fredmiranda.com
Login

Moderated by: Fred Miranda
  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Canon Forum | Join Upload & Sell

1       2       3       4              6       end
  

Archive 2011 · Canon's native ISO

  
 
Mr Joe
Offline
• • • •
Upload & Sell: Off
p.5 #1 · Canon's native ISO


I seem to remember an article on Vincent Laforet's site that said the native ISO for video is different than stills on the 5D Mark II. For the lowest noise with video, ISO increments of 160 are best. 160, 320, 640, etc.

For stills, the native ISO is 100 -- so stick to 100, 200, 400, etc.

Many people have erroneously applied the ISO 160 recommendation to stills.



Nov 08, 2011 at 05:04 PM
outlawyer
Offline
• • •
Upload & Sell: Off
p.5 #2 · Canon's native ISO


This is an enormously informative thread. Thanks.
Thomas Hardy's classic literary work, "The Return of the Native ISO," seems dated now.
(Couldn't resist, sorry. This is a great thread.)



Nov 08, 2011 at 06:37 PM
aborr
Offline
• • •
Upload & Sell: Off
p.5 #3 · Canon's native ISO


There are differences between the way different Canon cameras handle intermediate ISO settings.

For example, on the 'prosumer' 5D, ISO 160 has less noise than ISO 100, ISO 320 has less noise than ISO 200, and ISO 640 has less noise than ISO 400. On these cameras, it might make sense to choose ISO 160/320/640 over 100/200/400.

This is not true for 'pro' cameras such as the ID3. On that body, ISO 100 is better than ISO 160, ISO 200 is better than ISO 320, and so on. On the 1D3, it makes sense to choose 100/200/400 over 160/320/640.

Here are some links to the data for the 5D and 1D3. I am not sure if anyone has collected similar data for the latest Canon models.

http://www.pages.drexel.edu/~par24/rawhistogram/5DTest/5DTest.html
http://www.pages.drexel.edu/~par24/rawhistogram/Mk3Test.html

The explanation for why the 'prosumer' and 'pro' cameras behave differently is at

http://theory.uchicago.edu/~ejm/pix/20d/tests/noise/noise-p2.html#readandshot

Al



Nov 08, 2011 at 06:42 PM
cgardner
Offline
• • • • •
Upload & Sell: Off
p.5 #4 · Canon's native ISO


gdanmitchell wrote:
And it isn't even that simple.

When it comes to white balancing, making white white is not always the best approach - though there are some situations in which in can be the target. Let me offer one example and may be point to some related ones.


Agree and expound on this in my tutorials on using a gray card and color process control in the photographic workflow.

Setting Custom WB off a gray card is simply a process control baseline; a way of putting the color into a "neutral corner" like a boxer after a knock-out so the as not to interfere with the referee. It is useful to shoot to a "by the numbers" baseline at capture in most, but not all situations because human color vision is so adaptive our eyes and brain are lousy color evaluation devices.

For example if shooting under a tree canopy daylight will take on a green bias. The photographer's brain knowing the subject is wearing a white shirt and what their skintone looked like under the clear sky will adapt his vision to make both look "normal" — as expected.

The RAW file values will represent the light color as seen by the camera in all cases, but how that color is initially displayed on the monitor and first seen by the photographer is controlled by the WB metadata tag in the file header. If the photographer, unaware of the green bias, were to shoot using Daylight WB on the camera the first impression of the file would be one which has the same green bias.

But then a quirk of human perception occurs. The photographer who didn't see the green bias in person will not see it on the monitor either because his brain still expects the white shirt and face to look normal. His brain will adapt his color perception to make it so.

I've seen this occur hundreds of times in outdoor photos taken under trees or next to foliage. A green bias renders skin, which is dominant in the red channel, with a dull flat look as if shot through a gray veil. That's because green added to red shifts it towards neutral gray. But at the same time the green bias makes the foliage hyper-saturated shifting its hue from a deep emerald green more towards lime green. Most providing C&C on an image like that will make comments that the lighting was too flat — lacked contrast because it fools the eyes and brain.

There's a reason for that on a deeper physiological level. The human eye has two types color sensing cone cells concentrated around the optic nerve (fovea) covering about 2% of the retina. They sense blue/yellow and green/magenta similar to how color is recorded in a RAW file and Lab file format. The remaining 98% of the retina is covered with rod cells which are only sensitive to a narrow band of light in the green region of the spectrum. The rods are also about 3000x more sensitive to light than the cones in the center — why we can see better out of the corners of our FOV at night. It is also why Bayer sensors have twice as many green sites as blue and red and why color spaces are larger in the green region — they all model the physiology of the eye and now the brain processes its signals.

Vision is a chemical process over the synapses in the cells of the eyes. Staring for any length of time at bright colors will fatigue color vision because the sensor cells in the eye quite literally "run out of gas". Due to the green sensitivity of the rods, looking bright green will tire out the eyes and affect color perception rapidly...

I managed offset printing plants for a living doing thousands of color OKs over the years and attended many training classes. An illustration similar to the one below was used in a class on color management I attended at the Graphic Arts Technical Foundation (GATF). GATF is often called on to arbitrate disputes between printers and customers. In the case below a customer had designed a cover in bright green with a neutral gray 50% tint of black grid pattern, and complained the printer had instead printed with a magenta bias...

http://super.nova.org/TP/GreenBoxGray.jpg

If you stare at it for awhile your perception will shift as the green maxes out the cells in the eyes and the perception of the 128,128,128 gray bars will change. You may even see darker spots at the four nodes the lines intersect. The same thing happens to some degree when you open and look a scenic photo where most of the content is green foliage.

The take-away from all that background information on human perception is that you can't trust your eyes to judge color accurately. If you want to see color from a objective baseline you need to adjust it "by the numbers". A gray card and Custom WB allow us do that in our photos.

Stop and consider that your current trusted baseline for "accurate" color is most likely your calibrated monitor. But way if you made a mistake while calibrating? How would you ever know? You have put blind faith in that hockey-puck as your process control baseline.

Let's say you did screw up your calibration process and the monitor has a red bias. You take a shot at noon with Daylight WB, open it on screen and guess what, it looks green, the opposite of the red bias on the monitor. So what do you do? Add red. That's how color goes off the tracks into the ditch very quickly. The color was correct out of camera, but was screwed up based on the how it appeared on screen.

As the green square example illustrates even when your monitor is perfectly calibrated the shift in your color perception will cause you to make incorrect color judgements if you sit and stare an image for a long time. That's one of the reasons a neutral gray walls around the editing space with D65 white point illumination is recommend, so you can look away from the monitor periodically and recalibrate your color perception to the known neutral baseline of the wall.

Custom WB works in much the same way IF after setting it you shoot the target again as an editing reference. For example if shooting under trees in green light I would start with my camera set to Daylight WB as a starting baseline, shoot the card, set custom WB, then shoot the card again...

http://super.nova.org/TP/GreenCast.jpg

Relative to Daylight WB the first shot has a green balance. Custom WB adds magenta — not to the RAW file but to the WB recipe for display — and the card is recorded as neutral in the second reference shot. That way every shot in the session, when first seen on the monitor is seen from that neutral baseline.

Toggling between the two frames on the camera my brain, intellectually knowing the second frame represents the baseline I should trust, can now see the bias in the lighting relative to Daylight WB I can't see in person because my eyes adapt.

Back on the computer I open the first test shot of the subject under the trees holding the gray card. Because I set Custom WB I know the card is neutral "by the numbers". If clicked with the color eyedropper is will not change. If it doesn't look neutral by eye what does it tell me? That my monitor isn't calibrated correctly. By using Custom WB you change your brain's trusted process control baseline from the hockey-puck to the camera that generates the color.

Is technically neutral WB the best color perceptually? That depends on the context of the environment seen in the photo. If the other clues in the photo say it was taken in the middle of the day then neutral WB will be perceptually correct. If there are long sideways shadows seen which tell the viewer the shot was taken near sunset the neutral WB will look unnaturally cool. If taken on a winter day where a cool bias conveys the impression it is cold outside (because skin turns blue when cold) the neutral baseline will look unnaturally warm. In other words "normal" color is a relative thing — relative to the other clues in the photo and what is expected.

A portrait taken at sunset is an interesting perceptual paradox. The overall impression of the lighting in person will be warmer than the camera's Daylight WB baseline and if shot with Daylight WB the sunset will appear in the photo as it was remembered in person. But put a portrait subject in the foreground and the shot taken with Daylight WB will render a in the sun face too warm to the point of looking jaundiced, or if lit by the skylight cooler than expected relative to how it was remembered in person. Why? Because in person color vision adapts to whatever content the brain is focused on and shifts to make that content "normal" per expectations. So regardless of the ambient color temp the brain will make faces look "normal". Not technically gray card neutral, but different than the camera will record it.

What is the solution to that dilemma? What I'd do face the subject east with their back to the setting sun and illuminate the front with flash, with custom WB set off the flash/diffusers. My diffusers warm the flash about 600K making it a close match to Daylight WB, so in effect I'd be shooting with Daylight WB.

When I open the test file with the subject holding the gray card the card will look "normal" — neutral as expected — the sunset in the background will look similar to my impression in person of its warmth, but the face will look unnaturally cold with the technically neutral WB. So in that reference test file I will adjust the color warmer from that neutral starting point until the face looks right — per the context of the scene and my in person impression of the natural lighting on it. Making the face warmer will also make the background warmer, but that will seem perfectly normal given the time of day.

Capturing from a neutral baseline is just the starting point not the end goal. The end goal is to make the color in the photo the same as my impression of the natural light. Knowing that shooting in natural light alone will not allow the camera to render it that way, and that in any lighting my brain will shift perception to normalized the face, I know from experience the best strategy to overcome the limitations of the camera is to "fake" the lighting on the face with flash to normalize it —first to the technically neutral baseline, then by eye on the monitor perceptually to the context of the environment.

The face in the photo will wind up warmer than I'd render it at noon or in the studio where more neutral WB is expected, and cooler that the camera set to Daylight WB would record it with just the ambient light. In that situation I wouldn't want to use Custom WB with just the ambient light because it would make the sunset beach look like high-noon color-wise. The face would look more normal in terms of WB in absolute terms, but that neutrality wouldn't match all the other clues in the environment surrounding the face.

There are other ways to handle a situation like that, such as setting WB to Flash then using a straw gell on the flash to warm it up a bit. But knowing how much gel to add is a judgement call that isn't easy to make on the beach with adapting color vision. That's why I prefer to shot with the face "technically" neutral at capture, the background perceptually accurate, then tweek the balance on the face to match the background on the monitor where I can just move the blue/yellow slider back and forth and find what looks "right" by comparison.

In a situation like shooting under trees I would not add flash for this reason...

http://super.nova.org/TP/GreenCastFlashMix.jpg

Setting custom WB off a gray card from the ambient light causes the camera to shift WB from Daylight to + magenta. Relative to that adjusted baseline any flash added will have a magenta bias due to the custom WB and a slight blue bias due to the fact flash is a bit cooler than daylight. The result in the photo will be a mixed lighting muddle where the shadows illuminated by the ambient fill will be neutral, but the flash lit highlights will have a magenta cast. On the computer when editing if color is adjusted to get rid of the magenta bias in the flash highlights the ambient shadows will shift from neutral to green.

The solutions to that color balance paradox, in order of preference are:

1) Don't shoot portraits under trees: An ounce of prevention is worth a pound of cure in PP and mixed lighting is an incurable problem. The same is true for any situation where the environment bounces a color cast onto the skin: brick walls, red barn, cars, etc.

2) Use reflectors, not flash: If a reflector is used for a portrait in the woods it will bounce the same green light the Custom WB has neutralized so there will be no mixed lighting. The reflectors can be used both as "key" sources to model 3D shape and as fill to lift the overall scene to record shadow detail. Keep the fill reflector in front of the case about even with the nose and the "key" reflector above the head where you would place a key light flash to mimic the direction of the key and fill components of the natural lighting.

3) Set Custom WB to ambient then gell the flash green to match: While technically feasible its not practical because its will be difficult to judge by eye from the playback when the balance of green in the flash is an exact match.

Making decisions like that require experience to understand the cause and effect and need to start with the goals for the shot.

When there is face in a photo the goal is to make it seem "normal" similar to seen by eye and there's a relatively narrow range of what will seem normal in in a photo because of the way the eye adapts in person.

For landscapes and scenics there is a wider latitude for what color balance looks normal. Generally speaking the clues to color balance come from the angle of the shadows. If the shadows lead the viewer to understand it is the middle of the day they will expect technically neutral WB. But if the shadows are long and sideways they will expect a warm glow in the lighting. Either way all you need to do is keep the camera set to Custom WB.

I never use AWB which changes WB shot-to-shot based on content. With Custom WB or a pre-set like Daylight outdoors or Tungsten / Fluorescent indoors I can correct my test shot with the reference gray card by eye to taste, then batch copy the settings to all the other RAW files. AWB will get the WB perceptually correct some of the time but doesn't allow for batch adjustment which is huge time saver.

The advantage of a gray card vs. filter-type WB tool is the ability to put the card in a reference shot with the subject holding it next to their face. If custom WB is set with an over-lens device it must be pointed at the light source, not the subject, and can't be used in the shot as editing reference.




Nov 09, 2011 at 09:52 AM
jaybird555
Offline
• • •
Upload & Sell: On
p.5 #5 · Canon's native ISO


Does anybody knows what the Native ISO for a Mark 4 is ? or where could I find the info. Thanks.


Nov 09, 2011 at 03:39 PM
gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
p.5 #6 · Canon's native ISO


No, I think it is true even with large prints inspected very closely. The purported differences between, say, 100, 160, and 200 are only "visible" in very carefully constructed tests with 100% crops and typically working with images that are anything but typical of actual photographs.

These differences will be utterly insignificant in not plain invisible in real prints.

Dan

MrAdventure wrote:
I agree with that...at normal viewing distances I'm hard pressed to see any difference between 100 and 160...matter of fact 200. I'm sure the mileage may vary under specific lighting conditions though.

Even for studio work the ISO 100 is the magic setting...I ramp that up to 160-400 alot with a huge distraction at normal viewing distances.




Nov 09, 2011 at 07:50 PM
PetKal
Offline
• • • • • • •
Upload & Sell: Off
p.5 #7 · Canon's native ISO


jaybird555 wrote:
Does anybody knows what the Native ISO for a Mark 4 is ? or where could I find the info. Thanks.


This is all you need.
Do you think Cartier-Bresson concermed himself with ISO matters ?



Nov 09, 2011 at 08:02 PM
BrianO
Offline
• • • • •
Upload & Sell: Off
p.5 #8 · Canon's native ISO


PetKal wrote:
...Do you think Cartier-Bresson concermed himself with ISO matters ?


Since there was no ISO in his day, no; but he did concern himself with film speed ratings, contrast levels, grain structure, etc.



Nov 09, 2011 at 10:27 PM
cgardner
Offline
• • • • •
Upload & Sell: Off
p.5 #9 · Canon's native ISO


PetKal wrote:
Do you think Cartier-Bresson concermed himself with ISO matters ?


No, like most in his era he understood how to judge exposure by eye. About the only thing you can do wrong with B&W is underexpose it.

My first exposure meter was the chart printed on the inside the yellow box: Sunny 16 / Shady 5.6 handles 90% of situations outdoors with B&W negative film. Noise reduction? That was a choice between Panatomic-X, Plus-X, or Tri-X



Nov 09, 2011 at 10:37 PM
PetKal
Offline
• • • • • • •
Upload & Sell: Off
p.5 #10 · Canon's native ISO


cgardner wrote:
No, like most in his era he understood how to judge exposure by eye.

And sometimes even that might have been off by a lot, like in this famous picture of his.
However, I think all he really cared about was getting the jumper and his reflection.



Nov 09, 2011 at 10:45 PM
pingflood
Offline
• • • •
Upload & Sell: On
p.5 #11 · Canon's native ISO


cgardner wrote:
No, like most in his era he understood how to judge exposure by eye. About the only thing you can do wrong with B&W is underexpose it.


Ah, just soup it in Diafine and you can get away with murder.



Nov 10, 2011 at 06:44 AM
dmacmillan
Offline
• • • • •
Upload & Sell: On
p.5 #12 · Canon's native ISO


PetKal wrote:
And sometimes even that might have been off by a lot, like in this famous picture of his.
However, I think all he really cared about was getting the jumper and his reflection.

What makes you think his exposure was off? Have your read something about that negative or do you think the exposure was off from looking at the image? BTW, I have seen a print from that negative. It had more info than you see in its reproduction.



Nov 10, 2011 at 10:08 AM
Pixel Perfect
Offline
• • • • • • •
Upload & Sell: Off
p.5 #13 · Canon's native ISO


PetKal wrote:
That is all true, Anders my friend, if a key photography objective was to have the least amount of noise in an image. However, that is often only an auxilliary technical aspect.
I often "expose to the left" where that suits my photographic quality objective better. True enough, here and there I thereby create pronounced noise which has to be dealt with, yet I gain much more in my own aesthetic sense.

For example, I must have seen thousands of images of birds with white colours in their plumage which have been exposed to the right following the internet mantra. Now, as
...Show more

+1
With white birds I err on the side of caution too and would rather a slightly underexposed image. I can bring up the background in PP and apply NR if needed, but the bird's feather detail is all intact. In the case of white subjects, the histogram can fool you as it doesn't have the resolution to show you the little tiny peak close to 255,255,255 that represents the highlight detail. You may look at it and think you can expose more to the right, when in fact that important little peak has now been pushed to blown out. So often I under expose enough that that little peak is to the right; this means the bulk of the histogram is much further left and it looks like you've wasted tonal levels, but it really does help obtain full detail in the feathers or fur etc.



Nov 10, 2011 at 04:56 PM
PetKal
Offline
• • • • • • •
Upload & Sell: Off
p.5 #14 · Canon's native ISO


Whayne, I guess you can relate because you also photograph distinctive targets which are often a smaller fraction of the whole metering area.

Truly, when photographing something like a snow white egret in flight against some sort of arboreal backdrop, not only that the histogram ranks in utility with the direct print button, it can also be destructive to the image if paid serious or any attention to.

The way I understand this is that histogram is a global representation of luminance values accross the entire metering area of the camera. However, all we want to do is expose our chosen discrete target properly as priority #1, and then deal with the background and the rest of the frame as priority #2.

That reminds me a bit of something I read on one of the Canon photography fora recently: " The new 500mm f/4 IS MkII is expected to be sharper than the MkI in the periphery of its image circle." There you go, that improvement in corner IQ might appeal to landscape photographers out there, however, I do not think that improvement will be one of the top ten reasons wildlife photographers will be upgrading their 500s to the MkII.
Incidentally, how many landscape photographers do we know of who use their 500s routinely ?

Edited on Nov 11, 2011 at 09:31 PM · View previous versions



Nov 10, 2011 at 06:05 PM
Pixel Perfect
Offline
• • • • • • •
Upload & Sell: Off
p.5 #15 · Canon's native ISO


I think the mk II's greatest benefits are in IS, mfd, weight and probably IQ and AF performance with extenders; still a long list, but bare IQ would differ little. I've used the 500 for the occasional landscape, but in between bird shots often the heat shimmer means I throw the shot out anyway.


Nov 10, 2011 at 06:45 PM
PetKal
Offline
• • • • • • •
Upload & Sell: Off
p.5 #16 · Canon's native ISO


Pixel Perfect wrote:
I think the mk II's greatest benefits are in IS, mfd, weight and probably IQ and AF performance with extenders; still a long list, but bare IQ would differ little. I've used the 500 for the occasional landscape, but in between bird shots often the heat shimmer means I throw the shot out anyway.


We are flying off tangent with this, but let me just say that my reasons for upgrading to 500 II are as follows:

(1) Weight reduction........70 %
(2) 4 stop IS....................20 %
(3) MFD reduction.............9 %
(4) IQ improvement............1 %



Nov 10, 2011 at 07:34 PM
gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
p.5 #17 · Canon's native ISO


PetKal wrote:
Truly, when photographing something like a snow white egret in flight against some sort of arboreal backdrop, not only that the histogram ranks in utility with the direct print button, it can also be destructive to the image if paid serious or any attention to.


This doesn't quite make the value of the histogram equivalent to that of the DP button, but it does point out that using the histogram display requires more sophistication than merely moving the end very close to the right edge of the display. For a subject like this I would - if I had time! - still consult the histogram display and, among other things, use it to set the exposure to a value that would not blow the beautiful detail of the near-white feathers. In other words, by the standards of some, I might choose to "under expose" to protect that detail.

That reminds me a bit of something I read on one of the Canon photography fora recently: " The new 500mm f/4 IS MkII is expected to be sharper than the MkI in the periphery of its image circle." There you go, that improvement in corner IQ might appeal to landscape photographers out there, however, I do not think that improvement will be one of the top ten reasons wildlife photographers will be upgrading their 500s to the MkII.
Incidentally, how many landscape photographers do we know of who use their 500s routinely ?


Just to tweak you a bit, I'm a big fan of shooting landscapes with long focal lengths. I don't own a 500mm prime at this point, but I use a 100-400mm zoom a lot for landscape work.

Dan

http://gdanmitchell.com/gallery/d/3837-5/GraniteCreekBridgeArchSurf20100102.jpg



Nov 10, 2011 at 08:08 PM
PetKal
Offline
• • • • • • •
Upload & Sell: Off
p.5 #18 · Canon's native ISO


That's a lovely shot, Dan, whatever the lens was you used.
However, that doesn't make 400 or 500mm lens commonly used landscape photography instruments.
Similarly, you will not see too many B&W BIF shots either.


Edited on Nov 11, 2011 at 09:31 PM · View previous versions



Nov 10, 2011 at 08:28 PM
snapsy
Offline
• • • • • •
Upload & Sell: On
p.5 #19 · Canon's native ISO


gdanmitchell wrote:
http://gdanmitchell.com/gallery/d/3837-5/GraniteCreekBridgeArchSurf20100102.jpg


That photo is great on so many levels but it's the tones that really stand out.



Nov 10, 2011 at 09:22 PM
gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
p.5 #20 · Canon's native ISO


PetKal wrote:
That's a lovely shot, Dan, whatever the lens was you used.
However, that doesn't make 400 or 500mm lens commonly used landscape photography instruments.
Similarly, you will not see too many B&W BIF shots either.


It was 320mm. The long FLs are "commonly used' by me! ;-) And longer FLs are used a lot more often than many folks who presume landscape is done with wide angle lenses realize. (A number of pretty highly regarded folks here in my neck of the woods regard the 70-200mm zoom as their favorite landscape lens.)

:-)

Dan



Nov 10, 2011 at 10:29 PM
1       2       3       4              6       end




FM Forums | Canon Forum | Join Upload & Sell

1       2       3       4              6       end
    
 

Welcome back
Log in to your account