Summary of my RF 200-800mm testing

[AlanF]

To me, it looks more like a slight mis-focus. I regularly use my EF 100-400IIL + 1.4x TC, fully racked out and it's super sharp. The main issue for me is the reduced AF speed and accuracy.
To put my comments in to some sort of perspective, I also use a EF 400mm f2.8 LIS mk II, which is known to be a one of Canon's sharpest lenses ever made. I also use a R8 and R6ii, whihc both share a very sharp 24mp sensor. Both my EF 400mm f2.8 LIS II, native, with a 1.4x or a 2x TC seem to still out resolving the Sensor on my two camera bodies. My EF 100-400mm f5.6 LIS II with or without a 1.4x TC are also out resolving my camera sensors. So between my 400 prime and my 100-400 zoom, wide open with say a 1.4x TC...there is no sharpness difference between these two optics. However on a R5, or a R7 which has a far higher pixel density, you might see more of a difference. On my R6ii, I get exceptional sharpness with either lens.

DocSmith, The comparision between the EF 100-400 f5.6 LIS II and the RF 100-500LIS are well documented.
There are slight benefits and deficiencies with both lenses. The EF lens is brighter natively, but less focal length. A good copy of the EF lens is slightly sharper (even with a 1.4x TC on the EF lens), but this is marginal and propbaly would not be seen in real world photos. The EF lens is reputedly slightly more robust and old skool build. However, the RF lens is lighter (due to it's newer build construction). It's hood is way better than the EF version, which seems to not work as well when reversed. The Rf lens gets a longer focal length at the long end for a reduced aperture rating. The Rf lens has a removable tripod collar (The EF's removable foot is an appaling design and prone of issues). The RF has a superior AF and IS system. At MFD, there is little between them and they both focus breath substantially. The Rf lens can capture samll erratic moving bugs way better than the EF version...it's AF just isn't in the same league. For many, the RF lens is the slightly better option, however if you already ahve a great copy of the EF version, one questions the rationale of side grading to the RF version. the SH value of the Ef version is plummeting into bargain status and the RF version is generally only available new and it's eye wateringly expensive.

AlanF, Your sharpness observations of your RF 200-800 LIS at 800mm are in line with what I've observed, reading the MFT charts. I think that the RF 200-800 LIS is an ideal partner on a R6ii. It's lover pixel density will allow this lens to shine and will be more than sharp enough for 100% crops with that combo.

I've had 3 copies of the EF 100-400mm II, and have compared the best with the RF 100-500mm on the R5. A lot of truth in what you write. A few points in addition. At mfd, the 100-500mm is significantly sharper and better for insects. The RF lens takes the 2x TC better than the EF. The RF is superior at 100mm end.

 

[AlanF]

Both lenses have the same physical aperture, 70mm, what makes you believe the EF one is brighter? If it's the aperture value reported in the EXIF, that is unrealiable. You can get it to show a 'brighter' value just by telling your camera to use 1/8 stops instead of 1/3 stops. Like the focal length on zooms, the reported aperture value needs to be taken with a grain of salt.

And please tell me you don't mean 'f/5.6 is brighter than f/7.1'.....

The actual measured value on an optical bench for the RF 100-500mm at 393mm is f/6.95. https://www.photonstophotos.net/Gen...ample02P.txt,figureOpacity=0.25,AxisO,OffAxis

As you imply for comparing the 400mm f/5.6 vs 500mm f/7.1 full open at wide end, both put the same number of photons on a duck.

 

[Dragon]

One way to interpret that is that you really can't go wrong either way.

I briefly tried out a friend's 200-800 with my R3 this past weekend, I was left with the impressions that there was not a meaningful difference in the image output compared to my RF 100-500L + 1.4x, and not much difference in the handling (the 200-800 weighs more, but it really didn't feel significantly heavier in use, to me.

One thing I didn't care for with the 200-800 was the long zoom throw. The 100-500 with the 1.4x has a much shorter range, being limited to 420-700mm, but the throw is very short over the limited range.

Personally, if I did not have the 100-500 and TCs already, I would be tempted by the 200-800, but having the 100-500 already (and the 600/4 if I really need top IQ and longer reach), I have no plans to get the 200-800. Of course, you're in a different spot. My guess is you'd be less pleased with the IQ at 800mm compared to your EF 100-400L II, but if you consider everything over 400mm as a bonus, you could live with it. Assuming you don't have the RF 1.4x, that plus the 100-500 would deliver similar IQ beyond 600mm but cost at least $1000 more.

Put another way, if you plan to keep your 100-400 II and swap to the 200-800 only when you need longer than 400mm, that makes sense. If you plan to replace your 100-400 II with an RF lens, selling the 100-400 II would likely offset the price difference between the 100-500 + 1.4x and the 200-800 and the former would give you as good/better IQ compared to the 100-400 out to 500mm and similar IQ to the 200-800 beyond that, for a similar price.

I find that the 200-800 works well as a push-pull zoom (particularly the push part). If you start wide, the lens is quite easy to simply push out to the long end without loosing your subject. For me it works much better that twisting the ring.

 

[Klas R]

The other day I took some photos to try to secure my possible identification of a Broad-billed Sandpiper. So the pictures were not "nice portait pictures". Actually the distance to the bird was something like 350 meters, and it measures only 15-18 cm , i.e. 5-6".

They were taken at about 4 PM in the afternoon and I had some severe problems with heat waves. First I took som pics at 800mm at f9. But as the bird was so small on the sensor I added my 2XTC. Took some pictures at 1600mm, f18 but wasn't impressed. Maybe it was the hest waves or maybe it is about the lens performing much better at 600mm, so I finally took some pictures at 1200mm, f16.

To me it looks like the lens performs very good at 1200mm with the 2XTC, but not as well at 1600 mm. But actually you get almost the same information at 800mm only.


Photos from top to bottom: 1200mm, 1600 mm and 800mm

 

[AlanF]

The other day I took some photos to try to secure my possible identification of a Broad-billed Sandpiper. So the pictures were not "nice portait pictures". Actually the distance to the bird was something like 350 meters, and it measures only 15-18 cm , i.e. 5-6".

They were taken at about 4 PM in the afternoon and I had some severe problems with heat waves. First I took som pics at 800mm at f9. But as the bird was so small on the sensor I added my 2XTC. Took some pictures at 1600mm, f18 but wasn't impressed. Maybe it was the hest waves or maybe it is about the lens performing much better at 600mm, so I finally took some pictures at 1200mm, f16.

To me it looks like the lens performs very good at 1200mm with the 2XTC, but not as well at 1600 mm. But actually you get almost the same information at 800mm only.

View attachment 217042View attachment 217043View attachment 217044
Photos from top to bottom: 1200mm, 1600 mm and 800mm

We had a couple of days ago, an extremely rare sighting for us a female Red-necked Phalarope, 18cm long as well, at about the twice the distance as yours away (bird about 120 px long compared with ~250px for yours at 1600mm). I couldn't see it with my x8 bins but I focussed on the indicated area. Not exactly publishable, but it was identifiable! Whether an extender adds anything, depends on whether the increase in resolution will theoretically see more. The photos just before then, posted in the Birds thread, of the Ospreys did give much more detail at 1600mm, but needed a lot of processing.

 

[Klas R]

We had a couple of days ago, an extremely rare sighting for us a female Red-necked Phalarope, 18cm long as well, at about the twice the distance as yours away (bird about 120 px long compared with ~250px for yours at 1600mm). I couldn't see it with my x8 bins but I focussed on the indicated area. Not exactly publishable, but it was identifiable! Whether an extender adds anything, depends on whether the increase in resolution will theoretically see more. The photos just before then, posted in the Birds thread, of the Ospreys did give much more detail at 1600mm, but needed a lot of processing.

View attachment 217048

I was impressed by your photos of the Ospreys! What kind of processing gave that result?
On my pictures I only used DxO PureRAW 4 (the 800mm was ISO 1600, 1200: ISO 5000 and 1600: ISO 6400),
and sharpened the pictures somewhat in LrC.

I am also impressed by the Phalarope-picture.
Did you process this one the same way as the Ospreys?
(The Red-necked is also very rare in Sweden, but I have seen it twice. Actually one of the sightings was at the same place as I took my Broad-billed pics)

 

[AlanF]

I was impressed by your photos of the Ospreys! What kind of processing gave that result?
On my pictures I only used DxO PureRAW 4 (the 800mm was ISO 1600, 1200: ISO 5000 and 1600: ISO 6400),
and sharpened the pictures somewhat in LrC.

I am also impressed by the Phalarope-picture.
Did you process this one the same way as the Ospreys?
(The Red-necked is also very rare in Sweden, but I have seen it twice. Actually one of the sightings was at the same place as I took my Broad-billed pics)

Hi Klas, for the Ospreys, I used DxO PL6 with DeepPrime but without lens sharpening. I then sharpened in Topaz AI. Auto tone in Photoshop then got the colours right. The Phalarope was just PL6 and no more. DxO PL does doesn't seem to sharpen the 100-500mm at 1000mm or the 200-800mm at 1600mm with 2xTC properly. The 200-800mm at 800mm often benefits from Topaz sharpening after tight cropping.

 

[Klas R]

Hi Klas, for the Ospreys, I used DxO PL6 with DeepPrime but without lens sharpening. I then sharpened in Topaz AI. Auto tone in Photoshop then got the colours right. The Phalarope was just PL6 and no more. DxO PL does doesn't seem to sharpen the 100-500mm at 1000mm or the 200-800mm at 1600mm with 2xTC properly. The 200-800mm at 800mm often benefits from Topaz sharpening after tight cropping.

Thanks for taking time! Checked Topaz at their site. I imagine you talk about "Photo AI", the latest version seem to be no 3. It is quite expensive at USD 199.

Have you checked how it together with R5 and RF 200-800, with or without TC's compares to
- DxO PureRAW (I have got the latest version i.e. 4), as well as
- Adobes own functions of noise and sharpness handling implemented in the latest versions of LrC and Ps

For instance I find Adobes noise reduction tool in LrC almost as good as DxO PureRAW, and it is "free of charge" if you anyway have Adobes license

 

[AlanF]

Thanks for taking time! Checked Topaz at their site. I imagine you talk about "Photo AI", the latest version seem to be no 3. It is quite expensive at USD 199.

Have you checked how it together with R5 and RF 200-800, with or without TC's compares to
- DxO PureRAW (I have got the latest version i.e. 4), as well as
- Adobes own functions of noise and sharpness handling implemented in the latest versions of LrC and Ps

For instance I find Adobes noise reduction tool in LrC almost as good as DxO PureRAW, and it is "free of charge" if you anyway have Adobes license

I haven't tried Pure Raw. I'll have a play with Adobe when I have time.

 

[Dragon]

Thanks for taking time! Checked Topaz at their site. I imagine you talk about "Photo AI", the latest version seem to be no 3. It is quite expensive at USD 199.

Have you checked how it together with R5 and RF 200-800, with or without TC's compares to
- DxO PureRAW (I have got the latest version i.e. 4), as well as
- Adobes own functions of noise and sharpness handling implemented in the latest versions of LrC and Ps

For instance I find Adobes noise reduction tool in LrC almost as good as DxO PureRAW, and it is "free of charge" if you anyway have Adobes license

Hi Klas,
I have the full Adobe package, all the individual Topaz programs (as well as a year-old version of Photo AI), and DXO Photolab PL7. My experience with the noise reduction features is "it depends", meaning that some images at some ISOs and lighting conditions are better recovered by one program and others by a different program. Starting from RAW with a typical image, the LR "Enhance" and Photolab "Deep Prime XD" produce similar results. "Deep Prime XD" will typically show a bit more detail, but also will have more artifacts than LR "Enhance". I don't use Topaz directly from raw, but rather as an adjunct to LR. I have found that first using LR "Enhance" at 45% to 50% and following up with a light pass of either Topaz Denoise (for images that are already decently sharp) or Topaz Sharpen (for images that have some loss of focus or motion blur) produces the best results. Further, Topaz Gigapixel is generally more effective than the LR double resolution "Enhance" feature and it allows you to use the LR NR "Enhance" feature first, whereas Adobe does not. As an aside, I found the version of Topaz Photo AI that I have is less flexible that the individual programs, but Alan has indicated that Topaz has improved the flexibility of Photo AI. I have also found that DXO Photolab is particularly useful when using small sensor cameras and sharpness challenged lenses because the sharpening algorithm used is mapped to the character of the lens. For example, I almost always use Photolab with the Nixon P1000 or Olympus TG6 if I am trying to get the most out of the image. One last point. Topaz Sharpen has a very good masking feature that allows you to apply aggressive sharpening to the subject without messing up the bokeh region of the image. Hope that helps.

 

[AlanF]

Hi Klas,
I have the full Adobe package, all the individual Topaz programs (as well as a year-old version of Photo AI), and DXO Photolab PL7. My experience with the noise reduction features is "it depends", meaning that some images at some ISOs and lighting conditions are better recovered by one program and others by a different program. Starting from RAW with a typical image, the LR "Enhance" and Photolab "Deep Prime XD" produce similar results. "Deep Prime XD" will typically show a bit more detail, but also will have more artifacts than LR "Enhance". I don't use Topaz directly from raw, but rather as an adjunct to LR. I have found that first using LR "Enhance" at 45% to 50% and following up with a light pass of either Topaz Denoise (for images that are already decently sharp) or Topaz Sharpen (for images that have some loss of focus or motion blur) produces the best results. Further, Topaz Gigapixel is generally more effective than the LR double resolution "Enhance" feature and it allows you to use the LR NR "Enhance" feature first, whereas Adobe does not. As an aside, I found the version of Topaz Photo AI that I have is less flexible that the individual programs, but Alan has indicated that Topaz has improved the flexibility of Photo AI. I have also found that DXO Photolab is particularly useful when using small sensor cameras and sharpness challenged lenses because the sharpening algorithm used is mapped to the character of the lens. For example, I almost always use Photolab with the Nixon P1000 or Olympus TG6 if I am trying to get the most out of the image. One last point. Topaz Sharpen has a very good masking feature that allows you to apply aggressive sharpening to the subject without messing up the bokeh region of the image. Hope that helps.

Useful info there, and it overlaps with my experience. I too don't use Topaz on RAW. DxO PL usually does work sufficiently well for sharpening but not for the 2xTC on the RF 100-500mm or 200-800mm, where I tend to use Topaz. The RF 200-800mm at 800mm can benefit from turning off the PL lens sharpening and using Topaz. I'm not sure what Topaz are trying to achieve by having Photo AI and it's 3 elements available also separately.

 

[Klas R]

Hi Klas,
I have the full Adobe package, all the individual Topaz programs (as well as a year-old version of Photo AI), and DXO Photolab PL7. My experience with the noise reduction features is "it depends", meaning that some images at some ISOs and lighting conditions are better recovered by one program and others by a different program. Starting from RAW with a typical image, the LR "Enhance" and Photolab "Deep Prime XD" produce similar results. "Deep Prime XD" will typically show a bit more detail, but also will have more artifacts than LR "Enhance". I don't use Topaz directly from raw, but rather as an adjunct to LR. I have found that first using LR "Enhance" at 45% to 50% and following up with a light pass of either Topaz Denoise (for images that are already decently sharp) or Topaz Sharpen (for images that have some loss of focus or motion blur) produces the best results. Further, Topaz Gigapixel is generally more effective than the LR double resolution "Enhance" feature and it allows you to use the LR NR "Enhance" feature first, whereas Adobe does not. As an aside, I found the version of Topaz Photo AI that I have is less flexible that the individual programs, but Alan has indicated that Topaz has improved the flexibility of Photo AI. I have also found that DXO Photolab is particularly useful when using small sensor cameras and sharpness challenged lenses because the sharpening algorithm used is mapped to the character of the lens. For example, I almost always use Photolab with the Nixon P1000 or Olympus TG6 if I am trying to get the most out of the image. One last point. Topaz Sharpen has a very good masking feature that allows you to apply aggressive sharpening to the subject without messing up the bokeh region of the image. Hope that helps.

Hi Dragon & Alan,
Very interesting reading, from both of you! I will have to sit down and read your comments once more in detail. Who could think that denoising and sharpening is such a science on its own...? Thanks!

 

[Dragon]

Useful info there, and it overlaps with my experience. I too don't use Topaz on RAW. DxO PL usually does work sufficiently well for sharpening but not for the 2xTC on the RF 100-500mm or 200-800mm, where I tend to use Topaz. The RF 200-800mm at 800mm can benefit from turning off the PL lens sharpening and using Topaz. I'm not sure what Topaz are trying to achieve by have Photo AI and it's 3 elements available also separately.

One last piece of advice to newbies would be don't try any of this stuff if you are not willing to choke up for enough computer to run it. I had an i9-9900 with GTX 1080 which worked fine until the "AI" stuff started to drop and then it was suddenly glacial. I built a new computer with an i9-13900 and RTX 4070ti and it keeps the compute time to a minimal level. Also, you need substantial storage space and a good internet connection (particularly for Topaz). These programs all get updated regularly and the updates are several gigabytes every time. I think PL is the least bandwidth consumptive and and Topaz the most. Adobe is somewhere in the middle (unless you have the full Adobe package and then it is about the same as Topaz). I recently got a 2Gb bi-directional fiber connection (courtesy of our local power company) and the downloads are now trivial, but on the previous connection, they were tedious.

 

[AlanF]

One last piece of advice to newbies would be don't try any of this stuff if you are not willing to choke up for enough computer to run it. I had an i9-9900 with GTX 1080 which worked fine until the "AI" stuff started to drop and then it was suddenly glacial. I built a new computer with an i9-13900 and RTX 4070ti and it keeps the compute time to a minimal level. Also, you need substantial storage space and a good internet connection (particularly for Topaz). These programs all get updated regularly and the updates are several gigabytes every time. I think PL is the least bandwidth consumptive and and Topaz the most. Adobe is somewhere in the middle (unless you have the full Adobe package and then it is about the same as Topaz). I recently got a 2Gb bi-directional fiber connection (courtesy of our local power company) and the downloads are now trivial, but on the previous connection, they were tedious.

My MacBook Air (M2) has no problems.

 

[Dragon]

My MacBook Air (M2) has no problems.

That is a very contemporary computer so no surprise, but I can't deal with photo editing on a laptop. I use a 32 in. 4k screen and if there was one available with decent colorimetry, I would go for a 42 in. screen.

 

[OskarB]

My MacBook Air (M2) has no problems.

Alan, how much RAM do you use?

 

[AlanF]

Alan, how much RAM do you use?

8gb - minimum.

 

[OskarB]

8gb - minimum.

Thanks!

 

[Klas R]

I have a question regarding an other topic involving the R5 + RF 200-800mm. In this case not using a TC.
Recently during an excursion I/we had an encounter with a Red-footed Falcon. Initially we were not 100% sure about the identification, so we wanted proof.

The falcon was flying against the sky at quite some distance and flying in the direction from us. But I had no problem knowing where it was using my binoculars, so I knew I was pointing the camera in the correct direction and that the bird was there somewhere on the R5 screen.

The R5 had severe problems to get the bird in focus. I use back-button focusing and have copied the set-up that "Whistling Wings Photography" is using. I.e. Eye-AF using the whole field with one button, Eye-AF using Zone with the second button, and with the third button Spot-AF (no Eye-AF).
Trying all three buttons one at a time I still had these severe problem for the R5 to focus. I tried to focus at 200mm, 500mm and 800mm, same problem. Sometimes it would find the bird and for a moment the bird was possible to glimpse, but in a wink of time it was lost again. The end result: no sharp pictures of the falcon, some blurred pictures, and some where no bird was possible to see at all.

Have you or anybody else out there encountered a similar problem, and in such case do you have a solution?

 

[EricN]

The other day I took some photos to try to secure my possible identification of a Broad-billed Sandpiper. So the pictures were not "nice portait pictures". Actually the distance to the bird was something like 350 meters, and it measures only 15-18 cm , i.e. 5-6".

They were taken at about 4 PM in the afternoon and I had some severe problems with heat waves. First I took som pics at 800mm at f9. But as the bird was so small on the sensor I added my 2XTC. Took some pictures at 1600mm, f18 but wasn't impressed. Maybe it was the hest waves or maybe it is about the lens performing much better at 600mm, so I finally took some pictures at 1200mm, f16.

To me it looks like the lens performs very good at 1200mm with the 2XTC, but not as well at 1600 mm. But actually you get almost the same information at 800mm only.

View attachment 217042View attachment 217043View attachment 217044
Photos from top to bottom: 1200mm, 1600 mm and 800mm

Some people would harshly criticize these photos for the heat waves, but actually, I love them!