What will Canon bring to the table with the EOS R1?

[Chig]

R1 will exceed the tech specs of A1 and be in stock at BH within 12 months.

Canon will offer a (free?) program to modify Series III EF 400mm and 600mm lens to RF mount.

Why modify them when they work perfectly with the adapter and Canon will soon release RF versions anyway ?

 

[navastronia]

I just came in here to say that the A1 isn't very much camera for your money and that without a vertical grip and heavy duty build, it's not competing with a future R1 at all, and is instead an overpriced (albeit high-spec'd) R5 competitor.

IDK who Sony is fooling, but the A1 isn't a sports camera and shouldn't be priced like one.

 

[Chig]

Continue reading...

Be nice if Canon use a stacked sensor to match Sony's dynamic range which is only aspect where they lag behind Sony , hopefully with sensible 20-24 mp size for the R1 - the sports pros don't want massive files to slow down their workflow.

Would love them to start using round sensors to match RF lenses image circle so you can crop portrait/landscape/square at any angle in post instead of fussing around turning your camera around .
Need to have a circular evf too of course , just seems more natural like our own eyes , after all the rectangular shape is a relict from the days of film so time to move on

Tilting evf would be cool too , especially for holding the camera down low for ground level shots

Another cool feature would be a red dot sight mounted on the camera aligned with your left eye to make aiming long telephoto lenses easier , I have one mounted on each of my long lenses which is extremely handy.

An option of automatic exposure compensation based on the target's colouring. For example if you're shooting a light coloured object like a white bird the camera could automatically compensate to avoid blowing out the highlights and vice versa for a dark object . I'm often shooting light coloured sea birds when a dark bird suddenly appears and it takes time to adjust the exposure compensation. I can see Canon using an A.I. program for this similar to their animal eye af system

 

[privatebydesign]

Yes 30/32MP would be nice. Contrary to much opinion on this site, most 1D aren't used blasting away at sports events. All my town's newspaper photographers use them because they're burstproof, shrug off the rain and and have a familiar UI. They don't need 14fps for the Mayor opening a new garden centre but it doesn't hurt to have it to hand.

Sure they could probably make do using a 5D4 but if your salary depends on your photos, why take a risk? A 1D shoots and shoots and focuses like the pro it is.

They don't need 20mp for newspaper, magazine or news site output either, 30/32MP would be counter intuitive for that market.

The last image I sold to a printed paper used a 280kb thumbnail as a 4 column header and the IQ was up to their printing. I certainly didn't need to have taken that with a 4mp 1D.

 

[HarryFilm]

Steady on Harry, you're really getting carried away now........unless you know something about the 1DXIII we don't

I fixed that IBIS statement.... it was a mental wish list my fingers typed out.

v

 

[Aussie shooter]

Why would it need a larger mount ? Even if it did then just make the square sensor slightly smaller to match the image circle or better still a round sensor of exactly the same size as the image circle .

Because you can only fit 36 wide when the senor is 24 high. If it is 36 high you could only fit 24 wide. Yes. You could probably go something like a 30x30 with the current mount

 

[usern4cr]

Because you can only fit 36 wide when the senor is 24 high. If it is 36 high you could only fit 24 wide. Yes. You could probably go something like a 30x30 with the current mount

Since you said 24 wide by 36 tall was OK, then my point (way up there somewhere) was to have a 36 x 36 sensor so that you could get a 36x24 or 24x36 out of it (without rotating for portaits), and anything inbetween on the 36x36 would be gravy. I'd certainly jump for joy to have that sensor in my camera!

Of course, Canon will never make this, so we're all just havin' fun talking about it.

 

[Film Runner]

(me, sitting in the back quietly, still pretty happy with my EOS R)

 

[Aussie shooter]

Since you said 24 wide by 36 tall was OK, then my point (way up there somewhere) was to have a 36 x 36 sensor so that you could get a 36x24 or 24x36 out of it (without rotating for portaits), and anything inbetween on the 36x36 would be gravy. I'd certainly jump for joy to have that sensor in my camera!

Of course, Canon will never make this, so we're all just havin' fun talking about it.

Ah. Ok. I thought you wanted 36x36 all usable. My mistake.

 

[Chig]

Because you can only fit 36 wide when the senor is 24 high. If it is 36 high you could only fit 24 wide. Yes. You could probably go something like a 30x30 with the current mount

Yes , you're right I just worked it out with a ruler and a compass .

Current image circle is about 43mm diameter to fit 24x36 sensor and it would also fit a 30x30 sensor

A 36 x36mm square sensor would give you 24x36 in either vertical or horizontal orientation or a 30x30 image but some of the sensor would be outside the 43mm diameter image circle.

But a 43mm diameter round sensor would be even better giving you 30x30 square or 24x36 at any angle and also 43mm circular image and the areas of these would be 864 mm for the 24x36 (same as current) , 900mm for the 30x30mm square and 1452mm for the 43mm diameter circular image (although a circular finished image would be a bit niche but I think astronomers would like it though).

Also need round EVF too but this could have cropping options to show various cropped images, this would have a control button or dial to toggle through various framing options :

• you can have it cropped in landscape mode but toggle this button to instantly show portrait mode in the EVF instead of turning your camera on it's side .
• whatever cropping option you choose could be self-leveling with the camera showing you the level image in the EVF whatever angle you happen to be holding your camera

Still 1.68 x more area is quite something and of course the sensor manufacturing cost would perhaps be double too but might be worthwhile for the versatility : imagine no twisting around for portrait/landscape or worrying about holding your camera level and no need for vertical grips with duplicated control buttons !

Could do the same for aps-c with a 27mm diameter version

Look at the image below and imagine the possibilities !

 

[SteveC]

The way to figure this without having to make a drawing, is to use Pythagoras. Square your height, square your width, add them, take the square root. That will give you the length of the diagonal, and your image circle needs to be at least that wide (especially if using IBIS).

 

[Chig]

The way to figure this without having to make a drawing, is to use Pythagoras. Square your height, square your width, add them, take the square root. That will give you the length of the diagonal, and your image circle needs to be at least that wide (especially if using IBIS).

Thanks Steve I've just done that too (before I saw your reply after remembering my high school maths) giving 43.27mm for Full Frame and 26.68mm for aps-c

 

[Talys]

I just came in here to say that the A1 isn't very much camera for your money and that without a vertical grip and heavy duty build, it's not competing with a future R1 at all, and is instead an overpriced (albeit high-spec'd) R5 competitor.

IDK who Sony is fooling, but the A1 isn't a sports camera and shouldn't be priced like one.

It also looks to have the grip ergonomics that is awkward (to put it mildly) on large lenses. To be honest, when I saw the specs and the photos, I thought it was more much more in line with the A7R series and a R5/5D competitor than a pro sports workhorse.

I'll be curious as to how well the autofocus works compared to the R5, which is really in a league of its own at the moment.

 

[Joules]

Be nice if Canon use a stacked sensor to match Sony's dynamic range which is only aspect where they lag behind Sony

I am not so certain about that.

Canon doesn't market their sensor technologies all that much, but they have brought some serious punch to the table with the R5.

I believe dynamic range is not a direct benefit of stacked sensor designs. My understanding is that it allows more intricate read out circuits which can help with the noise floor, but Canon seems to have achieved comparable results (beaten Sony, actually) by using a different approach (see photons to photos comparison)

I understood stacked sensors to be mainly about read out speed. Unfortunately, this is a number that's not often measured in reviews (despite being of interest, as it quantifies the rolling shutter), and of late it seems people also confuse it with the flash sync speed on top of this. I understand these to be related, but not the same.

Measurements you can find for the a9 and R5 ( here and here) indicate that the a9 takes 1/150 s to read the full 24 MP sensor in fully electronic shutter mode, and the R5 takes 2.5 times as long at 1/60 s for the full 45 MP sensor. So the R5 is clearly slower, but it is a major jump for Canon and at 1.9 times the amount of pixels (and 3.8 times the amount of photosites) you could argue that the R5 actually almost keeps up with (or surpasses) the a9 in purely technical terms.

This is not saying that Sony's stacked designs isn't impressive. I'm just pointing out that Canon does not have to follow Sony's direction 1:1 to compete or beat them. There are a lot of rumors that the R1 and potentially even the high res R may see new sensor technology again.

I think we've seen by now that Canon is comfortable taking their sweet time to come up with their own implementations (IBIS, eye AF, video features, sensors,...) instead of jumping on what ever Sony is doing. Since they have not openly adapted stacked sensors yet, I believe they either have something they view as superior in the pipeline or will bring them once they do all they want without compromise. And cost is a compromise in this context.

 

[usern4cr]

Yes , you're right I just worked it out with a ruler and a compass .

Current image circle is about 43mm diameter to fit 24x36 sensor and it would also fit a 30x30 sensor

A 36 x36mm square sensor would give you 24x36 in either vertical or horizontal orientation or a 30x30 image but some of the sensor would be outside the 43mm diameter image circle.

But a 43mm diameter round sensor would be even better giving you 30x30 square or 24x36 at any angle and also 43mm circular image and the areas of these would be 864 mm for the 24x36 (same as current) , 900mm for the 30x30mm square and 1452mm for the 43mm diameter circular image (although a circular finished image would be a bit niche but I think astronomers would like it though).

Also need round EVF too but this could have cropping options to show various cropped images, this would have a control button or dial to toggle through various framing options :

• you can have it cropped in landscape mode but toggle this button to instantly show portrait mode in the EVF instead of turning your camera on it's side .
• whatever cropping option you choose could be self-leveling with the camera showing you the level image in the EVF whatever angle you happen to be holding your camera

Still 1.68 x more area is quite something and of course the sensor manufacturing cost would perhaps be double too but might be worthwhile for the versatility : imagine no twisting around for portrait/landscape or worrying about holding your camera level and no need for vertical grips with duplicated control buttons !

Could do the same for aps-c with a 27mm diameter version

Look at the image below and imagine the possibilities !

View attachment 195584

Thanks for the diagram. I think it'll be a glorious day when Canon makes this circular sensor!
Of course, when they do, you better look up at the sky since you might be hit by one of those "flying pigs! "

 

[usern4cr]

I am not so certain about that.

Canon doesn't market their sensor technologies all that much, but they have brought some serious punch to the table with the R5.

I believe dynamic range is not a direct benefit of stacked sensor designs. My understanding is that it allows more intricate read out circuits which can help with the noise floor, but Canon seems to have achieved comparable results (beaten Sony, actually) by using a different approach (see photons to photos comparison)

I understood stacked sensors to be mainly about read out speed. Unfortunately, this is a number that's not often measured in reviews (despite being of interest, as it quantifies the rolling shutter), and of late it seems people also confuse it with the flash sync speed on top of this. I understand these to be related, but not the same.

Measurements you can find for the a9 and R5 ( here and here) indicate that the a9 takes 1/150 s to read the full 24 MP sensor in fully electronic shutter mode, and the R5 takes 2.5 times as long at 1/60 s for the full 45 MP sensor. So the R5 is clearly slower, but it is a major jump for Canon and at 1.9 times the amount of pixels (and 3.8 times the amount of photosites) you could argue that the R5 actually almost keeps up with (or surpasses) the a9 in purely technical terms.

This is not saying that Sony's stacked designs isn't impressive. I'm just pointing out that Canon does not have to follow Sony's direction 1:1 to compete or beat them. There are a lot of rumors that the R1 and potentially even the high res R may see new sensor technology again.

I think we've seen by now that Canon is comfortable taking their sweet time to come up with their own implementations (IBIS, eye AF, video features, sensors,...) instead of jumping on what ever Sony is doing. Since they have not openly adapted stacked sensors yet, I believe they either have something they view as superior in the pipeline or will bring them once they do all they want without compromise. And cost is a compromise in this context.

Thanks for the good post, Joules. One thing that surprises me is that Canon has been able to do so much with dual pixels and a 45MP sensor (which I assume is 45M pixel elements consisting of 90M half-pixel elements) considering they still use "front-side" illumination instead of the "back-side" illumination that Sony switched to. I'd think that as you push the resolution further (90M half-pixel elements is already a ton!) you'd run out of room for the light bucket to see around all the upper circuitry. So in the new R1 sensor, if they do have quad-pixels then my guess is that their sensor will be front-sided, QP, FF with 6K-60fps video, 4K-120 video with "sufficient" heat control. Sensor is 6144 x 4096 (~25MP bayer array or slightly higher). This size QP sensor (probably with more AF points) would be faster & much more reliable at eye-AF which is one of the best things (IMHO) that ever happened to a camera. EDIT: 30 minute video limit should not exist as the EU no longer has this legal issue, but we'll see!

 

[Joules]

30 minute video limit (legal only issue) may be there (if so, probably a license fee can remove it).

As far as I know, it's not crazy to expect future products to exceed the 30 min limit. The legal explanation given for it in the past was that the EU applies an extra tax on video devices, which are cameras that can record over 30 minutes. But that apparently was removed. Seems like on the R6 and R5 Canon just kept the 30 minute video for the lolz. Is the word crippled already censored?

 

[usern4cr]

As far as I know, it's not crazy to expect future products to exceed the 30 min limit. The legal explanation given for it in the past was that the EU applies an extra tax on video devices, which are cameras that can record over 30 minutes. But that apparently was removed. Seems like on the R6 and R5 Canon just kept the 30 minute video for the lolz. Is the word crippled already censored?

If it was removed, then why do you think the A1 still has a 30 minute limit? I assumed it was because of the legal issue, but maybe it's because of a heat issue and Sony picked the 30 minute number to make people think it was a legal issue and therefore better "hide" a heat issue (albeit not as bad) that they have as well?

 

[juststeve]

I am not so certain about that.

Canon doesn't market their sensor technologies all that much, but they have brought some serious punch to the table with the R5.

I believe dynamic range is not a direct benefit of stacked sensor designs. My understanding is that it allows more intricate read out circuits which can help with the noise floor, but Canon seems to have achieved comparable results (beaten Sony, actually) by using a different approach (see photons to photos comparison)

I understood stacked sensors to be mainly about read out speed. Unfortunately, this is a number that's not often measured in reviews (despite being of interest, as it quantifies the rolling shutter), and of late it seems people also confuse it with the flash sync speed on top of this. I understand these to be related, but not the same.

Measurements you can find for the a9 and R5 ( here and here) indicate that the a9 takes 1/150 s to read the full 24 MP sensor in fully electronic shutter mode, and the R5 takes 2.5 times as long at 1/60 s for the full 45 MP sensor. So the R5 is clearly slower, but it is a major jump for Canon and at 1.9 times the amount of pixels (and 3.8 times the amount of photosites) you could argue that the R5 actually almost keeps up with (or surpasses) the a9 in purely technical terms.

This is not saying that Sony's stacked designs isn't impressive. I'm just pointing out that Canon does not have to follow Sony's direction 1:1 to compete or beat them. There are a lot of rumors that the R1 and potentially even the high res R may see new sensor technology again.

I think we've seen by now that Canon is comfortable taking their sweet time to come up with their own implementations (IBIS, eye AF, video features, sensors,...) instead of jumping on what ever Sony is doing. Since they have not openly adapted stacked sensors yet, I believe they either have something they view as superior in the pipeline or will bring them once they do all they want without compromise. And cost is a compromise in this context.

You certainly raise valid points. Canon often does seem to be marching to its own drummers.

On the other hand, reports of patents on this site and CanonNews indicate Canon has been doing a great deal of work on stacked sensors. To my mind, some of these patents are detailed to point to productions efficiencies, yield, lowering cost. They actually make for fascinating reading.

Prior to IBIS showing up in the R5-6, a similar wealth of patents showed up detailing likely approaches to production of IBIS units.

Seems to me Canon is working on a stacked sensor, but working out their own methods of production and implementation. In doing so, perhaps they will find a way to have quad pixel AF and the torrent of data it creates and still have faster readouts like Sony. It seems to me global shutter is still a ways off, but something Canon is certainly working on and has managed to pull off in its Cinema line. At a very high price.

A clue to Canon's stacked sensor intentions is the recent deal where Canon licensed another companies tech (believe it was reported here) which would facilitate stacked sensor manufacture.

Who knows, the R5 turned out to be more than probably most of us expected. Maybe the R1 will take another big leap forward.

 

[Joules]

On the other hand, reports of patents on this site and CanonNews indicate Canon has been doing a great deal of work on stacked sensors.

Canon does so much with patents, it's really hard to draw conclusions on what is or isn't coming based on them. But they are certainly working on it and it will be exiting what the next round of Canon sensors brings to light. Certainly the R1 and high MP R (or APS-C, for that matter) are perfect candidates to show off with another advancement.