Setting to get the most OOC Jpeg

ChrisLumix said:

If those two shots were taken under identical conditions, but with those particular ISO values dialled in, then I can't answer your question.

Click to expand...

I already gave you the answer. The ambient conditions were identical, but the noisier image received less exposure.

ChrisLumix said:

Under identical conditions, the higher the ISO, the higher the noise.

Click to expand...

Under truly identical conditions (including identical exposure), the higher ISO will have the same amount of noise.

ChrisLumix said:

Look at ANY camera review (where they use the same scene to ramp up the ISO and show the effect) and tell me it's not so.

Click to expand...

If you check ANY camera review where they show higher noise at higher ISO, then you will see they are decreasing the exposure as they are increasing ISO. It's the exposure (less light) that causes an increase in noise, not the ISO setting. The ISO setting only controls the brightness of the final image, not the amount of light or the amount of noise.

Randy Benter said:

If you check ANY camera review where they show higher noise at higher ISO, then you will see they are decreasing the exposure as they are increasing ISO. It's the exposure (less light) that causes an increase in noise, not the ISO setting. The ISO setting only controls the brightness of the final image, not the amount of light or the amount of noise.

Click to expand...

You're just using sophistry. High ISO noise is a result of low light exposures. You wouldn't USE high ISOs otherwise. So it doesn't matter whether the noise comes from low light or high ISO, it's the same thing in the end, and you know it.

Your 'trick' comparison above, comes about because you gave the ISO 200 shot way under the correct exposure.

ChrisLumix said:

Randy Benter said:

If you check ANY camera review where they show higher noise at higher ISO, then you will see they are decreasing the exposure as they are increasing ISO. It's the exposure (less light) that causes an increase in noise, not the ISO setting. The ISO setting only controls the brightness of the final image, not the amount of light or the amount of noise.

Click to expand...

You're just using sophistry.

Click to expand...

Nothing could be further from the truth. I am genuinely trying to help you, the OP, and anyone else reading this thread.

ChrisLumix said:

High ISO noise is a result of low light exposures. You wouldn't USE high ISOs otherwise. So it doesn't matter whether the noise comes from low light or high ISO, it's the same thing in the end, and you know it.

Click to expand...

It's not the same thing and there is a clear benefit to understanding where the noise comes from. Understanding the cause of a problem is the biggest advantage one can have when attempting to solve it. I will explain in more detail below.

ChrisLumix said:

Your 'trick' comparison above, comes about because you gave the ISO 200 shot way under the correct exposure.

Click to expand...

It wasn't a trick. The point of giving the ISO 200 shot less exposure was not to deceive, but to demonstrate and educate. I clearly demonstrated that noise does not have the direct relation to ISO that you claimed. I showed that the image receiving less light will have more noise regardless of ISO.

The questions at hand are "what causes noise?" and "how do I minimize noise in an image?". The answers are simple: Visible noise in an image is increased when less light is gathered and the only way to minimize noise is to gather more light. Generally when ISO is decreased, exposure is increased, which creates the simple relationship you refer to, but it doesn't stop there.

Once one understands the goal is to gather more light, then there are a few different ways to achieve that and minimize noise:

• reduce the shutter speed to gather light for a longer period of time
• increase the size of the aperture to let more light in during exposure (smaller f-stop#)
• increase the surface area used to gather light
• use the highest amount of exposure possible without overexposing critical highlights
• add more light to the subject/scene (e.g. flash)

Someone who doesn't understand how these various factors contribute to noise might miss an opportunity to reduce noise and improve image quality.

The idea that high ISO causes noise is a common misconception. The idea that a larger sensor has less noise than a smaller sensor is equally misunderstood by many. A larger sensor will only have less noise if it is allowed to gather more total light than the smaller sensor. For example, a m43 camera with an f/1.8 lens will have less noise than a FF camera with an f/5.6 lens when both cameras are shot with the aperture wide-open (and equal SS).

I understand it can be hard to accept new information when it seems to go against what you previously believed. I don't think what I am saying here is really contradictory, but rather additive. I am taking a simple relationship between noise and ISO and expanding on it to help people understand why their high ISO images are noisier and that there are other ways to reduce noise. I hope that some people reading this will find it helpful.

MrFresno said:

Randy Benter wrote: I hope that some people reading this will find it helpful.

Click to expand...

Count me as one Randy!!!! Most helpful, and altogether simple to understand. I know my formula for shooting, which I have, for the most part, learned here. Now I can more easily explain it. Making things simple is the mark of genius! IMHO Of course,

Warm regards,

Randy Benter said:

If you check ANY camera review where they show higher noise at higher ISO, then you will see they are decreasing the exposure as they are increasing ISO. It's the exposure (less light) that causes an increase in noise, not the ISO setting. The ISO setting only controls the brightness of the final image, not the amount of light or the amount of noise.

Click to expand...

Of course! It's just like driving your car. When you press the accelerator down, it doesn't increase your speed, it just shortens the time it takes to get from A to B. Who on earth wouldn't understand that?!

morepix said:

Randy Benter said:

If you check ANY camera review where they show higher noise at higher ISO, then you will see they are decreasing the exposure as they are increasing ISO. It's the exposure (less light) that causes an increase in noise, not the ISO setting. The ISO setting only controls the brightness of the final image, not the amount of light or the amount of noise.

Click to expand...

Of course! It's just like driving your car. When you press the accelerator down, it doesn't increase your speed, it just shortens the time it takes to get from A to B. Who on earth wouldn't understand that?!

Click to expand...

I recognize your attempt to be facetious, but your analogy missed the mark. Pressing the accelerator down does increase your speed, but why? Do engines become more powerful when rubber is squished beneath a nearby shoe? Or does the accelerator trigger another reaction that actually increases speed, like increasing the amount of fuel? Understanding how something works is a good thing. More fuel equals more speed and it doesn't really matter if you press a pedal or roll a throttle, or dump it in with a squirt bottle. Light is fuel for our images.

Raw or JPG capture?

--
David
pbase.com/morepix

morepix said:

Raw or JPG capture?

Click to expand...

The images I posted were raw captures converted in LR6. ISO800 vs. underexposed ISO200 to demonstrate that it's not the high ISO image that is noisier, but the image that received less light. Another example that I didn't show here is that if you compare ISO800 to ISO200 and set the exposure the same then the noise will be the same.

Randy,

I've been living with noise for years -- ever since I renounced DSLRs seven years ago. And I've never properly bothered to find out where noise comes from, only how to try to get rid of it. Your claims pushed me into a closer look at DPR's two recent articles on sources of noise: shot noise and read noise. Having now perused these two articles I can see that you're at the very least half right -- more than I was willing to give you just a little while ago.

You claim that the too little light reaching the sensor is the problem. That's what the DPR articles called "shot noise." The second source, which you seem to dismiss, is electronic irregularity contributed by the camera. DPR has called that "read noise." I think that up to now I've been tacitly assuming that the camera is the whole culprit, which I now see is not quite true. How much of the noise in a particular image is of each of the two types would seem to depend on a whole lot of factors.

So now I can climb down from my high horse and admit that I've actually learned something! Thank you.

So let's take one more look at the situation. It would seem that the only way to reduce shot noise is to get more light to the sensor, and the only way to do that is with a brighter lens, a brighter light source, or a longer exposure. In any actual on-the-ground situation, we have only the lens we have, we often cannot control the light, and we have moving subjects to capture, limiting the length of our exposure. So up goes the ISO, up goes the electronic gain, and up goes the read noise. Ergo, from the perspective of that situation, the devil is in the increased ISO.

--
David
pbase.com/morepix

Ok. I do accept what you're saying, but at the same time I have to wonder why you're actually saying it. Here's why:

It's something of a truism to say "No-one ever died of cancer or MS". Technically it may be true, the sufferer dies of heart failure, or pneumonia, or an aneurism, or something equally not cancer or MS. However, that isn't to say that cancer and MS aren't also the prime underlying cause of death. So with ISO and light : we raise the one when the other decreases - they're symbiotically linked; there's no point looking at ISO in isolation, as a camera simply isn't calibrated to deliver a random ISO value for any given exposure. Most cameras' software will aim to keep the ISO as low as possible.

Take your massively underexposed ISO 200 shot above. Someone could easily do that e.g. in manual mode. But when they review the shot, they're not going to say "OMG, what a lot of noise!!" They're going to see first and foremost a very dark shot, at least 2 stops under exposed. They're going to realise their mistake and correct the exposure, and the resulting shot will have very little noise.

As for camera reviewers : they don't show the effect of different ISO by progressively under-exposing AFAIA, but by manually raising the ISO while maintaining correct exposure. I just did something similar on my LX100 : using P mode I set ISO first to 200 (f4 1/125) then to 25600 (f8 1/3200). According to what you say, as both shots were correctly exposed with the same amount of light reaching the sensor, they should have displayed equal amounts of noise. But no! Looking at 100%, the 25600 shot shows a very ugly degree of noise, far more than the 200 shot.

ChrisLumix said:

Take your massively underexposed ISO 200 shot above. Someone could easily do that e.g. in manual mode. But when they review the shot, they're not going to say "OMG, what a lot of noise!!" They're going to see first and foremost a very dark shot, at least 2 stops under exposed. They're going to realise their mistake and correct the exposure, and the resulting shot will have very little noise.

Click to expand...

The underexposed ISO200 shot was a demonstration used to prove the cause of noise is less light and not higher ISO. The problem we are discussing is not accidental underexposure; the problem is noisy images and how best to minimize them.

ChrisLumix said:

As for camera reviewers : they don't show the effect of different ISO by progressively under-exposing AFAIA, but by manually raising the ISO while maintaining correct exposure.

Click to expand...

If you look at the EXIF data for those test shots, you will see that each shot does in fact get progressively less exposure.

ChrisLumix said:

I just did something similar on my LX100 : using P mode

I set ISO first to 200 (f4 1/125) then to 25600 (f8 1/3200). According to what you say, as both shots were correctly exposed

Click to expand...

I didn't say if "both shots were correctly exposed".

I said "if both shots get the same exposure" (that means if they have the same aperture and shutter speed), then noise will be the same regardless of ISO setting.

ChrisLumix said:

with the same amount of light reaching the sensor, they should have displayed equal amounts of noise.

Click to expand...

The 2 shots in your test did not have the same amount of light reaching the sensor.

ChrisLumix said:

But no! Looking at 100%, the 25600 shot shows a very ugly degree of noise, far more than the 200 shot.

Click to expand...

Yes, because it received much less light due to the much shorter exposure time (shutter speed).

Randy Benter said:

MrFresno said:

I just did something similar on my LX100 : using P mode

I set ISO first to 200 (f4 1/125) then to 25600 (f8 1/3200). According to what you say, as both shots were correctly exposed

Click to expand...

I didn't say if "both shots were correctly exposed".

I said "if both shots get the same exposure" (that means if they have the same aperture and shutter speed), then noise will be the same regardless of ISO setting.

MrFresno said:

with the same amount of light reaching the sensor, they should have displayed equal amounts of noise.

Click to expand...

The 2 shots in your test did not have the same amount of light reaching the sensor.

MrFresno said:

But no! Looking at 100%, the 25600 shot shows a very ugly degree of noise, far more than the 200 shot.

Click to expand...

Yes, because it received much less light due to the much shorter exposure time (shutter speed).

Click to expand...

? Both shots DID have the same exposure. There are THREE factors in exposure in modern cameras : the aperture, the shutter speed, and the ISO. Had I increased the ISO to 25600, but left the other values the same (f4 1/125) the resulting shot would have been so massively over-exposed you wouldn't have been able to see any detail at all. You cannot compare different ISO values without also varying the other exposure values, or you wouldn't get a properly exposed comparison.

As I said above, camera manufacturers are well aware of this, and program their software to keep the ISO as low as possible.

ChrisLumix said:

Randy Benter said:

Yes, because it received much less light due to the much shorter exposure time (shutter speed).

Click to expand...

? Both shots DID have the same exposure. There are THREE factors in exposure in modern cameras : the aperture, the shutter speed, and the ISO.

Click to expand...

No, there are only 2 settings that impact exposure, aperture and shutter speed. ISO does not affect exposure as it does not increase or decrease the amount of light gathered. ISO is used to increase the image brightness after exposure by amplifying the image (analog or digital).

I am sure you understand how a larger aperture and longer shutter speed increase the amount of light gathered. It should be easy to understand that ISO can't affect the amount of light gathered during the exposure.

It seems that you've read about the "exposure triangle", but that is an overly simplified method of teaching basic photography. It is a misnomer and I wouldn't even use it for teaching beginners because they just need to be retaught exposure later.

I am writing all of this to help you understand (not to argue), please assume for a minute that I know what I'm talking about (because I do). Try to absorb it and if you have more questions, then just ask. It seems you are approaching this like an argument and you are attempting to prove me wrong. If that's your goal, then I will stop trying to help you. If you really want to learn then I will continue trying to assist.

morepix said:

Thank you.

Click to expand...

You're welcome. I am glad to help.

morepix said:

In any actual on-the-ground situation, we have only the lens we have, we often cannot control the light, and we have moving subjects to capture, limiting the length of our exposure. So up goes the ISO, up goes the electronic gain, and up goes the read noise. Ergo, from the perspective of that situation, the devil is in the increased ISO.

Click to expand...

Yes, essentially this is correct, though I would word it a bit differently to retain clarity of the goal. The devil is in the underexposure.

One key point to understand before I continue is that there are only 2 settings that impact exposure, aperture and shutter speed. ISO does not affect exposure as it does not increase or decrease the amount of light gathered. ISO is used to increase the image brightness after exposure by amplifying the image (analog or digital).

Using the list that I provided, we first want to try any method we can to increase the amount of light captured. Once you reach the limits of those options, then you've captured as much light as possible and achieved the goal. However, this might not result in a bright image, so we (or the camera if using auto) intentionally decide to underexpose the image and then brighten it after the exposure is complete.

Here is a real world example (as you described): I only have one camera with me so I can't change the sensor size. I cannot add light to the subject because I am in a museum. I have 2 lenses with me; I am using the one with the brightest aperture and I've set it wide-open. The only setting I have left to control light is shutter speed. According to the meter, the camera is recommending 1/15 as a shutter speed to achieve the metered exposure and an adequately bright image. Unfortunately I know that that I need a minimum shutter speed of 1/60 with this camera/lens to avoid camera shake and the museum doesn't allow tripods either.

I have 2 choices: 1) I can shoot at the metered SS of 1/15 and get a blurry image or 2) I can shoot the image 2-stops underexposed by using 1/60 SS. I can't really fix a blurry image after exposure, so the better choice (and the one my camera will make in auto modes) is to underexpose the image by 2-stops and then brighten it after the exposure is complete. By choosing to underexpose the image by 2-stops, I have gathered less light and the image will have more visible noise.

It doesn't really matter how the image is brightened following the underexposure. If done automatically, the camera will increase the ISO by 2-stops and after the exposure is complete, it will amplify the dark image (analog or digital) to compensate for the underexposure. If I manually set the underexposure and shot raw, then I would simply increase the exposure slider to +2 during conversion.

I was going to expand on read noise and signal-noise-ratio (SNR), but this post is long enough. Reading about SNR can really help you understand this.

Randy Benter said:

One key point to understand before I continue is that there are only 2 settings that impact exposure, aperture and shutter speed. ISO does not affect exposure ...

Click to expand...

If you insist on swimming upstream, that's your privilege. It's a matter of language you're quibbling about. The standard language refers to and elaborates on the "exposure triangle." I'm sure you're aware of that, and it's beyond me why you choose a different convention. You could call what everyone else calls a dog, a "zigbar" instead. But then you'd forever be justifying your idiosyncratic usage. And maybe no one would care.

Cheers,

morepix said:

Randy Benter said:

One key point to understand before I continue is that there are only 2 settings that impact exposure, aperture and shutter speed. ISO does not affect exposure ...

Click to expand...

If you insist on swimming upstream, that's your privilege. It's a matter of language you're quibbling about. The standard language refers to and elaborates on the "exposure triangle." I'm sure you're aware of that, and it's beyond me why you choose a different convention. You could call what everyone else calls a dog, a "zigbar" instead. But then you'd forever be justifying your idiosyncratic usage. And maybe no one would care.

Cheers,

Click to expand...

I understand that many people have learned the exposure triangle and I don't think it's important that about half the people on these forums refer to exposure as having three parts while the other half refer to it as having two parts. I somewhat agree with you that it in the end it doesn't really matter what you call it. The important part that should be understood by all is that while all 3 can contribute to image brightness, only 2 of the 3 control the amount of light gathered at the time the image is captured, which is the whole point of this discussion. ISO has zero impact on the amount of light, which is why it has zero impact on noise.

Randy Benter said:

morepix said:

... only 2 of the 3 control the amount of light gathered at the time the image is captured, which is the whole point of this discussion.ISO has zero impact on the amount of light ...

Click to expand...

Click to expand...

How did the amount of light become the whole point of this discussion? It seems to me that the discussion was / is / should be the brightness of the image. As I recall, the thread was about image quality, but it's easy to lose sight of that when we wander down these arcane byways.

Over and out.

--
David
pbase.com/morepix

morepix said:

Randy Benter said:

... only 2 of the 3 control the amount of light gathered at the time the image is captured, which is the whole point of this discussion.ISO has zero impact on the amount of light ...

Click to expand...

How did the amount of light become the whole point of this discussion? It seems to me that the discussion was / is / should be the brightness of the image. As I recall, the thread was about image quality, but it's easy to lose sight of that when we wander down these arcane byways.

Over and out.

Click to expand...

The topic of discussion is "to minimize noise" (from the OP). I demonstrated that the amount of light influences noise and the ISO setting does not. Gathering more light is the key to minimizing noise and you can't change the amount of light by changing ISO.

"Total light" is not an arcane byway, it is a basic principle of photography and the single biggest influence on image quality. It's why pros used MF and LF in the film era. It's why the majority of pros today use FF and MF. It's why people are willing to pay a premium for f/1.4 primes and f/2.8 zooms. AND...It's why a low ISO image with long exposure has less noise than a high ISO image with short exposure.

Randy Benter said:

No, there are only 2 settings that impact exposure, aperture and shutter speed. ISO does not affect exposure as it does not increase or decrease the amount of light gathered. ISO is used to increase the image brightness after exposure by amplifying the image (analog or digital).

I am sure you understand how a larger aperture and longer shutter speed increase the amount of light gathered. It should be easy to understand that ISO can't affect the amount of light gathered during the exposure.

It seems that you've read about the "exposure triangle", but that is an overly simplified method of teaching basic photography. It is a misnomer and I wouldn't even use it for teaching beginners because they just need to be retaught exposure later.

Click to expand...

Ok, I accept you know more about the nature of digital ISO. I think many of us may have been misled by years of shooting with 35mm film, where ASA (ISO) really was one of the factors in exposure : aperture, shutter speed, and the sensitivity of the emulsion. That emulsion had a particular 'noise' (grain) quality that was present in all conditions - good light, low light, etc.

To give an analogy, the triangle back then was like someone standing behind a screen, in front of which was a window with a blind. The shutter speed was how long the blind was opened for, the aperture was how much of the window was exposed by the blind, and the film speed was how transparent the screen was. All three factors affected how much light came through to the observer behind the screen.

When I came to digital I assumed (as I'm sure many did) that the ISO of a sensor was a variable sensitivity, much like being able to decide shot by shot what film you were shooting with. From what you say about amplification, it seems this is a false assumption, that sensors are not even remotely like film. I'm sure you appreciate how easy it was to fall into this kind of assumption without sufficient knowledge?

ChrisLumix said:

When I came to digital I assumed (as I'm sure many did) that the ISO of a sensor was a variable sensitivity, much like being able to decide shot by shot what film you were shooting with. From what you say about amplification, it seems this is a false assumption, that sensors are not even remotely like film. I'm sure you appreciate how easy it was to fall into this kind of assumption without sufficient knowledge?

Click to expand...

Yes, I appreciate and completely understand that. You are correct that sensor sensitivity doesn't increase when ISO is raised. When I choose a faster shutter speed and higher ISO, the sensitivity doesn't increase and I am gathering less light, so I capture a darker image. Immediately after capture, the underexposed image will simply be amplified by an amount relative to the ISO setting. Thanks for your reply.