Written by 4:53 pm Gear Reviews • 148 Comments

Is the Canon 6D Under-Exposing? UPDATE: No, It’s Not.

Final Update and Winners of the BorrowLenses.com Gift Certificate, Friday, December 7, 2012 11:35 AM

Okay, we found the cause of the D600 bodies’ overexposure. Turns out, it WAS damage, not a defect. In the damaged bodies, the little prong that actually pushes the aperture closed was bent, as you can see in the image below. The top one is of one of the damaged D600’s, while the bottom is of an undamaged D7000.

No idea what caused this, but there you have it.

Winners of the $50 BorrowLenses.com Gift Certificate: K.G. Wuensch, who left the suggestion that led to our discovery of the cause of the overexposure on the D600 bodies is, unfortunately, not based in the U.S., and so is unable to use the certificate I promised him. He has, instead, requested that his prize be entered into the pool for the general drawing. So we now have two gift certificates to give out.

I entered all the commenters’ names into a list randomizer at random.org and the two names at the top are our two winners.

The winners of the gift certificate drawing.

The winners of the gift certificate drawing.

Congratulations to David Johnson and Michael Clark! Please email your contact info to sohail.mamdani at borrowlenses dot com, so I can send them to you.

Once again, thanks to everyone for your fantastic support and feedback. 

Update Thursday, December 6, 2012 2:11 PM

Thanks to a suggestion from one of the folks who left a comment below, K.G.Wuensch, we found the issue that led to the big discrepancy in the images you saw from my test, and the issue turned out to be with the D600, not the 6D.

Take a look at these images. Both cameras have apertures set to f/8, and I had one of our staffers hold down the Preview button on both bodies. As you can see, the defective D600 on the right doesn’t close the aperture down as much.

We found this behavior on two separate bodies. It affects not just the Sigma lens, but also the Nikon 50mm f/1.4G lens, so it’s not a 3rd-party lens issue. We know we received at least two duds, and the warehouse is looking to see if there are any more. We will, of course, take those – and any others we find – out of circulation and send them back to Nikon.

One of the defective D600s we found on the right, the normal on the left

One of the defective D600s we found on the right, the normal on the left

On a personal note: I really have been taken aback by the level of positivity in all of our readers’ comments. When I first wrote this article, I was prepared for an avalanche of brickbats and accusations of everything from anti-Canon/anti-Nikon conspiracies, to suspicions about personal gain. By staying above the cruft, you guys are kinda restoring my faith in the interwebnets 🙂

As a personal thank-you, everyone who has written a comment below as of the time of this update is entered into a drawing for a $50 Gift Certificate from BorrowLenses.com. Consider it my way of expressing my gratitude for your support, encouragement, and suggestions.

One commenter is not eligible for this draw, though. That’d be you, Mr/Ms/Mrs K.G.Wuensch – because I’m also giving you a $50 Gift Certificate from BorrowLenses.com for hitting the nail right on the head, separately from the draw. Please email your contact info to sohail.mamdani at borrowlenses dot com, so I can send it to you.

One caveat is that you do have to meet BorrowLenses.com’s rental requirements to use the gift certificates. I wish I had a way around that, since we do have an international readership, but alas, that’s not something I can arrange.

Thank you to everyone who took the time to read this piece and offer constructive feedback. It’s a pleasure working on this stuff for you.

Update: We’ve done further testing and added to this article. Please scroll to the bottom for the latest info.

We took delivery of Canon’s newest body, the 6D, late last week, and I was assigned one for testing immediately. With sunlight a rarity these days in the Bay Area (we’ve had a spot of weather lately), I resorted to shooting some night-time long exposures to start with.

Canon 6d

Canon 6d

Earlier in the day, as I was picking up gear, one of my BorrowLenses colleagues remarked in passing that he thought the 6D might underexpose things a bit. I didn’t pay much attention to this; depending on the metering mode set and the part of the composition that the camera’s metering sensor is looking at, exposure in one of the automatic modes can vary wildly. I didn’t even bother making a mental note to check on it.

As part of the 6D test, I decided to take a Nikon D600 along with me as well to do a side-by-side comparison of images. And when I saw what was coming off the two cameras in my first test I am glad I did.

My first test was shot in RAW on the 6D, which turned out to be a mistake, as neither Lightroom nor Aperture support the 6D just yet (hence no images from that test to show here). What I was seeing on the LCD, however, was enough to make me head back to the BL office and swap out the 6D body. I was sure I just had a bad copy.

So I headed out again the next night. My subject this time was the San Francisco skyline. For the images you see below, I tried to match the output from both cameras as much as possible.

For starters, I captured in RAW+JPEG mode for both cameras. I set the picture styles to Neutral on both bodies, and shot the images at ISO 100. I also used the same make/model of lens on each camera, the Sigma 50mm f/1.4.

First, have a look at the image below. This one was shot on the Nikon D600.

Nikon D600, f/8, 30-second exposure at ISO 100

Nikon D600, f/8, 30-second exposure at ISO 100

As I’ve noted, the exposure in Manual mode was f/8, 30-seconds long at ISO 100. No adjustments – other than resizing for the web – were made on this image.

Now, here’s the image from the Canon 6D, exact same exposure settings.

Canon 6D, f/8, 30-second exposure at ISO 100

Canon 6D, f/8, 30-second exposure at ISO 100

Pretty big difference, no? Here are the images, side-by-side.

Canon 6D on the left, Nikon D600 on the right.

Canon 6D on the left, Nikon D600 on the right.

Now, I’ve seen variances of .3 or even .6 stops in cameras. Sometimes, the perceived difference in exposure can be because one camera captures mode dynamic range than another (and that may well be true for the D600 over the 6D here), or because of some other reason. Certainly, cameras of one brand do tend to meter differently from cameras of another. But here, the images do appear to indicate that for identical settings, the Canon 6D does tend to underexpose its images by about two stops.

In the image below, you’re looking at the Canon 6D on the left and the Nikon D600 on the right. The images look pretty similar, right?

Canon on left, Nikon on right.

Canon at f/4, 30 secs on left, Nikon on right at f/8, 30 secs.

Interesting, here, the Canon 6D’s aperture has been bumped up to f/4. That’s a full 2 stops brighter than the image from the D600. Below, you’ll see a 100% crop from these two images.

100% crop of the 6D (left) and D600 (right) images.

100% crop of the 6D (left) at f/4, and D600 (right) at f/8.

I’m also being told that the differences you see here may also be exacerbated by the fact that this is a long exposure. I don’t know that that’s accurate, so we’ll be doing a few more shots with test targets and more “normal” exposures to be certain. I suspect that the difference won’t be as drastically apparent, but it’ll still be there nonetheless.

It’s important to note that this behavior was observed on two separate bodies. The difference was also marked between exposures on the 6D and the 5D Mark II, though since those were all shot the previous evening on RAW and I don’t have a RAW converter for the 6D handy just yet, the comparisons will have to wait. I did, however, see what I perceived to be at least a 1.5-stop difference, if not more, between the 5D Mark II and 6D images.

Even if the difference is only 1.5 or 2 stops, that can have an impact on all kinds of shooting. If you’re using a Sigma 50mm f/1.4 lens, for example, and shooting at f/1.4 at 1/60th, you’d have to shoot at 1/15th of a second to get a normally exposed image with a 6D. Your flashes would have to work harder, and your night-time long exposures could take three times as long. That’s a pretty big impact for many shooters.

So what do you guys think? Is the Canon 6D underexposing? Or is the difference more perceptual than anything else? Is this more about dynamic range than exposure? Leave us your thoughts!

UPDATE:

A few folks have been asking about histograms, so here are the histograms from Lightroom.

Histogram from the 6D shot of the San Francisco skyline

Histogram from the 6D shot of the San Francisco skyline

 

Histogram from the D600 shot of the San Francisco skyline

Histogram from the D600 shot of the San Francisco skyline

 UPDATE:

I shot a bunch of new tests earlier this morning under different circumstances with a Canon 50mm f/1.4. Looking through the test results now, but it looks like at the very least, the Nikon vs. Canon differential is holding. More to follow shortly.

UPDATE 6:08 PM Wednesday, December 5, 2012

And just because this HAD to get more confusing, here is a grouping of the latest images, shot this morning.

Things you need to know:

  • A lot of you talked about the lens being a factor here. That might well be true, but I don’t yet have enough data to verify this. I did shoot the Canon images with a Canon 50mm f/1.4 lens.
  • We will test again later this week with native lenses. Someone in the comments said the Sigma might be to blame here, and right now, that’s not out of the realm of possibility.
  • These shots are in-camera JPEGs.
  • D-Lighting, Auto Lighting Optimizer, and Highlight Tone Priority were turned off.
  • Spot metering mode was used (not that this should matter at all).
  • Neutral picture style was used for all five shots.

With that, here they are.

Canon 6D

Canon 6D

Canon 5D Mark III

Canon 5D Mark III

Canon 5D Mark II

Canon 5D Mark II

Nikon D800

Nikon D800

Nikon D600

Nikon D600

And finally, all five shots, together.

Clockwise from top-left: D600, D800, 5D Mark II, 5D Mark III, 6D

Clockwise from top-left: D600, D800, 5D Mark II, 5D Mark III, 6D

So what does this mean, exactly?

Looking at this, it sure looks like it’s the D600, not the 6D, that’s the outlier, doesn’t it?

*sigh* Back to the testing room.

My initial conclusion that the 6D was under-exposing was based on its comparison with the D600. I assumed that the D600 was exposing correctly was based on my experience from the first night of shooting, where, on the camera’s LCD, it looked like the D800E and the 5D Mark II were putting out exposures more in tune with the D600 than the 6D. That may still be the case – I won’t know till I can convert those darned RAWs from the 6D.

The images you’re looking at seem to indicate the converse – that it’s the D600 that’s overexposing. I suspected this might be the case, which is why I’ve posed this entire piece as a question and an ongoing series of discoveries.

Which is also why I sincerely thank the folks commenting below. We’re figuring this out together, and I appreciate the input from all of you. I’ve taken a lot of your suggestions into account, so keep them coming. Something is amiss here, and what started as a question about the Canon 6D may yield answers about the Nikon D600 – or, indeed, the lenses used or about the JPEG processing engines in cameras and software. 

BTW, I still have had no luck converting the 6D RAWs. The Canon software refuses to install on my MacBook Pro 13″ Retina model running Mac OS 10.8.2, so I’m going to try doing so on a PC.

Tags: , , Last modified: May 25, 2020
Close