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

riker

5D4
Jan 19, 2015
125
64
riker.hu
What's your problem with STM? The RF lenses are all focus by wire anyway, is there really much of a difference between STM and other implementations if it is electronically coupled anyway?
STM is slow and noisy compared to USM. There's a reason why cheap lenses get STM while proper lenses get USM. Or why do you think both exist (even in RF lenses) if they are the same?
 
Upvote 0
30 FPS would render an EVF unusable. Even LiveView is 60 FPS on the current DSLRs.

It may be 60, but it doesn't add up with my observations as I can see a longer delay in my 5DIV's LiveView. As in my message above, any internal processing including exposure can't be taking longer than a frame duration. If it takes longer, you either lower the FPS number or execute some processing in parallel CPU threads, but there isn't many threads in DIGIC 6+. Having two threads means you can process data for a 60-fps EVF not in 1/60s but in 1/30s (best ideal case). Three threads give you 3/60s = 1/20s processing time at 60fps etc.
 
Upvote 0

Joules

doom
CR Pro
Jul 16, 2017
1,801
2,247
Hamburg, Germany
It may be 60, but it doesn't add up with my observations as I can see a longer delay in my 5DIV's LiveView.
I checked again, and it turns out I was wrong. In the German specification site they state 30 FPS for LiveView on the 80D and 5D IV at least.

I had read 60 on a third party site a while back, but apparently they might have gotten some information wrong.

On the R, it is 60 Hz in LiveView, unless energy saving mode is used. But I can't find anything official about the EVF itself, which I find weird.

In any case you are right in that there is a significant lag in the LiveView that seems to go beyond just the frame rate refresh.
 
  • Like
Reactions: 1 user
Upvote 0

MadScotsman

EOS R / RP
Sep 9, 2019
45
82
Am I the only one who was hoping for a 24mm, 1.2?

Definitely not the only one.

In fact, I had hoped that "ultra-wide prime L" meant 14mm. But. I'll wait. May just grab one of the cheapo Samyangs to bridge me.

I've resolved to not but any new glass today isn't RF, so... I might be waiting a while.

:rolleyes:
 
  • Like
Reactions: 1 user
Upvote 0
Apr 25, 2011
2,510
1,885
Yep. The thing is, all these operations should be taking less than one frame duration (on average), otherwise you don't get the desired FPS number for EVF or LiveView.
Unless they use different parts of the hardware for different stages of the pipeline. Then they may have several frames "in flight" with no ability to accelerate processing by dropping some frames.

Also, there are cases (such as AWB) when full information about the frame is needed for per-pixel processing. When it's not such a case, it might be possible to process the image by, say, 32-line strips, but the logistics of doing it properly in both software and hardware can be a nightmare.
 
Upvote 0
Unless they use different parts of the hardware for different stages of the pipeline. Then they may have several frames "in flight" with no ability to accelerate processing by dropping some frames.

If you think about this pipeline as a black box, it should be spitting out 30 frames per second for EVF (or whatever required FPS). Now if there's just one pipe/thread inside this black box, it must be able to process the whole frame in 1/30s, no matter what hardware it uses internally and how many frames there are in the pipeline queue. Only second thread allows to delay the delivery and still render 30 frames a second. Ideal case (which won't be ideal in practice) - thread 1 processes even frames, 1/15s each, and thread 2 processes odd frames. So we have the lag of 1/15s and 30fps. But in this case, why wouldn't we just have both threads to process each frame 2x faster, so that each frame processing takes 1/30s, instead of laggy 1/15s?

There must be something I'm missing, but it may be a thing I don't know I don't know...
 
Upvote 0
Apr 25, 2011
2,510
1,885
If you think about this pipeline as a black box, it should be spitting out 30 frames per second for EVF (or whatever required FPS). Now if there's just one pipe/thread inside this black box,
Imagine you have two black boxes running at 30fps. One can only do debayering, another (due to different hardware) can only do DLO. If you need to use both boxes, the FPS is still 30, but the lag is 1/15.
 
Upvote 0
That's a beautiful lens, but its huge, heavy, expensive, and overkill for most people. A modest 50 f/1.4 IS would be amazing, and overdue I think.

50 f/1.4 IS with image quality similar to sigma 50 f/1.4 art will not be small and affordable. It will cost as much as other L primes. There are two option for canon to make 50mm affordable and small. Use old optical formula (double gauss) but it will sacrifice image quality or make it f/1.8 that compete directly to nikon z 50mm f1.8
 
Upvote 0

Fran Decatta

EOS R6
Mar 6, 2019
95
109
If I had the means to do both zooms and primes I would. But I'm not giving up my 1.4L primes. The zooms look incredible and it's VERY tempting (especially now that we will have IBIS moving forward) but I love my 1.4 too much. I'm far more tempted to get a 24-70 f2.8 IS. I need a quality all-arounder but I'm not gonna spend $3000 for one. For someone who is building a kit (no lens library yet) the f2 zooms make a LOT of sense. Buying a '35 and 50 is more than the f2 zoom (provided you are willing to sacrifice 1 stop on EF and 1.5 on RF)

Well, I already had 24, 35 and 50 1.4 + 85 1.8, I sold them all, after work with them for a few years. Was a hard decision, but lately I was stopping down to 1.8 - 2.2 and only using the 1.4 when the light was really low. Also, working as a wedding photographer, I noticed so much difference in my last works, This plus to change focal lengs so fast having a great aperture, was a step ahead on the work without losing quality, and just a little bit of bokeh / light. In a balance, it deserves the change. Of course, this is only my personal experience and everyone must look for his needs :D
 
  • Like
Reactions: 1 users
Upvote 0
Imagine you have two black boxes running at 30fps. One can only do debayering, another (due to different hardware) can only do DLO. If you need to use both boxes, the FPS is still 30, but the lag is 1/15.

I've already given the example with multithreading where you can have 30fps but the lag of 1/15s.

Your black boxes are the same as multithreading as they require DLO and debayering to run in parallel. Say they'll run DLO for frame 777 and debayering for frame 778, then 778 and 779 respectively etc.

But as far as I understand the Canon's architecture, DIGIC is the one who executes all of those DLO, debayering etc. steps. As a side note, is DLO even applied to EVF/LV?
 
Upvote 0
Weird to have 1.4/2x TC without lenses to use them on ie only the 100-500mm would be available and @1000mm f/14 (2x) would have few use cases.

Can we confirm that the RF TCs are RF to RF? If they are RF to EF TCs then it would make some sense ie to avoid RF-EF adapter + EF 1.4/2x TC and use the existing big whites (and EF 70-200mm :) )RF 1.4x TC.JPGEF-EOS R mount adapter.JPG
 
Last edited:
  • Like
Reactions: 3 users
Upvote 0
Apr 25, 2011
2,510
1,885
But as far as I understand the Canon's architecture, DIGIC is the one who executes all of those DLO, debayering etc. steps.
But DIGIC is not (just) a general purpose CPU. It contains DSP modules optimized for particular kinds of work.

As a side note, is DLO even applied to EVF/LV?
RF 24-240.
 
Upvote 0
size: XL
Weight: XL
Price: XL
Well smaller than a traditional 70-200 f/2.8 which I don't regard as XL, price will be high but will it be dearer than a 70-200 f/2.8 and size will be much shorter than a 70-200 f/2.8 and with smaller front element. I'm not sure it'll be heavier than 28-70 f/2 which would be a more complex design.
 
Upvote 0
Jan 30, 2020
410
513
I know that the R5 is hot stuff right now, but where's the dirt-cheap lens collection for the RP? It would make masses finally move to full frame mirrorless. And yes, I speak about RF lenses, not EF + adapter.
Yes, we need regular f/4 non "£" lenses . Because they are not sharing the mount info with sigma and tamron, we likely won't get 100% compatibility if they decide to build RF mount lenses (ie slower af, and perhaps issues with IBIS).
 
Upvote 0
50 f/1.4 IS with image quality similar to sigma 50 f/1.4 art will not be small and affordable. It will cost as much as other L primes. There are two option for canon to make 50mm affordable and small. Use old optical formula (double gauss) but it will sacrifice image quality or make it f/1.8 that compete directly to nikon z 50mm f1.8

While ‘affordable’ would always be nice, nobody asking for a 50/1.4 has mentioned affordability as the main reason they want one.
I would rather a 50/1.4 with the look of the Sigma 50/1.4 EX (or the old Zuiko 55/1.2) than the 50 Art. It doesn’t need to have ultimate corner resolution or be free of all aberrations, or have a flat plane of focus, or have minimal vignetting wide open, and it doesn’t need to be cheap. Just make it SMALL, nice and sharp in the centre, smooth bokeh without nauseating swirls, and give it modern AF.
 
Upvote 0
But DIGIC is not (just) a general purpose CPU. It contains DSP modules optimized for particular kinds of work.
I believe DIGIC is just ARM with extended instruction set and any DSP modules are in its firmware; but wherever the DSP modules are, the execution should be parallelised/threaded if the the EVF lag is longer than 1 frame.
 
Upvote 0
I believe DIGIC is just ARM with extended instruction set and any DSP modules are in its firmware; but wherever the DSP modules are, the execution should be parallelised/threaded if the the EVF lag is longer than 1 frame.
DIGIC is most definately a SoC - system on chip more than just an ARM processor.
 
Upvote 0