Patent: Canon RF 90mm f/2.8L IS Macro & RF 100mm f/2.8L IS Macro

Moving the IS forward? What would that help with?

I don't know, but it might work better with IBIS for macro ;-) ;-)

I swear if they update the MP-E for the RF mount I’ll jump off EF so fast.

Just for once, I’d like to see a lens actually meet the published specs in length and speed. It’s always shorter for longer lenses, and longer for shorter lenses, and it’s always slower. So it figures that this 90mm f2.8 is really an 86mm f2.87.

A bit off topic but I don’t get it why am actual 86mm lens will be named 90mm instead of 85?

The formulas are demonstrating a possible lens configuration and seldom give a focal length of a actual production lens. All of them have tested to match specifications that I've seen, a few people test them to other specs, such as distance to subject and find different values, but those are invalid tests, even if they sound reasonable. For example, a test at MFD for a Macro sounds reasonable, but lens focal length measurements are always at infinity.

I think it's better to specify that both lens (90mm and 100mm for approximation) were 1:1 Macro lens

From my experience, they usually round up. Or they choose a focal length that is better for marketing. In my opinion "85mm" is generally marketed as a portrait lens, whereas macro lenses over the years have tended to be 90mm or 100mm.

You could line up 5 different "50mm" lenses, and they'll likely all be a slightly different focal length.

I am just glad to see some more RF lenses and ones that again are very useful.

What's with the "back focus" of 16mm while the flange-to-sensor distance is 20mm?
Does one design's last lens stick 4mm deep into the body (unlike the other design)?
Or was the patent filed when the flange distance was not finalised yet?

Please. I’ve seen hundreds of reviews of lenses in 5+ decades, as well as having benched more than a few myself. Perhaps a handful has matched published specs.

When benching focal length I presume you either measure the distance from the front element to the sensor plane, or measure field of view and back calculate, but how in the world would you bench f number? Somehow gauge the diameter of the aperture as viewed through the front element?

Also add 180mm macro to that list, I just need a good aps-c RF based camera with 5x macro and 180mm macro just to jump ship from EF for good.

There are so many different things that specific users want, but so little time to do them all. I expect Canon to try and hit as many specialized uses as they can in the next 2 years, and a true Macro lens should be one of them. The 35mm is not a true 1:1 macro, and 180mm is probably to physically large. 90 - 120 mm might be the sweet spot.

I love the MP-E, but I think it could definitely use an update. I don’t really think anything should change other than the low end of the magnification range. If it started at .5x I think it would be a huge benefit to field shooters. As far as the top end, you can always add extension tubes. In my experience, 3.5x-4x is usually the upper limit in field use.

I’m constantly presented with the dilemma of using the MP-E or a 100L with extension tubes. If I think I’m going to see anything larger than a paper wasp (and they’re even hard to frame at 1x) I take the 100L. The 100L is great, but is nowhere near as sharp as the MP-E. Autofocus at higher mags is inconsistent and nearly useless, so I use it basically the same way I do the MP-E, moving back and forth until I hit where I want.

This all said, a 90-120mm that went to 2x could be interesting. Add a couple of tubes and you’re into the 3x-4x range. Make a proper white L macro and include a drop-in magnifying element a-la the Raynox 250 and similar. I’m sure this would be a huge lens though... Ok, enough pipe-dreaming

How about focus at infinity, take a picture of a small light source a known distance from the sensor plane and measure the diameter of the resulting bright circle?

If you cannot get hold of a true point light and don't know the expected size (because you don't know the real focal length), take two identical ones at a known distance from each other and you have a scale reference so you can compensate for the size of the light. Two 3mm LEDs on some perf board would suffice.

If focused at true infinity, the horizontal image angle should be 2*atan(26/(2*f)) for a rectilinear lens and close to image centre, magnification of oof light sources at 50*f distance become small enough (roughly 5% of image height for f/2) that distortion effects on reasonably modern lenses become smaller than the precision you get from counting pixels.

Ah. This was a macro lens. Then everything gets really tricky. The above is using the assumption that at a reasonable distance, the lens behaves in a way close enough to that of an ideal lens, where 1/f=1/o+1/i

Mea culpa

I wonder how many MP-E 65 users actually use it primarily in the field? I don't think it was really designed as a field lens, it really is far more at home clamped to a serious tripod at home. Once you get down to 5x you're pretty much required to stick to f/2.8 because of diffraction which means your depth of field almost doesn't exist, so unless you're shooting perfectly flat things you absolutely need to stack, and that's no fun out in the field

So I’m guessing there is some reason why it seems all 1:1 macro lenses are no faster than f 2.8? For me if the 90mm was even f2 it would double as a portrait lens, making it far more versatile.