We often hear cameras with small pixels are noisy, and the 7D has a reputation
for producing noisy images. But it is a reputation not deserved, especially
among wildlife photographers who often need more telephoto reach.
Example:
"Anyone who actually takes pictures knows the difference between the two cameras is huge. It is a no brainier"
Let's look at some of the issues. I'll include questions from others.
Question: "I thought that the pixels on the 7D were not only smaller but packed
more densely and thus generate more noise, I know Canon has done
wonders to suppress the noise in modern cameras but as the pixels are
larger and less closely packed in the 1D wouldn't there be less noise? "
Answer:
There are two dominant noise sources in our digital camera images.
1) Photon Noise: in all but the deepest shadows, the noise we see is due to
photon counting statistics from the random arrival of photons in a given
exposure time. The noise is the square root of the number of photons collected
in the pixel. What we perceive as noise is really signal to noise ratio (S/N).
The S/N is also the square root of the number of photons. Thus, to produce
noise free images, we simple need to collect as many photons as possible,
which is easier said than done.
2) Readout noise and analog to digital (A/D) converter + amplifier noise.
This noise is at a very low level so only affects the deepest shadows.
There are other types of noise which sometimes show. Fixed pattern noise
sometimes shows in shadow areas, especially in older model cameras.
This factor alone would make me tend to choose a newer camera model
over an older one. Another type of noise is thermal noise that builds up
with longer exposure times. This can be a factor in long exposure
(e.g. night images) lasting many minutes.
The dominant noise factors bird and other wildlife photographers need to
be concerned with are the first two above: photon noise and read + electronics
noise.
Pixel size does not generate noise; small pixels do not "generate" more noise.
To put into perspective the two dominant noise sources, consider a 7D operating
at ISO 400. On a normal exposure, and 18% gray card, the 7D will capture
about 1500 photons, which generates 1500 electrons. Thus the
S/N = square root 1500 = 39. The read + electronics noise is only 4.9 electrons.
Thus, the dominant noise is photon noise.
The 1DIV in comparison will record 2500 photons on the same gray card,
giving a S/N = 50. The read + electronics noise is only 6.6 electrons.
Case closed: the 1DIV has lower noise than the 7D. This is what leads to
the perception that the 7D is noisy, but that is not the whole story
and is in fact incorrect that the 7D is noisier.
One other thing to be aware of: ISO does not change sensitivity, ISO only
changes what range of the signal gets digitized, and instructs the camera
to change its shutter speed.
So we want the most photons for our images. We collect photons by:
1) The lens. The larger the diameter the lens, the more photons we collect,
much like a larger bucket in a rain storm collects more water.
2) The exposure time. Longer exposure time lets more light in, like leaving
the bucket out in the rain storm longer.
3) The lens focal length spreads out the light. The longer the focal length,
the more the light collected by the lens is spread out. This is why when you
add a TC, you must increase your exposure time.
4) The pixel. A larger pixel collects the light streaming from the lens.
Just like the bucket in the rain storm, the pixel is like a little bucket collecting
the light from the lens.
5) Pixel efficiency. This is the fraction of photons incident on the pixel and
that are trensmitted through the overlying filters that actually generate
electrons in the pixel. It is the transmission factor of the blur and Bayer filters,
times the transmission of the micro lens, timesthe quantum
eiffciency of the pixel. For a given era of camera sensors, the efficiencies
tend to be very close. The 7D and 1DIV efficiencies measure very close
in lab test, with the 7D coming out a few percent higher. The much older
1DII in contrast has a pixel efficiency about 2 times lower.
Now, there is one other factor of concern to photographers and a key factor
in noise and detail on a subject. It is the combination of pixel size (pixel
spacing) and focal length that delivers detail. But more detail means
less light per pixel. If one wants to double the pixels on subject, you can
double the focal length (add a 2x TC) or halve the pixel spacing. In one case
(focal length) we spread the light out more so there is less light per pixel,
and in the other case we chop it into finer pieces so there is less light
per pixel. There is no difference in total light per pixel per time interval
in either of these methods of getting more pixels on the subject.
So, in the case of a 7D (4.3 micron pixels) versus 1D Mark IV (5.7 micron
pixels), if we equalize the pixels on the subject by changing focal length on
one camera, then the light delivered to each pixel is the same. If the light is
the same, the S/N is the same (with minor differences due to the relative pixel
efficiencies).
So how to do this in practice? The ratio of the pixels is 5.7/4.3 = 1.33.
So the 1DIV needs 1.33 times more focal length than the 7D to give the
same pixels on subject. Next the aperture diameter of the lens must
remain the same, as that equalizes the light per time interval.
Finally, use the same exposure time.
So here is a proposed test to prove my point (this is basic physics so
I am confident in this prediction). (I use this kind of model prediction in
understanding the signals from spacecraft systems.)
Set up two tripods and telephoto lenses side by side looking at the same
target. For example, 500 f/4 lenses. On one lens put a 7D.
On the other put a 1.4x TC and the 1D4. The 1D4 will have
about 6% more pixels on the subject, so very close.
Next set the camera to manual and the same exposure time on
both cameras. Finally set the aperture diameter the same.
That means the f/ratios are different. For example.
f/4 on the 7D setup and f/5.6 on the 1D4 setup, or
f/5.6 on the 7D and f/8 on the 1D4 setup.
Note, with exposure time the same, depending on light, the 7D
may need to be at a lower ISO so it doesn't saturate.
An interesting additional factor with the differing focal lengths,
pixel sizes and f/ratios is that the depth of field will be identical.
So my challenge is for two photographers with the same model lenses,
one with a 7D and one with a 1D4 try the above test. Process the images
identically and show the images. I predict it will be very difficult
to tell them apart (except the 1D4 image will have slightly more
pixels on subject). But the noise difference will be difficult
to tell apart.
For those who want to get more into the math, google Etendue and read
the wikipedia page.
Another implication of this equalization of systems (the equalization of the
Etendue), is that the camera with smaller pixels needs less focal length to
achieve the detail on the subject. This leads to several other implications.
If one needs a 2x TC on the 1DIV, one only needs a 1.4x TC on the 7D. That
means a slower f/8 AF performance can be avoided! If one needs a 1.4x TC
on the 1D4, one needs no TCs on the 7D, reducing optical elements,
potentially improving image quality further.
Bottom line: understanding the management of pixels on subject and
exposure time for a given lens, one can use a wider range of cameras
and achieve high resolution quality images on distant subjects.
There are many reasons to choose a 1D4 camera over the 7D, including
the 45 point AF system, and weather sealing, but noise and detail on
subject and noise performance can be equalized. These concepts hold for
other cameras being compared when the pixel size varies between models.
If one does not have the money for a 1D series, the 7D is an excellent
performer.
Finally, large pixel size is not the metric for low noise as several other factors
control how much noise we see in our images. But knowing these concepts
I believe will enable people to manage their photography in the field
to obtain better S/N regardless of camera and its pixel size.
Here is a demonstration of the effects described in the above post on Etendue. Three cameras, with pixels sizes ranging by almost a factor of 2, same lens, same exposure time, same ISO (1600), same lens aperture diameter. Can you tell what the cameras are? I'll give the answers, but a hint: they are mentioned above. Another hint. The signal-to-noise ratios on a 16x16 pixel block on the gray patch below the 1 in "1px" are (in no particular order): 8.4, 10.5, 11,8, again in the order I would predict. Which camera has the highest S/N, and which has the lowest? The images are from the raw data, converted with the same settings in ACR, then a very small white balance correction (less than 1%) applied in photoshop. Statistics on 16-bit data, Adobe RGB. Presented here, of course is 8-bit sRGB jpeg.
Roger
Last edited by Roger Clark; 01-12-2012 at 03:08 PM.
The middle image has the best detail and sharpness, to my eye, and the outer two have more noise, but seem about equal. I have no idea which camera is which.
I tried the Wikipedia reference but did not get very far...
A couple of questions:
When you say that 7D will capture about 1500 photons and the 1DIV will capture 2500 photons on the same 18% gray card, is that total or by pixel? Is signal-to-noise ratio measured at the pixel level or at some other level? Also, how does pixels on subject relates to camera crop factor? Sorry if I am missing something obvious, but perhaps others have similar questions.
Great thread as usual. Back in the days when the Mark IV was just coming out, I proposed a similar test on this forum in a focal-length-limited scenario comparing the 7D and Mark IV using the same lens (not adding a TC to the Mark IV) and cropping both images to frame the subject identically. By doing so, the Mark IV's image would require heavier cropping when compared to the 7D. (1.3x vs 1.6 crop factor). I never had any takers, so I ended up buying a Mark IV and doing the test myself.
As it ended up, I did a three-way comparison photographing a feather from a set distance using the same lens (600mm) with the 5D Mark II (no crop), Mark IV (1.3x crop) and 7D (1.7x crop). To frame/crop the feather identically, I had to do extensive cropping with the 5D Mark II, moderate cropping with the Mark IV and very little cropping with the 7D.
So what were the results? The 5DM2 and Mark IV looked very close w/o any sharpening and noise reduction. The 7D's image was slightly noisier but showed more detail. I then took all three test images, applied noise reduction, then sharpened. The bottom line was that all three of the images were virtually identical once processed.
Occasionally I give presentations to photography clubs and I always show the side-by-side comparison with all three cameras. The bottom line is that when you are photographing wildlife from a set distance using the same lens, the 7D, Mark IV and 5DM2 all have very similar noise and resolution once you crop the images identically.
When I am asked which camera someone should buy for bird photography, my suggestion is either the 7D or Mark IV. I feel both of these bodies have very good autofocus and fps, with the Mark IV being slightly better. Image quality from both cameras in a focal-length-limited scenario (where you can't move closer to the bird) is very, very close even at 100%.
One final comment...my ideal choice for a birder photography camera is a full-frame sensor with 10 fps and enough resolution to do fairly heavy cropping. With a full-frame sensor, there's less chance of me clipping a wing out-of-frame. Is the 1Dx the camera? I'm not quite sure. 18 MP is somewhat limited resolution-wise when having to do a heavy crop. I am anxiously awaiting RAW samples from this camera to do some testing.
When you say that 7D will capture about 1500 photons and the 1DIV will capture 2500 photons on the same 18% gray card, is that total or by pixel? Is signal-to-noise ratio measured at the pixel level or at some other level? Also, how does pixels on subject relates to camera crop factor? Sorry if I am missing something obvious, but perhaps others have similar questions.
Enrique,
That is 1500 photos per pixel and the noise is on a per pixel basis. Pixels on subject has no relation to crop factor when the subject is small in the frame, like a distant bird. How many pixel high a bird is in the image is set only by the focal length and pixel size. Crop factor only sets field of view.
Here is a demonstration of the effects described in the above post on Etendue. Three cameras, with pixels sizes ranging by almost a factor of 2, same lens, same exposure time, same ISO (1600), same lens aperture diameter. Can you tell what the cameras are? I'll give the answers, but a hint: they are mentioned above. Another hint. The signal-to-noise ratios on a 16x16 pixel block on the gray patch below the 1 in "1px" are (in no particular order): 8.4, 10.5, 11,8, again in the order I would predict. Which camera has the highest S/N, and which has the lowest? The images are from the raw data, converted with the same settings in ACR, then a very small white balance correction (less than 1%) applied in photoshop. Statistics on 16-bit data, Adobe RGB. Presented here, of course is 8-bit sRGB jpeg.
Roger
Here are the detail on the comparison. Left: 7D + 500 mm lens at f/4. Center: 1D4 + 500 mm + 1.4x TC at f/5.6, Right: 1D2 + 500 mm + 2x TC at f/8. So the 7D image has the highest S/N, the 1D4 next and the 1D2 is the lowest. But there is a slight bias. The 7D pixels are 4.3 microns, and 1D4 are 5.7 microns, for a ratio of 1.33. As we only have 1.4x TCs, the 1D4 image size (pixels on subject) is 5.2% more magnification so 5.2% more pixels on subject. But that also means the signal is lower by 1.052 squared, so the S/N on the 7D would be 5.2% higher. I measured the S/N of the 7D (on the gray patch noted above) to be 11.8, the 1D4 S/N = 10.5 and the 1D2 S/N = 8.4. Thus, the 7D came out 12% higher. But the measured efficiency of the 7D is about 6% higher than the 1D4 (see Figure 10 at: http://www.clarkvision.com/articles/...ary/index.html ). Thus with 5% and 6% more we should have seen a ratio of 7d to 1D4 S/N's of 11% amd we measured 12%, which is impressively close!
Now the 1D4 to 1D2 pixel ratios is 8.2/5.7 = 1.44, so the jump from 1.4x on the 500 with the 1D4 to the 2x on the 1d2 is virtually a perfect match. If the 1D2 were the same pixel efficiency as the 1D4, the images would be identical (assuming no loss from the 2x TC). But the 1D2 pixel are less efficient by over a factor of 2 (see Figure 10 referenced above), so the S/N on the 1D4 should be higher by about 1.45x but it measured as 10.5/8.4 = 1.25, so the 1D2 came out 16% better than expected, which is still very close. My 1D2, deing quite old with a lot of shutter actuations shows a little variation from exposure to exposure so the difference is likely due to variation in shutter speed. Note this is only 0.2 stop.
So the results of the test are very close to predicted and well explained by Etendue. Equalizing the cameras, especially 1D4 and 7D produces nearly identical images. But wait, the 1D4 image looks a little better than the 7D image. That is because the focal length is slightly greater so slightly better resolves the detail. It someone has a 1.3x TC, we could make the images look even closer.
The bottom line for photographers is the 7D is not a noisy camera compared to other cameras. it is all in how it is used. 1D series cameras have many advantages but not in lower noise when one equalizes pixels on subject and exposure trme with the same lens. If I couldn't afford a 1D series in the canon line, I would choose the 7D for bird photography. The same principles apply to all other cameras, especially those made in the same generation. Smaller pixels give more pixels on subject with a shorter focal length lens, and larger pixels do not have an advantage concerning noise.
Another factor: f/8 AF is slower. If one needs 2X TCs and f/8 a lot, it would probably be better to get a camera with smaller pixels and work at f/5.6 for the faster AF.
I'm really looking forward to a full frame camera with pixels the size of the 7D (=46.7 megapixels) at 10 frames /sec.
Roger,
Here in Florida we can get closer to our subject. Many times I can start out with a 1.4 TC and move close enough to remove the TC and have the same magnifaction.
My question is, given the above which would you prefer, an image from the7D with a 500mm or a Mark IV with a 500mm lens? The subject would fill the frame equally, you would just be a little closer with the Mark IV.
Thanks
David, what James means is if subject size in the frame is same in both situations, with the same lens, he'd be closer to the bird with 1D mIV as it is 1.3x than he wud be with 7d..as it is 1.6x. He is not saying he will get closer because he is using 1D. BTW, he uses Nikon :-)
I know what James said Kaustubh, but he's not making sense because the image will not be the same size with the Mk IV and the 7D because in the real world of shooting birds we get as close as we can, usually, and we take the best image we can with the camera we have in hand. He's asking for a comparison of two cameras' performance in a theoretical situation that does not exist in the real world of shooting birds, at least for most of us. Unless James is often in the situation where he takes a picture of a bird with his MkIV and then says to himself, "How far do I back up to take the same image with a 7D?", then it's a irrelevant comparison for most of us shooting birds.
I shoot with a 7D and 500mm in Colorado, often with my 1.4xTC attached. I feel really lucky when I get close enough that I don't need to crop the image some in PP. The more appropriate question for Roger is, if I crop a MkIV image so the subject is the same size as a 7D image, which would you (Roger) prefer? I think Roger has already answered that.
There are good reasons to own the MkIV. Fighting my AF with an eagle approaching at relatively slow speed when I've got the 1.4x TC mounted, is just one example. I'm sticking with my 7D for now, but waiting to see real world reviews of the 1DX and 5D3 before I might change bodies, if I do.
I know what James said Kaustubh, but he's not making sense because the image will not be the same size with the Mk IV and the 7D because in the real world of shooting birds we get as close as we can, usually, and we take the best image we can with the camera we have in hand. He's asking for a comparison of two cameras' performance in a theoretical situation that does not exist in the real world of shooting birds, at least for most of us. Unless James is often in the situation where he takes a picture of a bird with his MkIV and then says to himself, "How far do I back up to take the same image with a 7D?", then it's a irrelevant comparison for most of us shooting birds.
I shoot with a 7D and 500mm in Colorado, often with my 1.4xTC attached. I feel really lucky when I get close enough that I don't need to crop the image some in PP. The more appropriate question for Roger is, if I crop a MkIV image so the subject is the same size as a 7D image, which would you (Roger) prefer? I think Roger has already answered that.
There are good reasons to own the MkIV. Fighting my AF with an eagle approaching at relatively slow speed when I've got the 1.4x TC mounted, is just one example. I'm sticking with my 7D for now, but waiting to see real world reviews of the 1DX and 5D3 before I might change bodies, if I do.
David I think you have trouble understanding bird photography. Who do you mean by most of "us" ? who is "us"? Who said bird photography is getting as close as you can and filling the frame with the bird? That's how you shoot postage stamps. I don't photograph like that...Doug Brown doesn't photograph like that, Jim Neiger doesn't photograph like that, Jim Shadle doesn't photograph like that...Ofer Levy doesn't photograph like that...so be careful before generalizing your ideas.
You need to adjust your distance to the bird according to the situation, FL and a lot of other things, if you can't or don't know how you have to go back home and try to another day...
"David, what James means is if subject size in the frame is same in both situations, with the same lens, he'd be closer to the bird with 1D mIV as it is 1.3x than he wud be with 7d..as it is 1.6x. He is not saying he will get closer because he is using 1D. BTW, he uses Nikon :-)"
Obviously, I did not say it as well as Kaustubh . He nailed it.
I use a D7000 1.5 DX sensor and a D700 Full Frame FX sensor.
The difference in magnification between a 1.6 crop sensor (7D) and a 1.3 crop sensor (Mark IV) is less than what I normally face.
Last edited by James Shadle; 01-18-2012 at 07:28 PM.
How does carrying a Mk IV in Florida make it possible to get closer to the subject than when carrying a 7D?
One, I'm very good at getting as close to my subject. Second, the birds in Florida are habituated to humans and very approachable.
No, No No, James, you didn't get the question. The answer is clear. The birds in Florida are experienced with photographers and are more enamored with those carrying pro cameras and let them come closer Or how about this one? The birds know the difference between the pro photographer (with the pro camera) and know the photographer just wants to take their picture, so lets them closer. The amateur and tourists are unpredictable so the birds stay further away.
Roger,
Here in Florida we can get closer to our subject. Many times I can start out with a 1.4 TC and move close enough to remove the TC and have the same magnifaction.
My question is, given the above which would you prefer, an image from the7D with a 500mm or a Mark IV with a 500mm lens? The subject would fill the frame equally, you would just be a little closer with the Mark IV.
Thanks
Hi James,
This is a simple answer. Because you are used the same lens but moved closer and reduced its focal length (removing the TC), the signal-to-noise ratio goes up with the 1DIV at the closer range. So the 1DIV closer with no TC will be the better image.
(I had a similar question on a graduate exam many years ago when I was teaching radiative transfer at the University of Hawaii--most got the question wrong, although it was a little more complex involving a spacecraft approaching a planet.)
The detailed answer:
When one changes distance, one is changing the Etendue. Using the same lens means the lens collects the same amount of light, but since you moved closer (1/r squared law) there is more light per pixel, and since you took the TC off to concentrate that light onto the pixels, the light goes up. So in the same exposure time, assuming you weren't constrained by depth of field limits, you would have more light so could reduce ISO and get a better image. Conversely, if the subject were moving, you could use a faster exposure to better freeze the action and get the same signal-to-noise ratio.
This, of course ignores issues like which is the better background/foreground and depth of field, which may trump all of this.
Now what if I have a 7D with 500mm lens and I shoot stopped down to F8 or F11 let's say. I also have a Mark IV with a 500mm and a TC1.4 shooting at the same F8 or F11 aperture. (I understand that with the larger sensor and being closer to the subject the Mark IV is going to offer shallower DOP than the 7D rig)
Shouldn't the Mark IV produce a superior image with it's better signal to noise ratio?
Can two camera's with identical pixel size have different noise characteristics due to the way they are "wired" and the way the camera processes the image?
The bottom line seems to be, get as close as possible with as little lens as possible.
Last edited by James Shadle; 01-18-2012 at 11:07 PM.
Now what if I have a 7D with 500mm lens and I shoot stopped down to F8 or F11 let's say. I also have a Mark IV with a 500mm and a TC1.4 shooting at the same F8 or F11 aperture. (I understand that with the larger sensor and being closer to the subject the Mark IV is going to offer shallower DOP than the 7D rig)
Shouldn't the Mark IV produce a superior image with it's better signal to noise ratio?
The 1DIV with a 1.4x TC has almost the same angular pixel size as the 7D on the same lens with no TC. So regardless of how close you move, the S/N will remain the same per pixel (assuming both systems at the same aperture diameter and same shutter speed). But if you move closer with the 1DIV system, you are resolving more detail. People have grown conditioned to think larger pixels = less noise, but pixels are only part of the system as it is the lens that collects the light. So you said working at the same f/8 (let's say f8 and not 11). F8 on the 1DIV and 700 mm = 700/8 = 87.5 mm diameter aperture, and the 7D with no TC at f/8 = 500/8 = 62.5 mm diameter aperture. So the 7D system is collecting less light to deliver to the pixels. In that case, the 1DIV collects more light in a given exposure time, but not because of larger pixels, but because of the larger aperture diameter lens. Photography has become confusing, especially in the digital era, because of f/ratios.
But it is also true that to collect more photons, one must have larger pixels, like a larger bucket in a rain storm is needed to collect more water. But while the pixel collects the light, it is the lens that delivers the light, and the larger the diameter of the lens, the more light you pump to the pixel. The larger pixel simply enables more light to be collected without overflowing.
Originally Posted by James Shadle
Can two camera's with identical pixel size have different noise characteristics due to the way they are "wired" and the way the camera processes the image?
Yes, but it is usually (always so far?) because of improving technology. For example the new Canon 1DX will have about the same pixel size as the Canon 1D Mark II (both close to 8 microns), but the 1DX will collect more light per pixel with a given lens because the pixels will be more efficient (better transmission on the blur and Bayer filters, better micro lenses, and perhaps higher quantum efficiency of the pixel itself). The 1DX should be at least 2 times more efficient than the 1DII pixels from the 2004 era.
Originally Posted by James Shadle
The bottom line seems to be, get as close as possible with as little lens as possible.
But a little lens collects less light. More like get as close as possible with as big a lens as possible to collect the most light, with the caveat of depth of field may require a small aperture diameter.
Roger,
What is the difference in light transmission between a 600mm F4 @ F8 compared to a 300mm F2 @ F8?
How about the difference in light transmission between a 800mm F5.6 @ F8 compared to a 400mm F2.8 @ F8?
Do the smaller lenses collect less light in this example?
Roger,
What is the difference in light transmission between a 600mm F4 @ F8 compared to a 300mm F2 @ F8?
How about the difference in light transmission between a 800mm F5.6 @ F8 compared to a 400mm F2.8 @ F8?
Do the smaller lenses collect less light in this example?
Hi James,
I assume you mean the total light collected by the lens, not the percentage transmitted through the glass.
Total light would be proportional to the area of the lens aperture, so:
600 f/4 at f/8: 600/8 = 75 mm diameter = 4418 sq mm area
300 f/2 at f/8: 300/8 = 37.5 mm diameter = 1104 sq mm area
800 f/5.6 at f/8: 800/8 = 100 mm diameter = 7854 sq mm area
400 f/2.8 at f/8: 400/8 = 50 mm diameter = 1963 sq mm area
If you pointed these lenses at the north star at night on a tripod and made 30-second exposures, the 800 f/8 combination would collect the most light and record the faintest stars. But if you opened each lens to wide open, the 600 mm lens has the largest diameter aperture (150 mm) where the 400 and 800 mm lenses have a diameter of 142.8 mm and the 300 f/2.8 only 107 mm. Thus, the 600 mm lens would record the faintest stars.
If you pointed these lenses at the north star at night on a tripod and made 30-second exposures, the 800 f/8 combination would collect the most light and record the faintest stars. But if you opened each lens to wide open, the 600 mm lens has the largest diameter aperture (150 mm) where the 400 and 800 mm lenses have a diameter of 142.8 mm and the 300 f/2.8 only 107 mm. Thus, the 600 mm lens would record the faintest stars..
What if you pointed them at a bird?
I assume you mean the total light collected by the lens, not the percentage transmitted through the glass.
The light transmitted through the glass is what hits the sensor?
Help me with this - using ISO 400, I expose a white subject on a full sunny day @ F8 - 1/2000 SS.
That is my setting regardless of the lens I use. How does one lens allow more light to the sensor at F8 than another one does?
If that was the case, I would need to adjust my ISO or shutter speed to compensate.
If you pointed these lenses at the north star at night on a tripod and made 30-second exposures, the 800 f/8 combination would collect the most light and record the faintest stars. But if you opened each lens to wide open, the 600 mm lens has the largest diameter aperture (150 mm) where the 400 and 800 mm lenses have a diameter of 142.8 mm and the 300 f/2.8 only 107 mm. Thus, the 600 mm lens would record the faintest stars..
What if you pointed them at a bird?
I assume you mean the total light collected by the lens, not the percentage transmitted through the glass.
The light transmitted through the glass is what hits the sensor?
Help me with this - using ISO 400, I expose a white subject on a full sunny day @ F8 - 1/2000 SS.
That is my setting regardless of the lens I use. How does one lens allow more light to the sensor at F8 than another one does?
If that was the case, I would need to adjust my ISO or shutter speed to compensate.
Good question. The answer is a little complex but is the basis of why a small-sensor point and shoot camera produce images that appear much noisier than DSLRs.
First, changing ISO does not change the amount of light the camera collects. ISO only sets the range that is converted to digital.
Second, true exposure on cameras with different sized pixels is not the same. Exposure for digital cameras is relative to how large the charge capacity (electrons generated by photons in each pixel). Cameras with larger pixels can collect more photons (electrons) in the pixel. Thus to properly expose a large pixel (say 18% gray card) requires more photons per pixel than the exposure on a tiny pixel. But generally with cameras (think small point and shoot to full frame DSLR), the lens also scales with the sensor size, so the larger lens collects more light for the larger pixel, all in proportion. Photographers see it as the same "exposure" but it is actually quite different. It's like saying 3 people have filled a container with water half way. But we don't know how much water each has because they could be holding different sized containers. Same with pixels and light.
More details. Consider two cameras, each 10 megapixel, one 35mm full frame, and one 1/4 the size of the first. Put a lens on each camera that gives the same field of view: 100 mm f/8 on the large camera and 25 mm f/8 on the small camera. The resolution and pixels on subject are the same. Take a picture with both cameras with the same camera exposure at the same ISO. Same shutter speed, same ISO, same f/ratio, same field of view, same resolution on subject. But different light per pixel! Different Etendue, and different true exposure The exposure only looks the same to the photographer because the pixels are filled to the same fraction of full capacity. The lenses collect different amounts of light (a factor of 16 less light for the small camera), but the pixels can hold the same proportionally less light (factor of 16). Exposure in digital cameras is a proportion of full capacity of the pixel, not an absolute amount. The largepixel camera has a 100/8 = 12.5 mm diameter lens and the small camera 25/8 = 3.125 mm diameter lens. The larger lens collects 16 times more light during the same shutter speed, so the larger camera has 16 times more light per pixel resulting in 4 times the signal-to-nise ratio. (Small sensor P&S cameras can never match large sensor DSLR S/N due to these basic physical facts).
The star test reveals these differences. Stars, being point sources, do not get magnified with different lenses like a bird or other close subject would. So for a given f/ratio the diffraction spot size is the same regardless of lens focal length, but the lens aperture controls the collection of light. So in the star test, the different diameter lenses record stars to different degrees of faintness. For example, if one has a 500 mm f/4 lens, make a 30-second exposure of the north star at f/4, then put on a 50 mm lens at f4 and do the same exposure. The 500 mm lens will show many fainter stars. even though the "camera exposure" is the same. The Etendue is quite different and tells the true story, and so does an assessment of the faintest stars recorded.
Back to your original question. On an extended subject (not stars), like birds or a general scene, the constant f/ratio scenario has two things balancing: longer focal length lens collects more light, but also spreads the light out more. So all f/8 lenses deliver the same pixels per square micron at the sensor (assuming the have the same transmittance) on any given scene, making exposure calculation for standard scenes (not point sources like stars) simpler.
So the bottom line after all this off topic stuff is there is no need to fork out for a 1D unless you want or need to use AF extenders with your big white lenses? Is that what you are getting at?
So the bottom line after all this off topic stuff is there is no need to fork out for a 1D unless you want or need to use AF extenders with your big white lenses? Is that what you are getting at?
Hi Tony,
I would put it differently: There is no need to get a 1D camera if you think it will be lower noise in focal length limited situations like bird photography. There are plenty of reasons to go for a 1D series, including faster response, 45-point AF better sealing, and little known: the 1D series has tighter specification on the circle of confusion so focus is more accurate. Also, in my testing the AF on the same lens is faster on the 1DIV. But noise should not be a consideration. Nether should pixels on subject for subjects small in the frame be a factor because it is almost the same, even considering TCs.
My original choice for 1DIV versus 7D was a 1DIV and I later bought a 7D (and I already had a 1DII and 5DII) . Both 1DIV and 7D have advantages. 7D is also great for macro.
"First, changing ISO does not change the amount of light the camera collects. ISO only sets the range that is converted to digital."
Of course I understand that. My point was, if lenses with different length/maximum apertures allowed different levels of light to reach the sensor at F8, I would need to compensate by changing the ISO or SS. From my experience, I exposure is the same @ F8 using my 300mm F5.6 or My 600 F4
"There is no need to get a 1D camera if you think it will be lower noise in focal length limited situations like bird photography."
I tend to disagree that bird photography is focal length limited. Getting close to your subject is also a bird photography "technique".
If I understand correctly, the 7D noise with a 500mm F4 lens should be similar to the noise of a Mark IV with a 500mm F4 and a 1.4 TC(in the same light and subject size in frame).
So if the Mark IV image quality is degraded, it will have noise close to that of a 7D. Likewise, the 7D with a 500mm F4 lens will have more noise than a Mark IV with a 500mm F4.
Are your tests done with in camera noise reduction turned on?
Does pixel size affect dynamic range?
Can lower dynamic range cause an perception of increased noise(by trying to pull additional details out of the shadows)?
Why Nikon D3s have a one stop (+/-) advantage over the Nikon D3/D700 when they have the same or nearly the same size pixels?
Thanks!
Last edited by James Shadle; 01-19-2012 at 10:33 PM.
James Shadle:
"First, changing ISO does not change the amount of light the camera collects. ISO only sets the range that is converted to digital."
Of course I understand that. My point was, if lenses with different length/maximum apertures allowed different levels of light to reach the sensor at F8, I would need to compensate by changing the ISO or SS. From my experience, I exposure is the same @ F8 using my 300mm F5.6 or My 600 F4
Hi James (I'll do my response in red)
In moving from 600 mm at f/8 to 300 mm at f/8, you've done two things: 1) the 300 f/8 has half the lens diameter so collects 1/4 the light. 2) You've decreased the focal length by 2x so increased the angular size of the pixel by 2x (4x angular areal extent) thus compensating for the loss in light gathering. Thus the light gathered per pixel for an extended object for the same shutter speed stays the same. Stars, not being extended objects will not, however record the same: the 600mm at f/8 will record fainter stars compared to the 300 mm at f/8.
"There is no need to get a 1D camera if you think it will be lower noise in focal length limited situations like bird photography."
I tend to disagree that bird photography is focal length limited. Getting close to your subject is also a bird photography "technique".
I agree with you on this. I should have said something like:
There is no need to get a 1D camera if you think it will be lower noise in focal length limited situations that sometimes occur in bird photography
If I understand correctly, the 7D noise with a 500mm F4 lens should be similar to the noise of a Mark IV with a 500mm F4 and a 1.4 TC(in the same light and subject size in frame).
Yes, and show close to the same detail given the same exposure time and lens diameter.
So if the Mark IV image quality is degraded, it will have noise close to that of a 7D.
I would say the same, not degraded.
Likewise, the 7D with a 500mm F4 lens will have more noise than a Mark IV with a 500mm F4.
Correct, because the 500 f/4 lens delivers the same amount of light, but the 7D pixels divide the image into finer detail. So the 7D pixels collect less light per pixel and will be noisier but the image will be more detailed.
Are your tests done with in camera noise reduction turned on?
No, and recorded as raw and all processed the same.
Does pixel size affect dynamic range?
Yes. Larger pixels will have the potential for greater dynamic range, but one can compensate. For example, if the 1DIV image were at ISO 800 and an exposure time of 1/1000 second, and the 7D were at ISO 400 and 1/1000 second, the dynamic range would be nearly identical (10.7 versus 10.8 stops) and the noise would still be the same, and the total light collected would be very close to the same (small differences in pixel efficiency).
Can lower dynamic range cause an perception of increased noise(by trying to pull additional details out of the shadows)?
Yes, particularly at high ISO. At higher ISOs, sensor read noise will be an increasing factor in the shadow noise.
Why Nikon D3s have a one stop (+/-) advantage over the Nikon D3/D700 when they have the same or nearly the same size pixels?
I haven't seen the sensor data for the D3s so can't really comment for sure. The D3s is a newer generation so probably has more efficient pixels and better electronics.
Roger
I understand the 7D has an extra anti-alias filter compared to the 1D IV. Is it possible that that extra filter degrades the image or makes it appear less sharp than a 1D IV? I own a 7D and have never used a 1D IV but I hear other say they feel you can get a sharper image from the 1D IV and it sounds like it is not all due to the focusing system. Anything to that?
Do you have to have extra anti-alias filters when you get to a certain size of pixel to prevent a moire pattern? I think I've seen discussion that thins might be able to be removed in post processing.
I understand the 7D has an extra anti-alias filter compared to the 1D IV. Is it possible that that extra filter degrades the image or makes it appear less sharp than a 1D IV? I own a 7D and have never used a 1D IV but I hear other say they feel you can get a sharper image from the 1D IV and it sounds like it is not all due to the focusing system. Anything to that?
Do you have to have extra anti-alias filters when you get to a certain size of pixel to prevent a moire pattern? I think I've seen discussion that thins might be able to be removed in post processing.
Thanks for the great info.
Hi Dave,
Interesting question. It is also possible the dual anti-alias filter improves image quality over a single one. The image I posted show the 7D looks a little less detailed, but the image scale is also slightly different. If I get some time this weekend, I'll reshoot the test with the 7D slightly closer to equalize the pixels on the subject and then be able to better tell if there is any effect due to anti-alias filter. I predict the 7D and 1DIV will be very very close.
Smaller pixels should need less anti-alias filtering because diffraction and other aberrations are also doing that. As one moves to larger pixels, a stronger anti-aliasing filter is required, thus impacting fine detail more. This shows in the 1DII image in the test I posted. But it is a small effect.
Another issue with some of the newer cameras is they have 2 green pixels, with with different spectral response. Thus reconstruction of the RGB image is more complex and we saw some software produced mazing effects in 7D images. But it seems photoshop and other raw converters have that worked out pretty well now.
A very interesting thread, Roger. Thanks for sharing. I'm still trying to absorb your explanation in pane 26.
A slightly off-topic question, though: In the comparison image in pane 3, shouldn't the 7D's pixel density offset the 1D4 setup's focal length advantage in terms of resolving detail?
A very interesting thread, Roger. Thanks for sharing. I'm still trying to absorb your explanation in pane 26.
A slightly off-topic question, though: In the comparison image in pane 3, shouldn't the 7D's pixel density offset the 1D4 setup's focal length advantage in terms of resolving detail?
Hi Mark,
The 7D pixels are 4.3 microns and the 1DIV 5.7, for aa ratio of 5.7/4.3 = 1.33. On the image in pane 3, the 1DIV had a 1.4x TC on so the 1DIV image should appear about 1.4/1.33 = 1.05 times larger and that is what we observe. So the 7D pixel density almost bit not quite compensates for the 1DIV with a 1.4x TC added. Does that answer your question?
If you were to take an identical image from both cameras at the same set distance and then crop them to the same field of view, the quality of the images are very, very close. The reason for this is that you will have to crop heavier with the Mark IV's image vs the 7D's image. The heavier you crop an image, the greater the noise level will be. Conversely, the less you crop an image, the impact of perceived noise will be less. This somewhat equalizes the quality of images from both cameras.
If you have both cameras, do the test and see for yourself. Identically cropped from a set distance, the Mark IV's image will still be a tad less noisy but pretty close. The 7D's image will show a tad greater resolution. Equalling them out with noise reduction, and the image will appear almost identical even at 100%. The reason for this is that you are cropping the image heavier with the Mark IV than with the 7D to obtain the same post-processed field-of-view.
With that stated, the Mark IV still has faster fps and better autofocus, but the quality difference is not that great when comparing the two cameras from a set fixed distance.
the quality difference is not that great when comparing the two cameras from a set fixed distance.
Your statement may or may not be true Alan; I think the noise is substantially greater on the 7D, especially at ISO 800 and greater. But what I said was that it's very seldom that I can't get a little closer to my subject if necessary. I try to keep my subject roughly the same size in the viewfinder no matter what body/lens/TC combination I'm using.
Hi Mark,
The 7D pixels are 4.3 microns and the 1DIV 5.7, for aa ratio of 5.7/4.3 = 1.33. On the image in pane 3, the 1DIV had a 1.4x TC on so the 1DIV image should appear about 1.4/1.33 = 1.05 times larger and that is what we observe. So the 7D pixel density almost bit not quite compensates for the 1DIV with a 1.4x TC added. Does that answer your question?
Roger
Thanks, Roger. I got it now. I just got the variables mixed up.
A competent bird photographer does not chain themselves to a tree and photograph a far subject and then try to crop the image. Heavy crops and farming noisy pixels is the number 1 reason for those lousy IQ images that we often see posted. To think bird photography is generally "focal length limited" is ridiculous IMO and it indicates a short coming on the photographer side. If the birds are too far and you can't get close enough you should not waste those electrons taking a picture at first place...
We will soon put together an article that not only scientifically compares noise based on RAW data and spectral analysis it will also provide field-relevant visual examples to see what differences there are. Please stay tuned, the article will be posted soon and many of these issues will be addressed.
Last edited by arash_hazeghi; 01-21-2012 at 08:30 PM.
Hi Doug,
Is it possible to include SS, Fstop, and subject distance for the two images?
Also, can we assume the same exposure adjustment for both in LR?
thanks!
Enrique
Arash, I'm wondering what you are hopping to accomplish with the comments you made. I can't imagine this would lead to respectful discussions with others. I know I would really appreciate it if you could change the harshness of the language.
Originally Posted by arash_hazeghi
A competent bird photographer does not chain themselves to a tree and photograph a far subject and then try to crop the image. Heavy crops and farming noisy pixels is the number 1 reason for those lousy IQ images that we often see posted. To think bird photography is generally "focal length limited" is ridiculous IMO and it indicates a short coming on the photographer side. If the birds are too far and you can't get close enough you should not waste those electrons taking a picture at first place...
We will soon put together an article that not only scientifically compares noise based on RAW data and spectral analysis it will also provide field-relevant visual examples to see what differences there are. Please stay tuned, the article will be posted soon and many of these issues will be addressed.
Doug Brown, I believe what Roger is implying that if you are the same distance from the bird using both cameras then to get the bird to fill the same same of the final frame you will be adding an extender, increasing the f-stop by 1 stop and then increasing ISO by 1 stop also to create the same shutter speed. My understanding of this is that the 1D IV then is at least 1 stop better in noise when comparing ISOs.
In your scenario if you are always able to get as close as needed without adding another 1.4 extender then it would make sense the the 1D IV would be less noisy. Probably by at least 1-stop.
Arash, I'm wondering what you are hopping to accomplish with the comments you made. I can't imagine this would lead to respectful discussions with others. I know I would really appreciate it if you could change the harshness of the language.
Doug, I have to disagree, I don't think my language was harsh I just tried to make it obvious . I am honest and blunt in expressing what I think is right, I have learned it from Artie Morris himself Too bad he is not around to weigh in on this thread.
My comments above were general and not directed to any particular individual at all-I respect all other opinions but express mine as well-Thanks.
Last edited by arash_hazeghi; 01-21-2012 at 10:34 PM.
Examples of focal-length-limited scenarios in which I cannot move closer to the bird:
Example 1: I am in Florida photographing birds. Yes, the birds may be tame but I am photographing a bird from the boardwalk at Wakodahatchee Wetlands. I cannot move any closer to the bird or I will end up in the drink.
Example 2: I am photographing a Blue-Winged Warbler that is moving from branch to branch 30' off of the ground. I can only move closer if I climb the tree.
Example 3: I am photographing a Bald Eagle from Lock and Dam 14 in Le Claire, IA. I have my gear set up on the I-wall of the dam. I cannot move closer to the eagle unless I end up in the lock.
Example 4: I am photographing Short-Eared Owls hunting over a grassland. Although I may attempt to move closer, I am at the mercy of where the owl is flying at the time that I shoot. By the time I would attempt to move closer, I would miss my shot.
I could continue with examples of focal-length-limited scenarios, but I'm sure everyone is already bored with my post. There are plenty of times in which you are chained to a tree as you put it.
7D or 1DIV: better noise?
Originally Posted by Enrique Patino
. Second, the birds in Florida are habituated to humans and very approachable.
. He nailed it.
