Opinion: This patent identifies my ongoing issue with Canon

[Dragon]

Where does the f/9 come from? My understanding is that it would have the depth of field and light gathering ability of a 400 f/4.5 but the reach of an 800. (Not counting the built-in 1.25x TC of course.) In other words, f/4.5 is f/4.5 regardless of what the sensor size is. But a 400mm lens has greater depth of field than an 800.

But maybe I’m not understanding correctly.

You already have two answers, but I will give you another perspective. DOF is a function of both f stop and magnification. For equivalency purposes we will assume that both the micro 4/3 camera and the FF camera have the same pixel count. This means the micro 4/3 camera will have 1/2 the pixel pitch of the FF camera and thus is twice as demanding (in a linear sense) as the FF sensor with regards to focus, so for any image projected on the sensor, the focus accuracy will have to be twice as good to appear to have the same sharpness. In the limit, there is only one distance that is actually in focus, so the small sensor is twice as sensitive to change in distance away from that point as the FF sensor. Thus the f/4.5 lens results in the same effective DOF on the M4/3 sensor as the f/9 lens on the FF sensor. Hope that helps.

 

[Dragon]

there's something to be said about this (in before canon's marketshare quotes).

but - Sony is ALWAYS in the camera sites, lenses being reviewed for it, and so on. When I'm looking for instance for an article to write up on a new lens review, all my favorite review sites? they haven't done a canon lens in months - but they have done 10+ for sony E mount. It's this sigma lens reviewed on a sony or tamron, etc,etc.

sooner or later, that incessant display isn't going to help Canon.

And almost always when there is an obvious reporting bias spread over a large group, there is some money changing hands. Just sayin'. Quite possible your favorite review sites are wh**es.

 

[neuroanatomist]

If you look at physical dimensions, both OM and Canon lenses have 95mm front filter, indicating similar light gathering capabilities, so why is OM f4.5 and Canon f9? It is because the OM reduces all that light into a m4/3 image circle size of 22.5mm diameter while the Canon spread the light over 43.3mm (approximately 2x crop factor on OM).

No. Just…no.

First off, there was no actual Canon lens being discussed, rather the hypothetical FF equivalent of the OM 150-400/4.5. Second, with telephoto designs the image circle isn’t limiting. The OM lens projects a larger circle than the sensor needs.

The reason the OM 150-400/4.5 and the real Canon 200-800/6.3-9 both take 95mm filters is because 400mm / 4.5 = 800mm / 9 = 89mm. The focal length and f/number are intrinsic properties of a lens, independent of the sensor behind it.

 

[David_B]

For me this is the patent why I won't give up EF-M:

Canon Patent Application: Canon EF-M 11-22mm F4-5.6

In this patent application, it looks like Canon is revisiting the excellent EOS-M lens for the Canon EF-M mount; The EOS-M 11-22mm F4-5.6 which is one of the mainstays of the EF-M ecosystem. Where there are similarities to the EF-M 11-22, the element design is certainly different. Could this...

www.canonnews.com

The EF-M 11-22 isn't even a high end lens but it is better than every Rf-M lens equivalent.

Canon are hell bent on creating new RF-S/RF lenses that don't create an image circle to cover the sensor because they can use firmware to stretch the image.
In other words, Canon are requiring APS-C shooters on RF-S to accept image compromises.
I can't wait until Sigma/Tamron/others get into the game on RF, even if it is just RF-S, to save us from Canon's small mindedness.

 

[David_B]

There should be a couple of APS-C L level zooms and a prime. Maybe with a new colour for the ring.

They don't even need to be L-level zooms, just not crippled by design. It is really distresing to see where Canon has gone with RF-S.

 

[EricN]

You already have two answers, but I will give you another perspective. DOF is a function of both f stop and magnification. For equivalency purposes we will assume that both the micro 4/3 camera and the FF camera have the same pixel count. This means the micro 4/3 camera will have 1/2 the pixel pitch of the FF camera and thus is twice as demanding (in a linear sense) as the FF sensor with regards to focus, so for any image projected on the sensor, the focus accuracy will have to be twice as good to appear to have the same sharpness. In the limit, there is only one distance that is actually in focus, so the small sensor is twice as sensitive to change in distance away from that point as the FF sensor. Thus the f/4.5 lens results in the same effective DOF on the M4/3 sensor as the f/9 lens on the FF sensor. Hope that helps.

I like the idea of explaining with the same pixel count.
I'll try to add something that can demonstrate it with a zoom lenses: That is take a photo at the same aperture but twice the focal length. then crop the wider angle to the appropriate size to match M4/3. Next, resize the uncropped longer focal length to match the M4/3 size. Finally, compare the images.

 

[Dragon]

For me this is the patent why I won't give up EF-M:

Canon Patent Application: Canon EF-M 11-22mm F4-5.6

In this patent application, it looks like Canon is revisiting the excellent EOS-M lens for the Canon EF-M mount; The EOS-M 11-22mm F4-5.6 which is one of the mainstays of the EF-M ecosystem. Where there are similarities to the EF-M 11-22, the element design is certainly different. Could this...

www.canonnews.com

The EF-M 11-22 isn't even a high end lens but it is better than every Rf-M lens equivalent.

Canon are hell bent on creating new RF-S/RF lenses that don't create an image circle to cover the sensor because they can use firmware to stretch the image.
In other words, Canon are requiring APS-C shooters on RF-S to accept image compromises.
I can't wait until Sigma/Tamron/others get into the game on RF, even if it is just RF-S, to save us from Canon's small mindedness.

How does that patent affect EF-M? It was never built and the application was over 4 years ago so it never will be built, at least for EF-M. The RFS 10-18 is not as nice a lens as the 11-22, but it is sharp and much smaller and lighter, which seems to be the real drive behind many of Canon's decisions. The EF-M 11-22 is one of the chunkiest of the EF-M lenses, but it is quite good even on an M6 II.

 

[EricN]

No. Just…no.

First off, there was no actual Canon lens being discussed, rather the hypothetical FF equivalent of the OM 150-400/4.5. Second, with telephoto designs the image circle isn’t limiting. The OM lens projects a larger circle than the sensor needs.

The reason the OM 150-400/4.5 and the real Canon 200-800/6.3-9 both take 95mm filters is because 400mm / 4.5 = 800mm / 9 = 89mm. The focal length and f/number are intrinsic properties of a lens, independent of the sensor behind it.

I think what confuses some people is that the relationship between focal length and sensor size is much more obvious.

 

[EricN]

For me this is the patent why I won't give up EF-M:

Canon Patent Application: Canon EF-M 11-22mm F4-5.6

In this patent application, it looks like Canon is revisiting the excellent EOS-M lens for the Canon EF-M mount; The EOS-M 11-22mm F4-5.6 which is one of the mainstays of the EF-M ecosystem. Where there are similarities to the EF-M 11-22, the element design is certainly different. Could this...

www.canonnews.com

The EF-M 11-22 isn't even a high end lens but it is better than every Rf-M lens equivalent.

Canon are hell bent on creating new RF-S/RF lenses that don't create an image circle to cover the sensor because they can use firmware to stretch the image.
In other words, Canon are requiring APS-C shooters on RF-S to accept image compromises.
I can't wait until Sigma/Tamron/others get into the game on RF, even if it is just RF-S, to save us from Canon's small mindedness.

Stretching the image with software will reduce adding aberrations, weight, size, and price associated with stretching via increasing the number of lens elements.

 

[Dragon]

I like the idea of explaining with the same pixel count.
I'll try to add something that can demonstrate it with a zoom lenses: That is take a photo at the same aperture but twice the focal length. then crop the wider angle to the appropriate size to match M4/3. Next, resize the uncropped longer focal length to match the M4/3 size. Finally, compare the images.

For equivalency in that demo, the wide shot needs to be taken 2 stops brighter than the long shot. Also have to be careful with the resize algorithm in that many add a lot of sharpening that would affect the result.

 

[EricN]

For equivalency in that demo, the wide shot needs to be taken 2 stops brighter than the long shot. Also have to be careful with the resize algorithm in that many add a lot of sharpening that would affect the result.

I was focused on depth of field, and didn't think about these two points.

 

[Dragon]

I was focused on depth of field, and didn't think about these two points.

In your original proposal, the DOF would be greater in the wide shot, but if you open the lens up 2 stops for the wide shot, then the DOF should be same for the crop as for the long shot, just like OM 400 f/4.5 on M4/3 and the Canon 800 f/9 on FF. Just have to be careful with the resizing for the reason mentioned. AI software can cause confusion.

 

[mxwphoto]

No. Just…no.

First off, there was no actual Canon lens being discussed, rather the hypothetical FF equivalent of the OM 150-400/4.5. Second, with telephoto designs the image circle isn’t limiting. The OM lens projects a larger circle than the sensor needs.

The reason the OM 150-400/4.5 and the real Canon 200-800/6.3-9 both take 95mm filters is because 400mm / 4.5 = 800mm / 9 = 89mm. The focal length and f/number are intrinsic properties of a lens, independent of the sensor behind it.

You are right. I got myself a bit mixed up with equivalents and aperture calculation.

I was trying to get at why 400 f4.5 on m43 is equivalent to 800 f9 on full frame and that they have same DOF.

 

[David_B]

Stretching the image with software will reduce adding aberrations, weight, size, and price associated with stretching via increasing the number of lens elements.

Summary: stretching with software allows Canon to make lenses cheaper while selling them for the same price as before.

The EF-M 11-22 is small enough and light enough to go in a coat pocket.

 

[David_B]

The RFS 10-18 is not as nice a lens as the 11-22, but it is sharp and much smaller and lighter, which seems to be the real drive behind many of Canon's decisions. The EF-M 11-22 is one of the chunkiest of the EF-M lenses, but it is quite good even on an M6 II.

RF-S vs EF-M = smaller, lighter, cheaper, less quality. That is why we need Sigma and Tamron and others being able to make RF-S lenses: to fill the product quality gap that Canon doesn't want to. If there are market segments that Canon doesn't want to address (and that is quite clear now), then Canon should green light the 3rd party manufacturers to fill it.

 

[EricN]

Summary: stretching with software allows Canon to make lenses cheaper while selling them for the same price as before.

The EF-M 11-22 is small enough and light enough to go in a coat pocket.

Your summary is misinformation. The ef 14mm f/2.8L ii was originally sold for about ten times the price of the rf 16mm f/2.8 stm. Even now, it sells for almost seven times more new, but a used one is closer to two or three times. Besides the price difference, the rf lens is approximately one third the weight (a portion of this is because of metal versus plastic) and less than half the volume of the ef lens.
The only noticeable difference in image quality is the ef lens has chromatic aberrations towards the sides while they are not visible at all.
You can say the rf lens doesn't have weather sealing, but I believe if you are using aps-c lenses Canon's never offered one with weather sealing.

Are there any aps-c lenses that don't fit in your coat pocket?

 

[EricN]

RF-S vs EF-M = smaller, lighter, cheaper, less quality. That is why we need Sigma and Tamron and others being able to make RF-S lenses: to fill the product quality gap that Canon doesn't want to. If there are market segments that Canon doesn't want to address (and that is quite clear now), then Canon should green light the 3rd party manufacturers to fill it.

It's possible Canon will make something that meets your standards. Rome wasn't built in a day, as they say.

 

[David_B]

Your summary is misinformation. The ef 14mm f/2.8L ii was originally sold for about ten times the price of the rf 16mm f/2.8 stm.

Is this an APS-C lens? No.

Are there any aps-c lenses that don't fit in your coat pocket?

Yeah, quite a lot of then actually. All those EF-S lenses for starters.

 

[David_B]

It's possible Canon will make something that meets your standards. Rome wasn't built in a day, as they say.

And if Canon doesn't want to build Rome, it should let someone else that wants to.

 

[EricN]

Is this an APS-C lens? No.


Yeah, quite a lot of then actually. All those EF-S lenses for starters.

Obviously a full frame lens can be used on an apsc camera
you must have small coat pockets...