fredmiranda.com Login
This topic is in the 2015 archive. You can read it and like its posts here. It is read only.

Archive 2015No DR improvement = no 5Ds for me

  
 
RustyBug
Offline
Details
p.11 #1No DR improvement = no 5Ds for me
takowasa wrote:
but the electronic noise can vary quite a bit


This was part of my concern @ entry consumer vs. semi-pro / pro
Feb 08, 2015 at 06:18 PM
takowasa
Offline
Details
p.11 #2No DR improvement = no 5Ds for me
takowasa wrote:
but the electronic noise can vary quite a bit


RustyBug wrote:
This was part of my concern @ entry consumer vs. semi-pro / pro


Variations in electronic noise have nothing to do with build quality of the camera body.


RustyBug wrote:
5D, 6D & 7D are (imo) same grade of build ... just different feature sets. Different from XXD or XXXD series.


Again, all non-sequiturs with regards to sensor tech.
Feb 08, 2015 at 06:30 PM
RustyBug
Offline
Details
p.11 #3No DR improvement = no 5Ds for me
takowasa wrote:
takowasa wrotebut the electronic noise can vary quite a bit
RustyBug wroteThis was part of my concern @ entry consumer vs. semi-pro / pro
Variations in electronic noise have nothing to do with build quality of the camera body.

RustyBug wrote5D, 6D & 7D are (imo) same grade of build ... just different feature sets. Different from XXD or XXXD series.
Again, all non-sequiturs with regards to sensor tech.


Honest question ... nothing else in the design influences noise (power, heat, etc.)

Feb 08, 2015 at 06:59 PM
takowasa
Offline
Details
p.11 #4No DR improvement = no 5Ds for me
takowasa wrote:
Again, all non-sequiturs with regards to sensor tech.


RustyBug wrote:
Honest question ... nothing else in the design influences noise (power, heat, etc.)


There are other factors in noise besides the quantum efficiency and electronic noise which can be important in specific circumstances (e.g. thermal noise in long exposures), but nothing systematic that would differentiate sensors in "pro" bodies vs "consumer" bodies.
Feb 08, 2015 at 09:02 PM
RustyBug
Offline
Details
p.11 #5No DR improvement = no 5Ds for me
Okay, setting that aside ...

snapsy wrote:
at base ISO the D3s has 12EV of DR vs the D800's 14.3EV but at ISO 25,600 the D3s has 8EV of DR vs the D800's 7EV.


Are these actually same base ISO?

Base ISO for D3s @ ISO 200
Base ISO for D800 @ ISO 100

IF they were of the same base, how would that change the relationship?

Edited on Feb 08, 2015 at 10:15 PM · View previous versions
Feb 08, 2015 at 10:01 PM
snapsy
Offline
Details
p.11 #6No DR improvement = no 5Ds for me
RustyBug wrote:
Okay, setting that aside ...

snapsy wroteat base ISO the D3s has 12EV of DR vs the D800's 14.3EV but at ISO 25,600 the D3s has 8EV of DR vs the D800's 7EV.
Are these actually same base ISO?

Base ISO for D3s @ ISO 200
Base ISO for D800 @ ISO 100


The D3s's base ISO is 200 but its DR for both 100 and 200 are identical. ISO 100 is implemented as a software pull.
Feb 08, 2015 at 10:08 PM
takowasa
Offline
Details
p.11 #7No DR improvement = no 5Ds for me
snapsy wrote:
at base ISO the D3s has 12EV of DR vs the D800's 14.3EV but at ISO 25,600 the D3s has 8EV of DR vs the D800's 7EV.


RustyBug wrote:
Are these actually same base ISO?

Base ISO for D3s @ ISO 200
Base ISO for D800 @ ISO 100


There's a lot of confusion about what role the ISO setting plays in digital photography. The reason the base ISO of the D800 is one stop lower than the D3s is because the D800 sensor can absorb twice the light as the D3s sensor.
Feb 08, 2015 at 10:14 PM
RustyBug
Offline
Details
p.11 #8No DR improvement = no 5Ds for me
takowasa wrote:

the D800 sensor can absorb twice the light as the D3s sensor.


Okay ... how is that possible and yet it be the same ISO 200. It would seem that it either is ISO 100 or it is ISO 200. To state that it can absorb twice as much light, but it isn't really ISO 100 ... starting to go here.

I tend to take things pretty straightforward physics ... please explain how it is, but isn't, but is ... I'm starting feel like this is going in a giant circle.





Edited on Feb 08, 2015 at 10:21 PM · View previous versions
Feb 08, 2015 at 10:17 PM
snapsy
Offline
Details
p.11 #9No DR improvement = no 5Ds for me
takowasa wrote:
snapsy wroteat base ISO the D3s has 12EV of DR vs the D800's 14.3EV but at ISO 25,600 the D3s has 8EV of DR vs the D800's 7EV.
RustyBug wroteAre these actually same base ISO?

Base ISO for D3s @ ISO 200
Base ISO for D800 @ ISO 100
There's a lot of confusion about what role the ISO setting plays in digital photography. The reason the base ISO of the D800 is one stop lower than the D3s is because the D800 sensor can absorb twice the light as the D3s sensor.


It's the combination of FWC and QE that determines base ISO. In this case the QE of both are the same so in this instance it's correct to say the D800 has 1EV lower base ISO due to twice the aggregate FWC. But both should be quoted when evaluating how base ISO is arrived at. As an interesting side note, starting with the D4 sensor the FWC is effectively doubled through the addition of selectable FD capacitor switch on each pixel, which brought the D4's base ISO down to 100.
Feb 08, 2015 at 10:21 PM
takowasa
Offline
Details
p.11 #10No DR improvement = no 5Ds for me
takowasa wrote:

the D800 sensor can absorb twice the light as the D3s sensor.


RustyBug wrote:
Okay ... how is that possible?


Newer design. The Canon 5D3/6D sensors can absorb twice the light as the original Canon 5D sensor, but they also have twice the quantum efficiency (which means they record twice as much light for a given exposure), so the two cancel each other out and the base ISO remains the same.

While the quantum efficiency of the D3s and D800 is the same, the D800 pixels have double the relative saturation (i.e. three D800 pixels can hold twice as much light as one D3s pixel).
Feb 08, 2015 at 10:24 PM
snapsy
Offline
Details
p.11 #11No DR improvement = no 5Ds for me
RustyBug wrote:
takowasa wrotethe D800 sensor can absorb twice the light as the D3s sensor.
Okay ... how is that possible and yet it be the same ISO 200. It would seem that it either is ISO 100 or it is ISO 200. To state that it can absorb twice as much light, but it isn't really ISO 100 ... starting to go here.

I tend to take things pretty straightforward physics ... please explain how it is, but isn't, but is ... I'm starting feel like this is going in a giant circle.


Think of a sponge - FWC establishes how much total water the sponge can hold and QE establishes how fast the sponge absorbs water. So for example if you double the QE but keep FWC the same then the base ISO goes up by 1EV, since for a fixed unit of exposure time the sponge will have absorbed twice as much water.. If you double the FWC but keep QE the same then the base ISO does down by 1EV; the sponge is absorbing at the same rate but can now hold twice as much water and thus it takes 2x the exposure time to fill it up, thus its rated sensitivity goes down by half. If you double both QE and FWC then base ISO stays the same.
Feb 08, 2015 at 10:26 PM
takowasa
Offline
Details
p.11 #12No DR improvement = no 5Ds for me
snapsy wrote:
Think of a sponge - FWC establishes how much total water the sponge can hold and QE establishes how fast the sponge absorbs water. So for example if you double the QE but keep FWC the same then the base ISO goes up by 1EV, since for a fixed unit of exposure time the sponge will have absorbed twice as much water.. If you double the FWC but keep QE the same then the base ISO does down by 1EV; the sponge is absorbing at the same rate but can now hold twice as much water and thus it takes
...Show more →

That analogy is full of holes.
Feb 08, 2015 at 10:34 PM
RustyBug
Offline
Details
p.11 #13No DR improvement = no 5Ds for me
Okay, so you're saying that Canon 5D ISO 100 does not equal Canon 5D III / 6D ISO 100 @ design approach variance.

What has the corresponding design change from the 5D to the 5D III / 6D achieve by changing the FWC / QE in offsetting relationships to maintain each @ ISO 100?
Feb 08, 2015 at 10:50 PM
takowasa
Offline
Details
p.11 #14No DR improvement = no 5Ds for me
RustyBug wrote:
Okay, so you're saying that Canon 5D ISO 100 does not equal Canon 5D III / 6D ISO 100 @ design approach variance.


Just the opposite.

What has the corresponding design change from the 5D to the 5D III / 6D achieve by changing the FWC / QE in offsetting relationships to maintain each @ ISO 100?

The quantum efficiency of the 5D3/6D is double the quantum efficiency of the 5Dc, so the sensor fills up twice as fast. However, the 5D3/6D sensors can absorb twice as much light as the 5Dc sensor, so they fill up in the same time for the same exposure. Thus, the same base ISO.

The sensors in the D3s and D800 both have the same quantum efficiency, so they fill at the same rate for a given exposure. But the D800 sensor can absorb twice as much light, so the D800 sensor fills up in half the time as the D3s sensor (again, for a given exposure). Hence the D800 base ISO is one stop lower than the D3s base ISO.
Feb 08, 2015 at 10:53 PM
RustyBug
Offline
Details
p.11 #15No DR improvement = no 5Ds for me
takowasa wrote:
The quantum efficiency of the 5D3/6D is double the quantum efficiency of the 5Dc, so the sensor fills up twice as fast. However, the 5D3/6D sensors can absorb twice as much light as the 5Dc sensor, so they fill up in the same time for the same exposure. Thus, the same base ISO.


So, what's the advantage to be derived from the design change if they offset to retain the same ISO? (i.e. Math of FWC / QE).
Feb 08, 2015 at 11:00 PM
takowasa
Offline
Details
p.11 #16No DR improvement = no 5Ds for me
takowasa wrote:
The quantum efficiency of the 5D3/6D is double the quantum efficiency of the 5Dc, so the sensor fills up twice as fast. However, the 5D3/6D sensors can absorb twice as much light as the 5Dc sensor, so they fill up in the same time for the same exposure. Thus, the same base ISO.


RustyBug wrote:
So, what's the advantage to be derived from the design change? (i.e. Math of FWC / QE).



Lower noise for a given exposure (the 5D3/6D collect twice as much light for a given aperture and exposure time), greater base ISO DR (the 5D3/6D can absorb twice as much total light resulting in a one stop increase in base ISO DR).

In short, ISO really doesn't mean a lot. It's how much light falls on the sensor and how much electronic noise the sensor and other hardware add.
Feb 08, 2015 at 11:05 PM
RustyBug
Offline
Details
p.11 #17No DR improvement = no 5Ds for me
takowasa wrote:
takowasa wroteThe quantum efficiency of the 5D3/6D is double the quantum efficiency of the 5Dc, so the sensor fills up twice as fast. However, the 5D3/6D sensors can absorb twice as much light as the 5Dc sensor, so they fill up in the same time for the same exposure. Thus, the same base ISO.
RustyBug wroteSo, what's the advantage to be derived from the design change? (i.e. Math of FWC / QE).
Lower noise for a given exposure (the 5D3/6D collect twice as much light for a given aperture and exposure time), greater base ISO DR (the 5D3/6D can absorb twice as much total light resulting in a one stop increase in base ISO DR).

In short, ISO really doesn't mean a lot. It's how much light falls on the sensor and how much electronic noise the sensor and other hardware add.


At the same ISO for a given exposure (aperture / time) ... the same amount of light is falling on the sensor, so that is an unchanged variable.



Edited on Feb 08, 2015 at 11:09 PM · View previous versions
Feb 08, 2015 at 11:08 PM
snapsy
Offline
Details
p.11 #18No DR improvement = no 5Ds for me
RustyBug wrote:
takowasa wroteThe quantum efficiency of the 5D3/6D is double the quantum efficiency of the 5Dc, so the sensor fills up twice as fast. However, the 5D3/6D sensors can absorb twice as much light as the 5Dc sensor, so they fill up in the same time for the same exposure. Thus, the same base ISO.
So, what's the advantage to be derived from the design change? (i.e. Math of FWC / QE).


The photon shot noise SNR is equal to the square root of the total electrons captured, thus the higher the aggregate FWC the better the SNR. So there's two ways to realize an IQ gain based on the combination of FWC and/or QE improvements. If only FWC is increased, then base ISO drops by 1EV and so you have to double the exposure time to capture 2x the amount of light to realize the noise improvement. If both FWC and QE are doubled then you'll get that IQ gain without having to double the exposure time (ie, at the same ISO you'll see reduced noise).
Feb 08, 2015 at 11:09 PM
takowasa
Offline
Details
p.11 #19No DR improvement = no 5Ds for me
takowasa wrote:
The quantum efficiency of the 5D3/6D is double the quantum efficiency of the 5Dc, so the sensor fills up twice as fast. However, the 5D3/6D sensors can absorb twice as much light as the 5Dc sensor, so they fill up in the same time for the same exposure. Thus, the same base ISO.


RustyBug wrote:
So, what's the advantage to be derived from the design change? (i.e. Math of FWC / QE).


takowasa wrote:
Lower noise for a given exposure (the 5D3/6D collect twice as much light for a given aperture and exposure time), greater base ISO DR (the 5D3/6D can absorb twice as much total light resulting in a one stop increase in base ISO DR).

In short, ISO really doesn't mean a lot. It's how much light falls on the sensor and how much electronic noise the sensor and other hardware add.


RustyBug wrote:
At the same ISO for a given exposure ... the same amount of light is falling on the sensor, so that is an unchanged variable.



The same amount of light is falling on the sensor for a given exposure (regardless of the ISO), but that does not mean that the same amount of light is recorded by the sensor or that the sensor can absorb all the light that does fall on it.

For example, if we took a photo of a scene at f/2.8 1/100 ISO 100 with a 5Dc and with a 5D3/6D, the 5D3 would record twice as much light as would the 5Dc due to the fact that the 5D3/6D have twice the quantum efficiency as the 5Dc.

So, for the same exposure and ISO, all sensors will have filled up to the same percentage of their capacity, but sensors with larger capacities will absorb more light in total and thus have less noise.
Feb 08, 2015 at 11:16 PM
RustyBug
Offline
Details
p.11 #20No DR improvement = no 5Ds for me
Borrowing from here

alundeb wrote:
Say at ISO 100, at average 10000 photons are converted to electrical charge in each pixel. The photon shot noise is then sqrt(10000) = 100 photons, and the signal/noise ratio is 10000/100 = 100


If the FWC is doubled to 20,000, then the sqrt = 141.4. Then 20,000 / 141.4 = 141.4 ... for the FWC aspect, correct?
Feb 08, 2015 at 11:23 PM
FM ForumsCanon ForumJoin Upload & Sell
  
 

You are not logged inLog inRegister