Are these the 7 RF lenses Canon will be announcing in 2020? [CR1]

[chasingrealness]

I'd keep the 600mm 1.2, then after the thief is gone, I'll sell it and buy every other 1.2 prime.

If that's cheating, I'll keep the 50mm. 85mm is too long to use indoors at Christmas gatherings, and I have some relatives that desperately need to be prettied up with a 1.2.

Haha good choice! Imagine the depth of field on a lens with these specs...

 

[chasingrealness]

The RF 50mm f/1.2L and RF 85mm f/1.2L are true wonders. I would say that the 85mm is the better of the two for portraits. Both are insanely sharp at f/1.2. I have sold my RF 50mm f/1.2L only because my 28-70 covers 50mm just fine for me (also a truly fantastic lens). I just do not have much use for a 50mm prime for my portrait work. Really looking forward to the RF 70-135mm f/2, though an RF 135mm f/1.4L would also be extremely tempting. The autofocus on all three are as near perfect as one can get. With the R's eye-AF my keeper rate (tack sharp focus on the iris) is at least 95%. That is far and away more than what I got with EF lenses on the 5D Mark III. Hope this helps with your decision. I have seen zero CA with the 50 or the 85 in my work.

Thank you for the response. This is my feeling as well. I really love the 85 focal length and think it’s worth the few extra bucks over the 50. And the 28-70 would be better than the 50 since I could get both 35 and 50 in one package with only a slight sacrifice in light gathering capability.

 

[Ozarker]

Thank you for the response. This is my feeling as well. I really love the 85 focal length and think it’s worth the few extra bucks over the 50. And the 28-70 would be better than the 50 since I could get both 35 and 50 in one package with only a slight sacrifice in light gathering capability.

In my opinion, you would not be disappointed in the 28-70mm or the 85mm. They are both worth every penny. Honestly, as good as the EF 135mm f/2L was for me in the bokeh dept., an f/2 zoom from 70-135 would probably do it for me. I could be really really happy with just those three lenses. I was never satisfied with the bokeh of the EF 70-200mm f/2.8L IS II though it was a great lens otherwise.

 

[navastronia]

Yeah, probably same. I would want to go for 24, 35, 50, 85, and then cap off with a 135 f/1.4 if they end up releasing that one per the patent spec posted not too long ago on CR.

I'm thinkin' 35 and 85, and then a 14mm f2-2.8 (?? depending on what Canon offers), plus this just-announced 100-500. That would keep me happy for a long time!

 

[davidespinosa]

And the 28-70 would be better than the 50 since I could get both 35 and 50 in one package with only a slight sacrifice in light gathering capability.

It's not a slight sacrifice -- it's more than a full stop.
The zoom is f/2, the primes are f/1.2.

 

[derpderp]

give me a 24mm f1.2 that out performs Sony GM equivalent and i'd ditch my sony kit in a heartbeat

 

[Kit.]

It's totally the opposite, I said the time passed since the beginning of exposure. Obviously it includes exposure itself, readout etc.

But your estimation does not include any number related to exposure itself etc.

Why?

But what I said in the previous message was, should we calculate the lag not from the beginning of exposure, but earlier, when some physical event of interest happened. In which case, as you can see below, if the event happens right after the exposure, there's no way to see it until the next exposure is fully processed.

          Event                                             Event is shown
            |                                                      |
-<Exposure>-+-<Processing>-<EVF>---------<Exposure>-<Processing>-<EVF>---------------> time

Why can't you start the new exposure right after the data from the previous one is read out from the sensor? It's not like the sensor is busy with something at that moment. "Processing" and "EVF" is not what the sensor does.

 

[telemaque]

This is an amazing list. But why release the converters without well-suited primes to use them with?

I would think a 70-135mm f2 plus an extender is quite interesting.
If x2 extender this gives a 140-270mm f4, quite interesting in my personal opinion.

Obviously depends on the price of the 70-135mm f2...

 

[telemaque]

What percentage of the time is a 70-200mm f/2.8 too dark? 1%? Maybe 3% on a really bad shoot?

Your comment is correct for photography. For videography, the situation is a bit different.
You are stuck with a shutter speed of 1/50 s at best, so f2.8 can be quite dark in a church during a wedding...

In those moments, 1 stop, so as said 100% more light, can be a big difference.
Reason why I bought from Rokinon their cinelens with T1.5, this changes your life in those moments!

However, this is a pure videography point of view...

 

[telemaque]

No love for us... sniff...

I would bet when the R1 type of body will be launched a bunch of big white lenses should follow.
Than you will feel the love of Canon !
and your bank account feel very bad all of a sudden...

 

[telemaque]

Sorry to finish your dreams but I think prices will be slightly above the 500-550 $ / EUR for the RF35 because large focal lengths at the same aperture need larger (diameter) and thicker lenses and AF motors / IS have to move more weight. And I am shure they will use advanced lens designs like RF 35 (and not that of the current nifty fifty).

But my RF35 is worth every buck: good wide open, excellent above @f/2.8 and very very flexible due to its image stabilization and MACRO FUNCTION !!!

Interesting comment from you

Have you also used in the past the EF 35mm f2 IS?

If yes, how do you compare the image quality between the two lenses?

I am asking as I own the EF 35mm and I am very happy with it, almost an "L" quality in my opinion...

 

[telemaque]

I notice you are very focused on vignetting. While it may be an issue for you, for the vast bulk of images made it is a non-issue.
I use the EF 11-24 on a daily basis in high end architectural applications along with my 17TS-E and vignetting is not a problem.
Actual use of most lenses is not wide open but stopped down where vignetting becomes invisible.
Real world use is where I evaluate lenses and my experience informs me that lenses like the 28-70 f2 used wide open are brilliant, the EF 11-24 is superb and the 17TS-E is a classic high performer.
Canon is the leader in the optical world at the moment. Not Zeiss, not Leica, not Sigma, not Nikon. Each of those manufacturers has a few gems but not the number and momentum that Canon has.

If I may, I think Zeiss and Leica are not targetting the same type of customers.
What they produce is pleasing their customers A LOT.

If you look at comments from Leica and Zeiss owners, they speak about their lenses and cameras as if God had produced them.

I love Canon quality but I see Leica and Zeiss as high end producers and would not rank them below Canon.

Maybe being a European influences my point of view...

 

[Quarkcharmed]

But your estimation does not include any number related to exposure itself etc.

Why?

Why would I need to include the number? Remember we were simply trying to answer the question whether the lag can be longer than frame duration. As in my previous diagram, depending on how you define the lag, it can be up to two times longer. But that's not because of some tricky parallel processing etc.

In terms of the numbers, obviously the exposure itself must be shorter than one frame. It must also leave a room for readout and processing/render. I believe the readout goes at 10 or 12 bits to speed it up, but I suspect the exposure + readout take most of the time as all the processing happens in memory and should be fairly quick.

Why can't you start the new exposure right after the data from the previous one is read out from the sensor? It's not like the sensor is busy with something at that moment. "Processing" and "EVF" is not what the sensor does.

I think you can do some processing (from the previous frame for example) during the exposure. The readout - I'm not so sure, maybe it takes both in-sensor processing and CPU who controls the reading ow by row.

Anyway, you can't start a new exposure right away. In order to get smooth video on the screen, you have to take exposures in regular intervals (and render them using the same intervals). Say your frame is 1/30s and the exposure is 1/60s. It makes no sense to take two exposures one after another. The beginning of each exposure should be aligned with 1/30s intervals.

Also, if you take the second exposure right after the readout, where do you read the data to? It should be using some fast memory, do they have enough memory? I don't know.

 

[Steve Balcombe]

I would think a 70-135mm f2 plus an extender is quite interesting.
If x2 extender this gives a 140-270mm f4, quite interesting in my personal opinion.

Obviously depends on the price of the 70-135mm f2...

... and the performance, both AF and optical, with the x2 attached. And more fundamentally, if the 2x will physically fit which I would guess is unlikely.

The best route to that combination of focal length and aperture would be the 70-200/2.8 with a 1.4x, giving you a 98-280 f/4. Slightly more range and only a 1.4x so very likely better performance. Also has the advantage of not being a rumour! Again the problem might be compatibility - Bryan Carnathan notes here: "the RF 70-200's rear lens element is quite shallow when the lens is retracted, leaving little space for an extender to be inserted into the back of the lens." He wrote this before the RF Extenders were announced, and we now know they have quite significant projections into the lens. Surely, you might say, Canon wouldn't take away the Extender compatibility we had with the EF 70-200s? Let's hope not, but I wouldn't bank on it :-(.

 

[riker]

(Partially) wrong. USM is a bit faster, but STM is more silent and smoother in movement. For video you will probably prefer STM. For photography STM is almost as good as USM.

Well, that is simply not what I experience, having STM lens, old USM lens and new USM lens. I'm afraid your statement is more from marketing materials and official articles than actual real life use.

 

[Daner]

That 70-135 could be really interesting, especially if it would work with an extender. Have a Tamron 35-150mm F/2.8-4 that is a great "all around" lens without sacrificing much. Pairs really well with 17-35mm F/2.8-4 when the bigger lenses just can't make it in the bag. (Wish the 15-30mm F/2.8 would make it in the backpack more often!)

One of the benefits of the RF mount is that it enables the positioning of the rearmost glass element to be closer to the sensor than with the EF designs. If I understand things correctly, every RF lens that has been released so far makes use of this, and is thus unsuitable for use with extenders. The upcoming 100-500 is the exception among announced lenses.

Making the most of this optically has made it so the RF 70-200 F/2.8 does not work with extenders. For this reason, I am pretty sure that the 70-135 F/2 will not either.

 

[Kit.]

Why would I need to include the number? Remember we were simply trying to answer the question whether the lag can be longer than frame duration.

Why do you equate this "frame duration" (whatever that means) to the EVF refresh rate and not include the values for other important delays into it? That's what I am asking.

It seems that your "frame duration" is actually the lag (a tautology) and you falsely assume that the EVF refresh rate somehow equates to it.

Anyway, you can't start a new exposure right away. In order to get smooth video on the screen,

That's not our goal. Our goal is to get decent tracking in live view, not "smooth video on the screen". With full electronic shutter, we may as well start next exposure for the current line right after we read it (if the scene is dark enough), not waiting to finish reading the whole sensor.

Also, if you take the second exposure right after the readout, where do you read the data to? It should be using some fast memory, do they have enough memory? I don't know.

The sensor. It has a lot of unused analog memory after the readout. Where do you think all those photoelectrons are stored during exposure?

 

[koenkooi]

... and the performance, both AF and optical, with the x2 attached. And more fundamentally, if the 2x will physically fit which I would guess is unlikely.

The best route to that combination of focal length and aperture would be the 70-200/2.8 with a 1.4x, giving you a 98-280 f/4. Slightly more range and only a 1.4x so very likely better performance. Also has the advantage of not being a rumour! Again the problem might be compatibility - Bryan Carnathan notes here: "the RF 70-200's rear lens element is quite shallow when the lens is retracted, leaving little space for an extender to be inserted into the back of the lens." He wrote this before the RF Extenders were announced, and we now know they have quite significant projections into the lens. Surely, you might say, Canon wouldn't take away the Extender compatibility we had with the EF 70-200s? Let's hope not, but I wouldn't bank on it :-(.

Have a look at https://www.canon.co.uk/lenses/rf-70-200mm-f2-8l-is-usm-lens/specifications/ and scroll down to "Extender Compatibility".

 

[Steve Balcombe]

Well, that is simply not what I experience, having STM lens, old USM lens and new USM lens. I'm afraid your statement is more from marketing materials and official articles than actual real life use.

There are two types of STM mechanism. The 'lead screw' type, as used on the zooms, is pretty quiet, but there isn't room for this in physically shorter lenses, which in practice means the primes. They have a gear train similar to the old micromotor which is inclined to be noisier.

 

[Quarkcharmed]

Why do you equate this "frame duration" (whatever that means)

It's the time period between two distinct sequential images rendered in the EVF/LiveView.

to the EVF refresh rate

I tried to avoid this term. EVF rate can be different than actual frames-per-second. Say an EVF can run at 120Hz but camera only feeds it with 30 frames per second. It's similar to any LCD monitor. It shouldn't matter for our discussion though because most likely the system will feed EVF with new frames just before refreshing. How it works exactly I don't know as it's very specific technical documentation we don't have.

and not include the values for other important delays into it? That's what I am asking.

Which values? It's all inside <processing> in this diagram.
If you claim your camera does 30fps, you should capture frames 30 times a second. The delta time between the beginnings of exposures must be 1/30s. Your system should be able to process the raw data and feed the rendered frame to the EVF before the next capture. Rendering should also be at 30fps exactly in regular intervals. Irregularities cause so called stuttering and you normally want to avoid them at all costs especially in the EVFs. Stuttering may cause headaches quite literally. Also it's better to have 60fps, not 30, but we use 30 just as an example.

The sensor. It has a lot of unused analog memory after the readout. Where do you think all those photoelectrons are stored during exposure?

Ok. So you suggest to do the exposure but hold it in the sensor because there's previous capture in memory. So we can keep two exposures taken one right after another. Right. And why would we need to do the exposures in pairs?