The Canon EOS R6 Mark III is “definitely” coming in Q1 of 2025

The original R6 had the same sensor as their flagship 1DX mark iii
I have read that, and it seemed to be the case except...in a recent discussion of AF points/areas I noted that the 1D X III has 3,869 selectable AF points for stills in Live View, while the R6 has 6072 selectable AF points for stills. So...is it the same sensor? I'm no longer certain.
 
  • Like
Reactions: 1 user
Upvote 0
It's truly remarkable how often, @SvO2, the following DR plots have to be posted here to show your point.
I wasn't actually gonna reply to his post, but good thing this is the world wide web, and we all have access to the information.

First of all, you've just compared a camera shooting at 12 bit RAW (R6 Mark II) with another shooting at 14 bit RAW (R5 Mark II).

But that's not your fault, because your source claims that my camera shooting at 12 bit RAW (ES) has pretty much the same dynamic range as when shooting at 14 bit RAW (MS or EFCS), from ISO 1600 and beyond, and that couldn't be farther from the truth. I've been shooting the R6 for almost four years, and I specifically began changing to MS or EFCS when shooting at high ISO, once I discovered how much better the performance is (almost two f-stops better).

And they state the same for other models that drop to 12 bit on ES as well. That tells how reliable that source is.


The dxomark values definitely match a lot more the experience I have with these cameras. Those don't.


I said before, in this forum, that I've been waiting for a R6 with the ability to shoot 14 bit RAW in ES, and the dynamic range on high ISO with electronic shutter is the biggest reason (and rolling shutter, of course). In addition to my desire, I recently began shooting at a venue that requires silent cameras, the light there is crap, and it's been a PITA having to shoot at high ISOs needing high dynamic range, due to the lights in the ceiling being pointed directly at the musicians. I wouldn't have issues with mechanical or EFC shutters, and sometimes there's one or two spots where I can use them, but generally they don't allow me to make noise there.
 
  • Like
Reactions: 1 user
Upvote 0
First of all, you've just compared a camera shooting at 12 bit RAW (R6 Mark II) with another shooting at 14 bit RAW (R5 Mark II).
That's true, the R5II has 14-bit ES mode, however the original R5 with 12-bit ES still does better than the R6 in ES:
1733015930157.png

I've been shooting the R6 for almost four years, and I specifically began changing to MS or EFCS when shooting at high ISO, once I discovered how much better the performance is (almost two f-stops better).
How have you measured those two stops? It's technically impossible. The ES can be in theory up to 2 stops worse than EFCS at ISO 100 - in practice is a bit less than that due to misc. reasons. But at high ISOs the difference becomes less because of how amplification works - there's simply no information to fill 12 bits per pixel with required granularity, so 12 or 14 makes very little difference.

The dxomark values definitely match a lot more the experience I have with these cameras. Those don't.
DxOMark don't even measure in ES mode as far as I can tell.
 
Upvote 0
DxOMark don't even measure in ES mode as far as I can tell.
the original R5 with 12-bit ES still does better than the R6 in ES
I never found that to be true in any mode whatsoever. I've shot the R5 many times and I always found the files to be very similar, but slightly inferior to those of the R6.

How have you measured those two stops? It's technically impossible. The ES can be in theory up to 2 stops worse than EFCS at ISO 100 - in practice is a bit less than that due to misc. reasons. But at high ISOs the difference becomes less because of how amplification works - there's simply no information to fill 12 bits per pixel with required granularity, so 12 or 14 makes very little difference.
Sorry, I didn't mean to say that I measured two extra stops of DR, I can't measure that, you're right.
I meant that I have about the same performance at say ISO 8000 using MS or EFCS as with ES at about 3200, or 6400 vs 1600 on each side. Noise levels drop dramatically when lifting shadows/pulling highlights, and I can go much further. There's a lot more information on the file at 14 bit RAW. It's nothing like Photons to Photos says it is.
I wouldn't say it's exactly 2 extra f-stops of additional performance, but it's pretty close to that.
 
Last edited:
Upvote 0
Well, I really like my R6ii. It's by far the most rounded and capable camera body I've ever owned. 24mp, high DR, great iso, fast 12fps 1st curtain and 40fps electronic....what's not to love? It's about the photos not the specifications. I've been really impressed with the R6ii, I've taken so many portfolio worthy photos this year. Some of my best work for sure.
Exactly the same conclusion here: The sum of single specs and ergonomics is good, but the system as a whole is excellent! Including the sheer image quality which is superb. 24 million pixels which have excellent single pixel quality!
 
Upvote 0
R6 ii is ~1800€ including an additional battery and +20€ will add an adapter EF→RF!
I think it is due to replacement.
I would like to buy a 2nd body but I think I am in a relaxed situation:
→ Buy a 2nd R6 ii while the R6 iii is out at similar prices or
→ Buy a 1st R6 iii if it is ergonomically compatible and
gives me some additional value (global shutter ... just kidding)
in terms of video speed (faster readout), higher 4k frame rates,
off standard frame rates (would be really great).
 
Upvote 0
R6 ii is ~1800€ including an additional battery and +20€ will add an adapter EF→RF!
I think it is due to replacement.
I would like to buy a 2nd body but I think I am in a relaxed situation:
→ Buy a 2nd R6 ii while the R6 iii is out at similar prices or
→ Buy a 1st R6 iii if it is ergonomically compatible and
gives me some additional value (global shutter ... just kidding)
in terms of video speed (faster readout), higher 4k frame rates,
off standard frame rates (would be really great).
That's a really great price for a really great camera!
 
  • Like
Reactions: 1 user
Upvote 0
I will buy the R6iii if it has a top screen. Otherwise I am waiting a few years and will get an r5ii, used.

The r5ii is perfect for me, apart from the price, but my 5d4 is still okay. The only upgrade that would encourage me to purchase much sooner is 1.3x and 2x crop raw files, so I can compose different shots with prime lenses, e.g. 85 f1.4 becomes 170mm f2.8 equivalent, or 35mm f1.4 can work as 35, 45, 56 and 70mm at the push of a button, similar to the Leica Q series.
 
Upvote 0
I will buy the R6iii if it has a top screen. Otherwise I am waiting a few years and will get an r5ii, used.

The r5ii is perfect for me, apart from the price, but my 5d4 is still okay. The only upgrade that would encourage me to purchase much sooner is 1.3x and 2x crop raw files, so I can compose different shots with prime lenses, e.g. 85 f1.4 becomes 170mm f2.8 equivalent, or 35mm f1.4 can work as 35, 45, 56 and 70mm at the push of a button, similar to the Leica Q series.
As long as you're already shooting raw, what prevents you from cropping in post? Also, there's already 1.6 crop mode in the R5 and R5II.
 
  • Like
Reactions: 1 user
Upvote 0
I never found that to be true in any mode whatsoever. I've shot the R5 many times and I always found the files to be very similar, but slightly inferior to those of the R6.

I don't have an R6, but this comparison tool from DPR ("comp" view) shows they're basically identical
1733056952562.png

if you compare them 1:1 ("full" view), yes the R5 will show more visible noise
1733057057294.png

but that's because the R5's image is magnified more in this view. But because the cameras have different resolution, the comparison should be done in the "comp" view which normalises the images to the same target size.
Sorry, I didn't mean to say that I measured two extra stops of DR, I can't measure that, you're right.
I meant that I have about the same performance at say ISO 8000 using MS or EFCS as with ES at about 3200, or 6400 vs 1600 on each side. Noise levels drop dramatically when lifting shadows/pulling highlights, and I can go much further. There's a lot more information on the file at 14 bit RAW. It's nothing like Photons to Photos says it is.
That would have been a 2-stop difference from your description, but it doesn't match my observations. Tbh I'm too lazy to shoot the tests myself but again you can find samples for the R6 (ES and EFCS) on DPR.

At ISO 1600, the noise is almost identical (ES is slightly worse)
1733058170960.png

The issue is that writing raw files to a container with 14 bits per pixel doesn't mean there's actually 14 bits of usable information per pixel.
 
  • Like
Reactions: 1 users
Upvote 0
I never found that to be true in any mode whatsoever. I've shot the R5 many times and I always found the files to be very similar, but slightly inferior to those of the R6.


Sorry, I didn't mean to say that I measured two extra stops of DR, I can't measure that, you're right.
I meant that I have about the same performance at say ISO 8000 using MS or EFCS as with ES at about 3200, or 6400 vs 1600 on each side. Noise levels drop dramatically when lifting shadows/pulling highlights, and I can go much further. There's a lot more information on the file at 14 bit RAW. It's nothing like Photons to Photos says it is.
I wouldn't say it's exactly 2 extra f-stops of additional performance, but it's pretty close to that.
I love the smell of confirmation bias in the morning. No, actually I prefer coffee but this morning I have both.
 
Upvote 0
I will buy the R6iii if it has a top screen. Otherwise I am waiting a few years and will get an r5ii, used.
I doubt it will


I love the smell of confirmation bias in the morning. No, actually I prefer coffee but this morning I have both.
Glad to fill your day with joy, sir! But we're past lunch here already :ROFLMAO:

Test charts and sensor tests are nice-to-have references, but unless someone is using their devices improperly, I'll take real world usage experience anyday, specially if my own. After all, that's what clients are paying for, test charts don't sell much.

That would have been a 2-stop difference from your description, but it doesn't match my observations.
Yeah, like I said, I don't think the difference reaches 2 stops, but it's close.

The issue is that writing raw files to a container with 14 bits per pixel doesn't mean there's actually 14 bits of usable information per pixel.
I know, I understand that.

After replying to you, actually I thought of something else, which is...does this dynamic range difference come from the fact that a given camera drops the file to 12 bit, or is it due to the nature of the electronic shutter?
If it is the latter, then shooting a 14 bit RAW with electronic shutter, on the same given camera, would make no improvement. Testing that would require taking to field a R3, R5 II, R1 or maybe this future R6 Mark III, of which I own none.
 
  • Like
Reactions: 1 user
Upvote 0
Test charts and sensor tests are nice-to-have references, but unless someone is using their devices improperly, I'll take real world usage experience anyday, specially if my own. After all, that's what clients are paying for, test charts don't sell much.
+2. On other platforms I've been insisting "worse than competition" in test charts doesn't mean the camera is bad, it's still more capable than you can handle. But spec warriors will just label me as xxx-fanboy because I "defend" a product.
 
  • Like
Reactions: 1 user
Upvote 0
Test charts and sensor tests are nice-to-have references, but unless someone is using their devices improperly, I'll take real world usage experience anyday, specially if my own. After all, that's what clients are paying for, test charts don't sell much.
Real word usage experience is good until you claim that "there's a two stop difference between ES and EFCS at ISO 1600". Then you have to tell us how exactly those two stops were measured.

Yeah, like I said, I don't think the difference reaches 2 stops, but it's close.

So how did you measure it exactly? I provided samples above that show almost the same noise at ISO 1600 with EFCS vs ES. There is a minor difference but definitely not 2 stops, not even close to two stops.

After replying to you, actually I thought of something else, which is...does this dynamic range difference come from the fact that a given camera drops the file to 12 bit, or is it due to the nature of the electronic shutter?
There's electronic shutter working in all shutter modes, including full mechanical. It's always 'electronic' because readout goes sequentially row by row (or by small groups of rows). In the pure electronic mode without the mechanical curtains the readout speed may be increased which may result in an increased read noise. However, the major contributor is the reduced bit depth.
If it is the latter, then shooting a 14 bit RAW with electronic shutter, on the same given camera, would make no improvement. Testing that would require taking to field a R3, R5 II, R1 or maybe this future R6 Mark III, of which I own none.
Those tests are typically done not in the field but in a controlled environment, because you need to capture the same scene with the same illumination, using the same angle of view, preferably the same lens and exact same exposure settings. That's required to eliminate or reduce the contribution of unknown variables on your measurement.
 
Last edited:
  • Like
Reactions: 1 user
Upvote 0