Canon RF 200-800mm IS USM Previews / Reviews

[Dragon]

Some comments and questions the next morning.
1. RF 200-800mm on FF versus RF 100-500mm on R7
He complains that the AF of the 200-800 on the R7 isn't good, and I wouldn't want that lens on an APS-C anyway, so let's consider its designed for FF, and what it does as is in my last post is to give a similar field of view range as the RF 100-500mm on APS-C. So, I'd like to see a), a comparison of the IQ of 100-500mm on an R7 or lower Mpx APS-C vs 200-800mm on the R5 and lower Mpx FF. b) a comparison of the AF at 800mm on the R5 etc with AF on the R7 etc.
2. He doesn't recommend the use of extenders on the 200-800 but does show some good images. However, how do the images at 1120mm and 1600mm compare with that at 800mm in terms of additional real resolution. I would guess not much in practice.

I think his conclusion that it wont replace his 100-500 but he will keep it as a useful extra is fair comment. But, what I really want to know is it even significantly useful if you have an R7 if that plus the RF 100-500 covers what the R5 or R6 Plus 200-800 can do in practice for IQ and AF for flying birds? Lugging around and packing a larger lens, the poorer AF, poorer IQ zoomed in and lower magnification at mfd are negatives. But, the R7 isn't my first choice for BIF, and from what I see from him the 100-500 on the R5 or other FF are better than 200-800mm

He did note that the R7 AF with the lens was much better in good light and that is just the nature of the R7. One observation didn't quite make sense. He was discussing "brightness bleed" and assigning the problem to the lens, but then went on to say that he had to decrease exposure to resolve the problem. If decreasing the exposure resolves the problem, the lens is not the issue other than that it may report lower brightness than it actually delivers so the camera overexposes. I find that with both the R5 and R7 pulling the exposure down 1/3 to 2/3 stop from default generally results in a more flexible raw file with most lenses unless I am using spot exposure. That is to say that both cameras tend to overexpose highlights more than I would like with evaluative metering set at default exposure.

Lastly, the teleconverter results are encouraging.

 

[AlanF]

He did note that the R7 AF with the lens was much better in good light and that is just the nature of the R7. One observation didn't quite make sense. He was discussing "brightness bleed" and assigning the problem to the lens, but then went on to say that he had to decrease exposure to resolve the problem. If decreasing the exposure resolves the problem, the lens is not the issue other than that it may report lower brightness than it actually delivers so the camera overexposes. I find that with both the R5 and R7 pulling the exposure down 1/3 to 2/3 stop from default generally results in a more flexible raw file with most lenses unless I am using spot exposure. That is to say that both cameras tend to overexpose highlights more than I would like with evaluative metering set at default exposure.

Lastly, the teleconverter results are encouraging.

For bird photography, it's better to use full manual and judge the exposure to avoid overexposing highlights and its pretty well essential if you have a bird flying between a background of clear sky and darker woods etc. The sunny 16 rule is rather useful and on a mirrorless what you see is close enough to what you get in order not to go far wrong.

 

[Dragon]

For bird photography, it's better to use full manual and judge the exposure to avoid overexposing highlights and its pretty well essential if you have a bird flying between a background of clear sky and darker woods etc. The sunny 16 rule is rather useful and on a mirrorless what you see is close enough to what you get in order not to go far wrong.

I wholly agree with you for a planned shoot, but sometimes with opportunistic shots you just don't have the time for such planning, particularly when walking through the woods where lighting can change dramatically in a few dozen steps. Spot exposure works well for stationary subjects, but not well at all for BIF. Spot exposure tracking the AF point would be delightful. That feature could inspire me to buy an R1 as that will likely be the only place we will see such capability in the reasonable future.

 

[AlanF]

I wholly agree with you for a planned shoot, but sometimes with opportunistic shots you just don't have the time for such planning, particularly when walking through the woods where lighting can change dramatically in a few dozen steps. Spot exposure works well for stationary subjects, but not well at all for BIF. Spot exposure tracking the AF point would be delightful. That feature could inspire me to buy an R1 as that will likely be the only place we will see such capability in the reasonable future.

I have 3 different settings programmed into the mode dial/button for likely situations and the 4th position is Fv or manual. The 3 programmed are all auto iso.

 

[Quack]

He did note that the R7 AF with the lens was much better in good light and that is just the nature of the R7. One observation didn't quite make sense. He was discussing "brightness bleed" and assigning the problem to the lens, but then went on to say that he had to decrease exposure to resolve the problem. If decreasing the exposure resolves the problem, the lens is not the issue other than that it may report lower brightness than it actually delivers so the camera overexposes. I find that with both the R5 and R7 pulling the exposure down 1/3 to 2/3 stop from default generally results in a more flexible raw file with most lenses unless I am using spot exposure. That is to say that both cameras tend to overexpose highlights more than I would like with evaluative metering set at default exposure.

Lastly, the teleconverter results are encouraging.

Thanks for pointing out that his struggles with the AF on the R7 were not a product of the lens, but of the R7, and that it was mainly in low light. It seems many folks are latching on to the negatives and ignoring Jan's comments that make those negatives much less severe. It also makes me wonder if he was shooting at 30 fps on the R7, as it seems that it has been established that the slower read-out speed is partly to blame for the AF difficulties that some folks have. I always shoot at 15 FPS or slower on the R7 and find the AF works very well. Better in some situations than my R5, especially if the bird is flying over water and water is in the background of the shots.
My impression was that that Jan was pleasantly surprised by how good the lens is and that it exceeded his expectations. I got the same impression from Duade Paton's video. As a RF 100-500 owner, I may not buy this lens, but I plan on renting it at some point to see how it does. The main drawback for me is the size and weight - although I know that for it's focal length, it is actually lighter than one might expect.

 

[Klas R]

I suddenly understood that you can achieve 600 and 700 mm both with the RF 100-500 mm equipped with the 1.4X TC and the 2.0X TC.

Has anyone out there compared the results for each one of these focal length combinations?
Which one gave the sharpest pictures? Which combination gave the best contrast?

Differences that should seriously be taken into consideration?

 

[AlanF]

I suddenly understood that you can achieve 600 and 700 mm both with the RF 100-500 mm equipped with the 1.4X TC and the 2.0X TC.

Has anyone out there compared the results for each one of these focal length combinations?
Which one gave the sharpest pictures? Which combination gave the best contrast?

Differences that should seriously be taken into consideration?

According to Canon's MTF charts, the 1.4x is sharper and more contrasty at both.

 

[Klas R]

According to Canon's MTF charts, the 1.4x is sharper and more contrasty at both.
View attachment 213625

Thank you Alan !
Very much appreciated, and as well excellent knowledge and input for future photographic excursions.

 

[Dragon]

According to Canon's MTF charts, the 1.4x is sharper and more contrasty at both.
View attachment 213625

Canon MTF charts also suggest the 200-800 will have slightly better sharpness and virtually identical contrast at 800mm as the 100-500 with 1.4x at 700mm. More testing is needed to find where statistical reality lies (given sample variation), but in any case, the lens performance looks very close, with the 200-800 having an extra 100mm and 1/3 stop advantage, which may give it a slight edge for those working near the practical limits of ISO.

 

[AlanF]

Excellent!

However, I notice that Jan W has already received 331 comments related to his review.

He will have a lot to do just reading them. Though I think he takes notes on interesting questions and comments.
Hopefully we will get some answers in the next presentation.

He has now attached a heart to all the comments except for my questions.

 

[AlanF]

I suddenly understood that you can achieve 600 and 700 mm both with the RF 100-500 mm equipped with the 1.4X TC and the 2.0X TC.

Has anyone out there compared the results for each one of these focal length combinations?
Which one gave the sharpest pictures? Which combination gave the best contrast?

Differences that should seriously be taken into consideration?

I've checked my back tests. On the R5, the RF 2x on the 100-500mm at 600mm outresolves the bare lens at 500mm. At 726mm, it slightly outresolves the lens with the 1.4xTC at 700mm. I conducted a long series of tests with extenders in the past few days and I am having a new 1.4x delivered tomorrow. There's a huge cash back on it just announced. What has surprised me, is that the 1.4x is consistently more impressive when used on the R7. Anyway, if anything interesting turns up in my testing of the new TC, I'll report it.

 

[Dragon]

I've checked my back tests. On the R5, the RF 2x on the 100-500mm at 600mm outresolves the bare lens at 500mm. At 726mm, it slightly outresolves the lens with the 1.4xTC at 700mm. I conducted a long series of tests with extenders in the past few days and I am having a new 1.4x delivered tomorrow. There's a huge cash back on it just announced. What has surprised me, is that the 1.4x is consistently more impressive when used on the R7. Anyway, if anything interesting turns up in my testing of the new TC, I'll report it.

Curious to see what you find. Even the spendy stuff has sample variation.

 

[Quack]

I've checked my back tests. On the R5, the RF 2x on the 100-500mm at 600mm outresolves the bare lens at 500mm. At 726mm, it slightly outresolves the lens with the 1.4xTC at 700mm. I conducted a long series of tests with extenders in the past few days and I am having a new 1.4x delivered tomorrow. There's a huge cash back on it just announced. What has surprised me, is that the 1.4x is consistently more impressive when used on the R7. Anyway, if anything interesting turns up in my testing of the new TC, I'll report it.

Are you testing on a tripod or hand held? And what distance to your subject/test chart? I think the majority of users will be hand-holding, so very curious, as my not particularly scientific, hand holding comparisons with many lenses in the past has led me to believe that extenders are usually not worth it compared to cropping the images without extender on anything but really fast aperture lenses, typically primes - especially for long distance shots. Jan mentions this in his video - and I would agree - that contrary to what most folks want the extenders for - they work best for small subjects that are close to middle distance away, not to give additional benefit for resolving very distant subjects.

 

[Dragon]

Are you testing on a tripod or hand held? And what distance to your subject/test chart? I think the majority of users will be hand-holding, so very curious, as my not particularly scientific, hand holding comparisons with many lenses in the past has led me to believe that extenders are usually not worth it compared to cropping the images without extender on anything but really fast aperture lenses, typically primes - especially for long distance shots. Jan mentions this in his video - and I would agree - that contrary to what most folks want the extenders for - they work best for small subjects that are close to middle distance away, not to give additional benefit for resolving very distant subjects.

Aside from magnifying small subjects, things that are closer will inherently have more opportunity for increased resolution most of the time and that is simply because of the air. In most places, there are very limited times when long distance shots are not mucked up by various air issues including fog, haze, smoke, and most often thermal distortion. This is particularly true if the distant object is at a similar elevation as the camera. OTOH, on those rare occasions where the air is truly clear, then all the magnification you have available can be useful for long shots, but a lot of patience and constant observation of atmospheric conditions is required.

 

[AlanF]

Are you testing on a tripod or hand held? And what distance to your subject/test chart? I think the majority of users will be hand-holding, so very curious, as my not particularly scientific, hand holding comparisons with many lenses in the past has led me to believe that extenders are usually not worth it compared to cropping the images without extender on anything but really fast aperture lenses, typically primes - especially for long distance shots. Jan mentions this in his video - and I would agree - that contrary to what most folks want the extenders for - they work best for small subjects that are close to middle distance away, not to give additional benefit for resolving very distant subjects.

I test all my lenses hand held under conditions I use for them in the wild (I don't use a tripod there) and on a tripod for careful comparisons. I use three distances: close to minimum focus distance for my use with insects; and standard distances of mainly 19m and also 12m, all outside in natural light. There is always a Bob Atkins chart Blutacked to a wall so I can nip out and test. I test under gloomy conditions and in bright light. A couple of days ago, I tested the 100-500mm and 800mm/11 with the 1.4x and 2xTCs on a tripod on gloomy light at speeds of 1/125s and 1/400s. I wanted to check the consistency of focus and then resolution of the zoom at 700mm at 16.75m with the 800mm at 19m where both images would be the same size on the sensor. I did 20 shots for each. The results were very disappointing as most shots were affected by shake at the pixel level. Yesterday, I repeated them in reasonable light, not direct sun, and there was excellent consistency, and much better resolution and contrast. I don't like going below 1/1000s hand held in real life. Whatever they say about IS and the charts you see on test sites, you need high shutter speeds to get tack sharp with telephotos - their tests are just for larger visible shake and not for microscale movements that soften and lose contrast.

As for when extenders are useful, it is too easy to make simplistic remarks and also it depends on what you are doing. An extender will always put more pixels on an image, give you a larger image that requires less magnification to view at the same size or the ability to blow up larger. I use an extender to get more resolution, which is not necessarily the same as putting more pixels on the image. That works only if the detail you want to see is just too fine to be resolved without the extender but the extra focal length will just bring it out. Nowadays, you can upresolve a low resolution image with Gigapixel etc to put more Mpx on the image and so if the image with an extender has no more resolution than without, post processing enlargement is just about as good as using an extender. If the bird, in my case, is sufficiently close that the bare lens lens resolves sufficient detail, then the extender is redundant and may make things worse if lowers IQ. For me in most cases, if I can have 1500 px along the length of the bird, the image is good enough and 2500 super sharp. Even as low as 800 px with a sharp lens will get me something good enough for a greetings card. For very long distances, an extender may bring the image within a range I can upresolve it. The same reasoning applies to focal length - a super sharp image at 500mm may give better results than a softer one at 800mm that resolves no more detail.

In a nutshell, I blow hot and cold about longer lenses and extenders as they are advantageous in specific situations and "reach" is only one factor.

 

[AlanF]

Aside from magnifying small subjects, things that are closer will inherently have more opportunity for increased resolution most of the time and that is simply because of the air. In most places, there are very limited times when long distance shots are not mucked up by various air issues including fog, haze, smoke, and most often thermal distortion. This is particularly true if the distant object is at a similar elevation as the camera. OTOH, on those rare occasions where the air is truly clear, then all the magnification you have available can be useful for long shots, but a lot of patience and constant observation of atmospheric conditions is required.

Where I come from, heat haze is the least of my worries and fog though more common still not too much of a problem...

 

[Quack]

I test all my lenses hand held under conditions I use for them in the wild (I don't use a tripod there) and on a tripod for careful comparisons. I use three distances: close to minimum focus distance for my use with insects; and standard distances of mainly 19m and also 12m, all outside in natural light. There is always a Bob Atkins chart Blutacked to a wall so I can nip out and test. I test under gloomy conditions and in bright light. A couple of days ago, I tested the 100-500mm and 800mm/11 with the 1.4x and 2xTCs on a tripod on gloomy light at speeds of 1/125s and 1/400s. I wanted to check the consistency of focus and then resolution of the zoom at 700mm at 16.75m with the 800mm at 19m where both images would be the same size on the sensor. I did 20 shots for each. The results were very disappointing as most shots were affected by shake at the pixel level. Yesterday, I repeated them in reasonable light, not direct sun, and there was excellent consistency, and much better resolution and contrast. I don't like going below 1/1000s hand held in real life. Whatever they say about IS and the charts you see on test sites, you need high shutter speeds to get tack sharp with telephotos - their tests are just for larger visible shake and not for microscale movements that soften and lose contrast.

As for when extenders are useful, it is too easy to make simplistic remarks and also it depends on what you are doing. An extender will always put more pixels on an image, give you a larger image that requires less magnification to view at the same size or the ability to blow up larger. I use an extender to get more resolution, which is not necessarily the same as putting more pixels on the image. That works only if the detail you want to see is just too fine to be resolved without the extender but the extra focal length will just bring it out. Nowadays, you can upresolve a low resolution image with Gigapixel etc to put more Mpx on the image and so if the image with an extender has no more resolution than without, post processing enlargement is just about as good as using an extender. If the bird, in my case, is sufficiently close that the bare lens lens resolves sufficient detail, then the extender is redundant and may make things worse if lowers IQ. For me in most cases, if I can have 1500 px along the length of the bird, the image is good enough and 2500 super sharp. Even as low as 800 px with a sharp lens will get me something good enough for a greetings card. For very long distances, an extender may bring the image within a range I can upresolve it. The same reasoning applies to focal length - a super sharp image at 500mm may give better results than a softer one at 800mm that resolves no more detail.

In a nutshell, I blow hot and cold about longer lenses and extenders as they are advantageous in specific situations and "reach" is only one factor.

Thanks Alan for your reply. Glad you are testing under real-life situations. I agree that the number of variables makes generalizations a bit too simplistic as results can differ depending on camera, lens, amount of light, distance, atmospheric conditions and ability to hand-hold steadily. While many folks seem to mention heat haze when they talk about atmosphere, humidity is a huge factor in my environment (New York State, where heat haze is minimal). If there is one generalization that may be accurate for newcomers to photography and extenders, it is that you can pretty much always say, "lower your expectations!" Shall I get an extender may be the number one question that people ask about on the Facebook groups I follow, as many folks just do the math and assume they are getting 1.4x or 2x the resolution if they get an extender.

I took some comparison shots this morning on my R5 with the RF 100-500, hand held (but with elbows resting on a ledge), in dark, humid, overcast conditions. For practical purposes, the images are virtually identical, but with a heavy crop - looking at the subject of both images at the same size (thus at 100% and 140% on the computer screen) - the non-extender shots were slightly sharper and had slightly more contrast.

 

[Dragon]

Thanks Alan for your reply. Glad you are testing under real-life situations. I agree that the number of variables makes generalizations a bit too simplistic as results can differ depending on camera, lens, amount of light, distance, atmospheric conditions and ability to hand-hold steadily. While many folks seem to mention heat haze when they talk about atmosphere, humidity is a huge factor in my environment (New York State, where heat haze is minimal). If there is one generalization that may be accurate for newcomers to photography and extenders, it is that you can pretty much always say, "lower your expectations!" Shall I get an extender may be the number one question that people ask about on the Facebook groups I follow, as many folks just do the math and assume they are getting 1.4x or 2x the resolution if they get an extender.

I took some comparison shots this morning on my R5 with the RF 100-500, hand held (but with elbows resting on a ledge), in dark, humid, overcast conditions. For practical purposes, the images are virtually identical, but with a heavy crop - looking at the subject of both images at the same size (thus at 100% and 140% on the computer screen) - the non-extender shots were slightly sharper and had slightly more contrast.

I think the term "heat haze" is a bit misleading. The actual problem with thermal distortion is temperature gradients. I live on the Oregon coast were it rarely gets hot, but all summer we have temperature inversion (i.e. increasing temperature with increasing altitude). That condition also causes eddy currents of warm and cool air to form lenticular structures in the air. On a casual glance, the air looks perfectly clear, but if you throw an 800mm lens into the mix, the subject will be very distorted. Video really shows up the problem as the subject will wiggle and squirm like a worm, when actually not moving at all. I normally don't save the distorted images, but the below example will give the idea. The first image was shot at 400mm with an R7 and a Tamron 18-400 and the second with an SL2 and an EF800mm f/5.6L with 2x TC. The lenses are in no way comparable, but you can see the the lens is not the limiting factor in the first image. The power wires are wiggly and even the pole is not entirely straight. You can also see the areas of fuzziness, particularly on the dangling arc suppressors. That was a bad air day, but I have seen many worse. The second image shows just how clear the image can be on a good day with a good lens, whereas in the first case, a better lens would have only served to show up the wiggles and fuzz to greater extent. The distortion problem is worst when the air is still and the sun is shining. If there is a breeze blowing, it tends to break up the lenslets that form in the air and clears the image, at least to a degree. For reference, the power pole is about a mile away about 800 ft higher than the camera in an area of often heavy thermal inversion.

 

[kaihp]

I don't like going below 1/1000s hand held in real life. Whatever they say about IS and the charts you see on test sites, you need high shutter speeds to get tack sharp with telephotos - their tests are just for larger visible shake and not for microscale movements that soften and lose contrast.

Not going below 1/1000s is bit of a luxury.

When shooting sports like motorcycle racing, you want to show the speed and motion - and using the slow shutter is paramount. If you shoot at below 1/1000s and make the wheels completely still, it will look like the bike is propped up with a stick.

My limit is currently around 1/160s, where I get around 15-20% keepers.

 

[AlanF]

Not going below 1/1000s is bit of a luxury.

When shooting sports like motorcycle racing, you want to show the speed and motion - and using the slow shutter is paramount. If you shoot at below 1/1000s and make the wheels completely still, it will look like the bike is propped up with a stick.

My limit is currently around 1/160s, where I get around 15-20% keepers.

I was discussing the technique of photographing birds with @Quack leading on from the bird photographer Jan Wegener's review.