1DX vs 5D3

[bdunbar79]

That is very interesting.

The D800 fans say how much more they can push the shadows - perhaps we are seeing that against the 5DIII that is true - but against the 1DX that may not be true or not so obvious in the field. +2 is as much as I would expect.

What's interesting too, is that I can "fix" the 1DX image very easily, whereas the 5D3 I cannot do so that easily. Notice how slightly "darker" the 1D bodies expose vs. the 5D3, which leads me to believe at least better metering accuracy. Whether this means much in the field I doubt, especially for outdoor shots.

 

[drjlo]

"Notice the difference between what the 5D3 says and what the 1DX says:"

Hmm, "says" what, where?

Do you have something to contribute? Or are you just bored and lonely?

Jeez, there's no need to take my head off for asking for clarification for what you were trying to say. You wrote "what 5D3 says and what the 1DX says," so I thought your post or photo somehow/somewhere had a EXIF info or some other info tags I wasn't finding that "says" some data info.

If you meant for us to just look at the photo itself and interpret the IQ subjectively, not looking for some literal numeric info that "says" something in numbers, then that's fine.

 

[bdunbar79]

Sorry, I've been dealing with real smartass posts on another forum. My apologies, really.

Again, sorry, I'll provide the information you ask:

For the overexposure, it was very interesting. I set both cameras to auto ISO, f/8, 1/30s, and the 5D3 read 10,000 for a +2EC whereas the 1DX read ISO 6400 for the +2EC. This is very, very fascinating to me. Do you know what it means? I'm not sure.

Why does ISO 6400 on the 1DX read +2EC but it takes ISO 10,000 on the 5D3 to read +2EC?

 

[wickidwombat]

Sorry, I've been dealing with real smartass posts on another forum. My apologies, really.

Again, sorry, I'll provide the information you ask:

For the overexposure, it was very interesting. I set both cameras to auto ISO, f/8, 1/30s, and the 5D3 read 10,000 for a +2EC whereas the 1DX read ISO 6400 for the +2EC. This is very, very fascinating to me. Do you know what it means? I'm not sure.

Why does ISO 6400 on the 1DX read +2EC but it takes ISO 10,000 on the 5D3 to read +2EC?

Do the shots expose the same? Probably just the better metering on the 1dx

 

[drjlo]

I'll have some photos I'd like you to look at and evaluate. What this may mean is that the 5D Mark III is a much, much better daylight camera despite the high ISO capabilities than a low-light camera regarding color accuracy. Time will tell.

My issue lately with 5DIII has been unpredictable auto white balance under low light, mixed flourescent/tungsten settings. So many reviews have said the 5DIII auto WB is quite improved, i.e. compared to Nikon D800 auto WB, etc, but I often get shockingly off colors and saturation in auto WB. For example, below shot:


DZ3C5505C by drjlo1, on Flickr

Post-Processed.


DZ3C5505B by drjlo1, on Flickr

 

[canon816]

Ok, I went down and did the FIRST test for EXPOSURE ONLY. I set up the same conditions, 105mm, f/8, 1/100s. The 5D3 read the correct exposure at ISO 6400. The 1DX gave the correct exposure at ISO 5000. I shot with the camera set to Fluorescent WB.

If I could point out... different camera bodies will register different exposure readings. Not just from one body to the next, but there is also a small amount of variance between camera bodies of the same model.

I have a 1DIV, and a buddy of mine has two 1DIV's. Line all three up with the same lens at the same focal length and aperture.... and there are three different exposure readings. Between the three bodies there is a variance of over half a stop. I also noticed this when I used to shoot with 7D's... I'm not sure that canon has done anything fishy here but rather there is a certain amount of tolerance variability with the hardware itself.

I'm not a camera tech guy... just noticing what happens in the field.

 

[bdunbar79]

You're right. Except the photos look different. They are lighted differently, when they all said exposure was the same, the 5D3 registered 10,000 vs. 6400 and lo and behold, it was brighter. I don't know why, but it appears as though the 1D bodies are metering more accurately. I'm not sure if Canon is putting some technology in the 1D bodies that they are NOT putting in the 5D bodies, or what. You're right though, 6400 vs. 10,000 is not a full stop. I'm just pointing out that there is a difference.

 

[bdunbar79]

drjlo,

That's EXACTLY what I was talking about. Perfect example. I'll try to do similar shots with both cameras this weekend.

 

[canon816]

You're right. Except the photos look different. They are lighted differently, when they all said exposure was the same, the 5D3 registered 10,000 vs. 6400 and lo and behold, it was brighter. I don't know why, but it appears as though the 1D bodies are metering more accurately. I'm not sure if Canon is putting some technology in the 1D bodies that they are NOT putting in the 5D bodies, or what. You're right though, 6400 vs. 10,000 is not a full stop. I'm just pointing out that there is a difference.

Just a quick followup question: Is it that the 1D bodies are metering more accurately or just differently? After all they are just tools for us to use. If you are able to nail the exposure by over-riding the exposure while in AV mode or in manual when you understand the nuances of your particular body.... does it really matter?

 

[bdunbar79]

You're right. Except the photos look different. They are lighted differently, when they all said exposure was the same, the 5D3 registered 10,000 vs. 6400 and lo and behold, it was brighter. I don't know why, but it appears as though the 1D bodies are metering more accurately. I'm not sure if Canon is putting some technology in the 1D bodies that they are NOT putting in the 5D bodies, or what. You're right though, 6400 vs. 10,000 is not a full stop. I'm just pointing out that there is a difference.

Just a quick followup question: Is it that the 1D bodies are metering more accurately or just differently? After all they are just tools for us to use. If you are able to nail the exposure by over-riding the exposure while in AV mode or in manual when you understand the nuances of your particular body.... does it really matter?

No, it doesn't matter at all. You just set the 5D3, for instance, to expose at -2/3? Something like that. I've had to do this with the 7D. The only thing I saw was that I was able to push the shadows on the 1DX shot, but that's not conclusive because that was just one shot in a controlled environment. But no, it's just a differing of metering. Why it does what it does is beyond the scope of what I'm doing.

 

[Chris Burch]

Just a quick followup question: Is it that the 1D bodies are metering more accurately or just differently? After all they are just tools for us to use. If you are able to nail the exposure by over-riding the exposure while in AV mode or in manual when you understand the nuances of your particular body.... does it really matter?

The 1DX uses the focus point to meter, which the 5D3 does not (so I'm told). After the last shoot I did with both cameras, I can with certainly say the 1DX meters much more accurately on the subject than the 5D3 does. The 5D3 was getting thrown off by bright backgrounds or lights on the out edges of the frame, whereas the 1DX was pretty much spot-on for the actual subject in the photos.

If you're trying to compare the exposures of one camera to the other, these examples aren't all that useful because it seems automated setting are in use. However, from what I have seen on other reviews, with everything set to manual, the 5D3 is about 1/3 stop brighter than the exact same exposure settings on the 1DX. Check out the review someone posted for nighttime photography. I have heard that attributed to sensor programming, but I think that was just an educated guess.

 

[bdunbar79]

Just a quick followup question: Is it that the 1D bodies are metering more accurately or just differently? After all they are just tools for us to use. If you are able to nail the exposure by over-riding the exposure while in AV mode or in manual when you understand the nuances of your particular body.... does it really matter?

The 1DX uses the focus point to meter, which the 5D3 does not (so I'm told). After the last shoot I did with both cameras, I can with certainly say the 1DX meters much more accurately on the subject than the 5D3 does. The 5D3 was getting thrown off by bright backgrounds or lights on the out edges of the frame, whereas the 1DX was pretty much spot-on for the actual subject in the photos.

If you're trying to compare the exposures of one camera to the other, these examples aren't all that useful because it seems automated setting are in use. However, from what I have seen on other reviews, with everything set to manual, the 5D3 is about 1/3 stop brighter than the exact same exposure settings on the 1DX. Check out the review someone posted for nighttime photography. I have heard that attributed to sensor programming, but I think that was just an educated guess.

No. Everything was set to manual. And I used center AF point for all shots. Where are you getting that anything was automatic? Nothing was automatic. And metering center-weighted average is the SAME on all cameras. How can you center-weight average through any other area but the CENTER? Spot metering yes, but all the others no. Do you know what I'm saying? How is this not useful?

 

[Chris Burch]

If everything is set to manual, then the camera isn't metering anything. The only time metering would come into play is when you're in an AV, TV, or auto ISO mode and the camera is deciding what settings to choose. If everything is on manual, then you are doing the metering not the camera. Am I missing something?

If you do want to test how the cameras each meter (and that is a valid question) you need to put them both in the same auto mode and shoot with identical focus points. When you do that you'll see the 5D3 and 1DX behave quite differently, specifically because of the focal point-weighted metering on the 1DX.

 

[bdunbar79]

If everything is set to manual, then the camera isn't metering anything. The only time metering would come into play is when you're in an AV, TV, or auto ISO mode and the camera is deciding what settings to choose. If everything is on manual, then you are doing the metering not the camera. Am I missing something?

If you do want to test how the cameras each meter (and that is a valid question) you need to put them both in the same auto mode and shoot with identical focus points. When you do that you'll see the 5D3 and 1DX behave quite differently, specifically because of the focal point-weighted metering on the 1DX.

I'm lost. The camera meters out of the same point on both cameras. The 1DX only meters out of a focus point on spot metering. I didn't use spot metering, I used center weighted average. Second point I'm lost on, is that when the camera tells me it's properly exposed (the meter on the right hand side on 1DX and bottom on 5D3) the brightness is different on each camera. 0 EV on 1DX is different than 0 EV on 5D Mark III. This is a very simple test. You can shoot in auto ISO at 1/100, f/8, and get ISO 6400 on the 5D3 and ISO 5000 on the 1DX. That is very similar. However, upon inspection of the photos, the exposure is brighter on the 5D3. Well of course, it's ISO is higher. The point is that it gave correct exposure at a brighter image than the 1DX did. Again, a very simple thing.

And I don't know why people can't understand that AF-point metering is for spot metering. CWA comes out of the CENTER. It doesn't matter which camera you are using. Besides, why is it so hard to understand that I used the center AF point anyways in all of my shots?

I cannot explain this again, so I'm sorry if I'm not being clear.

 

[Chris Burch]

So you're using auto ISO -- that's what I wasn't clear on since that means everything wasn't set to manual. As I said earlier, it already been demonstrated that the 5D3 produces a brighter image with the exact same settings at the 1DX. If you set everything to manual (including ISO) with the exact same settings, the resulting image will be brighter on the 5D3 than the 1DX.

The real-world application of this is that with a fixed shutter and aperture, you can shoot a properly exposed image with a 1DX and the same properly exposed image on the 5D3 will have a lower ISO. So when it comes to comparing high ISO image quality, it's not a valid comparison to look at the same ISO side by side. In real-world shooting you would need to compare a 6400 1DX image with a 5000 5D3 image. This I believe is exactly what you showed in your testing. Given that balance, the 1DX may not have any edge in high ISO over the 1DX. I've heard someone propose Canon did this on purpose to give the 1DX an apparent edge in high ISO.

 

[bdunbar79]

So you're using auto ISO -- that's what I wasn't clear on since that means everything wasn't set to manual. As I said earlier, it already been demonstrated that the 5D3 produces a brighter image with the exact same settings at the 1DX. If you set everything to manual (including ISO) with the exact same settings, the resulting image will be brighter on the 5D3 than the 1DX.

The real-world application of this is that with a fixed shutter and aperture, you can shoot a properly exposed image with a 1DX and the same properly exposed image on the 5D3 will have a lower ISO. So when it comes to comparing high ISO image quality, it's not a valid comparison to look at the same ISO side by side. In real-world shooting you would need to compare a 6400 1DX image with a 5000 5D3 image. This I believe is exactly what you showed in your testing. Given that balance, the 1DX may not have any edge in high ISO over the 1DX. I've heard someone propose Canon did this on purpose to give the 1DX an apparent edge in high ISO.

Ok Chris, I understand you now. Got it. I agree, probably no real ISO advantage.

 

[bdunbar79]

I think then we can eliminate high ISO comparisons. They don't seem to be relevant. I think we should look at underexposure/overexposure and shadow detail recovery. Also low ISO color rendition.

 

[nightbreath]

... Also low ISO color rendition.

I looked closer to the samples you posted and saw slight difference in color tones. I'm not sure if it was affected by ISO difference or camera sensor color rendition, so I wouldn't comment here. Just look at the 5D Mark III image that has some slight purple / green color tints comparing to the 1D X image.

 

[bdunbar79]

I am and have been convinced that you are on to something here. Let me get back with you this afternoon with some test shots of real things. I am going to do a portrait shot with 135 f/2L lens this evening. The perfect test shot will be in tungsten lighting. Stay tuned. Do you want the RAW or TIF files? Let me know, I have your email addy.

 

[bdunbar79]

Did some on purpose bad lighting shots this evening. In this case, I took each camera with the same settings, only I shot in auto ISO for all shots. I did each lighting scenario with CWA metering and spot metering. If a highlight was blown out, I could NOT get the 1DX to focus properly. One shot I must have done 10 times and it's still blurry, whereas the 5D3 just locked on and I shot. Here's the first four: