rf 16mm 2.8 - Raw Images (Normal Distance + Closeup)

Hi,

I just shot some images of a wall with a lot of detail to see the sharpness level of each aperture. I also did this closeup with minimum focus distance (13cm).

The pictures are shot with my EOS R.

https://1drv.ms/u/s!Ap8Y3G8kon1Zgo52hcV6f4EDNVItyQ?e=Npt5wu

I dont't know if its normal that you have this black circle vignette at closeup distance?

My opinion so far:

- heavy barrel distortion

- sharp corners around f5.6 at normal distances

- good closeup sharpness in the center area

Seems like people dont understand that this lens, RF 24-240 and RF 14-35 do require lens correction due to design that projects larger FOV than adverted

you need to use DPP4 to get rid of black corners

I do not understand why people are so concerned about what raw images look like. They are an intermediate step and the complained-about effects they show are expected from them. When made into jpegs, they are cropped to produce the required angle, i.e. 16mm in this case. This also helps remove the vignetting and correct barrel distortion. Other operations are also carried out automatically in the process of producing a final result that can be saved as a jpeg or tiff file. (See other threads.)

When one buys something that has been processed, does one care about what intermediate steps the product went through before one uses it?

In designing rf lenses, Canon accepts that distortions are corrected in preparing the jpeg image. If you dont like this philosophy, why not stay with older lenses and systems?

tkbslc wrote:

SergioMPS wrote:

Seems like people dont understand that this lens, RF 24-240 and RF 14-35 do require lens correction due to design that projects larger FOV than adverted

you need to use DPP4 to get rid of black corners

I think Canon could do a better job of marketing this fact. They sort of pretend like everyone is going to shoot corrected JPEGs.

Everyone who shoots the 16, the 14-35, the 24-240, or the cheaper 24-105 is, or they're going to correct the shot after shooting RAW.

Anyone who doesn't know that, hasn't read much about those lenses.

tkbslc wrote:

Eddie Rizk wrote:

tkbslc wrote:

SergioMPS wrote:

Seems like people dont understand that this lens, RF 24-240 and RF 14-35 do require lens correction due to design that projects larger FOV than adverted

you need to use DPP4 to get rid of black corners

I think Canon could do a better job of marketing this fact. They sort of pretend like everyone is going to shoot corrected JPEGs.

Everyone who shoots the 16, the 14-35, the 24-240, or the cheaper 24-105 is, or they're going to correct the shot after shooting RAW.

Anyone who doesn't know that, hasn't read much about those lenses.

So what if they haven’t? Some people shop by seeing the lens on the shop or just by casually browsing marketing materials. Should be an asterisk on the box about RAW with these lenses.

What?

It is pretty much impossible to view a raw image without having specific software (e.g. DPP, Photolab4, etc) to do so, and the default mode of that software is to make the necessary corrections automatically. In any case, a casual purchaser, such as you describe, would most likely only save jpeg files in the camera, and thus would never encounter any problem.

I have seen nothing about anyone having this hypothetical problem in connection with the rf24-240 or other rf lenses that rely on correction in software.

I just ordered this lens. It will be a while before I get it. I'm hoping that LR will have a profile for it by then.

Alastair Norcross wrote:

I just ordered this lens. It will be a while before I get it. I'm hoping that LR will have a profile for it by then.

You can always use first DPP4 and than import .tiff into LR

tkbslc wrote:

davidwien wrote:

tkbslc wrote:

Eddie Rizk wrote:

tkbslc wrote:

SergioMPS wrote:

Seems like people dont understand that this lens, RF 24-240 and RF 14-35 do require lens correction due to design that projects larger FOV than adverted

you need to use DPP4 to get rid of black corners

I think Canon could do a better job of marketing this fact. They sort of pretend like everyone is going to shoot corrected JPEGs.

Everyone who shoots the 16, the 14-35, the 24-240, or the cheaper 24-105 is, or they're going to correct the shot after shooting RAW.

Anyone who doesn't know that, hasn't read much about those lenses.

So what if they haven’t? Some people shop by seeing the lens on the shop or just by casually browsing marketing materials. Should be an asterisk on the box about RAW with these lenses.

What?

It is pretty much impossible to view a raw image without having specific software (e.g. DPP, Photolab4, etc) to do so, and the default mode of that software is to make the necessary corrections automatically. In any case, a casual purchaser, such as you describe, would most likely only save jpeg files in the camera, and thus would never encounter any problem.

I have seen nothing about anyone having this hypothetical problem in connection with the rf24-240 or other rf lenses that rely on correction in software.

No software has profiles for the 16mm except DPP. Same with 14-35. If you use industry standard software and shoot RAW, you may be unpleasantly surprised.

Then the software should probably have an asterisk / show a pop up warning people they haven’t caught up with the lens being used. Being so called industry standard doesn’t make something infallible

IMO, people who shoot RAW without understanding the implications of PP requirements should be OK to learn it with surprise. LR also has crappy default profiles for new canon cameras, so unless they have done something about it, they are getting poor colours from their industry standard software anyways

PicPocket wrote:

tkbslc wrote:

davidwien wrote:

tkbslc wrote:

Eddie Rizk wrote:

tkbslc wrote:

SergioMPS wrote:

Seems like people dont understand that this lens, RF 24-240 and RF 14-35 do require lens correction due to design that projects larger FOV than adverted

you need to use DPP4 to get rid of black corners

I think Canon could do a better job of marketing this fact. They sort of pretend like everyone is going to shoot corrected JPEGs.

Everyone who shoots the 16, the 14-35, the 24-240, or the cheaper 24-105 is, or they're going to correct the shot after shooting RAW.

Anyone who doesn't know that, hasn't read much about those lenses.

So what if they haven’t? Some people shop by seeing the lens on the shop or just by casually browsing marketing materials. Should be an asterisk on the box about RAW with these lenses.

What?

It is pretty much impossible to view a raw image without having specific software (e.g. DPP, Photolab4, etc) to do so, and the default mode of that software is to make the necessary corrections automatically. In any case, a casual purchaser, such as you describe, would most likely only save jpeg files in the camera, and thus would never encounter any problem.

I have seen nothing about anyone having this hypothetical problem in connection with the rf24-240 or other rf lenses that rely on correction in software.

No software has profiles for the 16mm except DPP. Same with 14-35. If you use industry standard software and shoot RAW, you may be unpleasantly surprised.

Then the software should probably have an asterisk / show a pop up warning people they haven’t caught up with the lens being used. Being so called industry standard doesn’t make something infallible

IMO, people who shoot RAW without understanding the implications of PP requirements should be OK to learn it with surprise. LR also has crappy default profiles for new canon cameras, so unless they have done something about it, they are getting poor colours from their industry standard software anyways

Come on, guys!  If you don't know that many lenses need software correction, you're probably not shooting RAW anyway, so you don't know that it's being corrected, and it looks fine.

crusliq wrote:

Hi,

I just shot some images of a wall with a lot of detail to see the sharpness level of each aperture. I also did this closeup with minimum focus distance (13cm).

The pictures are shot with my EOS R.

https://1drv.ms/u/s!Ap8Y3G8kon1Zgo52hcV6f4EDNVItyQ?e=Npt5wu

I dont't know if its normal that you have this black circle vignette at closeup distance?

My opinion so far:

- heavy barrel distortion

- sharp corners around f5.6 at normal distances

- good closeup sharpness in the center area

Thanks for giving links to the RAW files!!!

My impressions:

• The RF16mm barrel distortion looks nearly identical to the amount of barrel distortion the RF24-240 has at 24mm. I compared the curvature of the top of the wall versus mine of a test chart (after scaling my R5 shot), both using Adobe RAW with correction disabled.  I also looked at the RF15-35, and it is vastly better (as one should hope with an L lens - some barrel but much, much less). 
• The "normal" photo looks about like I would expect with that much distortion. You will lose about 1/2 the resolution for the pictures in the corners due to the amount of correction required. 
• I notice there is very little improvement in sharpness in the corners with the "normal" distance photo between F2.8 and f8. I notice this with the RF24-240 as well. With that much image correction in the corners, the distortion and not focus is the main limiting factor. 
• I did notice a big difference in illumination uniformity between f2.8 and f8. And yes, I double-checked, and DPP4 had peripheral illumination correction on, but in both cases, I used the "default" amount, and obviously, the f2.8 needs more.  
• Wow, the closeup picture looks horrible from about 1/2 way out from the center. You don't even have to look at the picture magnified to tell it is a mess.  It might be "field curvature," as in that focus varying from the center to the outside. It would be interesting to see what happens if your focus point was in one of the corners.  Seeing this, I ran a quick check on my RF24-240, and when focused at the minimum focus distance and 24mm, and sure enough, it too looks horrible compared to "normal" distance shots. If the RF24-240 at 24mm focuses in the corner, the corner looks much better, but the center gets blurry.  Probably best not to use either lens for closeups. 

Karl_Guttag wrote:.

• Wow, the closeup picture looks horrible from about 1/2 way out from the center. You don't even have to look at the picture magnified to tell it is a mess. It might be "field curvature," as in that focus varying from the center to the outside. It would be interesting to see what happens if your focus point was in one of the corners. Seeing this, I ran a quick check on my RF24-240, and when focused at the minimum focus distance and 24mm, and sure enough, it too looks horrible compared to "normal" distance shots. If the RF24-240 at 24mm focuses in the corner, the corner looks much better, but the center gets blurry. Probably best not to use either lens for closeups.

Surely looks like field curvature to me, based on how it’s behaving with successive smaller f stops. Not surprising for a compact wide angle at close focus distance

tkbslc wrote:

im not saying people don’t know about RAW and lens profiles I’m saying many are getting a nasty surprise that the profiles are not out yet and may not be for a long time. I feel like if your lens requires a lens profile to be normally usable, that needs to at least be in the fine print in the specs or on the box. Also, no Canon lenses needed anywhere near this kind of correction before mirrorless. So it wasn’t as big of a deal if you had no profile. People who have been at this a long time may not even think to check.

but keep sticking up for Canon and Adobe. I’m sure they need your support!

No, I guess they just need the nagging. I’m sure you are taking it up with Adobe so they quickly provide the required updates on their industry standard software so their users don’t have surprises. And also provide camera matching profiles for canon cameras made in last few years which is only a big deal if you want good colours out of the box without any surprises. But somehow, it seems you want Canon to put an asterisk for Adobe? Keep sticking it to them, I’m sure that will fix the software

All the while, users are shooting RAW being oblivious to what it takes and are surprised why the colours are bad and their brand new lenses have distortion and vignetting. Because they just saw a new lens in a store and bought it. The same place where they bought their camera, which came set to shoot JPG but they somehow figured out they should change it to RAW instead

Karl_Guttag wrote:

crusliq wrote:

Hi,

I just shot some images of a wall with a lot of detail to see the sharpness level of each aperture. I also did this closeup with minimum focus distance (13cm).

The pictures are shot with my EOS R.

https://1drv.ms/u/s!Ap8Y3G8kon1Zgo52hcV6f4EDNVItyQ?e=Npt5wu

I dont't know if its normal that you have this black circle vignette at closeup distance?

My opinion so far:

- heavy barrel distortion

- sharp corners around f5.6 at normal distances

- good closeup sharpness in the center area

Thanks for giving links to the RAW files!!!

My impressions:

• The RF16mm barrel distortion looks nearly identical to the amount of barrel distortion the RF24-240 has at 24mm. I compared the curvature of the top of the wall versus mine of a test chart (after scaling my R5 shot), both using Adobe RAW with correction disabled. I also looked at the RF15-35, and it is vastly better (as one should hope with an L lens - some barrel but much, much less).
• The "normal" photo looks about like I would expect with that much distortion. You will lose about 1/2 the resolution for the pictures in the corners due to the amount of correction required.
• I notice there is very little improvement in sharpness in the corners with the "normal" distance photo between F2.8 and f8. I notice this with the RF24-240 as well. With that much image correction in the corners, the distortion and not focus is the main limiting factor.
• I did notice a big difference in illumination uniformity between f2.8 and f8. And yes, I double-checked, and DPP4 had peripheral illumination correction on, but in both cases, I used the "default" amount, and obviously, the f2.8 needs more.
• Wow, the closeup picture looks horrible from about 1/2 way out from the center. You don't even have to look at the picture magnified to tell it is a mess. It might be "field curvature," as in that focus varying from the center to the outside. It would be interesting to see what happens if your focus point was in one of the corners. Seeing this, I ran a quick check on my RF24-240, and when focused at the minimum focus distance and 24mm, and sure enough, it too looks horrible compared to "normal" distance shots. If the RF24-240 at 24mm focuses in the corner, the corner looks much better, but the center gets blurry. Probably best not to use either lens for closeups.

In Lightroom, use the RF 24-240 lens profile for the RF 16mm. It works quite well.

tkbslc wrote:

No software has profiles for the 16mm except DPP. Same with 14-35. If you use industry standard software and shoot RAW, you may be unpleasantly surprised.

But from my perspective lack of profiles is not a Canon problem but software providers issue, canon give the hardware and developed by own software, that works symbiotically with their lenses. this is our choice: we will use it or waiting for a other companies.

tkbslc wrote:

Eddie Rizk wrote:

PicPocket wrote:

tkbslc wrote:

davidwien wrote:

tkbslc wrote:

Eddie Rizk wrote:

tkbslc wrote:

SergioMPS wrote:

Seems like people dont understand that this lens, RF 24-240 and RF 14-35 do require lens correction due to design that projects larger FOV than adverted

you need to use DPP4 to get rid of black corners

I think Canon could do a better job of marketing this fact. They sort of pretend like everyone is going to shoot corrected JPEGs.

Everyone who shoots the 16, the 14-35, the 24-240, or the cheaper 24-105 is, or they're going to correct the shot after shooting RAW.

Anyone who doesn't know that, hasn't read much about those lenses.

So what if they haven’t? Some people shop by seeing the lens on the shop or just by casually browsing marketing materials. Should be an asterisk on the box about RAW with these lenses.

What?

It is pretty much impossible to view a raw image without having specific software (e.g. DPP, Photolab4, etc) to do so, and the default mode of that software is to make the necessary corrections automatically. In any case, a casual purchaser, such as you describe, would most likely only save jpeg files in the camera, and thus would never encounter any problem.

I have seen nothing about anyone having this hypothetical problem in connection with the rf24-240 or other rf lenses that rely on correction in software.

No software has profiles for the 16mm except DPP. Same with 14-35. If you use industry standard software and shoot RAW, you may be unpleasantly surprised.

Then the software should probably have an asterisk / show a pop up warning people they haven’t caught up with the lens being used. Being so called industry standard doesn’t make something infallible

IMO, people who shoot RAW without understanding the implications of PP requirements should be OK to learn it with surprise. LR also has crappy default profiles for new canon cameras, so unless they have done something about it, they are getting poor colours from their industry standard software anyways

Come on, guys! If you don't know that many lenses need software correction, you're probably not shooting RAW anyway, so you don't know that it's being corrected, and it looks fine.

im not saying people don’t know about RAW and lens profiles I’m saying many are getting a nasty surprise that the profiles are not out yet and may not be for a long time. I feel like if your lens requires a lens profile to be normally usable, that needs to at least be in the fine print in the specs or on the box. Also, no Canon lenses needed anywhere near this kind of correction before mirrorless. So it wasn’t as big of a deal if you had no profile. People who have been at this a long time may not even think to check.

but keep sticking up for Canon and Adobe. I’m sure they need your support!

I'm not sticking up for anything but common sense.

Canon, of course, has the correction profiles on release.

Adobe is not a subsidiary of Canon.  Certainly, it would be nice, if they worked together before the release date, but they don't, and they never have.  Adobe never has the correction profiles for any lens the day it comes out.  Adobe generally figures it out pretty quickly.

No doubt, someone out there has the numbers to do the corrections manually, for those who can't bring themselves to use DPP.  Anyone who understands processing can probably figure it out for themselves.

It's a temporary inconvenience.

I am almost schocked that some expensive L lenses require software correction in postprocessing to make decent RAW images. Only for JPEGs it is corrected in-camera, so why not for RAW ?
I thought Canon does nice work in optical design which allows doing most correction optically which all older lenses do.
Software is just an extra tool to correct minor / leftover optical errors, but now it appears that modern RF lenses depend heavily on software, and, worse, that software (lens profiles) is not available yet completely.