Kenko 1.4X TC/5D Mark III Questions...

[John Chardine]

Very interesting discussion. I think I'm getting it but I have a question- why do light meters give you shutter speed and f-stop for a particular ISO and not ask you for the focal length of the lens?

[Don Nelson]

John
Whether a modern digital camera lens or an old Globe or Petzvel formula from the dawn of daugerrian/wetplate, f stop is just a dimensionless number determined by dividing the focal length by the aperature diameter (although older lenses use a slightly different formula for the old US system)

Millions of real world photographers don't need to know aperature diameter while making a photograph..,yet can make simple tradeoffs between shutter speed and fstop(==aperature) in the field. Indeed, even without a meter using the old "sunny f16" rule.

Don

[Roger Clark]

Lots to discuss here.

Very interesting discussion. I think I'm getting it but I have a question- why do light meters give you shutter speed and f-stop for a particular ISO and not ask you for the focal length of the lens?

Hi John,
This is one of the simplifications that has been with us for 100+ years.
ISO assumes a change in sensitivity. Digital cameras fake it; they do not change sensitivity with ISO.
F/ratio is also a simplification as it has both lens aperture diameter and focal length folded into one parameter. For extended objects (general scenes) the f/ratio gives constant light level per unit area in the focal plane. This worked with film because film doesn't bin signals by pixel size. Pixels in different digital cameras, however, are different sizes so collect different amounts of light with a given lens. Thus the 7D pixels collect different amounts of light than 5D3 pixels. Effects of this will become clearer with answers to other questions (I hope).

I do wonder though and appreciate your discussions, but it seems to me that you are trying to convince a few of the best bird photographers in the world that they do not know what they are doing!
All the discussion about star photos and the like are fine except that they have nothing to do with taking a great photo of a diving Tern, or am I being simplistic?
It is interesting, but I am not sure how a fairly new photographer benefits from a lot of this.
It is entertaining, I just don't understand a lot of it.

Good questions Dan. First, the relevance to this thread. People were trying to figure out how to get a 5D3 to AF at f/8, presumably to get more pixels on their subjects. I pointed out that f/8 AF is generally slow, will likely cause hunting in this case, and a better alternative is a 7D. People then objected saying the small pixels of the 7D are poor performers. So what I am talking about has direct in the field real-world applications to bird photography. Understanding the effects of Etendue will allown one to make a better choice in purchasing a camera and in what camera+lens combination to use in the field. Unfortunately, digital cameras with the varying pixel sizes, sensor sizes, and faked ISO has led to many misconceptions. I used the star examples as a simple way anyone can make simple exposure that prove my points. The same facts apply in bird photography (and all other photography) just one needs to calculate photon levels the prove the details.

I go to the field see an owl in evening light, my exposure is f/5.6 1/2000sec ISO 400. It is set by the scene. I put a 500mm lens on a 7D and shoot. Then I attach 1.4X TC to my 5D3 and shoot again. My exposure remains fixed. I compare the two RAW files. The 5D3 image has received 2.56X times as many photons delivering a cleaner and sharper file. Do I care the aperture diameter changed? No.

Arash,
You are mistaken that the 5D3 collected more light because of sensor size. You get, in your example, 2x more light on your subject because the aperture diameter was increased by 1.4x. If the subject was small enough to fit in the frame of the 7D with the 1.4x TC on, then the 7D+1.4x would have double the light as 7D with no TC. That is because the aperture diameter is changing in your example.

Roger suggests I should stop down to f/8 after adding the TC to equalize aperture diameter and then raise the ISO until I get the same brightness because I need to normalize the system transfer function. Why in the world should I photograph like this? It makes no sense to me.

Raising ISO does NOT change the light received, and ISO has nothing to do with the "system transfer function."

Aperture diameter may be the all important factor in the science behind light gathering on the sensor, but it has zero use in the field. I can't say that I know anyone who is confused by f/ratio, in spite of your insistance to the contrary. You advocate a number that requires the use of a calculator to figure out.

I think you are really misunderstanding what I am saying. The question has solely to do with can two cameras with different sized pixels produce equally good images. I do not go to the field with a calculator and compute aperture diameter every time I change something while photographing birds or other wildlife. The question is only how to properly compare two different cameras. The 7D has been unfairly maligned because comparisons have not been equal. The usual error is the 7D gets less light in the comparison due to a smaller lens diameter. In some cases I've seen people use different exposure times, and others different pixels on subject. Unequalized test conditions leads to wrong results and is an indication that the testers are misunderstanding something.

That's not how people take pictures. Sometimes I think science gets the best of you. In your comparison of noise between the 7D and the 1D Mark IV, you devised a test that compared the 7D at ISO 400 to the Mark IV at ISO 800; this was done to normalize the aperture diameter.
As I pointed out then, this type of comparison doesn't pass the sniff test; I believe you were blinded by science.

Doug, that is incorrect. The science is correct. I did not change ISO to normalize aperture diameter. ISO does NOT control how much light is collected by any digital camera. Calling ISO sensitivity, as the manufacturers do, is absolutely incorrect. It is false advertising by the camera manufacturers. try the star test as a function of ISO to prove this yourself.

In photography, we contend with available light. When comparing high ISO performance, what's relevant to photographers is ISO performance at equivalent exposures as dictated by available light. Equalizing the number of photons reaching each pixel is not relevant. That's a lab test, not a field test.

If the exposure is the same, the photon count, by definition, is the same. So equalizing the number of photons is equalizing the exposure.

In real field conditions you send to and collect more photons with the FF sensor per given integration time than you do to with a smaller sensor. The number of collected photons is not equal. That's where the advantage comes from. Normalizing system transfer function is good for evaluating some aspects of sensor performance, but it does not resemble how a photographer makes photos in the field.

This is not correct. Let's do an example of sensor sizes and show how much light gets to the sensor.

Let's say you have two sensors: A = full frame, and B= 1.67x crop. Both have the same number of megapixels. The common view is that if one uses a 500 mm f/4 lens on camera A and a 300 mm lens on camera B one gets the same pixels on subject, and same field of view. The larger sensor gets more light when the two cameras are exposed for the same exposure time and the same f/ratio. But the light is different BECAUSE the lens aperture is different, NOT the sensor size. Camera A with the 500 mm lens has an aperture diameter of 125 mm, while camera B has only 75 mm diameter lens.

Now make the lens aperture diameters the same. Camera A with the large sensor has a 125 mm aperture. So put a 125 mm aperture 300 mm focal length lens on camera B (that would be a 300 mm f/2.4 lens). Then the light with camera B would be equal to the light collected by camera A. The two cameras would then have the same pixels on the subject, the same depth of field and the same light in each pixel given equal exposure times.

Indeed, if someone had difficulty hand holding with the weight of a 500 mm lens with a full frame camera, switching to a 300 mm f/2.8 lens with a camera with 1.6x smaller pixels would give almost the same performance. This is the power of knowing the factors of Etendue, as it enables one to make intelligent decisions based on facts and not perceptions.

More on Etendue at:
http://www.clarkvision.com/articles/...m.performance/
The graphs show that some camera + lens combinations will produce equivalent results.

Roger

[Tom Graham]

FWIW, quickly reading some of the latest posts it seems to me that sensor size is sometimes being confused with sensor PIXEL size, or vice-versa? (Except by Roger).
Tom

[Arthur Morris]

With regards to Pane 53. I have tried to understand Etendue. I do not get it. I cannot follow any of the scientific lines of reasoning in Pane 53. When I use a teleconveter, I am never thinking of pixels on the subject. I have no idea in the world how to normalize aperture. And I have no idea in the world what photon count is. In spite of all of the above, I have been making images that sell and do well in the big contests for more than 2 1/2 decades. I do very much agree with Dan Kearl. I do agree with Doug's feeling the science is getting in the way Roger's way. I think that I agree with Arash but I am not smart enough to know for sure :).

[Arthur Morris]

Peter, Peter, Pumpkin-eater,

Thanks.

re:


It looks like the 5D3 might have some sort of conflict with the Kenko Pro300.

I am not sure what is going on there.... Or if either the TC or my 5D III has a problem. And nobody has been able to answer my specific question. And even Patrick will not be able to answer it directly because he is buying a different version.

Reading through all this, I see the following options:

Kenko 1.4 Pro300, non reporting, possible issues with 5D3, price US$259
Kenko 1.4 MC4, non reporting, seems to be performing just under the Pro300, price US$149
Tamron F AF 1.4x, non reporting, no issues with 5D3, from B&H only available in 'Used Store' ?
Tamron SP AF 1.4x reporting, no apparent issues with 5D3, price US$224
Canon 1.4x EF III, well known, priced at US$449

@Arthur: do you see performance differences between the Kenko Pro300 and the regular Canon TC ?

I did not use it much with the 800; I can only say that I was amazed by the sharpness. But then folks need to ask themselves, "What does he know?"


have you tried taping the Canon TC on your (pane 1 described) combo ? Just to see if you get this '00' problem as well

I have taped the pins with the Canon 1.4X III TC and the 5D III and the 800. AF was painfully slow (and jumpy) but accurate.

It looks like you could give the Kenko MC4 a try.

I will try to exchange it but only after I investigate a bit more.

@All: what is your experience with Kenko vs. Tamron TC's. Anything that prefers the one over the other ?

I know nothing about Tamron.

It (Tamron or Kenko) looks like a much better deal than the (expensive) Canon TC, except for those using the new type II lenses.

No clue yet.

[Joel Eade]

Artie, with your non-scientific method, do you think if you spent an equal amount of time with a 7D and a 600mm version II lens you could produce images just as well as you do now with 1D MKIV and your 800mm?

[Arthur Morris]

Artie, with your non-scientific method, do you think if you spent an equal amount of time with a 7D and a 600mm version II lens you could produce images just as well as you do now with 1D MKIV and your 800mm?

In the course of writing the hugely popular 7D User's Guide I used the camera a lot with a variety of lenses. With the 800mm f/5.6L IS lens it was often far too long (effective 1280) and offered zero flexibility, i.e., just one effective focal length. I liked the files made in full sun but not those made in low light despite the fact that I know how to expose to the right. With the 600 IS II you would have a bit more flexibility with two focal lengths (effective 960 &1344) available but the 600 alone with the 7D would often still be too much lens. With the 600 II and a Mark IV for example you would have the following effective focal lengths: 780, 1092, and 1560.

Please do not get me wrong. Competent photographers using a 7D with any lens can create lots of great images. Can you say Dan Cadieux?

[Roger Clark]

With regards to Pane 53. I have tried to understand Etendue. I do not get it. I cannot follow any of the scientific lines of reasoning in Pane 53. When I use a teleconveter, I am never thinking of pixels on the subject. I have no idea in the world how to normalize aperture. And I have no idea in the world what photon count is. In spite of all of the above, I have been making images that sell and do well in the big contests for more than 2 1/2 decades. I do very much agree with Dan Kearl. I do agree with Doug's feeling the science is getting in the way Roger's way. I think that I agree with Arash but I am not smart enough to know for sure :).

Hi Artie,
I do not know what pane 53 is, because it is relative to that the system presents and thus dependent on user settings. For example, I only see 50 entries on page 1, but only 2 on page 2. The others are hidden in a line that says "more replies below current depth" and do not get counted. When I click on the more relpies, the numbering starts over from 1. If I go to page 2, numbers starts from 1.

Let me try another way to describe Etendue. But first let me say that with your experience you are likely doing things so second nature that you don't have to think about it. So when you add a TC, why? To get the subject larger in the frame? But that gives you more pixels on the subject. So indirectly you do think about more pixels on the subject.
Why is that you want to add another TC to your 5D3, getting to f/8? More reach? That means more pixels on subject.

Etendue describes the total system. That means pixels (detail) on subject as well as total light gathered in each pixel. So when comparing two cameras and lenses, if exposure is different or the detail is different, it biases our perception and the test is not equal. Etendue equalizes that comparison so the results are unbiased.

I mentioned the 7D as an alternative to 5D3 plus TCs pushing to f/8. We've seen in this thread the (false) perception that larger pixels and larger sensors deliver better images.
it is false because sensors do not deliver the light. It is the lens and exposure time that delivers the light. And not ISO either. Etendue tells the real story.

I understand that these are hard concepts, as the prevailing view is completely different. I have 36+ years in professional digital imaging experience, and I'm obviously having difficulty communicating these concepts that I see as straightforward. let me try a couple of other things.

Exposure: there are 2 things that control true exposure: the lens and exposure time. Note ISO is NOT a factor in true exposure.

The sensors in digital cameras have only one sensitivity: the quantum efficiency. ISO is, in effect, faked sensitivity to get around electronics limitations in the current and previous crops of digital cameras. Scientific sensors do not have ISO settings because the electronics are better (including 16-bit A/D converters).

Here is an analogy of ISO with bank accounts. Say you can count only 100 units. If I choose dollars, I can count from 1 to 100. If I use 10 dollar bills, I can count by tens from 10 to 1000. I can count in pennies, but then my range is only $0.01 to $1. Which level one chooses to count will be dictated by the amount of money in the bank account if one wants to know the full amount of money in the account. If I have $623.47 in the account, I must count by tens, and I would determine I have $620 in the account. I could count by twenties, and still find I have $620. If I count by fifties, I would find $600. I can never get a precise amount in the account, but I can get a close approximation.

So too with digital cameras and ISO. A given lens delivers so much light per second to a pixel in a given scene, and the exposure time sets the total amount of light. We choose the ISO as a range in which to digitize the signal from light the sensor captured. Low ISO is like choosing counting by tens with the bank account. High ISO is like counting by pennies in the bank account. Changing ISO is only changing the way we count, not how much light the sensor actually captured.

The next significant innovation in DSLRs will be better electronics with better A/D converters (e.g. good 16-bit A/Ds) and then ISO can be "chosen" at the time of raw conversion.

So we are back to Etendue and the remaining variables: lens aperture collects the light, focal length spreads out the light (except for point sources), diffraction spreads out the light, exposure time determines how long light is collected, the optics transmission and quantum efficiency of the sensor determines how much of that light gets recorded, and the pixel determines the area in the focal plane over which to collect the light.

Where I think people get tripped up is in focussing (pun) on the focal plane. The system (the math) can be simplified by instead of considering the focal length and pixel size (e.g. in microns on the sensor), to combine them into the angular size of the pixel. After all, as photographers (and even scientists doing science), we could care less about the pixels. We want our subject and adequate detail on the subject. So instead of focusing only on linear pixel size, or focal length (e.g. the obsession with pushing the telephoto to f/8 with TCs), we are really only concerned with detail on the subject. That detail is given by the plate scale (I have discussed this all before). There are many ways to get the detail with equal true exposure: longer focal length with larger pixels, versus shorter focal length with smaller pixels, and a continuum in between. As long as the lens aperture diameter and exposure time are held constant, then the true exposure and detail on the subject remains constant, including depth of field. That is Etendue.

So when peope choose a camera to buy, or select a camera to use in the field. One chooses a camera and lens to frame the subject and get the detail on the subject. The large pixel camera may not be the best choice. For example, pushing to f/8 with TCs to get the detail, but slowing down AF on the big pixel camera is not the choice I would do. I would choose a camera with smaller pixels so I still had the detail, but worked at f/4 or f/5.6 and had faster AF. And with the same lens aperture diameter (e.g. 600 mm wide open) and the same exposure time, I will get the detail, the same noise, the same depth of field with the smaller pixel camera, and have better AF to boot. That is Etendue.

Roger

[Arthur Morris]

Roger, While viewing in VB4 skin, this clearly shows as Pane 53:

Lots to discuss here.


Originally Posted by John Chardine
Very interesting discussion. I think I'm getting it but I have a question- why do light meters give you shutter speed and f-stop for a particular ISO and not ask you for the focal length of the lens?



Hi John,
This is one of the simplifications that has been with us for 100+ years.
ISO assumes a change in sensitivity. Digital cameras fake it; they do not change sensitivity with ISO.
F/ratio is also a simplification as it has both lens aperture diameter and focal length folded into one parameter. For extended objects (general scenes) the f/ratio gives constant light level per unit area in the focal plane. This worked with film because film doesn't bin signals by pixel size. Pixels in different digital cameras, however, are different sizes so collect different amounts of light with a given lens. Thus the 7D pixels collect different amounts of light than 5D3 pixels. Effects of this will become clearer with answers to other questions (I hope).


Originally Posted by dankearl
I do wonder though and appreciate your discussions, but it seems to me that you are trying to convince a few of the best bird photographers in the world that they do not know what they are doing!
All the discussion about star photos and the like are fine except that they have nothing to do with taking a great photo of a diving Tern, or am I being simplistic?
It is interesting, but I am not sure how a fairly new photographer benefits from a lot of this.
It is entertaining, I just don't understand a lot of it.



Good questions Dan. First, the relevance to this thread. People were trying to figure out how to get a 5D3 to AF at f/8, presumably to get more pixels on their subjects. I pointed out that f/8 AF is generally slow, will likely cause hunting in this case, and a better alternative is a 7D. People then objected saying the small pixels of the 7D are poor performers. So what I am talking about has direct in the field real-world applications to bird photography. Understanding the effects of Etendue will allown one to make a better choice in purchasing a camera and in what camera+lens combination to use in the field. Unfortunately, digital cameras with the varying pixel sizes, sensor sizes, and faked ISO has led to many misconceptions. I used the star examples as a simple way anyone can make simple exposure that prove my points. The same facts apply in bird photography (and all other photography) just one needs to calculate photon levels the prove the details.


Originally Posted by arash_hazeghi
I go to the field see an owl in evening light, my exposure is f/5.6 1/2000sec ISO 400. It is set by the scene. I put a 500mm lens on a 7D and shoot. Then I attach 1.4X TC to my 5D3 and shoot again. My exposure remains fixed. I compare the two RAW files. The 5D3 image has received 2.56X times as many photons delivering a cleaner and sharper file. Do I care the aperture diameter changed? No.



Arash,
You are mistaken that the 5D3 collected more light because of sensor size. You get, in your example, 2x more light on your subject because the aperture diameter was increased by 1.4x. If the subject was small enough to fit in the frame of the 7D with the 1.4x TC on, then the 7D+1.4x would have double the light as 7D with no TC. That is because the aperture diameter is changing in your example.


Originally Posted by arash_hazeghi
Roger suggests I should stop down to f/8 after adding the TC to equalize aperture diameter and then raise the ISO until I get the same brightness because I need to normalize the system transfer function. Why in the world should I photograph like this? It makes no sense to me.



Raising ISO does NOT change the light received, and ISO has nothing to do with the "system transfer function."


Originally Posted by Doug Brown
Aperture diameter may be the all important factor in the science behind light gathering on the sensor, but it has zero use in the field. I can't say that I know anyone who is confused by f/ratio, in spite of your insistance to the contrary. You advocate a number that requires the use of a calculator to figure out.



I think you are really misunderstanding what I am saying. The question has solely to do with can two cameras with different sized pixels produce equally good images. I do not go to the field with a calculator and compute aperture diameter every time I change something while photographing birds or other wildlife. The question is only how to properly compare two different cameras. The 7D has been unfairly maligned because comparisons have not been equal. The usual error is the 7D gets less light in the comparison due to a smaller lens diameter. In some cases I've seen people use different exposure times, and others different pixels on subject. Unequalized test conditions leads to wrong results and is an indication that the testers are misunderstanding something.


Originally Posted by Doug Brown
That's not how people take pictures. Sometimes I think science gets the best of you. In your comparison of noise between the 7D and the 1D Mark IV, you devised a test that compared the 7D at ISO 400 to the Mark IV at ISO 800; this was done to normalize the aperture diameter.
As I pointed out then, this type of comparison doesn't pass the sniff test; I believe you were blinded by science.



Doug, that is incorrect. The science is correct. I did not change ISO to normalize aperture diameter. ISO does NOT control how much light is collected by any digital camera. Calling ISO sensitivity, as the manufacturers do, is absolutely incorrect. It is false advertising by the camera manufacturers. try the star test as a function of ISO to prove this yourself.


Originally Posted by Doug Brown
In photography, we contend with available light. When comparing high ISO performance, what's relevant to photographers is ISO performance at equivalent exposures as dictated by available light. Equalizing the number of photons reaching each pixel is not relevant. That's a lab test, not a field test.



If the exposure is the same, the photon count, by definition, is the same. So equalizing the number of photons is equalizing the exposure.


Originally Posted by arash_hazeghi
In real field conditions you send to and collect more photons with the FF sensor per given integration time than you do to with a smaller sensor. The number of collected photons is not equal. That's where the advantage comes from. Normalizing system transfer function is good for evaluating some aspects of sensor performance, but it does not resemble how a photographer makes photos in the field.



This is not correct. Let's do an example of sensor sizes and show how much light gets to the sensor.

Let's say you have two sensors: A = full frame, and B= 1.67x crop. Both have the same number of megapixels. The common view is that if one uses a 500 mm f/4 lens on camera A and a 300 mm lens on camera B one gets the same pixels on subject, and same field of view. The larger sensor gets more light when the two cameras are exposed for the same exposure time and the same f/ratio. But the light is different BECAUSE the lens aperture is different, NOT the sensor size. Camera A with the 500 mm lens has an aperture diameter of 125 mm, while camera B has only 75 mm diameter lens.

Now make the lens aperture diameters the same. Camera A with the large sensor has a 125 mm aperture. So put a 125 mm aperture 300 mm focal length lens on camera B (that would be a 300 mm f/2.4 lens). Then the light with camera B would be equal to the light collected by camera A. The two cameras would then have the same pixels on the subject, the same depth of field and the same light in each pixel given equal exposure times.

Indeed, if someone had difficulty hand holding with the weight of a 500 mm lens with a full frame camera, switching to a 300 mm f/2.8 lens with a camera with 1.6x smaller pixels would give almost the same performance. This is the power of knowing the factors of Etendue, as it enables one to make intelligent decisions based on facts and not perceptions.

More on Etendue at:
http://www.clarkvision.com/articles/...m.performance/
The graphs show that some camera + lens combinations will produce equivalent results.

Roger

Last edited by Roger Clark; Today at 01:12 AM.

Roger N. Clark

[Arthur Morris]

Hi Artie,
I do not know what pane 53 is, because it is relative to that the system presents and thus dependent on user settings. For example, I only see 50 entries on page 1, but only 2 on page 2. The others are hidden in a line that says "more replies below current depth" and do not get counted. When I click on the more relpies, the numbering starts over from 1. If I go to page 2, numbers starts from 1.

Let me try another way to describe Etendue. But first let me say that with your experience you are likely doing things so second nature that you don't have to think about it. So when you add a TC, why? To get the subject larger in the frame? But that gives you more pixels on the subject. So indirectly you do think about more pixels on the subject.
Why is that you want to add another TC to your 5D3, getting to f/8? More reach? That means more pixels on subject.

Etendue describes the total system. That means pixels (detail) on subject as well as total light gathered in each pixel. So when comparing two cameras and lenses, if exposure is different or the detail is different, it biases our perception and the test is not equal. Etendue equalizes that comparison so the results are unbiased.

I mentioned the 7D as an alternative to 5D3 plus TCs pushing to f/8. We've seen in this thread the (false) perception that larger pixels and larger sensors deliver better images.
it is false because sensors do not deliver the light. It is the lens and exposure time that delivers the light. And not ISO either. Etendue tells the real story.

I understand that these are hard concepts, as the prevailing view is completely different. I have 36+ years in professional digital imaging experience, and I'm obviously having difficulty communicating these concepts that I see as straightforward. let me try a couple of other things.

Exposure: there are 2 things that control true exposure: the lens and exposure time. Note ISO is NOT a factor in true exposure.

The sensors in digital cameras have only one sensitivity: the quantum efficiency. ISO is, in effect, faked sensitivity to get around electronics limitations in the current and previous crops of digital cameras. Scientific sensors do not have ISO settings because the electronics are better (including 16-bit A/D converters).

Here is an analogy of ISO with bank accounts. Say you can count only 100 units. If I choose dollars, I can count from 1 to 100. If I use 10 dollar bills, I can count by tens from 10 to 1000. I can count in pennies, but then my range is only $0.01 to $1. Which level one chooses to count will be dictated by the amount of money in the bank account if one wants to know the full amount of money in the account. If I have $623.47 in the account, I must count by tens, and I would determine I have $620 in the account. I could count by twenties, and still find I have $620. If I count by fifties, I would find $600. I can never get a precise amount in the account, but I can get a close approximation.

So too with digital cameras and ISO. A given lens delivers so much light per second to a pixel in a given scene, and the exposure time sets the total amount of light. We choose the ISO as a range in which to digitize the signal from light the sensor captured. Low ISO is like choosing counting by tens with the bank account. High ISO is like counting by pennies in the bank account. Changing ISO is only changing the way we count, not how much light the sensor actually captured.

The next significant innovation in DSLRs will be better electronics with better A/D converters (e.g. good 16-bit A/Ds) and then ISO can be "chosen" at the time of raw conversion.

So we are back to Etendue and the remaining variables: lens aperture collects the light, focal length spreads out the light (except for point sources), diffraction spreads out the light, exposure time determines how long light is collected, the optics transmission and quantum efficiency of the sensor determines how much of that light gets recorded, and the pixel determines the area in the focal plane over which to collect the light.

Where I think people get tripped up is in focussing (pun) on the focal plane. The system (the math) can be simplified by instead of considering the focal length and pixel size (e.g. in microns on the sensor), to combine them into the angular size of the pixel. After all, as photographers (and even scientists doing science), we could care less about the pixels. We want our subject and adequate detail on the subject. So instead of focusing only on linear pixel size, or focal length (e.g. the obsession with pushing the telephoto to f/8 with TCs), we are really only concerned with detail on the subject. That detail is given by the plate scale (I have discussed this all before). There are many ways to get the detail with equal true exposure: longer focal length with larger pixels, versus shorter focal length with smaller pixels, and a continuum in between. As long as the lens aperture diameter and exposure time are held constant, then the true exposure and detail on the subject remains constant, including depth of field. That is Etendue.

So when peope choose a camera to buy, or select a camera to use in the field. One chooses a camera and lens to frame the subject and get the detail on the subject. The large pixel camera may not be the best choice. For example, pushing to f/8 with TCs to get the detail, but slowing down AF on the big pixel camera is not the choice I would do. I would choose a camera with smaller pixels so I still had the detail, but worked at f/4 or f/5.6 and had faster AF. And with the same lens aperture diameter (e.g. 600 mm wide open) and the same exposure time, I will get the detail, the same noise, the same depth of field with the smaller pixel camera, and have better AF to boot. That is Etendue.

Roger

Thanks Roger. With all due respect: in your 13 paragraph-response above you are missing my main point: I do not know what Etendue is, and I do not need to know what it is. How do I know that? Because I am and have been making great images without knowing.

[P-A. Fortin]

I don't think Roger's point is that one _needs_ to know about Etendue or such stuff to make great images. I don't know much about gravity, yet I'm a pretty decent walker (!).

He is simply providing a scientific and technical explanation about how stuff works. Taking this information into account or not is not going to make much of a difference in the quality of the images one shoots. Yet the fact that probably most photographers do not know about such stuff nor take it into account does not diminish the value of the information he is providing.

I might be the only one, but I find it a very interesting reading and will keep that in mind whenever I decide to upgrade my gear.

[Tom Graham]

Take another art, music. A musician makes a CD recording that is beautiful and is popular. How much does the artist need to know about digital theory by Fourier, Shannon and Nyquist? Does the artist care, should they care, and could they even begin to understand it? No, not necessarily. But their recording is still dependent on the scientists and engineers who do understand and make electronic reproduction possible. If understanding some of science helps the artist make music, satisfies some curiosity, that's fine but it is not the heart of their art.
Tom

[adrian dancy]

I am not a scientist and forgive me if my question seems a little absurd or demonstrates ignorance but there is something which has always puzzled me. It concerns the manner in which photons hit different sensor types. As I understand it, there are sensors with gaps between pixels and those where there is little or no gap. My question is this. What happens to a photon that hits the edge of a pixel? Are any photons lost completely in the process by being bounced out or off? Assuming that no or very few photons are lost then will the distribution of the photons hitting the edge of pixels cause any image quality differences, and if so, in what respect?

[Ofer Levy]

Hi all,
Since this thread is about "Kenko 1.4X TC/5D Mark III Questions..." and since this topic is very important for me and many others with f5.6 lenses like the 400 and 800 please lets stay on track and maybe leave the scientific stuff to a new thread?
I really want to know whether the Kenko 1.4 will work with the 800 and I don't like to go through all this long scientific information which for me is totally meaningless. It should be very simple - does it work or it doesn't and how well if it does.
Thanks,
Ofer

[Don Nelson]

Ofer
I have one on order that will be received by the weekend, and will try it on the 800mm.
regards
Don

[Patrick Sparkman]

Well the race is on for all of us to get our Kenko teleconverters and test them. I should get mine Friday, and will test with the 800 and 5DIII.

[Ofer Levy]

Thanks Don and Patrick - that would be awesome!!

[Patrick Sparkman]

Well, I am saddened to report that the Kenko MC4 DGX 1.4 tele plus did not work with my 5DIII and 800 f5.6. I does the same thing that Artie reported. The Aperture goes to 0.0 and the camera locks up requiring the battery to be removed to get it operational again. It works great on my 70-300 5.6L, focusing very well at f8.0. It even reports the aperture correctly at 5.6 at 70mm, and f8.0 at 300mm. I will report on the sharpness, but it focuses well. Also works on the 24-105, but I put it on the 85 f1.2 and it locks up the same as the 800. So it does not appear to be an issue with aperture, but rather certain lenses. I guess I will get the Tamron and give it a try.

[Doug Schurman]

Well, I am saddened to report that the Kenko MC4 DGX 1.4 tele plus did not work with my 5DIII and 800 f5.6. I does the same thing that Artie reported. The Aperture goes to 0.0 and the camera locks up requiring the battery to be removed to get it operational again. It works great on my 70-300 5.6L, focusing very well at f8.0. It even reports the aperture correctly at 5.6 at 70mm, and f8.0 at 300mm. I will report on the sharpness, but it focuses well. Also works on the 24-105, but I put it on the 85 f1.2 and it locks up the same as the 800. So it does not appear to be an issue with aperture, but rather certain lenses. I guess I will get the Tamron and give it a try.

Interesting Patrick. It works well with my 400 f5.6 and my 500 f4 IS plus Canon 1.4 II extender. I'll bet it would work well on 600 f4 IS too.

[Ofer Levy]

Thanks Patrick. Bad news for the 800 f5.6 owners...

[Raymond Lee]

If anyone has a Canon EOS-5D MIII and a Kenko 1.4X TC (the top of the line one: C-AF 1.4X TELEPLUS PRO 300) and (preferably) an 800mm f/5.6L IS (or a 400mm f/5.6L) I ask:

Have you used any of the above combos successfully getting AF when you are not supposed to get AF (i.e., at f/8)?

Hi Artie,

I kind of skimmed through some but not all of the discussions here. Anyway, not sure if it's within the original topic or your question was answered, but I have the Kenko 1.4x TC (C-AF 1.4x TELEPLUS PRO 300 DGX), and I could get autofocus with my 7D and 400/5.6L without taping any pins. However, the AF doesn't lock very well in low light or on low contrasting subjects compare to the bare 400/5.6L lens. Recently, I've also been using the combo Canon 7D + Kenko 1.4x TC + Canon 1.4x TC II + 300mm f/2.8L IS to get a bit more reach as birds are quite skittish in our area, and I have quite good autofocus speed (as fast as just a single Canon 1.4x TC II attached to the 300/2.8L) with fairly sharp captures, even with tiny subjects like Killdeer in flight. Hope the info helps!

Regards,
Raymond

[Arthur Morris]

Hi Artie,

I kind of skimmed through some but not all of the discussions here. Anyway, not sure if it's within the original topic or your question was answered, but I have the Kenko 1.4x TC (C-AF 1.4x TELEPLUS PRO 300 DGX), and I could get autofocus with my 7D and 400/5.6L without taping any pins. However, the AF doesn't lock very well in low light or on low contrasting subjects compare to the bare 400/5.6L lens. Recently, I've also been using the combo Canon 7D + Kenko 1.4x TC + Canon 1.4x TC II + 300mm f/2.8L IS to get a bit more reach as birds are quite skittish in our area, and I have quite good autofocus speed (as fast as just a single Canon 1.4x TC II attached to the 300/2.8L) with fairly sharp captures, even with tiny subjects like Killdeer in flight. Hope the info helps!

Regards,
Raymond

Thanks Ray but my question specifically asked about the 5D MIII :),

[Patrick Sparkman]

Ok, I finally got back to my house (was visiting my parents) and did a little playing around with the Kenko MC4's on 5DIII. As mentioned the 1.4 locks up the 5DIII, but works beautifully on the 5DII - 800 f5.6 combo. Autofocuses faster than I remember the 1DIV did with the Canon 1.4III and reads aperture properly. Then I tried the 2X with the 5DIII on the 800, and it works! It even autofocuses through the viewfinder. You have to use the autofocus area that selects all points, but it will lock in. Of course it focuses fine on the Liveview screen as well and reads aperture properly along with the metadata reporting 1600mm. I will do some sharpness testing tomorrow, but it looks comparable so far to the Canon 2XIII. I also tried the Kenko 2X on the 85 f1.2 and it works fine. So it looks like there is a bit of a software bug in the 5DIII that is preventing it's working properly with the Kenko 1.4 and some lenses.

I wish they (Canon) would just come to their senses and remove the dumb f8 restriction. Obviously the autofocus system works fine at f8, and even f11. Maybe it is not perfect with all lens light combo's, so just put a line of code in that allows the user to unblock the f8 restriction. Like you have to do with the high and low ISO settings. And while they are at it, don't slow down the focus with the teleconverters. The Kenko definitely focuses quicker, so they are intentionally slowing the autofocus with teleconverters. Canon surely has a lot more resources than Kenko, yet the Kenko fits on any lens, cost $300 less, weighs less, seems to be as sharp (although I will get some better tests of that in the next few days.) Artie - Go get 'em!

[Arthur Morris]

Good detective work Patrick. I will follow up with Chuck Westfall and report back.

[Paul Granone]

Artie,

I ordered one of the Kenko TelePlus MC4 AF 1.4X DGX Teleconverter. I also ordered a Canon 400mm F5.6 L lens. The 400 F5.6 and the Kenko TelePlus MC4 AF 1.4X DGX Teleconverter do autofocus on the 5D mark III. I was surprised how quickly the combination does focus even in dimmer light. This was due to rainy weather here in New Mexico. So no nice bird shots to add. Just a confirmation that autofocus works with the Kenko TC , 400mm F5.6, and the 5D III. Thank you for the tip. I am excited to get out and use the combination. -- Paul

[Arthur Morris]

Thanks Paul. But that is the info that Doug provided in Pane 13 :). I am glad that the combo focuses well. However, your reminder unveils what seems to be a new mystery: Patrick: if you have this TC, the TelePlus MC4 AF 1.4X DGX Teleconverter how in the world can it AF with the 5D III on a 400 f/5.6 and not on the 800 f/5.6??? That would seem to point to the lens rather than the 5D III.... Anybody got a 1DX, a Kenko TC, and either the 400 f/5.6 or the 800 f/5.6???

[Arthur Morris]

Thanks Paul. But that is the info that Doug provided in Pane 13 :). I am glad that the combo focuses well. However, your reminder unveils what seems to be a new mystery: Patrick: if you have this TC, the TelePlus MC4 AF 1.4X DGX Teleconverter how in the world can it AF with the 5D III on a 400 f/5.6 and not on the 800 f/5.6??? That would seem to point to the lens rather than the 5D III.... Anybody got a 1DX, a Kenko TC, and either the 400 f/5.6 or the 800 f/5.6???

[Ofer Levy]

Patrick, WOW - this is some exciting and excellent news for Canon 800 f5.6 owners! Just confirming - the 5D 2 will AF just fine with each of the Kenko converters - the x1.4 and the x2. The 5D III won't AF with the Kenko x1.4 but gets the AF to work fine with the Kenko x2.....?
I wonder whether this Kenko x2 will AF properly on the 1D Mark IV.....

Some comparison images from the 5D 2 with the 800 plus the Canon and Kenko converters would be great!

[Patrick Sparkman]

It is definitely a Camera, Lens, Teleconverter combination issue. Not a f-stop issue. To recap;

The 5DIII will lock-up and show 0.0 for aperture with the Kenko MC4 1.4 Tele and the Canon 800 f5.6 along with the 85 f1.2.
The 5DIII works well with the MC4 1.4X and the Canon 70-300L 5.6, 70-200L II 2.8, 24-105L 4
The 5DIII works well with the MC4 2X and all the above lenses. 800, 85, 70-300L, 70-200L II, and 24-105L 4

The 5DII works well with the MC4 1.4X and all of the above lenses 800, 85,70-300L, 70-200L II, and the 24-105L

So this is a 5DIII issue with at least two (very different) lenses that I own and the 5DII works fine. I would call that a bug. And yes, it would be nice to test with the 1DX! Anyone have one I could borrow for a couple of days?

I have ordered both Tamron Teleconverters to see if they work. They are probably identical to the Kenko's, but it is worth a shot.

[Humberto Ramos]

I have ordered both Tamron Teleconverters to see if they work. They are probably identical to the Kenko's, but it is worth a shot.

You also have the sigma 1.4x and 2x tc...

[Arthur Morris]

Tried this one: Kenko 1.4X TC (the top of the line one: C-AF 1.4X TELEPLUS PRO 300) with the 300 f/2.8L IS II and the 5D III. Aperture read correctly. Pushed shutter button. Shutter released but the camera froze up. Had to remove the battery and re-insert. No image was created.... I wrote the Kenko folks but they have not bothered to answer. And I am pretty sure that Canon will day, "Not our product..."

[Patrick Sparkman]

I started a new thread here http://www.birdphotographers.net/for...converter-Test to give some results of my testing with the Kenko and Tamron teleconverters.

[Gary Kinard]

Wow, interesting. I am just a hack birder. I spend most of my time in the sticks of Asia. But, most of the guys here have sold there 1DlV's and go with the 7D. It is kinda weird. They all had the 7D and bought the lDl4 then sold the later and kept the 7D? Maybe they are all just not experienced bird photographers as you guys in America and Canada. Maybe I will try it again. I must be missing something.
Great thread guys. I enjoy reading discussions like this immensely. Thanks for starting it Authur.

[Arthur Morris]

YAW Gary. What are you going to try again?

[Gary Kinard]

Hay, Mr Morris. Kinda off subject, getting old :)
Anyway, a friend of mine from China bought the 1DlV and of course I could not stand it so I bought one, but returned it.
He was telling me not to buy one, that for the small birds here it is not as good as the 7D. And I returned mine for the same reason. I THINK
I did not give it much of a chance. He and a couple other friends I shoot with here have the 1D1V but use the 7D most of the time for birding.
So I just got off line with my dealer in BKK and he said I could use the 1D1V again to see if I liked it.
I am catching a flight tomorrow morning back to BKK. So if no problems I will try it this coming weekend.
Still cannot see the full frame for birding, but everyone is different. :) Although my bud from China said he was going to try it out.

Hey, great site Authur, I spread the word about this site every chance I get. Most of my time is in Asia, and most of the folks over there do not follow a forum. As far as discussing equipment. This site is a bonanza for me. I will let the folks know about it.

Really I think your contributors are great also. You have some smart cookies here.

Gotta go, will be back in a year or two.

Reguards, Gary