Doug Brown Photography

INTRODUCTION

I just received my new Canon 600mm f/4L IS II courtesy of Gary Farber and John Duggan at Hunt’s Photo & Video.  I think I put down a deposit almost a year and a half ago!  My first impression is that this lens is light!  The new 600mm weighs about 3.2 pounds less than the model it’s replacing.  It’s almost identical in weight to the outgoing 500mm f/4, but because the weight is distributed over an extra 2.5″ of body length, the lens feels even lighter.  The good news is that Canon did not compromise on build quality to achieve the weight savings; the lens feels very solid.  The tripod collar is as smooth as butter compared to the one on the old 600mm, and the 4-stop IS works as advertised.  Since there are so few copies of the lens in circulation and so few images available for viewing on the internet, photographers don’t know what to make of this $13,000 beast.  For these reasons, I decided to share my initial impressions of the 600mm f/4 II with you.

TESTING METHODOLOGY

I wanted to focus my testing on the performance areas that are relevant to bird photographers, but these findings will be relevant to anyone who uses a supertelephoto lens.  All of the tests compare the new 600mm with the version I 600mm.  I devised 5 basic tests for this comparison.

Test 1 compares the sharpness of the 2 bare lenses at 3 common apertures (f/4, f/5.6, and f/8).  The camera/lens combination was mounted on a Wimberley II head sitting atop a sturdy Gitzo tripod.  I chose the 5D Mark III as the body for this test, and I upgraded to the latest firmware prior to testing.  I taped my trusty $20 bill on the wall at tripod height, and placed the tripod about 22 feet from the target.  I used Live View AF (with manual focus adjustment as needed) to focus on the area right between the eyes, and I also employed both mirror lockup and a cable release.  I took multiple frames at each setting and chose the best ones for comparison.  This same technique was used for all other tests except for the one testing IS.  I opened up the best frames in Lightroom 4 and did side-by-side comparisons at 100% magnification, and then created screen captures for display in this blog post.

Test 2 compares the 2 lenses with the 1.4x III at f/5.6, f/8, and f/11.

Test 3 compares the lenses in combination with the 2x III at f/8 and f/11 (I don’t tend to use apertures smaller than f/11 in the real world).

Test 4 looks at the effect of minimum focus distance on subject size in the frame.

And Test 5 removes the camera/lens from the tripod and compares image stabilization while hand holding.

Before we get to the testing, here’s a shot of the new 600 all dressed up in LensCoat Digital Camo, which does a great job of protecting expensive gear.

TEST 1

Here is a side-by-side comparison of the 2 lenses at f/4.  Version I is on the left and Version II is on the right, and I will maintain this orientation throughout the remainder of the comparison.  Sharpness is comparable between the 2 lenses in the center of the frame.  I might give a slight edge in contrast to the new 600mm, but there’s nothing here that would make me want to sell my Version I and upgrade to a Version II lens.

Next we’ll look at the area a little left of center at the bottom of the frame.  Sharpness and contrast are improved with the new lens.

Now for the bottom left corner.  Both sharpness and contrast are again better with the new lens.  The top left corner shows the same relative performance.  But again, the improvement is not worth the cost of admission in my opinion.

At the top and bottom right corners we find comparable sharpness, but the new 600mm has slightly better contrast.

Next we’ll stop down to f/5.6.  The results are comparable to those we saw at f/4, so I’ll spare you from having to view the 100% crops, except for the center of the frame.

For the last bare lens test I’ve stopped down to f/8.  Nothing noteworthy has changed at this aperture.

The last thing we’ll look at with the bare lenses is the step-up in image quality that occurs as you stop down from f/4 to f/8.  First we’ll look at the 600mm Version I. I would say that there’s not much difference in sharpness when you compare the old lens wide open (the image on the left) to the old lens stopped down 2 full stops, and that’s no big surprise.  My experience in the field confirms this; the Canon 600mm looks great wide open.

Now we’ll do the same comparison for the 600mm Version II.  To my eye it’s difficult to discern a difference between the two frames.

What about the new lens at f/4 vs. the old lens at f/8.  I give the nod to the new lens wide open (the image on the left) for its better contrast and slightly better sharpness.

*** Note that for this comparison the new lens is on the left, not the right***

TEST 2

For this test we add the 1.4x III into the mix.  As per Canon’s recommendation the TC was mounted on the lens prior to attaching the body.  At f/5.6 center sharpness and contrast are nearly identical between the old lens (left) and the new lens (right).  The results are the same at all 4 corners of the frame, so I won’t include images.  I’m sorry about the lack of color correction; this project is already taking up way too much time!

We’d be hard-pressed to claim that there’s a difference at f/8 either.  Both lens/TC combos produce excellent results at the center and the edges.

Switching to f/11 produces excellent-looking frames with both lenses.  I’d be lying if I picked a favorite.

As with the previous test, there’s not much difference in image quality as you move from f/5.6 (wide open, left) to f/11 (stopped down 2 stops, right).  This holds true for both lenses.  The sample below comes from the 600mm Version II.

The new lens at f/5.6 (left) is neck and neck with the old lens at f/11 (right) when using the 1.4x III.

*** Note that for this comparison the new lens is on the left, not the right***

TEST 3

For this test the 1.4x III is exchanged for the 2x III.  I shot sample frames at f/8 and f/11 only, because I don’t venture into f/16 territory in the field.  The frame you see below shows the Version I lens/2x TC combination on the left and the new one on the right; the aperture is f/8.  Center sharpness and contrast are identical between the two lenses.

But when you look at the bottom left corner, image quality starts to fall apart with the old Version I lens.  The Version II lens still looks great.

The top left corner tells a similar story.  The old 600mm, when combined with the 2x III, has significantly poorer sharpness at the corners than the new 600mm.

The corners on the right side of the frame demonstrate similar softness in the Version I 600mm lens, but I’m going to save you a little bandwidth by not including images.

At f/11 the story is the same as the one at f/8.  Center sharpness is identical at f/11 between the old (left) and the new (right) 600mm with a 2x III.

The f/11 aperture is not enough to correct for the corner softness in the 600mm Version I at 1200mm, although it is somewhat improved.  Corner sharpness with the Version II 600mm is still excellent.

TEST 4

This is a quick test to demonstrate the impact of the improved minimum focus distance found in the Version II 600mm lens.  For people who can get close to their subjects (e.g. photographers shooting from a blind), it’s like getting a free teleconverter but without the aperture penalty.

TEST 5

There’s no scientific way to test image stabilization.  That said, I specialize in hand held bird flight photography and the 600mm Version I has been my hand held lens of choice about 99% of the time.  I’ve been very pleased with the results I’ve gotten over the years, but hand holding an 11+ pound lens takes a toll on your body.  To say that I was anxiously awaiting the arrival of the new 600mm is the understatement of the year!!!  So do the weight savings and improved IS allow you to take steadier shots at slow shutter speeds?  For this test I hand held each lens (unsupported in any way) and shot between 5 and 10 frames through a progression of steadily slower shutter speeds.  I used the sharpest frames at each shutter speed for comparison purposes.  These are the slowest shutter speeds that I could get a usably sharp image at.  For the 600mm Version I lens, that shutter speed was 1/80.

Things were definitely better with the Version II 600mm lens.  I managed 1/20 second.  Of course shutter speeds this slow require considerable skill and a fair amount of luck to pull off.  But it’s nice to know that you can get sharp images at extremely slow shutter speeds if you need to.

CONCLUSIONS

The new Canon 600mm f/4L IS Version II is an impressive piece of glass to be sure.  But does it live up to the hype and to its $13,000 price tag.  It all depends on what kind of photographer you are and how deep your pockets are.  Let me start by saying that I was surprised at how well the Version I 600mm lens held up throughout the test.  The only significant sharpness difference between old and new lens was at the corners of the frame, and softness only became problematic with the 2x III added into the equation; use of the 2x is on the rise because of Canon’s penchant for full frame bodies of late.  But since center sharpness is comparable, you need to ask yourself how often your subject is placed in the corner of the frame.    In summary, kudos to the old 600mm for holding its own.  But the new 600mm did consistently best the old one optically.

The improved minimum focus distance of the new 600 should be a boon to setup/blind photographers.  You’ll now be able to get quite a few more pixels on your subject without having to resort to a TC.

The new IS delivers on its promise of 2 extra stops of performance.

However, it’s the weight savings that make the new 600mm lens worth every penny for me.  The lens is incredibly light….a joy to hand hold!  I can get on my subjects faster and track them more consistently.  I can stay on a stationary subject for considerably longer (helpful when waiting for a bird to take off from a perch).  Although I haven’t gotten to spend much time in the field with the new 600mm, the short time I did spend showed me improved autofocus accuracy when tracking birds in flight against a varied background.  Here are a few image posts from the new 600, both taken with the 1.4x III.

Link 1

Link 2

Anyone who can hand hold a 500mm lens can now step up to the rarified air of 600mm hand held photography.  When the new 600 was announced my initial thought was that if the lens was optically the same as the old 600 but the weight was trimmed by 3 pounds, I’d buy it in a heartbeat.  As it turns out, the lens was in such short supply that I bought it sight unseen.  And fortunately my prayers were answered.  Canon delivered an optically superior (albeit not by much) lens with a 27% reduction in weight!  My back (but not my wallet) owes Canon a debt of gratitude!

Thanks again to Gary Farber and John Duggan at Hunt’s for getting me an early copy of this lens.  The folks at Hunt’s are a pleasure to deal with and their customer service is second to none!

INTRODUCTION

Canon has just come out with a new full frame camera body, the 5D Mark III, and the 1D x is coming soon. Nikon has just released the D4 and the D800 bodies.  This is an exciting time for photographers!  Bird photographers are no exception; unfortunately, most conventional sample images and comparison tests are not designed with bird photographers in mind.

So how do we decide which body to spend our hard-earned money on?   The 5D III is touted as being a great high ISO performer.  Canon’s 5D Mark III features a full frame CMOS sensor with an imaging area approximately 1.7x larger than that of the 1D Mark IV; this allows the 5D III to collect more light and deliver better high ISO performance in theory.  But does the 5D III deliver on that promise?  I’ve designed a test that hopefully answers this important question in a way that’s relevant to bird photographers.

TESTING METHODOLOGY

This test involves 2 separate but related comparisons.  The first comparison (Comparison 1) is between the 5D Mark III and the 1D Mark IV with identical framing or field of view (you see the exact same thing when looking through the viewfinder of either camera).  The second comparison (Comparison 2) is between the 2 bodies with identical focal lengths and distance to subject (a focal length limited scenario in which the subject appears larger in the Mark IV’s viewfinder).

I chose a subject that’s covered in feathers.  I shot at a fixed distance to subject using the 70-300mm L lens and decided to push the ISO from 3,200 to 12,800 (in full stops) to make it a fairly extreme test.  I used a constant aperture of f/7.1, Live View manual focus, mirror lockup, and a cable release for all frames.  The 5D III images were acquired at 300mm focal length.  I then mounted the 1D Mark IV on the lens, zoomed out to the identical field of view I had on the 5D Mark III (221mm), and snapped a series of frames.  Finally I shot the Mark IV at the full 300mm zoom from the same distance.

I uploaded the RAW files from Lightroom 4.1 RC into Photoshop CS5, making no adjustments other than the Lightroom defaults.  For Comparison 1 I took a 100% crop of the 1D Mark IV image taken using the same FOV as the 5D Mark III; this crop measured 1024 x 683.  I cropped the same portion of the image from the 5D Mark III file and then downsized it using Photoshop’s bicubic method to 1024 x 683.  For Comparison 2 I took a 100% crop of the 5D Mark III image measuring 1024 x 683.  I then cropped the identical area in the Mark IV image taken at the same focal length and distance to subject and downsized it to 1024 x 683.

You’ll note that I shot 2 versions of the Mark IV image.  My reasoning is that some photographers feel that when comparing a full frame body to a crop body, an identical field of view should be used.  This scenario resembles a case where you can either get physically closer to your subject or use more focal length to equalize the field of view. The problem with this approach in our current comparison is that it neutralizes any advantage in image detail created by the increased pixel density of the 1D Mark IV over the 5D III.

The second version of the Mark IV image was shot with the same distance to subject and focal length as was used in the 5D III image.  Some photographers feel that this focal length limited scenario more accurately reflects bird photography.  Here we allow the Mark IV to put more pixels on the subject.  I’m hoping to avoid the debate over which testing methodology is the correct one.  You can choose whichever one you wish.

THE IMAGES

Here is a scaled-down version of the full-frame test image used in Comparisons 1 (5D Mark III and 1D Mark IV) and 2 (5D Mark III).  My apologies for the subject’s good looks!  It was the only bird I could get to hold still for my little project!

And here’s a scaled-down version of the entire test frame obtained from the Mark IV for Comparison 2.  The subject magnification (tighter field of view) is due to the 1.3x cropped sensor of the Mark IV.

COMPARISON 1: IDENTICAL FIELD OF VIEW

5D Mark III ISO 3,200 (300mm)

1D Mark IV ISO 3,200 (221mm)

5D Mark III ISO 6,400

1D Mark IV ISO 6,400

5D Mark III ISO 12,800

1D Mark IV ISO 12,800

COMPARISON 2: IDENTICAL FOCAL LENGTH AND DISTANCE TO SUBJECT

5D Mark III ISO 3,200 (300mm)

1D Mark IV ISO 3,200 (300mm)

5D Mark III ISO 6,400

1D Mark IV ISO 6,400

5D Mark III ISO 12,800

1D Mark IV ISO 12,800

CONCLUSIONS

Now for my impressions.  I was surprised at how well the Mark IV fared versus the 5D Mark III.  I was expecting the full frame 5D III to better the Mark IV by close to 2 stops.  While the 2-stop advantage may hold true for in-camera JPEGs, this test is all about RAW file ISO performance.  After analyzing the images (using the naked eye), I’d say that the 5D Mark III is in the ballpark of 2/3 to 1 stop better in the noise department than the 1D Mark IV, but no more when you use the same FOV for both bodies (Comparison 1).  The noise is almost identical in Comparison 2 (the one with identical focal length and distance to subject), but the Mark IV has better image detail (especially noticeable at ISO 12,800).  In other words there is no advantage in the 5D Mark III when not using the entire sensor area (i.e. cropping your files to any significant extent).

I use a 1D Mark IV as my primary body and have recently added the 5D III as a second body.  If I had to choose between the two bodies based solely on high ISO noise, I’d go with the 5D Mark III as long as I was not focal length limited (meaning that I could frame perfectly in camera and not resort to cropping in post).  In these conditions, the 5D Mark III has lower noise and slightly image detail.  However, if I’m going to be cropping, the Mark IV gets my vote.  It has similar noise levels and better image detail (even at ISOs as high as 12,800).

Of course there’s more to a camera than its high ISO noise performance.  Another purported advantage of the 5D III is the AF system lifted from Canon’s soon-to-be top-of-the-line 1D x.  My initial impression of the AF system is favorable, but I need to spend some more time photographing birds in flight before I can draw any meaningful conclusions.  But if you factor in things like f/8 autofocus, frame rate, buffer size, and build quality, the 1D Mark IV starts becoming more and more attractive.  And if you are focal length limited and will need to crop, the Mark IV’s higher pixel density gives it the edge, even in low light conditions.

A special thank you to Gary Farber of Hunt’s Photo and Video for getting me an early production copy of the 5D Mark III.  Please help support this site by purchasing your camera gear at Hunt’s.  Thanks also to Greg Basco and Arash Hazeghi for their assistance.

I hope you’ve enjoyed this post. Please leave a comment below to let me know what you think!

Best regards from the desert,

Doug Brown

Canon Professional Network has just posted a video highlighting the features of the AF system in the new EOS-1D x.  It’s nice to see some of the key features slowly being fleshed out.  Click on the image above to be taken to the video.

Canon Professional Network has just posted a video highlighting the features of the AF system in the new EOS-1D x.  It’s nice to see some of the key features slowly being fleshed out.  Click on the image above to be taken to the video.

Just noticed that SanDisk Extreme Pro memory is on sale at B&H.  The price reduction is huge, and these are great cards!!!

16 GB Extreme Pro CF: $59.95

32 GB Extreme Pro CF: $99.95

64 GB Extreme Po CF: $221.95

Yesterday morning I had the pleasure of speaking about the Canon EOS-1D x with Chuck Westfall, Canon USA’s Technical Advisor in the Professional Engineering and Solutions Division.  In my questioning I tried to focus on the design decisions relevant to nature and wildlife photographers.  A big thank you to Chuck for taking the time out from his busy schedule to answer my questions!

DB: Everyone is very excited about the new AF system in the EOS-1 Dx.  Among wildlife photographers, the big concern is the loss of f/8 autofocus.  Why did Canon choose to eliminate that feature?
CW: Canon felt that while it would have been nice to preserve AF at f/8, the overall improvement in AF achieved with the 1D x took precedence.  The new AF system is faster, more controllable, and more intelligent.  The improved algorithm means that AF won’t jump to the background as quickly, won’t be tripped up as easily by obstacles in front of the subject, and will handle subjects that are small in the frame much better.  As you mention, AF is unavailable on the EOS-1D X if the maximum aperture reported to the camera through the electronic lens mount is smaller than f/5.6. This is a lower specification than previous EOS-1 series DSLRs. There are a lot of trade-offs in life, and this is one of them. Essentially, Canon Inc. decided that the overall improvement of the new 61-point AF system compared to the older 45-point system took precedence over the need for that particular feature. Taking your f/4 lenses as an example, with the new 1D X you now have 41 cross-type AF points (or 21 cross-types with the f/5.6 lenses) plus color and face detection tracking when using those lenses whereas there were none of those features with previous 1D/1Ds cameras. The AI Servo system on the new camera is vastly better than earlier cameras as well, in terms of sensitivity and performance as well as controlability.

DB: Does that mean that it’s not possible to AF at f/8, or that it’s not reliable but could be implemented?  Could f/8 AF be enabled as a custom function?
CW: I have no further comment at this time, because Canon Inc. has not published answers to those questions.

DB: I’ve read that the new AF sensor covers a larger area than in previous Canon bodies; could you elaborate?
CW: The AF sensor is larger than in any earlier 1D/1Ds body, measuring 19mm from side to side.  This compares to a 15mm-wide AF sensor in both the 1D Mark IV and the 1Ds Mark III.

DB: As a follow-up question, why hasn’t Canon placed AF points at the compositional power points dictated by the rule of thirds?
CW: In a phase detect AF system, it is impossible to spread the AF points out any further.  If you were to place AF points at the compositional power points, they wouldn’t receive enough light to do their job properly.  You would need an entirely new AF technology to make that possible.

DB: Could you tell us about the improvements that have been made in low-light and low-contrast autofocus?
CW: Low-light autofocus has been improved by a full stop compared to the 1D Mark IV.  AF can be achieved at EV -2, the equivalent of locking focus under the light of a full moon.  Low-contrast AF has also been significantly enhanced.  Each AF point now has a double line of sensors in each direction (horizontal and vertical), and the lines are offset from one another.  The result is that the AF sensor is much more sensitive to small changes in contrast than ever before.  Lenses that can take advantage of the central f/2.8 sensors will see even better results.  In our older systems, we used single lines of sensors, so no offset was possible.

DB: With the 1D x Canon has introduced a feature called EOS intelligent Tracking and Recognition AF (EOS iTR AF).  What can you tell us about it?
CW: The AE system of the 1D x is able to detect the color of the subject, and can pass on this information to the AF system to improve subject tracking.  To quote the Canon website “By using the colour of the subject that was initially focused, the AF system can track the movement of that subject, both by contrast and colour across the frame, and automatically select the most appropriate focus point given the position of the subject within the frame. This frees you up to concentrate on composing your images rather than selecting the best AF point for focus.”

DB: I’ve read that the AE and AF systems also incorporate face detection.  Some wildlife photographers are concerned that face detection may confuse the AF system when photographing animals.  Is there any way to disable the feature?
CW: The default AF mode is phase detect AF.  Color and face detection is optional, but you can’t choose one or the other.  It’s PDAF plus Color and Face Detection or PDAF only.

DB: What information do you have on the size of the 1D x buffer?
CW: At this point, no specs have been released on buffer size.

DB: How did Canon decide upon an 18 megapixel sensor for the 1D x?
CW: The Canon engineers felt that 18 megapixels struck the best balance between image quality and speed.  Also they wanted the body to achieve 12 fps shooting, and that helped guide the decision on megapixels.

DB: Moving forward, has Canon dropped the APS-H sensor from its future plans?
CW: The 1.3 crop sensor is not being killed off.  Canon is keeping all options open for future products.

DB: One of the advantages of moving to a full-frame sensor with a relatively low megapixel count is an improvement in high ISO noise levels.  Canon is claiming a two stop gain in high ISO performance when compared to the 1D Mark IV.  Will we see the full two stops when shooting RAW files, or is this limited to JPEG files only?
CW: To get the full two stops of improved performance you’ll need to shoot in the JPEG format.  That’s not to say that high ISO performance isn’t significantly better when shooting in RAW, because it absolutely is.

DB: Will the new electronic first curtain replace mirror lock-up?
CW: The electronic first curtain only works in Live View, so it does not replace MLU.  It does however offer a significant reduction in camera vibration when used in conjunction with Live View.  The performance gains are most noticeable at very long focal lengths (800mm and longer).

DB: Is there a way to easily access the mirror lock-up function in the 1D x?
CW: The 1D x brings custom shooting modes (C1, C2, and C3) to the 1-series EOS bodies.  These modes allow you to store regularly used settings for quick recall.  Mirror lock-up can be programmed to any one of these custom shooting modes.  You can also place mirror lock-up in My Menu.  And don’t forget that Live View effectively provides one-touch mirror lock.

DB: Why the new battery, and what are its performance specs?
CW: The new battery was designed to meet some new safety requirements from the Japanese government.  The new battery has slightly more capacity than the 1D Mark IV/ 1Ds Mark III battery, and there is complete forward-and-backward compatibility for LP-E4 and LP-E4N battery packs with the 1D x and 1D/1Ds Mark III as well as 1D Mark IV cameras.  Canon has not released performance specs yet on the new battery.

DB: Some nature and wildlife photographers feel like Canon is ignoring their input in the design of new camera bodies.  How do you address these concerns?
CW: Canon is very cognizant of the importance of the nature and wildlife segment when designing new camera gear.

DB: Canon’s Digital Photo Professional software does an excellent job handling high ISO noise, where products such as Adobe’s Lightroom seem to struggle.  Why doesn’t Canon license their RAW processing engine to other software makers?
CW: Canon has an SDK that includes its RAW processing engine that is freely available to other software vendors.  To my knowledge no other software maker has chosen to utilize our NR algorithm.

DB: What about the bounding box around the histogram?!?!
CW: Canon is aware that this feature has been requested.  It was found in the original EOS-1D, but vanished in subsequent generations.  You can enable a live histogram that stands out very clearly from the background in Live View.

In a collaborative effort with Keith Bauer and Greg Basco, I take a look at how the new Canon TCs perform compared to their Series II counterparts.  You can find the review here.  In this age of increasingly expensive camera gear in a sinking economy, it’s more important than ever to spend your money wisely.  This review may help you save a few bucks!