Review: Canon EOS R3 final review by DPReviewTV

I like real world reviews more than spec reviews. honestly i just look at the high iso noise section in spec reviews. seeing a person use it means more. I pay for quality and convenience.
Yes, of course you still need real world, hands on experience for the qualitative aspects of the camera. But you shouldn’t really apply the same attitude to the aspects of the camera that can actually be measured more objectively. Hence me highlighting their ‘impression’ of the noise levels of the camera (which you also mentioned being a useful component of technical reviews).

example: many forum users here are seeing less (or equal) noise in the R3 files compared to R6 and S5 that were shown as points of comparison in the review. Who is right? Perhaps we would all have a better idea of the signal to noise ratio if it was actually measured.

Obviously DPReview is not aiming for this level of thoroughness in the technical aspects of their review… that’s fine… but I personally need more meaningful detail in something being called a ‘final review’…. Perhaps the written version will be much better…
 
Upvote 0
Maybe we will have such a feature in a R1 where we can switch between 80 vs. 20 MP?? Pure speculation, but would be a great feature to have.
The R1 wishlist includes quad pixel AF, global shutter (no mechanical shutter), fast eshutter sync speed, pixel shift high res, and~20mp on-the fly over sampled (no lossy compression/cRAW/S-RAW) at full 30fps unlimited buffer. Best of both worlds.
The question is whether am down-sampled image would be raw or only for jpgs. From my understanding, professional sports tend to output jpg and not raw for so this could be an option.
 
  • Like
Reactions: 1 user
Upvote 0

unfocused

Photos/Photo Book Reviews: www.thecuriouseye.com
Jul 20, 2010
7,184
5,483
70
Springfield, IL
www.thecuriouseye.com
...But you shouldn’t really apply the same attitude to the aspects of the camera that can actually be measured more objectively...
...example: many forum users here are seeing less (or equal) noise in the R3 files compared to R6 and S5 that were shown as points of comparison in the review. Who is right? Perhaps we would all have a better idea of the signal to noise ratio if it was actually measured...

I get what you are saying. But on the other hand, as long as it looks good to you, what difference do the measurements make? Disregarding scientific or forensic applications where absolutes might matter, if the measurements show differences that are not visible to users or those viewing the images, it doesn't really matter.
 
Upvote 0
Jul 21, 2010
31,099
12,863
The R1 wishlist includes quad pixel AF, global shutter (no mechanical shutter), fast eshutter sync speed, pixel shift high res, and~20mp on-the fly over sampled (no lossy compression/cRAW/S-RAW) at full 30fps unlimited buffer.
…and a partridge in a pear tree. It’s a lovely list, and I’m sure we all wants the one camera to rule them all, my Precious. I’d recommend you start preparing now for the forthcoming disappointment. ;)
 
  • Like
Reactions: 4 users
Upvote 0
I'm tired of these talking head reviews without backing up their statements. They should be more like Jared Polin who shows you pictures and videos. Heck, he even provides the RAW files. And I don't understand why DPReview can't get the equipment for more than a day or so. They even split usage over the course of a day! All these other reviewers had weeks with it.
 
Upvote 0
…and a partridge in a pear tree. It’s a lovely list, and I’m sure we all wants the one camera to rule them all, my Precious. I’d recommend you start preparing now for the forthcoming disappointment. ;)
Well, in my defense, I did say wishlist and whilst I am not in the target audience for it....

The R1 needs to be better than the R3 and the R5. How they determine that is the question but it is fun to speculate. :)

Neither the A1 or the Z9 can do high res @ 30fps in raw. Is this a processor limitation or bus speed or heat or?
Could dual Digic X processors provide enough grunt for downsampling 30fps @ 45mp to ~20MP jpg?
The Z9 has demonstrated removing the mechanical shutter so they are prepared to wear the DR hit (if that is the reason).
The R3 has a stacked sensor so it is reasonable to assume that a stacked sensor will also be used in a R1. A global shutter may effectively be in place with a fast enough stacked sensor.
Pixel shift could be a firmware update for R5/R6/R3. It isn't clear why Canon hasn't included it when implementation shouldn't be major challenge given the competition
QPAF is the logical extension to DPAF improving AF for horizontal lines - as you have mentioned. It may be impossible to implement or just expensive.

Ultimately, we are getting to minor differences at the pointy end. What would Canon marketing have to announce?
They would hope that there is at least one new feature not available on the A1/Z9/R3/R5.
What feature would tempt you to upgrade from your new R3 to a R1?
 
Last edited:
Upvote 0
Are we kidding with these new ads??? The site is next to impossible to navigate now. I could at least use the site a few days ago, but now I can’t even press the Menu without going in an add. The top and bottom ones are hopelessly in the way…. Jeez….
Forkert tråd å klage i, Viggo.
 
  • Like
Reactions: 1 user
Upvote 0

Del Paso

M3 Singlestroke
CR Pro
Aug 9, 2018
3,301
4,187
Yes, of course you still need real world, hands on experience for the qualitative aspects of the camera. But you shouldn’t really apply the same attitude to the aspects of the camera that can actually be measured more objectively. Hence me highlighting their ‘impression’ of the noise levels of the camera (which you also mentioned being a useful component of technical reviews).

example: many forum users here are seeing less (or equal) noise in the R3 files compared to R6 and S5 that were shown as points of comparison in the review. Who is right? Perhaps we would all have a better idea of the signal to noise ratio if it was actually measured.

Obviously DPReview is not aiming for this level of thoroughness in the technical aspects of their review… that’s fine… but I personally need more meaningful detail in something being called a ‘final review’…. Perhaps the written version will be much better…
Exactly what I meant.
"Real life" reviews by professional users PLUS a technical review by TDP's Brian. He is (my opinion) trying hard to remain objective, no matter which brand he is testing. Even though Canon friendly, he doesn't hesitate to praise a Nikon or a Sony.
 
Upvote 0

GoldWing

Canon EOS 1DXMKII
Oct 19, 2013
404
279
Los Angeles, CA
en.wikipedia.org
Saw that the other day when they posted. I'm a little underwhelmed, they titled it a Final Review, but it's not a review at all but a quick update to their earlier preview. 'll save y'all some time:

1. "I'd buy it over a 1DX III, but it's meh because 24mp".
2. "It's the only camera that will shoot full raw at 30fps, but it's meh because 24mp"
3. "It's got a great buffer, I like the AF performance, but it's meh because 24mp"
4. "It's expensive for what it is, because 24mp"

I usually find them reasonably on point in most instances. For me, this time around they come across as though if it ain't 50mp, it ain't sh*t.
They could have done a brief review in just two symbols

Z9
 
  • Like
Reactions: 1 user
Upvote 0
Apr 25, 2011
2,510
1,885
I'd hoped for that with the R3 but I'll save my money for the R1 but only if Canon doesn't release a 12mp R1 that shoots really fast poor quality images. It's funny how the competition can produce a high resolution camera but Canon can't?
DPAF is quite taxing when it comes to sensor data throughput.

For QPAF in R1, given the current state of the art, one could as well expect a 12-megaquadpixel sensor.
 
Upvote 0
Jul 21, 2010
31,099
12,863
DPAF is quite taxing when it comes to sensor data throughput.

For QPAF in R1, given the current state of the art, one could as well expect a 12-megaquadpixel sensor.
Might not be quad pixel, could be dual pixel with alternating rows split in orthogonal orientations. Canon has a patent on that the published a while back. Would accomplish the same goal with a lower computational load.
 
  • Like
Reactions: 2 users
Upvote 0
Apr 25, 2011
2,510
1,885
Might not be quad pixel, could be dual pixel with alternating rows split in orthogonal orientations. Canon has a patent on that the published a while back. Would accomplish the same goal with a lower computational load.
Not exactly the same goal: the focusing precision would be cut by half in at least one direction.
 
Upvote 0
Jul 21, 2010
31,099
12,863
I am suggesting that with DPAF the focusing precision is limited to the pixel pitch.
Aside from the fact that you do not seem to understand the difference between precision and accuracy (you’re not alone there, Canon’s marketeers use precision when they mean accuracy), DPAF does not determine the phase difference with a single pixel.
 
Upvote 0
Apr 25, 2011
2,510
1,885
Aside from the fact that you do not seem to understand the difference between precision and accuracy
What makes you think so?

DPAF does not determine the phase difference with a single pixel.
I have never claimed it does.

However, when you only sample every second pixel in the vertical direction for the vertical phase difference, you shouldn't expect to be able to achieve pixel-precise focusing on horizontal lines in the general case.
 
Upvote 0