Canon 1DX vs 5DIII Wildlife Comparison

[CarlTN]

RMC33, LOL !!...

Jrista, awesome owl pic...thanks for sharing it here...I want to shoot pics at least 80% that good! Wow, ISO 16,000...unbelievable! Frankly I could clean that image such that it could be printed at 24x36...with just ACR and CS5...

I want to pet that owl, and teach it not to be afraid of numbers higher than 1000...but I reckon it would prefer to peck me to death! ;D

Btw, one time I was out for a run in my front field, at night...and an owl flew right over my head...I never heard a sound...but the sight...scared the $tuff out of me...They truely make no sound at all in flight.

 

[jrista]

RMC33, LOL !!...

Jrista, awesome owl pic...thanks for sharing it here...I want to shoot pics at least 80% that good! Wow, ISO 16,000...unbelievable! Frankly I could clean that image such that it could be printed at 24x36...with just ACR and CS5...

Haha, yeah...I wish I could take photos HALF that good. A big part of it is having the gear...my 7D and 100-400 do ok...but they are definitely imposing a limit on what I can achieve in difficult scenarios.

 

[East Wind Photography]

Two shots using the 5DIII and 1DX. Both shot at ISO 1000. Both using a 600mm F4 IS with 1.4xIII extender. Both days were taken with complete overcast against a less than pleasing sky...Normally I would pass on any attempts in this condition but it makes for good comparisons in bad lighting. No post filters except 25 sharpness on both from the RAW files. 5DIII was at f/8 and 1DX was at F/10.

BE8 is the 1DX
NL7 is the 5DIII

Note that the 5DIII has smaller grain pattern than the 1DX due to pixel density. Chroma noise is about the same, I think maybe less with the 5dIII but I could be persuaded that it's the same. For those that say the 1DX has no chroma noise you are wrong. It may have less at ISO 25000 but you cant say it has none. Post with CS5 we can process out chroma noise pretty much at all ISO levels with both cameras.

Crops are 100% at 300dpi

 

[jrista]

Two shots using the 5DIII and 1DX. Both shot at ISO 1000. Both using a 600mm F4 IS with 1.4xIII extender. Both days were taken with complete overcast against a less than pleasing sky...Normally I would pass on any attempts in this condition but it makes for good comparisons in bad lighting. No post filters except 25 sharpness on both from the RAW files. 5DIII was at f/8 and 1DX was at F/10.

BE8 is the 1DX
NL7 is the 5DIII

Note that the 5DIII has smaller grain pattern than the 1DX due to pixel density. Chroma noise is about the same, I think maybe less with the 5dIII but I could be persuaded that it's the same. For those that say the 1DX has no chroma noise you are wrong. It may have less at ISO 25000 but you cant say it has none. Post with CS5 we can process out chroma noise pretty much at all ISO levels with both cameras.

Crops are 100% at 300dpi

(FYI, dpi only matters for print. On screen, you only have image dimensions in pixels and the PPI of the screen. It really wouldn't matter what dpi setting you used, on my screen it will always display at 103ppi. The average desktop screen these days has around 96ppi, most phone screens are 250-330ppi, and most tablets are 140-250ppi. People will see the image at the resolution of their device, regardless of what setting you choose when exporting. If you want everyone to see it at certain dimensions, there really isn't any way via the web, although I would imagine that if we actually could see those crops at 300ppi (my Lumia has a 336ppi resolution), they would look far less noisy than they do.)

Any chance you could post screenshots of the original histograms for each of those photos (without any post-process edits)? Given how much noise you have in those shots, and given the overall tone, I'd wager that a higher ISO setting with a greater amount of ETTR would produce better results from both cameras. I recently read some interesting info on exposure from one of my favorite bird photographers...Art Morris, who is a staunch advocate of ALWAYS exposing such that the rightmost ends of the histogram end up half way into the rightmost vertical box on the histogram display. Since I started doing that, my photos have been less noisy after post-process exposure correction (which, actually, tends to be minimal most of the time). Given what I've learned the last several months with my 7D, I would be willing to bet both the 5D III and 1D X could produce less noisy exposures for the same images you posted if you used both a higher ISO setting, and dropped the right-hand end of the histogram half way into the rightmost box.

It may not seem like "getting the exposure right in-camera", and if one was using film it would indeed be an entirely incorrect way to expose. But to maximize the potential of digital equipment, we have to think about it and use it differently than we think about and use film. I'm a staunch believer of ETTR, or expose to the right, when using modern digital cameras, and in my experience it does result in better dynamic range and noise performance out of any digital camera.

 

[bdunbar79]

For digital I agree with jrista. I always overexpose most of my shots and then when I bring them down in post, there is a lot less noise vs. going the other way. I agree, for film, not the way to go, but it's a trick I learned and take advantage of with digital. Most of my outdoor sports photos are at least +2/3 to +1 2/3 EV.

 

[East Wind Photography]

Yes I agree. These are not the best exposed images. As I said the lighting was awful and the shot with the 5DIII it had just started sleeting. It was overexposed by about 2 stops already. The 1DX was exposed about 1 stop over. I tried to find a couple of shots that were challenging in comparison.

Two shots using the 5DIII and 1DX. Both shot at ISO 1000. Both using a 600mm F4 IS with 1.4xIII extender. Both days were taken with complete overcast against a less than pleasing sky...Normally I would pass on any attempts in this condition but it makes for good comparisons in bad lighting. No post filters except 25 sharpness on both from the RAW files. 5DIII was at f/8 and 1DX was at F/10.

BE8 is the 1DX
NL7 is the 5DIII

Note that the 5DIII has smaller grain pattern than the 1DX due to pixel density. Chroma noise is about the same, I think maybe less with the 5dIII but I could be persuaded that it's the same. For those that say the 1DX has no chroma noise you are wrong. It may have less at ISO 25000 but you cant say it has none. Post with CS5 we can process out chroma noise pretty much at all ISO levels with both cameras.

Crops are 100% at 300dpi

(FYI, dpi only matters for print. On screen, you only have image dimensions in pixels and the PPI of the screen. It really wouldn't matter what dpi setting you used, on my screen it will always display at 103ppi. The average desktop screen these days has around 96ppi, most phone screens are 250-330ppi, and most tablets are 140-250ppi. People will see the image at the resolution of their device, regardless of what setting you choose when exporting. If you want everyone to see it at certain dimensions, there really isn't any way via the web, although I would imagine that if we actually could see those crops at 300ppi (my Lumia has a 336ppi resolution), they would look far less noisy than they do.)

Any chance you could post screenshots of the original histograms for each of those photos (without any post-process edits)? Given how much noise you have in those shots, and given the overall tone, I'd wager that a higher ISO setting with a greater amount of ETTR would produce better results from both cameras. I recently read some interesting info on exposure from one of my favorite bird photographers...Art Morris, who is a staunch advocate of ALWAYS exposing such that the rightmost ends of the histogram end up half way into the rightmost vertical box on the histogram display. Since I started doing that, my photos have been less noisy after post-process exposure correction (which, actually, tends to be minimal most of the time). Given what I've learned the last several months with my 7D, I would be willing to bet both the 5D III and 1D X could produce less noisy exposures for the same images you posted if you used both a higher ISO setting, and dropped the right-hand end of the histogram half way into the rightmost box.

It may not seem like "getting the exposure right in-camera", and if one was using film it would indeed be an entirely incorrect way to expose. But to maximize the potential of digital equipment, we have to think about it and use it differently than we think about and use film. I'm a staunch believer of ETTR, or expose to the right, when using modern digital cameras, and in my experience it does result in better dynamic range and noise performance out of any digital camera.

 

[RMC33]

For digital I agree with jrista. I always overexpose most of my shots and then when I bring them down in post, there is a lot less noise vs. going the other way. I agree, for film, not the way to go, but it's a trick I learned and take advantage of with digital. Most of my outdoor sports photos are at least +2/3 to +1 2/3 EV.

For snow sports I go -1/3 to -1 if it is a bright day. Overcast or snowing I could not agree more that over is better (under can make for some very very cool B&W)!

Certain water sports as well like kayaking benefit from your method too as the water can show up very dark and get a bit noisy.

 

[bdunbar79]

Problem with underexposing with snow is that it then has a gray tint, which you can fix in post I agree, so no big deal. At weddings I'd always shoot +1EV for the bride's dress to make it exactly white (with proper WB of course). With RAW now though, all that can be done if you don't get it right in cam.

EastWind, it's a tough world out there! I feel your pain with dark/light and mixed lighting scenes. I've battled that all winter. You'd think you could avoid it in outdoor sports, but not when the university schedules soccer games at 1pm in early September! You gotta just do the best you can.

Watch your 1DX metering. Mine both are too dark relative to my 5D3 and I had to set base AE up to 5/8 instead of 0 EV! I consulted with a few other 1DX owners and they had the same result. So you do have to watch that, as even metering is all relative.

 

[gary samples]

I have always had my best luck with
when shooting whites overexpose
when shooting blacks underexpose

 

[RMC33]

Just got my CPS 1Dx yesterday! GF hid it from me and it was on the table this AM with breakfast with a full charge and my 200 f/2 attached!

Can't wait to compare the two now!

 

[jrista]

Yes I agree. These are not the best exposed images. As I said the lighting was awful and the shot with the 5DIII it had just started sleeting. It was overexposed by about 2 stops already. The 1DX was exposed about 1 stop over. I tried to find a couple of shots that were challenging in comparison.

I am not sure I would use the word "overexposed", if I understand what you are getting at. If you are referring to the EC scale in the viewfinder that the camera meter updates, then I would say the meter gravely MISS-METERED those images. Keep in mind, the camera meter is rather dumb...it generally aims for an ~18% gray average tone. Given the sky in those photos, the meter was actually underexposing. I would actually say it probably under-exposed by two to three full stops!

We have exposure compensation for a reason, and if we have to boost exposure with EC, then that does not mean we are over exposing....it means we are correcting the meters incorrect automatic exposure settings. I would have happily blown the sky in those images entirely, if it allowed me to get better exposure on the eagle itself. I'd have pushed exposure as far as I could, to the point where the feathers of the eagles head were in the 240 RGB range, then corrected DOWN in post. By literally over-exposing, then pushing exposure down in post, you actually mitigate noise. I have some examples of this with a dragonfly I photographed a while back...I accidentally overexposed the original shots by some three stops, and corrected in post. Compared to the later shots, the corrected ones that were overexposed had almost zero visible noise, while one that was "correctly" exposed in camera had a plenty of visible noise. I'll see if I can dig those up.

Noise is not actually caused by high ISO...it is caused by too little light. The ISO setting simply changes the readout whitepoint, which intrinsically limits the maximum exposure level. If you push to ISO 1000, but then expose such that your whites are below an RGB value of 200...you are simply exacerbating the problem of not having enough light. If you cannot use a longer shutter speed, then the best way to maximize exposure is to increase ISO. It doesn't get any more light down the lens, but it reads out the exposure you have in the least-noisy way. Higher in-camera ISO will almost always trump post-process exposure push. Pushing a noisy exposure in post just makes the noise more apparent. Better to increase ISO, maximize the exposure in-camera (and even blow the sky such that it is entirely white when it is gray and overcast, since the sky isn't the subject and doesn't really matter), then pull the exposure down in post. You'll increase your signal to noise ratio, which is preserved with the post-process pull.

 

[East Wind Photography]

LOL! The one I had ended in serial 508.

Just got my CPS 1Dx yesterday! GF hid it from me and it was on the table this AM with breakfast with a full charge and my 200 f/2 attached!

Can't wait to compare the two now!

 

[RMC33]

LOL! The one I had ended in serial 508.

Just got my CPS 1Dx yesterday! GF hid it from me and it was on the table this AM with breakfast with a full charge and my 200 f/2 attached!

Can't wait to compare the two now!

Bummer 338 here. I have to say, I love the build quality already. The manual is a bit daunting but there are a lot of Cfn's I like that the 5d3 Lacks. Glad I have a bunch of time off in the next two weeks to shoot.

 

[East Wind Photography]

Interesting on the 1DX exposure adjustment. I wonder if it's due to the focusing screen they selected for them? One would think they would be calibrated based on a Canon standard.

On the samples I posted, I whole heartedly admit that the exposures were not the best. The point I was trying to make between the two shots is 1) the finer noise/grain patterns on the 5DIII and 2) the amount of Chroma noise. Yes more exposure would push the chroma on both shots below the noise floor but at least at ISO 1000 it shows they're either equal or (my opinion) the 5DIII actually shows very slightly less under similar conditions. Either way chroma is a non issue with the excellent post filters available today and at higher ISO's the 1DX would likely win in the chroma dept when comparing raw images and also the "grain" patterns at very high ISO's. Maybe not so much difference up to around 3200.

Problem with underexposing with snow is that it then has a gray tint, which you can fix in post I agree, so no big deal. At weddings I'd always shoot +1EV for the bride's dress to make it exactly white (with proper WB of course). With RAW now though, all that can be done if you don't get it right in cam.

EastWind, it's a tough world out there! I feel your pain with dark/light and mixed lighting scenes. I've battled that all winter. You'd think you could avoid it in outdoor sports, but not when the university schedules soccer games at 1pm in early September! You gotta just do the best you can.

Watch your 1DX metering. Mine both are too dark relative to my 5D3 and I had to set base AE up to 5/8 instead of 0 EV! I consulted with a few other 1DX owners and they had the same result. So you do have to watch that, as even metering is all relative.

 

[bdunbar79]

I don't know what it is. But set the 5D3 to 1/200s, f/3.2, ISO 1600, and then I set both of my 1DX's to the same settings, same lens, same tripod, and shot a target, and the 1DX images were darker. I moved the 1DX ISO up to 2000 and 2500 and the 2000 was still a tad dark and the 2500 was brighter. I played around with AE and set it at 3/8-5/8 and it seems to match up now to the settings I'm normally used to from previous cameras. I consulted many other 1DX owners who did the same thing. Again, I don't know what it is. One thing I did notice is that with the 5D3 you can have the EV scale at 0, and highlights still get blown. SAME situation and have EV at 0 with a 1DX, and none of the highlights are blown, ever. It's not aesthetically pleasing but I'm thinking Canon knows this and it is intentional, and has to do with the metering system accuracy over the 5D3. It's only a good thing.

 

[East Wind Photography]

Wondering of in that instance the highlights getting blown out are due to either the 5DIII overexposing based on AE +-0 or that the 1DX is underexposing? Obviously the 1DX is better at capturing the highlight detail but it is interesting and perhaps partially accounts for some of the loss. I'm sure Canon has a reasonable explanation. I didn't mess with it on the 1DX but I think I saw an AE micro adjust setting in the firmware. Just as you can adjust the AFMA you can tweak the metering. Is that where you set the +5/8?

I don't know what it is. But set the 5D3 to 1/200s, f/3.2, ISO 1600, and then I set both of my 1DX's to the same settings, same lens, same tripod, and shot a target, and the 1DX images were darker. I moved the 1DX ISO up to 2000 and 2500 and the 2000 was still a tad dark and the 2500 was brighter. I played around with AE and set it at 3/8-5/8 and it seems to match up now to the settings I'm normally used to from previous cameras. I consulted many other 1DX owners who did the same thing. Again, I don't know what it is. One thing I did notice is that with the 5D3 you can have the EV scale at 0, and highlights still get blown. SAME situation and have EV at 0 with a 1DX, and none of the highlights are blown, ever. It's not aesthetically pleasing but I'm thinking Canon knows this and it is intentional, and has to do with the metering system accuracy over the 5D3. It's only a good thing.

 

[bdunbar79]

Well, it's not just relative to the 5D3, it's all cameras other than the 1DX. Did the same thing with the 1D4. I'm not convinced that the 1DX isn't underexposing relative to 0EV, or the 5D3 is relative to some other value of EV than 0. It's my gut feeling that the 1DX slightly underexposes at lower ISO, and then actually slightly overexposes at high ISO. I ran a series of images at a night soccer game once and once the 1DX crossed from ISO 6400, to 8000, the image was much brighter overall than 1/3 stop. So it is something with the sensor and/or metering system that they did to reach cleaner, higher ISO's, and I don't know what it is exactly.

 

[East Wind Photography]

Well it's good to know CPS has more than one to loan out!

Please post some caparisons when you get a chance, particularly if you can take some in the snow. Lots of highlights and would be good comparison of the 5DIII and 1DX at reproducing such. We dont get a lot of snow here but Im planning a polar bear trip and would be nice to see if the 1DX does any better when everything is white except the eyes!

LOL! The one I had ended in serial 508.

Just got my CPS 1Dx yesterday! GF hid it from me and it was on the table this AM with breakfast with a full charge and my 200 f/2 attached!

Can't wait to compare the two now!

Bummer 338 here. I have to say, I love the build quality already. The manual is a bit daunting but there are a lot of Cfn's I like that the 5d3 Lacks. Glad I have a bunch of time off in the next two weeks to shoot.

 

[East Wind Photography]

I agree. When I said overexpose by 1 stop I meant over what the metering was telling me. Shooting into the sky will never give you an accurate metering unless you are shooting clouds and even then you might want to shoot under the metering value.

But I do agree that the histogram is the way to go if you have the time to make that comparison. Often with BIF images you cant predict where the bird will be and checking a histogram prior to it is impossible. You can be in the ball park though but maybe not an ideal setting. Experience usually will get you there as well.

Yes I agree. These are not the best exposed images. As I said the lighting was awful and the shot with the 5DIII it had just started sleeting. It was overexposed by about 2 stops already. The 1DX was exposed about 1 stop over. I tried to find a couple of shots that were challenging in comparison.

I am not sure I would use the word "overexposed", if I understand what you are getting at. If you are referring to the EC scale in the viewfinder that the camera meter updates, then I would say the meter gravely MISS-METERED those images. Keep in mind, the camera meter is rather dumb...it generally aims for an ~18% gray average tone. Given the sky in those photos, the meter was actually underexposing. I would actually say it probably under-exposed by two to three full stops!

We have exposure compensation for a reason, and if we have to boost exposure with EC, then that does not mean we are over exposing....it means we are correcting the meters incorrect automatic exposure settings. I would have happily blown the sky in those images entirely, if it allowed me to get better exposure on the eagle itself. I'd have pushed exposure as far as I could, to the point where the feathers of the eagles head were in the 240 RGB range, then corrected DOWN in post. By literally over-exposing, then pushing exposure down in post, you actually mitigate noise. I have some examples of this with a dragonfly I photographed a while back...I accidentally overexposed the original shots by some three stops, and corrected in post. Compared to the later shots, the corrected ones that were overexposed had almost zero visible noise, while one that was "correctly" exposed in camera had a plenty of visible noise. I'll see if I can dig those up.

Noise is not actually caused by high ISO...it is caused by too little light. The ISO setting simply changes the readout whitepoint, which intrinsically limits the maximum exposure level. If you push to ISO 1000, but then expose such that your whites are below an RGB value of 200...you are simply exacerbating the problem of not having enough light. If you cannot use a longer shutter speed, then the best way to maximize exposure is to increase ISO. It doesn't get any more light down the lens, but it reads out the exposure you have in the least-noisy way. Higher in-camera ISO will almost always trump post-process exposure push. Pushing a noisy exposure in post just makes the noise more apparent. Better to increase ISO, maximize the exposure in-camera (and even blow the sky such that it is entirely white when it is gray and overcast, since the sky isn't the subject and doesn't really matter), then pull the exposure down in post. You'll increase your signal to noise ratio, which is preserved with the post-process pull.

 

[garyknrd]

Thank's allot for the comparison. Well thought out and very informative. You mentioned shutter noise. I just got a Mark IV and it is the first time I can remember when a bird actually heard the noise and flew. It is very noisy also. My friend has a 5D III and it is so quiet compared to mine.
www.flickr.com/photos/avianphotos