ISO Invariance: an experiment

[FingerPainter]

The camera sensor adds shot noise.

No, shot noise in is the captured light.

What do you think is the source of the noise?? Whatever comes through the lens is the signal. It is just the light from the scene you are trying to capture.

Light contains its own noise due to the quantum nature of light. Light consists of photons and the photons arrive randomly. This randomness appears as noise in the image. Even a perfect camera (which introduces no noise itself) will still give noisy images because of the quantum nature of light.

So, what comes through the lens is not just the signal, it contains noise as well.

That's at another level. Make no mistake, the noise we are talking about in photography is optical noise from the process itself and is not the quantum noise nature of light.

This is quite incorrect. Shot noise, the noise present in light itself due to the quantum nature of the generation of photons, is the most significant form of noise in most photographs, including those taken by OP in his experiment.

In the analogy of a radio or radar receiver, noise that comes through the antenna would be considered interference noise. This, from the dpr article on 'ISO-invariance' is the noise we are talking about that contributes to the dynamic range of the sensor.

Only when talking specifically about how ISO invariant a camera is, is the amount of noise added by the camera, generally after the variable gain stage, a significant contributor to the difference in relative invariance.

"Sony sensors tend to show very high Raw dynamic range due to low noise characteristics, but we wanted to take a closer look at exactly how little noise the camera itself contributes to images".

https://www.dpreview.com/articles/7450523388/sony-alpha-7r-ii-real-world-iso-invariance-study

 

[FingerPainter]

...

Here's an example from the popular (but frequently incorrect) site Cambridge in Colour:



In the diagram, the "ISO Speed" is shown as directly affecting (only) image noise. In the text, the claim is that "ISO speed affects image noise", and that "ISO Speed: controls the sensitivity of your camera's sensor to a given amount of light".

This site is correct to say that ISO speed affects image noise because this is in the context of the exposure triangle.

The context is given here. It mentions nothing about the model applying only to a constant exposure or a constant lightness. It merely say you can use many combinations of the three settings to achieve the same exposure, not that the model only applies if you do.

I think you're misunderstanding the article's use of the word exposure.

No, I think I understand the article's mis-use of the word ":exposure". One of the many problems with the "exposure triangle model" is this mis-use. In photography and sensitometry, "exposure" has a defined meaning. What the so-called "exposure triangle" describes is not exposure.

But if you're going to understand what the article is saying, you need to keep its meanings in mind.

The article doesn't define exposure, so you cannot be certain from the article what it means by "exposure". However, "exposure" does have an actual definition in a photographic context, which is the amount of light falling on the sensitive medium per unit area. Furthermore, the article says:

Achieving the correct exposure is a lot like collecting rain in a bucket. ... You just need to ensure you don't collect too little ("underexposed"), but that you also don't collect too much ("overexposed").

This seems to indicate that it correctly means an amount of light collected (though it doesn't deal with the "per unit area" part of the definition. Adjusting the ISO setting without changing the aperture or shutter doesn't change the amount of light collected.

In the first paragraph, exposure appears to be defined as how light or dark an image the camera captures, and ISO is one of the variables involved:

"A photograph's exposure determines how light or dark an image will appear when it's been captured by your camera.

This sentence is more or less correct. The only problem with it is that how light or dark an image will appear depends not only on on both the exposure but also the ISO setting (and any changes in lightness made in development).

This may be what led to your confusion below.

I'm not the one ho is confused.

In this article, the ISO setting is part of the exposure.

ISO cannot be part of the exposure. ISO is a contributor to the image lightness.

Remember that exposure means how light or dark the image is that is captured,

No it doesn't, and teh article ins't clear taht is what the article incorrectly means by "exposure".

and that this exposure is determined by shutter speed, aperture, and ISO.

How light or dark the image is is affected by aperture , shutter and ISO settings. The exposure is affected by the first two of these. The exposure therefore affects how light or dark the image is, and so does the ISO setting.

Changing how light or dark the image is in development is not mentioned in the article.

Believe it or not, this is determined by just three camera settings: aperture, ISO and shutter speed (the "exposure triangle")."

And it is true that the three camera settings are the three settings that determine how light or dark an image will be. Another parameter that also affects how light or dark an image will be (scene luminance) is not a camera setting.

If you change the ISO setting but do not change the shutter speed and/or aperture to achieve the same lightness or darkness of the image, then you are not using the exposure triangle as it is set forth in the article.

Excuse me, but where does the article say that changing one parameter without changng either of the other two is not how to use the model?

By the very concept of exposure the article is using

How do you know what concept of exposure the article is using? It is quite imprecise on the matter.

and the sentence you quoted that says "One can therefore use many combinations of the above three settings to achieve the same exposure."

And it's treu, However, the ISO setting doesn't affect the exposure. SO amog the many combinations of the three settings that achieve the same exposure are
{1/100, f/8, ISO 100}, {1/100, f/8, ISO 200}, {1/100, f/8, ISO 400} and {1/200, f/f.5, ISO 800}.

If you remember that exposure determines the brightness of the image,

Exposure is one of two parameters that determines the lightness of the SOOC image. The other parameter is ISO setting.

that ISO is part of the exposure,

It is impossible to correctly remember this.

and that you are making "trade-offs" with respect to motion blur, depth of field, and noise

Except you don't make trade-offs with respect to exposure or noise by changing only the ISO setting.

when you make your camera setting choices, it is evident that a change in any one parameter requires a balancing change in one or more of the other parameters, otherwise you won't have the same exposure

I can change ISO without changing exposure. I can change ISO and maintain the same image brightness by making a reciprocal change to exposure.

(that is, image brightness).

Exposure is not image brightness

We repeatedly see beginners who have erroneous ideas about what they should do with their ISO setting because they have read such a descripton of the "exposure triangle", and have not seen any reference to the necessity to use it only in the context of constant lightness.

I have never met a beginner who would assume that anything other than constant lightness was even feasible.

Meet more beginners.

They therefore erroneously conclude that they shouln't raise ISO from base when they cannot get adequate brightness at max aperture and slowest acceptable shutter, because doing so will increase noise over what they'll get at basse ISO. OP's experiment correctly shows the opposite.

The OP's experiment could have shown that shooting that still life at 1/30 second, f/8, and ISO 200 would have resulted in a better picture, and anyone following exposure triangle thinking would likely have done something very similar to that.

That's not what the experiment was testing.

This means that the important text is: one can therefore use many combinations of the above three settings. This means that changing ISO changes exposure.

First, increasing the ISO while increasing the shutter speed or narrowing the aperture doesn't produce the same exposure, so the text you cite is incorrect.

If it produces the same lightness/darkness of the image, then it produces the same exposure, at least in the context of how "exposure" is used in the article.

Why do you support promoting a misunderstanding of what exposure is?

That's a great question. The conclusion I've come to is that it is not a misconception; instead, it is simply one way, a very popular way, of using the term,

A popular misuse of a technical term is still a misconception.

and I've seen too many people made happier with their hobby by adopting the exposure triangle concept.

People who don't dive very deeply into photography will not be significantly harmed by leaning this false model. A significant portion of people who go beyond basic SOOC JPEG photography will make errors if they learn this model. The correct model is not much more complicated. Why not teach it instead?

I have to admit that it still seems foreign to me to think of ISO as a variable. Growing up on slide film, I continue to think of ISO as a condition, along with the light falling on the scene, that I have to take into consideration when choosing my shutter speed/aperture combination.

Treating either as being necessarily a fixed condition is suboptimal in digital photography

For each of my particular digital cameras, I have found through testing the one ISO setting I find best in terms of image quality, then I typically just use that ISO setting for everything.

Then some of your photos have more noise than is necessary.

Second. the diagram and text don't say that it applies only to a constant exposure, or more properly a constant lightness. It says that changing ISO changes noise, full stop.

As I read the article, maintaining constant lightness while varying the three camera settings is, indeed, what the article is talking about.

Please quote where the article restricts itself to being about constant lightness.

Addressed above, but if you try to think in the author's terms, I think you'll find it self evident.

I have actually tried, but found his "how much water you want" bit to be confusing.

Third, increasing ISO does not cause an increase in noise, even in the context of constant exposure. If you increase ISO while holding exposure constant, you will get the same or less noisiness.

But increasing ISO while holding what you call exposure

FYP

You didn't fix my post, you're just showing an unwillingness to acknowledge that the term exposure can have more than one meaning.

It has only one correct meaning in this context.

 

[bobn2]

Why do you support promoting a misunderstanding of what exposure is?

That's a great question. The conclusion I've come to is that it is not a misconception; instead, it is simply one way, a very popular way, of using the term, and I've seen too many people made happier with their hobby by adopting the exposure triangle concept. I have to admit that it still seems foreign to me to think of ISO as a variable. Growing up on slide film, I continue to think of ISO as a condition, along with the light falling on the scene, that I have to take into consideration when choosing my shutter speed/aperture combination. For each of my particular digital cameras, I have found through testing the one ISO setting I find best in terms of image quality, then I typically just use that ISO setting for everything.

It seems that what you are proposing is that a central concept of photography should have two different definitions, which can be used interchangeably without the author specifying which one is to be used. This seems to be a state of affairs bound to cause confusion.

The word 'exposure' in photography is well defined. It occurs, with the same meaning, in every technical document on photography, from film and sensor specifications, to the ISO speed standards and to the textbooks which define the subject. I would propose that in the context of photography any other usage is ill founded and will lead to misunderstanding and confusion.

My suspicion is that the author of Cambridge in Colour did not know the definition of exposure and as a result is spreading confusion and misunderstanding.

 

[RedFox88]

To my eyes it looks like you made the developed pic brighter than the regular exposure. Adjust both to have the same overall brightness next time.

 

[bobn2]

In this article, the ISO setting is part of the exposure.

ISO cannot be part of the exposure. ISO is a contributor to the image lightness.

Moreover, ISO is defined in terms of exposure. If it were a part of the exposure, the definition of ISO would be infinitely self-referential and therefore meaningless.

 

[gollywop]

...

Here's an example from the popular (but frequently incorrect) site Cambridge in Colour:



In the diagram, the "ISO Speed" is shown as directly affecting (only) image noise. In the text, the claim is that "ISO speed affects image noise", and that "ISO Speed: controls the sensitivity of your camera's sensor to a given amount of light".

This site is correct to say that ISO speed affects image noise because this is in the context of the exposure triangle.

The context is given here. It mentions nothing about the model applying only to a constant exposure or a constant lightness. It merely say you can use many combinations of the three settings to achieve the same exposure, not that the model only applies if you do.

I think you're misunderstanding the article's use of the word exposure.

No, I think I understand the article's mis-use of the word ":exposure". One of the many problems with the "exposure triangle model" is this mis-use. In photography and sensitometry, "exposure" has a defined meaning. What the so-called "exposure triangle" describes is not exposure.

But if you're going to understand what the article is saying, you need to keep its meanings in mind.

The article doesn't define exposure, so you cannot be certain from the article what it means by "exposure". However, "exposure" does have an actual definition in a photographic context, which is the amount of light falling on the sensitive medium per unit area. Furthermore, the article says:

Achieving the correct exposure is a lot like collecting rain in a bucket. ... You just need to ensure you don't collect too little ("underexposed"), but that you also don't collect too much ("overexposed").

This seems to indicate that it correctly means an amount of light collected (though it doesn't deal with the "per unit area" part of the definition. Adjusting the ISO setting without changing the aperture or shutter doesn't change the amount of light collected.

In the first paragraph, exposure appears to be defined as how light or dark an image the camera captures, and ISO is one of the variables involved:

"A photograph's exposure determines how light or dark an image will appear when it's been captured by your camera.

This sentence is more or less correct. The only problem with it is that how light or dark an image will appear depends not only on on both the exposure but also the ISO setting (and any changes in lightness made in development).

This may be what led to your confusion below.

I'm not the one ho is confused.

In this article, the ISO setting is part of the exposure.

ISO cannot be part of the exposure. ISO is a contributor to the image lightness.

Remember that exposure means how light or dark the image is that is captured,

No it doesn't, and teh article ins't clear taht is what the article incorrectly means by "exposure".

and that this exposure is determined by shutter speed, aperture, and ISO.

How light or dark the image is is affected by aperture , shutter and ISO settings. The exposure is affected by the first two of these. The exposure therefore affects how light or dark the image is, and so does the ISO setting.

Changing how light or dark the image is in development is not mentioned in the article.

Believe it or not, this is determined by just three camera settings: aperture, ISO and shutter speed (the "exposure triangle")."

And it is true that the three camera settings are the three settings that determine how light or dark an image will be. Another parameter that also affects how light or dark an image will be (scene luminance) is not a camera setting.

If you change the ISO setting but do not change the shutter speed and/or aperture to achieve the same lightness or darkness of the image, then you are not using the exposure triangle as it is set forth in the article.

Excuse me, but where does the article say that changing one parameter without changng either of the other two is not how to use the model?

By the very concept of exposure the article is using

How do you know what concept of exposure the article is using? It is quite imprecise on the matter.

and the sentence you quoted that says "One can therefore use many combinations of the above three settings to achieve the same exposure."

And it's treu, However, the ISO setting doesn't affect the exposure. SO amog the many combinations of the three settings that achieve the same exposure are
{1/100, f/8, ISO 100}, {1/100, f/8, ISO 200}, {1/100, f/8, ISO 400} and {1/200, f/f.5, ISO 800}.

If you remember that exposure determines the brightness of the image,

Exposure is one of two parameters that determines the lightness of the SOOC image. The other parameter is ISO setting.

that ISO is part of the exposure,

It is impossible to correctly remember this.

and that you are making "trade-offs" with respect to motion blur, depth of field, and noise

Except you don't make trade-offs with respect to exposure or noise by changing only the ISO setting.

when you make your camera setting choices, it is evident that a change in any one parameter requires a balancing change in one or more of the other parameters, otherwise you won't have the same exposure

I can change ISO without changing exposure. I can change ISO and maintain the same image brightness by making a reciprocal change to exposure.

(that is, image brightness).

Exposure is not image brightness

We repeatedly see beginners who have erroneous ideas about what they should do with their ISO setting because they have read such a descripton of the "exposure triangle", and have not seen any reference to the necessity to use it only in the context of constant lightness.

I have never met a beginner who would assume that anything other than constant lightness was even feasible.

Meet more beginners.

They therefore erroneously conclude that they shouln't raise ISO from base when they cannot get adequate brightness at max aperture and slowest acceptable shutter, because doing so will increase noise over what they'll get at basse ISO. OP's experiment correctly shows the opposite.

The OP's experiment could have shown that shooting that still life at 1/30 second, f/8, and ISO 200 would have resulted in a better picture, and anyone following exposure triangle thinking would likely have done something very similar to that.

That's not what the experiment was testing.

This means that the important text is: one can therefore use many combinations of the above three settings. This means that changing ISO changes exposure.

First, increasing the ISO while increasing the shutter speed or narrowing the aperture doesn't produce the same exposure, so the text you cite is incorrect.

If it produces the same lightness/darkness of the image, then it produces the same exposure, at least in the context of how "exposure" is used in the article.

Why do you support promoting a misunderstanding of what exposure is?

That's a great question. The conclusion I've come to is that it is not a misconception; instead, it is simply one way, a very popular way, of using the term,

A popular misuse of a technical term is still a misconception.

and I've seen too many people made happier with their hobby by adopting the exposure triangle concept.

People who don't dive very deeply into photography will not be significantly harmed by leaning this false model. A significant portion of people who go beyond basic SOOC JPEG photography will make errors if they learn this model. The correct model is not much more complicated. Why not teach it instead?

I have to admit that it still seems foreign to me to think of ISO as a variable. Growing up on slide film, I continue to think of ISO as a condition, along with the light falling on the scene, that I have to take into consideration when choosing my shutter speed/aperture combination.

Treating either as being necessarily a fixed condition is suboptimal in digital photography

For each of my particular digital cameras, I have found through testing the one ISO setting I find best in terms of image quality, then I typically just use that ISO setting for everything.

Then some of your photos have more noise than is necessary.

Second. the diagram and text don't say that it applies only to a constant exposure, or more properly a constant lightness. It says that changing ISO changes noise, full stop.

As I read the article, maintaining constant lightness while varying the three camera settings is, indeed, what the article is talking about.

Please quote where the article restricts itself to being about constant lightness.

Addressed above, but if you try to think in the author's terms, I think you'll find it self evident.

I have actually tried, but found his "how much water you want" bit to be confusing.

Third, increasing ISO does not cause an increase in noise, even in the context of constant exposure. If you increase ISO while holding exposure constant, you will get the same or less noisiness.

But increasing ISO while holding what you call exposure

FYP

You didn't fix my post, you're just showing an unwillingness to acknowledge that the term exposure can have more than one meaning.

It has only one correct meaning in this context.

I fear you are wasting your breath. This is about an even dozen times I have seen jrtrent led through this exact same discussion over the years. It is clear he has no interest in learning what is correct.

--
gollywop
I am not a moderator or an official of dpr. My views do not represent, or necessarily reflect, those of dpr.

http://g4.img-dpreview.com/D8A95C7DB3724EC094214B212FB1F2AF.jpg

 

[Tom Axford]

To my eyes it looks like you made the developed pic brighter than the regular exposure. Adjust both to have the same overall brightness next time.

That wasn't the point of this particular experiment.

The point was to compare an image made at ISO 6400 with an image made at identical settings except ISO 200 and then processed with +5.0 exposure compensation (because ISO 200 and ISO 6400 are 5 stops different). For a fully ISO invariant system, the images should be identical. The fact that they are not indicates that the system is not fully ISO invariant.

 

[mamallama]

When you start with a "large signal" scene like your test scene (brightly lit - no dark shadows), noise will not show up regardless of ISO setting. The signal overwhelms the noise regardless. That answers the questions you asked about seeing more noise in the first image.

I think you are confusing the amount of signal coming from the scene with the amount of signal captured by the sensor.

A scene can be very well lit, but if your shutter is fast enough and your aperture narrow enough, only a small signal is captured. The shot noise of the captured image depends on the amount of signal captured, not the amount of signal in the scene.

True, but capturing enough signal for the test was no problem.

What do you mean by "enough signal for the test"? The point of the test seems to have been to capture a low enough signal that there would be visible shot noise and to see what effect different ISO settings had on total noise.

. The test scene was bright with no dark shadows

The test image that didn't need boosting in development was taken at settings of

1/1000 f/8 ISO 6400

That indicates the scene was EV 10, which is half way between day;light and a normally lit interior room at night. That is not part particularly bright. The fact that there were no shadows indicates a lack of strong direct light. It does not indicate bright lighting.

thus it was bright with no dynamic range challenges.

Bright light produces scenes with a greater dynamic range than low light.

The camera sensor adds shot noise.

No, shot noise in is the captured light.

What do you think is the source of the noise??

The source of shot noise is random variation in the light that is due to the quantum nature of the process which emits photons from a light source. The timing and wavelengh of released photons is the product a bounded random process, which results in random variation in light colour and intensity, but that has a particular distribution.

Whatever comes through the lens is the signal.

Whatever comes through the lens is a signal and its associated shot noise. It is not the same as the signal which is captured by the cameras because a significant portion of the light that comes through the lens is not captured, for a number of reasons, including:

• Some of the light coming through the lens does not fall on the sensor,
• If the sensor has a CFA, (most do), then only some of the photons, those whose wavelength is in the range passed by the element of the CFA they strike, get through to the sensor.
• Of those photons that get through the CFA, only some manage to release an electron when they strike the sensor.

As a result, the signal that is captured by the sensor is weaker than the light that came through the lens, and as a result, the SNR of the captured signal is lower than the SNR of the light passing through the lens (the SNR for shot noise is the square root of the signal).

For OP's images, the shot noise in the captured signal is still more significant than the camera-added noise.

How do you know?

It is just the light from the scene you are trying to capture.

No, it is not just a signal. The light has noise in it.

The camera adds several other sorts of noise, which can be grouped together under the term "camera-added noise".

ISO invariant sensors are low noise and add little noise which sets them apart for regular sensors.

Not necessarily. And it is cameras that are ISO Invariant, not sensors. ISO Invariant cameras trend to be ones that add little noise after the variable gain stage. They may add plenty of noise before that point in the processing chain.

Not so.

What exactly is "not so"?

There are two types of gain involved in the total process: analog gain and digital gain.

The digital manipulation which may occur is not properly termed "gain". When I talk about the "variable gain stage" I am talking about analog gain. Most cameras add more noise before the analog variable gain stage than after it.

Analog gain occurs in first stage in the sensor,

What do you mean by "first stage"?

digital gain occurs after the A/D process and, in this case, during the RAW processing.

Digital "gain" is a red herring. I am not talking about it.

The advantage of the ISO invariant sensor is that this analog gain is achieved by a very low noise amplifier

No. ISO invariance generally has noting to do with noise being added by an amplifier. In those cases where noise is added by an amplifier, if it is proportional to the amount of amplification, treat it as noise added before amplification, and if it is independent of the amount of amplification, treat it as noise added after amplification.

so the S/N is reasonably high even if the camera exposure is below that normal sensors. After the first amplification the signal is then high enough that the subsequent amplifiers do not have to be low noise.

Amplifiers used in digital cameras have little to no effect on SNR. The amplify both the signal and the noise by the same amount so the SNR after gain is roughly the same as before gain.

And after the signal is digitized, digital amplification adds no additional noise.

True, but after digitization is not what I'm talking about.

The resulting S/N of the output (picture) is basically set by the first stage in the ISO invariant sensor which has a low noise base.

You are omitting the effect of noise added after the gain stage but before digitization.

This whole process is exactly analogous to the amplification chain of a radio or radar receiver. Only the first stage needs to be low noise because after the first stage, the signal is much larger than any noise that would be added in subsequent stages.

In fact, on some cameras, the noise added at or after the variable analog gain stage is still significantly large enough that it plays a significant role in the final total SNR.

That's not how the ISO invariant sensor cameras work. Maybe so earlier rudimentary cameras worked that way. Read this explanation:

https://www.dpreview.com/articles/7450523388/sony-alpha-7r-ii-real-world-iso-invariance-study

When it is not amplified, but the signal and any noise present before the variable gain stage are are both amplified, the total SNR rises as the amplification increases.

That's not how it works. Read the article above. Once digitized the amplification is a digital operation which does not increase the SNR.

Generally, cameras in which post-gain-stage noise is insignificant are the ones which approach ISO invariancy.

So both test pictures started out with high S/N ratio.

Both pictures started out with about 5 stops more noise than they would have in direct sunlight. That means their SNR was was only about 17.7% of what it would have been in daylight at base ISO.

What does this have to do with the test?

It has to do with your claim that the SNR was high.

The test only shows the principle of the ISO invariant sensor works. No one is disputing that.

That was my point.

Your point seems to be without foundation.

 

[richarddd]

"Increasing ISO increases the noise" is generally false if the exposure (shutter speed and f-number) is kept constant.

Increasing ISO is increasing the analog amplification of the signal before converting to a digital value. Rising the analog amplification is rising the noise. Once "digital" you can "amplify" without noise (as increasing exposure in LR). (Electronics !)

There are some analog amplifiers (type A = HiFi !) and some of them have the noise very-very low, but exist all the time. Even "duplicating" an analog signal will result in some noise added both to initial electron flow as to the second (cloned) flow.

So increasing the ISO is also increasing the noise. TRUE!

Look at the two images I posted. Are you seriously saying that you can see a lot more noise in the first image (the one with the much higher ISO)? Really????

The effect he is describing is true for A/D converters which sensors are. The noise difference is not noticeable in high S/N situations which your test case represents because the signal overwhelms the noise in all cases.

If the signal overwhelms the noise, then why should anyone care about just the noise?

The signal overwhelms the noise for all ISO settings in high signal-to-noise cases like the test scene. It's a bad test scene to illustrate the advantage of that sensor.

Under what circumstances does increasing the ISO, but not changing the shutter speed or aperture, decrease the signal to noise ratio? Under what circumstances does it increase SNR?

Who said anything about increase or decrease?? I said it's high from the start and stays high throughout.

Who said anything about increase? The person who wrote "So increasing the ISO is also increasing the noise."

However, I asked my question in order to get an answer, even you did not say anything about increase, rather than to get an evasion.

 

[Iliah Borg]

For OP's images, the shot noise in the captured signal is still more significant than the camera-added noise.

How do you know?

You can know that, too. Take two shots 5 stops apart and check the progression of SNR between the shots.

 

[mamallama]

To my eyes it looks like you made the developed pic brighter than the regular exposure. Adjust both to have the same overall brightness next time.

That wasn't the point of this particular experiment.

The point was to compare an image made at ISO 6400 with an image made at identical settings except ISO 200 and then processed with +5.0 exposure compensation (because ISO 200 and ISO 6400 are 5 stops different). For a fully ISO invariant system, the images should be identical. The fact that they are not indicates that the system is not fully ISO invariant.

Not making them identical could be caused by human limitations. Humans are not perfect and are incapable of matching colors and brightness precisely. Moreover what your eyes see is not what other people eyes see.