Rob Galbraith Releases Mark IV AF Analysis

[Doug Brown]

For what it's worth, Rob Galbraith has published some initial impressions of the Mark IV. I don't think he was using firmware 1.0.6 though. You can find the review here.

[Axel Hildebrandt]

In the article he says that he used both firmware versions and didn't notice a difference. As for settings, the only one that I have differently is the tracking sensitivity, which I have at moderately slow.

I read a couple of times that some people were planning on making a decision based on his verdict. Maybe it will be easier for people interested in the camera to get one now. :)

[Rich Steel]

Mine came loaded with 1.0.6 and I have had no problems.

[Bill Foxworthy]

Well can someone explain to me how it work's great at one sport but not the other????

[Axel Hildebrandt]

Well can someone explain to me how it work's great at one sport but not the other????

I'm not sure, maybe different kind of light and some movements are more erratic?

[arash_hazeghi]

Well can someone explain to me how it work's great at one sport but not the other????

Light, movements and BGs are different, same is true in bird photogrpahy, tracking a duck or seagull is not the same as tracking a raptor.

[Desmond Chan]

Well can someone explain to me how it work's great at one sport but not the other????

I could be wrong, but it seems to me that Rob found that the Mark IV was able to lock on to the subject fast and was able to hold on to it very well as long as the subject did not change its pattern (speed, direction, etc) of movement.

[Bill Foxworthy]

That certainly makes sense. I wonder why he doesn't account for that in his views? Is he just a Nikon person?

[Robert Hardy]

I could be wrong, but it seems to me that Rob found that the Mark IV was able to lock on to the subject fast and was able to hold on to it very well as long as the subject did not change its pattern (speed, direction, etc) of movement.

yes, doesent sound to good for most birds as not many fly in a straight line .
Rob

[Jim Fenton]

The question is...is this really a problem?

Nobodies AF is perfect and I think we all expect some OOF frames...I certainly get them with my Nikon equipment.

[arash_hazeghi]

That certainly makes sense. I wonder why he doesn't account for that in his views? Is he just a Nikon person?

he mentions all of these issues thoroughly for both Nikon and Canon, including subject speed, time of day and type of lightning it's in his article.

[arash_hazeghi]

The question is...is this really a problem?

Nobodies AF is perfect and I think we all expect some OOF frames...I certainly get them with my Nikon equipment.

Of course, but it is relevant if you ask how many OOF frames you get from camera A versus camera B in the same shooting conditions.

[Desmond Chan]

Of course, but it is relevant if you ask how many OOF frames you get from camera A versus camera B in the same shooting conditions.

As somebody pointed out on another forum, it appears that Rob found the D3s a more reliable camera overall.

[chuck Shear]

In fact I've never had a camera that focused as well as the Mark IV. I must admit I'm a bit skeptical as Nikon buys add space on RG's website. I did have a Mark III and now have a 5D Mark II and 40D in addition to the Mark IV. No question that the Mark IV is in another league. I'm not a sports shooter so clearly my demands are different than are Gailbraith's . Didn't RG post a link to a friend of his (sports shooter) who claimed it was the best focusing camera he had ever used?

[Axel Hildebrandt]

I'm a bit surprised that he only used central sensor with surrounding points. This makes sense to me if the subject is small such as birds, for sports one would think more AF points might come in quite handy.

[arash_hazeghi]

I'm a bit surprised that he only used central sensor with surrounding points. This makes sense to me if the subject is small such as birds, for sports one would think more AF points might come in quite handy.

I think because human subjects are large it is easy to keep a single AF point on the subject at all times, thus eliminating the randomness of which point the camera chooses for more definitive results. With birds since they are small and fly erratically they can easily slip away from the center sensor and cause the focus to shift to BG. expansion AF points *if working as advertised* should maintain focus on BIF and hold on to it as it leaves the center point.

I have recently found that AF expansion mode works well with 5DMKII with 500 f/4 lens (there are 6 invisible AF points in the proximity of the center point) even against busy BG, but not so well in the case of 7D. I wonder how is it like in MKIV?

Here is an example the center AF point was completely off the subject but the assist point grabbed the neck. The bird is only 1200 pixels wide in RAW. 5D MKII

[Axel Hildebrandt]

This makes sense with a predictable movement, RG was mentioning soccer and I'm wondering if more AF points would have helped or hurt. I've never photographed sports events, though.

[arash_hazeghi]

This makes sense with a predictable movement, RG was mentioning soccer and I'm wondering if more AF points would have helped or hurt. I've never photographed sports events, though.

Most of the cases RG was originally and is still complaining about is when the AF sensor is dead on target yet the subject is completely OOF, one of my friend photographers calls it "full blown AF :D" why do you think expansion mode would help in these cases?

[Axel Hildebrandt]

Most of the cases RG was originally and is still complaining about is when the AF sensor is dead on target yet the subject is completely OOF, one of my friend photographers calls it "full blown AF :D" why do you think expansion mode would help in these cases?

To see if it works better. :)

[arash_hazeghi]

To see if it works better. :)

Fair enough, although there is no reason why it should, but I agree everything is worth trying!

[Axel Hildebrandt]

Fair enough, although there is no reason why it should, but I agree everything is worth trying!

Fortunately, the camera works really well for static birds and birds in flight, even if the BG is varying and in this instance, the tracking sensitivity seems to make a difference. RG didn't find differences when photographing sports.

[John Chardine]

RG mentions on the setup info page that he found AF expansion around the selected point to work the best, so clearly he used it.

[Ed Cordes]

I know I am biased as my MK4 has been outstanding. However, after re reading RG's article I am disappointed in the overall tone and style of his writing. His choice of adverbs and adjectives clearly show a bias in writing. For instance, he states that "At speedskating, the camera managed a healthy number of in-focus frames of skaters rounding the corner". The word "managed" clearly implies a negative bias.

The photographers on this site are among the best in the word; in one of the most difficult disciplines. Those using the MK4 have been very happy. How can there be this difference?

[John Chardine]

I found that too Ed. As he launched into a D3s paragraph suddenly the tone would soften. But this is exactly the point. We are all biased and can potentially influence results. It can even happen in well controlled experiments, which this was not. Having recognised that all things like this have bias, you have to see through to what is useful, and there was lots useful in the article. BTW I have a mk IV and I love it!

[Dave Courtenay]

I believe that RG is still a Canon user choosing the MK3 over the Nikon,I also remember when his site was covered in Canon sponsorship which seemed to end after the MK3 fiasco, He says he used 5 seperate MK4's and doesnt totally slate the camera saying the AF shows signs of brilliance, There are people having problems with their MK4's -just read Naturescapes,He also says that the D3s isnt perfect just more consistant than the Canon, I am sure Canon will take this on board and tweek the AF a bit more, It seems a big improvement over the MK3 so Canon are nearly there, I didnt read the test as a totally negative thing, It shows that where one AF system is srong the other isnt and vis versa

Dave

[arash_hazeghi]

Fortunately, the camera works really well for static birds and birds in flight, even if the BG is varying and in this instance, the tracking sensitivity seems to make a difference. RG didn't find differences when photographing sports.

Good to hear Axel, look forward to testing the camera with raptors.

[Desmond Chan]

There are people having problems with their MK4's -just read Naturescapes [snip]
Dave

I somehow found it interesting, too. Here's all happy happy. There...well, you go read the comments and be your own judge :)

[Desmond Chan]

This makes sense with a predictable movement, RG was mentioning soccer and I'm wondering if more AF points would have helped or hurt. I've never photographed sports events, though.

Depending on the sports but, generally speaking, humans are bigger and likely move slower than a flying bird. If your camera has no problem focusing on small birds, it should not have problems focusing on soccer players, for example.

Perhaps Rob should test the AF on free flying tree swallows :)

[Fabs Forns]

I didn't read the article and will not read it either. I rather rely on my experience.
The Mark III made me switch to Nikon. I'm back to Canon for a couple of month now and after trying Al's 1d4, I think it works as well if not better than the 1d2, my hero at the time.
I should add that I'm very happy they did not "improve" the AF on the 5D2 because it works great for birds in flight, only thing is the slow frame rate.

[Doug Brown]

I've barely gotten to use the Mark IV with my 500 or 600 to this point. I am going to the Bosque tomorrow afternoon, and will hopefully get plenty of frames under my belt. I've been very pleased with the AF in combination with my 400 f/5.6 lens. And my results in poorly lit gyms photographing my son's basketball games with the 85 f/1.8 at ISO 2000-3200 have been phenomenal!

Like Fabs, I prefer to draw my own conclusions about AF.

[Charles Glatzer]

I'm a bit surprised that he only used central sensor with surrounding points. This makes sense to me if the subject is small such as birds, for sports one would think more AF points might come in quite handy.

AxelL,

Single point, perhaps with expansion is the way to go for most sports that have a runners face on different plane than the chest, especially when shooting a long lens (400 f/2.8) wide open.

Chas

[arash_hazeghi]

I didn't read the article and will not read it either. I rather rely on my experience.
The Mark III made me switch to Nikon. I'm back to Canon for a couple of month now and after trying Al's 1d4, I think it works as well if not better than the 1d2, my hero at the time.
I should add that I'm very happy they did not "improve" the AF on the 5D2 because it works great for birds in flight, only thing is the slow frame rate.

Fabs, In your opinion how does MKIV compare against your ex D3 especially when tracking BIF against varied BG in expansion AF mode?

After I tried 5D MKII with 500 I agree 100% that is AF is good for flight, noticeably more consistent and predictable than the 7D. Wish it was faster :(

BTW, you don't need to read RGs review, you can just download more than 2GB of full-size images that he has with AF point superimposed and intact EXIF to evaluate for yourself. I find it a very useful.

He is also asking for paypal donations! That is hillarous :D

[Jay Gould]

Personally, I think I will rely on this review: http://www.the-digital-picture.com/R...ra-Review.aspx

[Doug Brown]

Initial impression after a day at the Bosque; the Mark IV does very well with Cranes flying against a varied BG. Better than either the 7D or the Mark III. I'm pretty pleased!

[Axel Hildebrandt]

Initial impression after a day at the Bosque; the Mark IV does very well with Cranes flying against a varied BG. Better than either the 7D or the Mark III. I'm pretty pleased!

Good to hear that it works for you, too. What tracking sensitivity and AF points did you use?

[Ken Watkins]

I never had any serios problems with the MkIII, I am just happy that AF works at all, it was certainly lots more difficult when everything was manual. Some people seem just to like to complain, if everything was perfect everytime then why would we bother!

I'll be picking up my MkIV next week, I am sure it will be as satisfactory as every other Canon I have used.

[Doug Brown]

What tracking sensitivity and AF points did you use?

I used center point with tracking sensitivity set to slow. I had a 90-95% keeper rate yesterday morning while photographing flying Cranes against varied BGs. Not the most challenging subject, but I know what kind of keeper percentage to expect with the 7D and the Mark III at the Bosque. As I said above, this was an improvement. Here's the kind of quality I was getting; this is a 100% crop of an unprocessed RAW file (save for Lightroom defaults).

[Doug Brown]

Same crop but with post-processing (my typical workflow).

[Charles Glatzer]

Doug,

You mention having a 90-95% keeper rate, but what percentage were tack sharp? The provided example does not appear tack to my eyes.

I have yet to test the camera with BIF.

Best,

Chas

[Doug Brown]

Chas, I didn't claim that this image was tack sharp; I stated that this was representative of my keepers. Could it be slightly sharper? Certainly. But this level of sharpness is quite high IMO when you're talking about 100% unprocessed crops of birds in flight. Remember too that Cranes don't have the fine feather detail of many other avian species.

It would help me if you would post an unprocessed 100% crop of a flight image (preferably a Crane) that you do consider to be tack sharp.

[Charles Glatzer]

Chas, I didn't claim that this image was tack sharp; I stated that this was representative of my keepers. Could it be slightly sharper? Certainly. But this level of sharpness is quite high IMO when you're talking about 100% unprocessed crops of birds in flight. Remember too that Cranes don't have the fine feather detail of many other avian species.

It would help me if you would post an unprocessed 100% crop of a flight image (preferably a Crane) that you do consider to be tack sharp.

I will have to dig through my Bosque files from two years ago. Will post later today, but a different camera, anti-aliasing filter, etc will not make the comparison relevant.

Download C1 Pro trial and use the Focus tool to judge sharpness out of camera. I think you will be amazed and/or disheartened at the results.

Best,

Chas

[Fabs Forns]

I never had any serios problems with the MkIII, I am just happy that AF works at all, it was certainly lots more difficult when everything was manual. Some people seem just to like to complain, if everything was perfect everytime then why would we bother!

I'll be picking up my MkIV next week, I am sure it will be as satisfactory as every other Canon I have used.

When I spend $4500 on a brand new body that is supposed to have the best AF ever and the thing is blind half of the time, and they tell you that it is operator error and 4 months later they do a big recalll, yes, I am likely to complain.
Wouldn't you?
If you did static photography, Mark III may have been good for you.
For BIF photography it was a joke.

[arash_hazeghi]

Same crop but with post-processing (my typical workflow).

Doug this sample is very nice, can you quantify what was your keeper rate with 7D in comparison with MKIV in similar conditions?

Did you use mostly single point or AF expansion?

[Peter Hawrylyshyn]

Doug -
The pic looks like a keeper to me. What ISO, f stop and shutter speed did you use?
The Canon Mark 4 AutoFocus PDF (pg 34-35) points out that because of the improved 16.1MP resolution of the MK 4, images at 100% will be much more likely to show minor motion blur or camera shake than the Mk3, as opposed to an "automatic" AF issue. They recommend using faster shutter speeds (1/1000 - 1/2000) at higher ISO's (which may not always be possible for BIF).

[Doug Brown]

I will have to dig through my Bosque files from two years ago. Will post later today, but a different camera, anti-aliasing filter, etc will not make the comparison relevant.

Download C1 Pro trial and use the Focus tool to judge sharpness out of camera. I think you will be amazed and/or disheartened at the results.

Best,

Chas

I've taken and processed many hundreds of thousands of BIF images Chas. I think it unlikely that I'd be disheartened with this image or any of my keepers from yesterday based on an interpretation by Capture 1's software. These images look great sized for web and they look great printed 13x19; they'd probably look great at 20x30 too. I have photographed with some of the best flight photographers in the world and I've seen what their RAW files look like. No one is cranking out close to 100% tack-sharp BIF images, and all would be very happy with the results I got yesterday.

For me if a camera produces excellent but not tack sharpness in more than 9/10 frames in the sizes that I care about, I'm a happy camper.

[Charles Glatzer]

Doug,

My reference to C1 was not image specific, it was suggested as a tool for comparative analysis for all viewing the post. Sometimes we think an image is sharp until we view another side by side. Distinct edge detail is the deciding factor for me when viewing a RAW and/or processed image, but as you know there are many factors that affect images apparent/perceived sharpness. I am admittedly anal with regard to what I consider definitively sharp with my own images, probably to my detriment.

I should not be so critical of images posted on the web least I put my foot in mouth (again), as I have personally stated before that trying to compare different images for sharpness on the web is nearly impossible unless taken from the same camera and lens, with similar exposure and ISO, environmental conditions, post-processing software, viewed monitor res, etc.

I agree, no one, myself included is capable of producing tack sharp images every time they depress the shutter. I have sat with hundreds of people since going digital who show me what they consider to be a very sharp image. Many photogs’ are unaware of what a critically sharp image even looks like, and when shown the difference many change their own impression of what is and is not acceptably sharp. In the end the whole thing is subjective anyway. And Visual Impact will win out every time.

Best,

Chas

[Doug Brown]

Doug -
The pic looks like a keeper to me. What ISO, f stop and shutter speed did you use?
The Canon Mark 4 AutoFocus PDF (pg 34-35) points out that because of the improved 16.1MP resolution of the MK 4, images at 100% will be much more likely to show minor motion blur or camera shake than the Mk3, as opposed to an "automatic" AF issue. They recommend using faster shutter speeds (1/1000 - 1/2000) at higher ISO's (which may not always be possible for BIF).

Hi Peter. This image was shot with the 600mm f/4, f/7.1, 1/1600, ISO 400, manual exposure, hand held.

[Fabs Forns]

The chances of making a critically sharp image with a hand held 600 are very slim IMO.
If you look at Chris Van Rooyen's work, his flights are never under 1/2500, and I don't even think he hand holds.

With the 1.3 crop factor, you are hand holding 900mm, 1/1600 hardly seems fast enough for a tripod photo, let alone handholding. And yes, a critically sharp focused image shines in the RAW file, even if we can't produce them every time we press the shutter.
Only my two cents, from one who can hand hold but chooses not to 95% of the time.

[Charles Glatzer]

This image above was taken with the best camera and lens in Canon's stable, it sets the gold standard for me regarding what my equipment is capable of producing under ideal conditions. Are other cameras and lens combinations capable of achieving similar results, and is it fair of me to judge my images against this one, maybe not as the variables are many. But, nonetheless sharp is sharp. As individuals we get to choose what we consider to be acceptably sharp according to our own set of standards. What one photog considers sharp, another may not, and all may be considered of “publishable quality”.

The image presented is a 100% crop of a RAW file w/o any in-camera or post sharpening applied, captured via SnagIt, converted to sRGB and jpeg compressed with BoxTop software in CS4. See image info for additional details.

I agree with Fabs, shutter speed has all to do with obtaining maximum image sharpness.

Best.

Chas

[Doug Brown]

And yes, a critically sharp focused image shines in the RAW file, even if we can't produce them every time we press the shutter.

I don't disagree with this statement Fabs. And I didn't call my Crane image tack sharp. We all get the occasional RAW file that knocks our socks off. But those frames are the exception, not the rule, in BIF photography. Chas seems to be insinuating that I don't know what a truly sharp RAW image looks like. He says he's going to show me a tack sharp Crane image from the Bosque, and instead shows me a Bald Eagle image shot at close range with a FF body and a 300 f/2.8 lens. Bald Eagles have crisp feather and beak detail that Cranes simply don't have. And distance to subject also has a big impact on image quality. Chas selected his gold standard image for BIF sharpness; I chose one out of hundreds of similarly sharp frames produced in a single morning at the Bosque.

I understand that there's a difference between critically sharp and just plain sharp. I'm more interested in my percentage of sharp images than my percentage of tack sharp images. How often do any of us get RAW files that look as good as Chas's Bald Eagle? Unless I'm doing something wrong, the answer is not very often. I base my camera's AF evaluation on how sharp the images are that I can consistently produce. If I get a Chas-quality image 1% of the time, that's icing on the cake as far as I'm concerned.