Canon RF 24-240mm f/4-6.3 IS USM sample gallery

[StoicalEtcher]

There is a lot of mechanical vignetting in the 24mm-32mm range, but only in the RAW images and not the jpegs.

I'm not an expert on optical design and nomenclature, but isn't vignetting "light fall-off" towards the corners of a shot? - What your shot shows looks more like something actually blocking the shot (like the edge of the lens!), in the same way that a poorly fitted hood or deep screw-on filter might.

Either way doesn't really matter ultimately - you get black corners, and the jpeg/correction to 24mm is ok - but I don't think "vignetting" is right - it looks like "obstruction".

 

[Bennymiata]

It does seem to be mechanical blocking rather than light fall-off in the corners.
The pic I showed was at F6.3. It will be worse wide open.
I'll try to take somemore shots at 24mm and wide open.

The lens is quite good apart from this problem, and worth the money.

 

[flip314]

It does seem to be mechanical blocking rather than light fall-off in the corners.
The pic I showed was at F6.3. It will be worse wide open.
I'll try to take somemore shots at 24mm and wide open.

The lens is quite good apart from this problem, and worth the money.

There are always compromises in super-zooms, but this one seems potentially excessive... Several people have recommended shooting no wider than 35mm, but the 24-35mm range is quite useful. No doubt it's probably still good enough in a pinch.

Are the full corrections applied in the EVF? Does the view look like the JPG or the RAW?

 

[neuroanatomist]

I'm not an expert on optical design and nomenclature, but isn't vignetting "light fall-off" towards the corners of a shot? - What your shot shows looks more like something actually blocking the shot (like the edge of the lens!), in the same way that a poorly fitted hood or deep screw-on filter might.

Optical vignetting is light fall off due to the lens design. Mechanical vignetting is when something physically blocks the periphery of the image, such as too thick a stack of filters. The latter is what appears to be happening with the 24-240, although it does seem that is by Canon’s design.

 

[SteveC]

Optical vignetting is light fall off due to the lens design. Mechanical vignetting is when something physically blocks the periphery of the image, such as too thick a stack of filters. The latter is what appears to be happening with the 24-240, although it does seem that is by Canon’s design.

How does Canon have any control over how many filters people put on?

Or is it (potentially) something that's part of the lens assembly (the actual casing) or the lens hood?

 

[neuroanatomist]

How does Canon have any control over how many filters people put on?

Or is it (potentially) something that's part of the lens assembly (the actual casing) or the lens hood?

“The latter” was in reference to mechanical vignetting versus optical vignetting (a stack of filters is merely an example of one cause of mechanical vignetting). .Something within the lens itself is physically blocking the light from filling the aperture. Almost certainly a necessary compromise of the super zoom design and price point of the lens.

 

[SteveC]

“The latter” was in reference to mechanical vignetting versus optical vignetting (a stack of filters is merely an example of one cause of mechanical vignetting). .Something within the lens itself is physically blocking the light from filling the aperture. Almost certainly a necessary compromise of the super zoom design and price point of the lens.

So basically, anything in the lens design that isn't the actual glass parts that could partially obstruct the light path could be mechanical vignetting...but if the optics design itself is causing the image to be too small in diameter (i.e., it won't cover the sensor) or makes it dark near the corners, that's optical vignetting.

I guess, either way--it was a decision by Canon, and they figured it could be counter-acted in software, hence JPEGs that look just fine. But they have had massive corrections done on them, and such will likely introduce loss of (for lack of my knowing the right term) fidelity of the image. I am thinking (at this point) Canon pushed the range of the zoom too far (10x) for this to be a truly <I>optically</I> high quality lens--the hardware can't by itself do what is being asked of it. (I tread on thin ice here and may be using the wrong words to convey my meaning, or both.)

 

[neuroanatomist]

So basically, anything in the lens design that isn't the actual glass parts that could partially obstruct the light path could be mechanical vignetting...but if the optics design itself is causing the image to be too small in diameter (i.e., it won't cover the sensor) or makes it dark near the corners, that's optical vignetting.

I guess, either way--it was a decision by Canon, and they figured it could be counter-acted in software, hence JPEGs that look just fine. But they have had massive corrections done on them, and such will likely introduce loss of (for lack of my knowing the right term) fidelity of the image. I am thinking (at this point) Canon pushed the range of the zoom too far (10x) for this to be a truly <I>optically</I> high quality lens--the hardware can't by itself do what is being asked of it. (I tread on thin ice here and may be using the wrong words to convey my meaning, or both.)

Agreed. I had the 28-300L — similar IQ to 24-105L lenses, but a big, heavy lens. Getting a 10x range in a package not much bigger than a 24-105 means compromising IQ.

 

[SteveC]

Agreed. I had the 28-300L — similar IQ to 24-105L lenses, but a big, heavy lens. Getting a 10x range in a package not much bigger than a 24-105 means compromising IQ.

I have the Tamron 18-200 (for M mount). It undoubtedly distorts, too, but not as badly. This is likely because of the fact that it's APS-C, though. Imagine applying a 1.6 crop to the RAW file above, the result wouldn't look nearly as bad because the worst distortion is towards the edges. (Incidentally, it's fatter than the Canon standard width for its M lenses...I think if Canon ever decides to widen the scope of EF-M lenses to bring in things that EF-S APS-C users are used to having access to, it's going to have to allow them to be fatter.)

 

[Act444]

There are always compromises in super-zooms, but this one seems potentially excessive...

It is...

Don't think a design like this would have worked on the EF system - the OVF view would be obscured as well.

Not sure what the advantage of doing it like this is TBH.
__________

I had the 28-300L for a little while and used it with a 6D (20MP). It was super convenient, but it was also big, heavy, attention-grabbing and the IQ was just....average. Not bad - especially for its class, but the difference was very visible when I went back to the 70-200. Ultimately I decided carrying all that weight just to get average IQ wasn't worth it, even with the convenience factor - and sold it. Still glad I got to experiment with it though, as it was a lens I dreamed of using for a while prior. That being said, I'd likely feel differently if I shot professionally, because I can't overstate how great it was to go from wide to super-tele in less than 2 seconds...made "getting the shot" that much easier.

 

[BbBbBbBbBb]

There are always compromises in super-zooms, but this one seems potentially excessive... Several people have recommended shooting no wider than 35mm, but the 24-35mm range is quite useful. No doubt it's probably still good enough in a pinch.

Are the full corrections applied in the EVF? Does the view look like the JPG or the RAW?
[/

I'm not an expert on optical design and nomenclature, but isn't vignetting "light fall-off" towards the corners of a shot? - What your shot shows looks more like something actually blocking the shot (like the edge of the lens!), in the same way that a poorly fitted hood or deep screw-on filter might.

Either way doesn't really matter ultimately - you get black corners, and the jpeg/correction to 24mm is ok - but I don't think "vignetting" is right - it looks like "obstruction".

I may post pictures later if I get a chance, but the lens shoots at 22mm with an image similar to the back side of a spoon meant to be corrected to 24mm. It’s hard to see when there are no people in the photo, but it really is a fisheye lens from 24-30 without corrections. With the corrections the photos are a bit flat and seem a bit 2 dimensional. Reminds me of the 18mm end of the 10-18mm. Lightroom and Photoshop have profile fixes that auto correct the image through crop and distortion correction that gets it to 24 and you can apply your own corrections using a slider.

The lens corrections are done very well, but the image is not the same as it would be if you were to take a few steps back and shoot at 35 instead. It’s not simply a photo with the corners blocked. It’s as if the photo was on a dome pushing out in the center. That’s undesirable if you are taking pictures of people.

In reality this lens is three lenses. Fisheye with corrections makes it an acceptable 24-30, excellent zoom 31-200, good telephoto zoom 201-240. Again, the 24-30 end can be corrected and it’s nice to have when you need it, but ideally you stay in the 31-200 range. If they had made this a 31-200 range lens it could be an L series. That said, I have gotten spoiled by the range and I wouldn’t trade this for a 24-105 any time soon.

 

[Kit.]

The lens corrections are done very well, but the image is not the same as it would be if you were to take a few steps back and shoot at 35 instead.

Isn't that true for every 24mm lens?

 

[SteveC]

I may post pictures later if I get a chance, but the lens shoots at 22mm with an image similar to the back side of a spoon meant to be corrected to 24mm. It’s hard to see when there are no people in the photo, but it really is a fisheye lens from 24-30 without corrections. With the corrections the photos are a bit flat and seem a bit 2 dimensional. Reminds me of the 18mm end of the 10-18mm. Lightroom and Photoshop have profile fixes that auto correct the image through crop and distortion correction that gets it to 24 and you can apply your own corrections using a slider.

The lens corrections are done very well, but the image is not the same as it would be if you were to take a few steps back and shoot at 35 instead. It’s not simply a photo with the corners blocked. It’s as if the photo was on a dome pushing out in the center. That’s undesirable if you are taking pictures of people.

In reality this lens is three lenses. Fisheye with corrections makes it an acceptable 24-30, excellent zoom 31-200, good telephoto zoom 201-240. Again, the 24-30 end can be corrected and it’s nice to have when you need it, but ideally you stay in the 31-200 range. If they had made this a 31-200 range lens it could be an L series. That said, I have gotten spoiled by the range and I wouldn’t trade this for a 24-105 any time soon.

Seems to me the best two lens solution is that 24-105, and the 100-400 L II on an adapter.

However if one can live with the correction "flatness" this is probably the best zoom range you're going to get out of one lens.

 

[BbBbBbBbBb]

Still working on a 24mm pic, but here are some other pics showing off the quality at different ranges between 32mm - 222mm. This is all from roughly the same vantage point. The fish photo is 240mm ISO 12800 F/6.3 1/125. The Turtle is ISO 12800 62mm F/5.6 1/125. If you want more details on the other photos let me know. These are unedited versions originally shot in RAW CR3 format with in unit noise correction turned off.

 

[BbBbBbBbBb]

As I previously stated, I am posting what was originally a RAW 24mm pic. It's not the best, but it will have to do. It is a pic completed inside the La Brea Tarpits museum at ISO 3200 24mm F/4.5 1/100 sec without in frame noise correction or any post production adjustments. With the IS I could have adjusted the shutter speed down to grab more light and lower the ISO or shot at F/8 or F11, but it was a quick shot in-between shooting a running moving object. (my kid) The first shot is without the lens correction. The second is with the default profile corrections in Lightroom:


As you can see in the first photo the corners are black and the center is curved due to distortion.


The second photo is a pretty massive difference. If you look closely the scene is slightly cropped on all sides. This is moving the shot from what I estimate is 22mm to 24mm. The bubble in the center of the shot is corrected and as you should notice the walls are no longer curved. With the profile corrections it's not terrible. Additional adjustments can be made for further correction if you don't like the level of distortion or vignetting correction. Overall I would say it does an acceptable job at 24-30 with the profile corrections. I had another lens with me, but it never left my bag because the size of the displays inside this particular museum were massive and I found myself constantly bouncing between closeup and wide shots in a period of seconds. I really am enjoying this lens.