1Dx M2 Sensor Resolution - Back of envelop estimate

[raptor3x]

tell me please

From an earlier post:

What's happening is that you can break the read noise down into two components: upstream of the amplifier and downstream of the amplifier (it should also be noted that some cameras use 2-stage amplifiers but the same approach is easily extensible to those as well). The resulting output referred noise becomes something like

No^2 = C^2*(N_us)^2 + (N_ds)^2

Where No is the output referred noise, N_us is the actual read noise of the upstream components, N_ds is the contribution of the downstream components, and C is the gain. To get the input referred noise, or how many electrons the read noise is equivalent to at the current gain setting, we simply use the relationship

Ni = (1/C)*No

where Ni is the input referred noise, which gives us

Ni^2 = (N_us)^2 + (1/C)^2*(N_ds)^2

Looking at both of these forms, you can see that as the gain becomes large, the relative contribution of the downstream noise becomes small; however, looking at the input referred noise should make it clear why you get an essentially constant read noise at high ISO in that table. The noise coming from the electronics doesn't change at all with ISO, but both their relative and absolute contributions to the final image absolutely changes.

1. you are not Neuro
2. do you agree that noise increase with higher iso as Neuro claims?

Noise absolutely does increase with higher ISO, not just noise-to-signal ratio. Do you think signal increases with higher ISO?

 

[rs]

I know one factor which definitely increases noise - posts my Mikael.

 

[Loswr]

Noise absolutely does increase with higher ISO...

Quick approximation of the tally. I stated that higher ISO means more noise. You and at least five others in this thread have stated the same. There have been no less than three separate paired image examples demonstrating more noise at higher ISO. Emil Martinec, Roger Clark, and Norman Koren – notable experts in the field – all state that noise increases with ISO. Eric Fossum – another expert in the field – first had his statement plagiarized but altered to be incorrect by 'filluppa', then answered only one part of a poorly-phrased bipartite question, and 'filluppa' misinterpreted his answer...importantly, Fossum did not disagree with the premise that noise increases with ISO/gain (and no doubt if asked that question directly, he'd agree). Even Mikael Residal stated that noise is higher at ISO 800 than at ISO 400 ('filluppa', if you're wondering who this unimportant Mikael dude is, find yourself a mirror).

Apparently lots of people, from CR forum members to digital signal experts to a dude named Mikael, all state that noise increases with ISO. But 'filluppa' says it doesn't. But then, I addressed that way back in post #84.

...he still believes he is correct, and will persist in doing so regardless of all evidence to the contrary.

 

[takesome1]

this is so interesting, I can not get a sensible answer from Neuro but a number of other people respond in his place, and none is about noise to signal ratio and Neuros claims that noise increase with higher iso
ps I know these people you are referring to
some are good friends since 2001
so Neuro if we take Canon as one example, where does the noise starts accelerating regarding noise

Now filluppa

In Neuro's last response he answered you. He Stated " I stated that higher ISO means more noise."

Where do we go from here?

 

[takesome1]

this is so interesting, I can not get a sensible answer from Neuro but a number of other people respond in his place, and none is about noise to signal ratio and Neuros claims that noise increase with higher iso
ps I know these people you are referring to
some are good friends since 2001
so Neuro if we take Canon as one example, where does the noise starts accelerating regarding noise

Now filluppa

In Neuro's last response he answered you. He Stated " I stated that higher ISO means more noise."

Where do we go from here?

tell me what you want
are you one of them who thinking that noise increase with higher iso?

Explain this to me.
What happens to the noise, data and signal when it is amplified by increasing ISO?
What does this amplification do?

 

[Loswr]

Understand?
ProbablyDefinitely not.

There, fixed that for you.

In Neuro's last response he answered you. He Stated " I stated that higher ISO means more noise."
Where do we go from here?

Oh that's quite obvious, in fact 'filluppa' preemptively addressed it. He doesn't like the truth – a truth also stated by several of his good friends since 2001 (and some goofball named Mikael) – so it's not a sensible answer.

Quite pathetic, but completely unsurprising. :

 

[takesome1]

the question was if noise increase with higher iso as Neuro claim
you have great difficulty keeping track of what's being discussed

Not all, were talking about higher ISO's.
So when I am adjusting the ISO how does that affect the signal, what happens to the noise and data.

Or explain why this doesn't apply.

Explain this to me.
What happens to the noise, data and signal when it is amplified by increasing ISO?
What does this amplification do?

 

[takesome1]

you are again name dropping, and it
doesn't make your answer more credible
so I ask you ones more, are noise increasing with higher iso?

Neuro said it did in post 241, go back and read it. Neuro wrote "I stated that higher ISO means more noise."

 

[3kramd5]

, and none is about noise to signal ratio and Neuros claims that noise increase with higher iso

I think I've figured it out: you're only reading your own posts. There is no other explanation for the above notion that the question hasn't been answered.

 

[Loswr]

, and none is about noise to signal ratio and Neuros claims that noise increase with higher iso

I think I've figured it out: you're only reading your own posts. There is no other explanation for the above notion that the question hast been answered.

I disagree...there are many other possible explanations, among which you're is likely the most flattering.

 

[takesome1]

who is stated what
neuro is not answering as usual
where are you Neuro who stated that that noise increased with higher iso



back to the issue

Overall noise is generally photon shot noise plus the losses/errors in the camera (electronic noise). Above ~ISO800 the losses in the camera are almost perfectly constant, so they neither increase or decrease. Then the determining factor for overall image signal/noise ratio is purely determined by the scene illumination, shutter speed and aperture. You can quite easily see this if you shoot a flat (low contrast) scene at ISO800-1600-3200 using the same camera settings. Make them identical in brightness by pushing the 800 file and pulling the 3200 file so that they match the 1600 file. They will be identical in signal to noise ratio. The only difference between the images will be in the white-clipping point, if you have something in the scene that goes "brighter than white" in the rendered output.

The reason a CR2 file increases in size as you increase ISO (even with identical settings!) is that the file compression Canon uses for their raws is more efficient with smaller contrasts. If you look at the signal [1,2,1,2,1,2] a CR2 will compress that to a fewer amount of bits than the signal [2,4,2,4,2,4] even though the actual content in the two signals are actually identical. You can transform signal 1 into signal 2 just by multiplying all numbers with "2". That means the actual content is functionally identical, but their compressed form might differ in size. And their signal to noise ratios are the same.

Who stated what; Neuro did. He answered your question. Probably not directly to you but he stated his belief. Go back and read it.

Now, back to the amplified signal.
Why is it amplified?

 

[takesome1]

, and none is about noise to signal ratio and Neuros claims that noise increase with higher iso

I think I've figured it out: you're only reading your own posts. There is no other explanation for the above notion that the question hast been answered.

I disagree...there are many other possible explanations, among which you're is likely the most flattering.

you disagree, tell me Neuro


back to the issue

Overall noise is generally photon shot noise plus the losses/errors in the camera (electronic noise). Above ~ISO800 the losses in the camera are almost perfectly constant, so they neither increase or decrease. Then the determining factor for overall image signal/noise ratio is purely determined by the scene illumination, shutter speed and aperture. You can quite easily see this if you shoot a flat (low contrast) scene at ISO800-1600-3200 using the same camera settings. Make them identical in brightness by pushing the 800 file and pulling the 3200 file so that they match the 1600 file. They will be identical in signal to noise ratio. The only difference between the images will be in the white-clipping point, if you have something in the scene that goes "brighter than white" in the rendered output.

The reason a CR2 file increases in size as you increase ISO (even with identical settings!) is that the file compression Canon uses for their raws is more efficient with smaller contrasts. If you look at the signal [1,2,1,2,1,2] a CR2 will compress that to a fewer amount of bits than the signal [2,4,2,4,2,4] even though the actual content in the two signals are actually identical. You can transform signal 1 into signal 2 just by multiplying all numbers with "2". That means the actual content is functionally identical, but their compressed form might differ in size. And their signal to noise ratios are the same.

Now, back to the amplified signal.
Why is it amplified?

 

[privatebydesign]

who is stated what, As I can see Neuro is hiding behind a lot of Neuro fans,
neuro is not answering as usual
where are you Neuro who stated that that noise increased with higher iso



so back to the issue and my answer 50

Overall noise is generally photon shot noise plus the losses/errors in the camera (electronic noise). Above ~ISO800 the losses in the camera are almost perfectly constant, so they neither increase or decrease. Then the determining factor for overall image signal/noise ratio is purely determined by the scene illumination, shutter speed and aperture. You can quite easily see this if you shoot a flat (low contrast) scene at ISO800-1600-3200 using the same camera settings. Make them identical in brightness by pushing the 800 file and pulling the 3200 file so that they match the 1600 file. They will be identical in signal to noise ratio. The only difference between the images will be in the white-clipping point, if you have something in the scene that goes "brighter than white" in the rendered output.

The reason a CR2 file increases in size as you increase ISO (even with identical settings!) is that the file compression Canon uses for their raws is more efficient with smaller contrasts. If you look at the signal [1,2,1,2,1,2] a CR2 will compress that to a fewer amount of bits than the signal [2,4,2,4,2,4] even though the actual content in the two signals are actually identical. You can transform signal 1 into signal 2 just by multiplying all numbers with "2". That means the actual content is functionally identical, but their compressed form might differ in size. And their signal to noise ratios are the same.

You obviously haven't actually done that, because it is not true.

 

[3kramd5]

who is stated what, As I can see Neuro is hiding behind a lot of Neuro fans,
neuro is not answering as usual
where are you Neuro who stated that that noise increased with higher iso



so back to the issue and my answer 50

Overall noise is generally photon shot noise plus the losses/errors in the camera (electronic noise). Above ~ISO800 the losses in the camera are almost perfectly constant, so they neither increase or decrease. Then the determining factor for overall image signal/noise ratio is purely determined by the scene illumination, shutter speed and aperture. You can quite easily see this if you shoot a flat (low contrast) scene at ISO800-1600-3200 using the same camera settings. Make them identical in brightness by pushing the 800 file and pulling the 3200 file so that they match the 1600 file. They will be identical in signal to noise ratio. The only difference between the images will be in the white-clipping point, if you have something in the scene that goes "brighter than white" in the rendered output.

The reason a CR2 file increases in size as you increase ISO (even with identical settings!) is that the file compression Canon uses for their raws is more efficient with smaller contrasts. If you look at the signal [1,2,1,2,1,2] a CR2 will compress that to a fewer amount of bits than the signal [2,4,2,4,2,4] even though the actual content in the two signals are actually identical. You can transform signal 1 into signal 2 just by multiplying all numbers with "2". That means the actual content is functionally identical, but their compressed form might differ in size. And their signal to noise ratios are the same.

You obviously haven't actually done that, because it is not true.

not true?
come again

Quick question: were it true, why did dpreview recently feel the need to introduce a metric which attempts to quantify how well pushing in post replicates amplification in camera, and why don't all cameras perform equally at said metric?

As you increase sensitivity you are indeed largely shot noise limited, so the higher you go, the more correct the statement. 800 isn't necesarily high enough.

 

[scyrene]

Has he been deleted?

 

[takesome1]

Has he been deleted?

Not only him but apparently all his posts

 

[takesome1]

its sad, because he had a point with noise and signal ratio and that noise doesent increase with higher iso

Gosh, why couldn't you have made your account just a few days earlier so you could have helped him with his point.

 

[Northstar]

its sad, because he had a point with noise and signal ratio and that noise doesent increase with higher iso

Gosh, why couldn't you have made your account just a few days earlier so you could have helped him with his point.

He has been booted from CR twice in one day. First as filluppa, then as Forsberg.

 

[rs]

its sad, because he had a point with noise and signal ratio and that noise doesent increase with higher iso

Gosh, why couldn't you have made your account just a few days earlier so you could have helped him with his point.

He has been booted from CR twice in one day. First as filluppa, then as Forsberg.

He'll be back

He might lie low for a bit, but Neuro will say something that makes complete sense which irks Mikael into appearing as another one of his many charecters.

In the meantime, we should all have enough time to stock up on popcorn.

 

[takesome1]

its sad, because he had a point with noise and signal ratio and that noise doesent increase with higher iso

Gosh, why couldn't you have made your account just a few days earlier so you could have helped him with his point.

He has been booted from CR twice in one day. First as filluppa, then as Forsberg.

He was back with at least a third name and was booted almost instantly.
Who knows, he is persistent. He could have been booted a dozen times and we just don't know.