great. But I wouldn't rely on it without conducting a test. Hopefully my tulips will bloom and I can get some good macro shots.

 

[Freida]

Since I have not seen the video in question and was only going by my own many years of experience using cameras, first film and then digital, I cannot answer your questions related to what you apparently gleaned from that video. My advice? Forget that video (not everything on the internet is accurate and correct!). Just get out there and experiment for yourself: shoot the same subject (preferably one which offers some fine detail, such as in feathers or patterns and textures ) with 100 ISO and then 1200 and take that into the computer, open up your editing program and see what happens.....

It was this one

And yeah, until now I've set up ISO manually myself (or had it on auto with a cap) but this was interesting.
I will test it out though

 

[Clix Pix]

Aha!!! Bingo! I hadn't realized that this video (click bait?) is one that was tossed on to the internet by Tony Northrup, who is not someone I consider at all a credible source when it comes to discussions of anything related to photography. I'll refrain from adding my further thoughts about where the "fake" terminology should really be attached..... Freida, no wonder you're confused!

On NikonCafe, a photography site where I hang out, a discussion was started on this very topic because of the video and strong reactions to it. Someone provided a link by someone who is much more respected in the photography world (he primarily uses Nikon gear but also writes about lots of aspects of photography which are not camera-brand-specific).

Check out this article by Thom Hogan:

https://dslrbodies.com/cameras/camera-articles/image-sensors/is-iso-fake.html

 

[kallisti]

Aha!!! Bingo! I hadn't realized that this video (click bait?) is one that was tossed on to the internet by Tony Northrup, who is not someone I consider at all a credible source when it comes to discussions of anything related to photography. I'll refrain from adding my further thoughts about where the "fake" terminology should really be attached..... Freida, no wonder you're confused!

On NikonCafe, a photography site where I hang out, a discussion was started on this very topic because of the video and strong reactions to it. Someone provided a link by someone who is much more respected in the photography world (he primarily uses Nikon gear but also writes about lots of aspects of photography which are not camera-brand-specific).

Check out this article by Thom Hogan:

https://dslrbodies.com/cameras/camera-articles/image-sensors/is-iso-fake.html

I had thought about commenting at just how far afield this thread has drifted. But the question of ISO invariance is an interesting one (and probably should have prompted a new thread a page or two back ).

The article by Thom Hogan is a good read (and I think he is correct), though the topic is more complex than he goes into there (and he acknowledges this).

In general it is still good advice to "expose to the right" of the histogram. Meaning increase your exposure until you are just shy of blowing out highlights. It is easy to apply some underexposure in post (either globally or in specific areas of the image) which won't introduce noticeable artifacts. Exposing to the left (meaning underexposing the image) will always introduce noise/artifacts (depending on how underexposed the image is).

For the record, my Nikon D850 is great to ISO 800. ISO 64 is optimal. ISO 1600 is acceptable. ISO 3200 or beyond will result in a very significant degradation in sharpness/detail and introduce very obvious noise. Underexposing at a lower ISO and then increasing exposure in post doesn't really fix this. I haven't tested this enough in-depth (say by using RawDigger) to know if there is a break point where underexposing by a stop is better than increasing the ISO by a stop.

 

[robgendreau]

Using aperture priority can have a side benefit, in that since many use it with the widest possible apertures they may be in a sort of inadvertent way exposing to the right. I agree with kallisti it's still the way to go since getting more light is usually a good thing.

Rawdigger has been mentioned a couple of times, and it's a great program to have if one wants to dig around with this more. And so is FastRawViewer, by the same developer. I think both are worth getting just to support Mr Borg; his tutorials and tech info on both of the sites below are great resources (Mr Hogan in the article cited above consulted with him). Mr Borg is also involved with libraw, a raw converter. You can use it within Graphic Converter; it's in the preferences. It might deal with the issues Mr Hogan mentioned in reference to Raw Photo Processor (I don't know what engine that program uses; I don't think it's been updated in years). The difference in using floating point with examples is discussed here: https://www.libraw.org/articles/channel-noise-and-raw-converters.html

https://www.fastrawviewer.com

https://www.rawdigger.com

 

[smirking]

It doesn't change that anything other than I-S-O sounds wrong to me, though .

I feel your pain. I always try to pronounce it the right way, but it sounds so wrong.

 

[mmomega]

I shoot only manual and find myself very confused trying to use anything else.

I don’t think program users are wrong for doing so, but my brain doesn’t work that way.

This me, I learned quite a while back when digital cameras began coming on to the scene. I was rather young and unknowledgeable with them and when I would ask advice, it was very immediately snobbish responses like if you can't use your camera in manual then you shouldn't waste your or my time.
So this notion made me learn manual because I did want to be good.
Now I am comfortable with it but I have a very different approach when newer photographers come up and ask my opinion.


I am teaching my wife now, she has gone from using a Sony a6000 to now an a7rii.
How I get her MORE used to manual is. And this is strictly my method because 'set in my ways'

I normally go full wide-open on my lens to learn it. This will also give you the most light coming in. So on the Sony, front dial at your pointer finger - spin it until lens is wide open.
Dial on the upper rear is shutter speed. Set it to roughly the focal length of your lens, roughly. So if you are at 30mm then just set the shutter speed to 1/30th. The lower the shutter also gives more light.
Then set your ISO to Auto and take a lot of shots and review them.

With the Sony's this is , to me, a more easy way to understand the 3 features that control light and depth.

So if movement is blurred slightly in photos then increase your shutter speed slightly. Doing that will freeze action more but reduce light coming in and because light is reduce, auto ISO will cause the ISO to increase to make up for it.

Higher ISO can also add some graininess to the photo, but if your other 2 settings are set to let in the most light already the only option you have left is to raise ISO anyway.

Now you know the only way at this point, to take this exact photo in this place, you will have to introduce more light another way to keep the ISO lower. So you either turn on more lights in the room or get in to adding a flash.

So for my wife, this has helped her. Mileage may vary and every person has a different beginning method to teach.
It isn't about who is right or wrong but finding the method that works for the person learning.

 

[Strider64]

I set the minimum ISO and the maximum ISO (The Maximum Noise I personally can tolerate) on my Sony A9. I have it on Aperture mode and the Sony A9 has an ISO minimum Shutter Speed feature that adds a nice touch to the mix. Most digital cameras be it DSLRs or MILC cameras take nice pictures at higher ISOs and I like I said setting the maximum ISO is a personal preference. This type of setup is nice for sports and action, in my case I use it for birds-in-flight.

For those using a large lens having a good tripod and gimbal head is nice wildlife photography, for wildlife doesn't always do want you want at a given moment. It beats having to stand there and hold the camera and lens for long periods of time.

Here's my current setup for taking pictures of birds-in-flight:


An here's an example that I did last Saturday -


Not the greatest for it was cloudy and cold outside that day, but still pretty good in my opinion. Aperture mode isn't meant for every situation, but for fast moving objects in the great outdoors - I think it is. I know manual pretty well, but manual mode for fast moving objects would be a little tricky. I know it would be for me, because by I time I get the right settings for everything the wildlife will be gone or the conditions would had change enough that the photo wouldn't come out like I expected it to. I might go with Shutter Priority on a nice Sunny day in order to be able to take a clean ISO pictures. Though this is why it is nice to have different modes on these cameras.

 

[kallisti]

I'm actually not concered about noise but the revelation that ISO is POSTPROCESSING feature.
Ie. basically it doesn't matter what setting on ISO you use, the result is the same.

(sure, you will need to tweak it in post if your ISO is wrong but essentially its the same image). Thats what I meant.
So getting ISO wrong in the first place will not degrade the image quality at all.
Unless I misunderstood something

There is a dramatic loss in useful image data (and thus IQ) by significantly underexposing at a lower ISO compared to correctly exposing using a higher ISO.

Nikon D850 pics of a color chart. All light coming from a strobe, output constant between the pics. f/11, 1/250th sec. The first pic taken at ISO 3200, the second at ISO 100. For the first image below I made no changes in post. For the second I increased exposure by 5 stops in LR. While they may look similar below, the actual files are radically different.

Sharing the RawDigger histograms of each pic. The area under the curves is the pixel level data for each pic. Gaps in the histograms represent absence of data for that EV value. It should be quite obvious that the files are not equal regarding the data they contain. The ISO 3200 file contains significantly more data than the ISO 100 file. The histograms reflect the data present in the RAW files, not JPEGs.

Again, the light is exactly the same between the two images. The only difference was the ISO at the time of capture.

[Edit: my initial image of the RawDigger histogram for the ISO 100 pic didn't have the same Y-axis scaling as the ISO 3200 pic. But the results are the same--significant gapping in the ISO 100 image that reflects absence of pixel data for those EV values. This means there are abrupt rather than smooth tonal transitions for many areas--especially in the shadows. Since the data do not exist, it is impossible to recover.]

Since the histogram is a graph of the EV value recorded by each pixel, where did all the pixels go in the ISO 100 image? The answer lies in the following graph. Massive numbers of pixels recorded identical EV values in the shadow areas (because the pic was effectively 5 stops underexposed at capture). Even good sensors struggle with maintaining even tonal transitions in the shadows. Somewhat analogous to "blown out" highlights which truly can't be recovered because there is no data there to recover.

Effects of different ISO values are *not* something that happens as part of RAW conversion, it is happening at the time of image capture. Pixels are reacting to the amount of light they receive and in an underexposed image (regardless of whether it is underexposed because of aperture, shutter speed, or ISO) you will lose smooth tonal transitions because the sensor can't adequately distinguish between true tonal variations that are present in the subject. One reason why people say to "expose to the right" at the time of capture--meaning increase exposure until you are just shy of blowing out important highlights and then decrease exposure as needed in post.

 

[Freida]

Wow, this is awesome and again very informative. Thank you for that. So basically is better to have a touch brighter image than darker. Anyway, I love your answer and practical example. Kinda makes things clearer
Really appreciate your time - thanks

There is a dramatic loss in useful image data (and thus IQ) by significantly underexposing at a lower ISO compared to correctly exposing using a higher ISO.

Nikon D850 pics of a color chart. All light coming from a strobe, output constant between the pics. f/11, 1/250th sec. The first pic taken at ISO 3200, the second at ISO 100. For the first image below I made no changes in post. For the second I increased exposure by 5 stops in LR. While they may look similar below, the actual files are radically different.

Sharing the RawDigger histograms of each pic. The area under the curves is the pixel level data for each pic. Gaps in the histograms represent absence of data for that EV value. It should be quite obvious that the files are not equal regarding the data they contain. The ISO 3200 file contains significantly more data than the ISO 100 file. The histograms reflect the data present in the RAW files, not JPEGs.

Again, the light is exactly the same between the two images. The only difference was the ISO at the time of capture.

[Edit: my initial image of the RawDigger histogram for the ISO 100 pic didn't have the same Y-axis scaling as the ISO 3200 pic. But the results are the same--significant gapping in the ISO 100 image that reflects absence of pixel data for those EV values. This means there are abrupt rather than smooth tonal transitions for many areas--especially in the shadows. Since the data do not exist, it is impossible to recover.]

Since the histogram is a graph of the EV value recorded by each pixel, where did all the pixels go in the ISO 100 image? The answer lies in the following graph. Massive numbers of pixels recorded identical EV values in the shadow areas (because the pic was effectively 5 stops underexposed at capture). Even good sensors struggle with maintaining even tonal transitions in the shadows. Somewhat analogous to "blown out" highlights which truly can't be recovered because there is no data there to recover.

Effects of different ISO values are *not* something that happens as part of RAW conversion, it is happening at the time of image capture. Pixels are reacting to the amount of light they receive and in an underexposed image (regardless of whether it is underexposed because of aperture, shutter speed, or ISO) you will lose smooth tonal transitions because the sensor can't adequately distinguish between true tonal variations that are present in the subject. One reason why people say to "expose to the right" at the time of capture--meaning increase exposure until you are just shy of blowing out important highlights and then decrease exposure as needed in post.

 

[jerwin]

If you analyze a properly exposed ISO 100 photograph with rawdigger, what does the histogram look like?

 

[kallisti]

If you analyze a properly exposed ISO 100 photograph with rawdigger, what does the histogram look like?

I started a thread examining the RawDigger histograms from ISO 64 to ISO 12800:

https://forums.macrumors.com/threads/test-effect-of-iso-on-raw-histograms.2172907/

 

[jerwin]

I started a thread examining the RawDigger histograms from ISO 64 to ISO 12800:

https://forums.macrumors.com/threads/test-effect-of-iso-on-raw-histograms.2172907/

So a properly exposed image at ISO 64 has a solid histogram? It's not spiky?

 

[kallisti]

So a properly exposed image at ISO 64 has a solid histogram? It's not spiky?

It depends on the subject. It may very well have peaks and valleys (corresponding to the actual tonal values present in the subject) but it shouldn't have gaps. Gaps indicate lack of pixel data for that EV value. A large spike with gaps around it implies that for a range of EV values, all of the pixels registered the same. So the sensor thought there was only one tone present while the subject likely had variation around that EV value.

As a practical example, pretend that there are shadow areas in a subject that have detail in them. Reviewing your image, these areas look black. So you decide to boost the shadows in post to bring out the detail.

If the RAW data for the portion of the image you want to boost is relatively free of gaps, you will be able to see that detail as you boost the shadows because there are pixels that registered the different tonal values in that EV range.

On the other hand, if the RAW data for that portion of the image is composed of large gaps then there is no pixel data for that range of EV values. So boosting the shadows makes the black area look lighter, but it is still without detail (because there is no detail present in the RAW image).

Something similar happens with JPEGs. The reason the files are so much smaller is because they are throwing out sensor data. So if RawDigger worked on JPEG files, the histogram would be filled with gaps along the entire EV range. JPEG compression algorithms try to do this intelligently so you can't notice these gaps. But they are there. Which is why people suggest shooting in RAW rather than JPEG--starting out with more data lets you manipulate the image more in post without introducing obvious artifacts (like banding).

 

[jerwin]

so, tony northrup can't stop talking about iso invariance-- and perhaps fstoppers can't either.

he did mention Photons to Photos, though.

I suggest you look up the Read Noise to ISO plot that's specific to your camera, and make note of any spikes in the graph. (especially if you use a Canon) Might be useful in the field.