Canon executives say a lot more coming in 2019

Quarkcharmed

EOS 5DMkIV
Feb 14, 2018
565
465
Australia
www.michaelborisenko.com
And you would do so with a Sony A73 or Nikon D850 as well. No currently shipping camera has the DR necessary to cover the most extreme scenes which can be found.
It doesn't have to cover most extreme scenes. Again better DR simply means more scenes to cover in one shot.

What experience? Show me the pair of images you shot on a Canon 5D4 and Sony where the Sony had more recoverable detail vs. a small difference in noise in the recovered detail while pixel peeping.
I'm not shooting with Sony, and even if I were, would you require me to shoot hundreds of images a year with both Canon and Sony side by side?

You don't understand the tests at those sites if you think 1ev difference in scoring means anything other than slight noise differences in recovered detail. Look again at the S1 and 7D examples and that's a 3.5ev difference in scoring.
All those cameras, even the old ones, are 14 bit, but it doesn't mean they all have 14-stop DR. Lower DR simply means more noise all over the image, not just shadows. Of course the noise affects the shadows the most, but in older cameras you can see it in the light shadows and even mid-tones while the cameras with better DR push the noticeable noise to the deeper shadows. 3.5ev is actually a lot, try pushing exposure by 3.5ev and the shadows will turn into a mess in old cameras.
 

Quarkcharmed

EOS 5DMkIV
Feb 14, 2018
565
465
Australia
www.michaelborisenko.com
If the difference is >2ev and if you shoot wide DR scenes and if you shoot RAW and if you know how to recover the shadows in post...you'll get a few shots that would require HDR techniques on the lower DR sensor.
2ev is a lot, I often use 2 stop grad ND filter and it makes a huge difference. If a next-gen sensor had 2-stop DR improvement, it'd be fantastic, I wouldn't need to use the ND grad or would significantly reduce its use.

If you're talking about a 1ev difference you'll see a noise difference while pixel peeping.
1ev is plenty, it's often a difference between an 'ok sky' and a 'blown out sky'. Yes I'd like to minimise the use of HDR techniques, stacking etc.

So finally to the level Nikon was at in 2012. Silly Sony and their old sensor designs, when will they offer us something new? :LOL:
I simply showed the statement that Sony hasn't improved the DR in years is false. Now you're trying to play it down and call the spirit of Nikon. What does Nikon have to do with the statement about DR improvements in Sony cameras?
 

QuisUtDeus

EOS 80D
Feb 20, 2019
115
80
2ev is a lot, I often use 2 stop grad ND filter and it makes a huge difference. If a next-gen sensor had 2-stop DR improvement, it'd be fantastic, I wouldn't need to use the ND grad or would significantly reduce its use.
Are you going to complain if (when) Sony's next-gen sensor doesn't have a 2-stop DR improvement? Or will that be understandable.
 

Don Haines

Beware of cats with laser eyes!
Jun 4, 2012
8,230
1,841
Canada
I love how people say things like “Canon was forced to rush the camera to market”!

For starters, it takes about 5 years from the start of a project to get a camera to the store. About the only part of the project that can be rushed is software development, and doing that means buggy software. If anything, Canon is known for the stability of their products and there is no way that they are going to risk their reputation (probably their best marketing friend), to speed up the design cycle.
 

3kramd5

EOS 5D MK IV
Mar 2, 2012
3,083
404
I don't care as I don't use Sony cameras. And for them it'd be hard to have a 2-stop improvement. They'd need to go from 14 bit to say 16 bit for that. Canon still has a room for improvement within the 14 bit sensor architecture.
Sony’s next gen sensors do use 16-bit ADCs.

Here’s a camera which uses one:
 
Last edited:

Quarkcharmed

EOS 5DMkIV
Feb 14, 2018
565
465
Australia
www.michaelborisenko.com
Sony’s next gen sensors do use 16-bit ADCs.

Here’s a camera which uses one:
That's a niche camera, but if such sensor really appears in A7RIV (as it's rumored), it'd be very nice, that might make me switching to Sony. 16 bit ADC is going to give an unbeatable image quality and presumably a very good DR.
 

dtaylor

Canon 5Ds
Jul 26, 2011
1,416
880
It doesn't have to cover most extreme scenes. Again better DR simply means more scenes to cover in one shot.
You say this but then you say...

I'm not shooting with Sony, and even if I were, would you require me to shoot hundreds of images a year with both Canon and Sony side by side?
The advantage is so huge that you need to randomly shoot hundreds of images over the course of the year, side-by-side, hoping to find one that shows the advantage?

This just tells me that you do not understand how the DR scores at various sites translate into real world performance. You just like a higher score and you're going to tout it as a big advantage.

Lower DR simply means more noise all over the image, not just shadows.
This is literally the opposite of what it means.

3.5ev is actually a lot, try pushing exposure by 3.5ev and the shadows will turn into a mess in old cameras.
Read page 4 of this thread and tell me again how much a difference of 3.5 in the DR score of a testing site means.
 

3kramd5

EOS 5D MK IV
Mar 2, 2012
3,083
404
That's a niche camera, but if such sensor really appears in A7RIV (as it's rumored), it'd be very nice, that might make me switching to Sony. 16 bit ADC is going to give an unbeatable image quality and presumably a very good DR.
It is niche. So is the zwo using it. Cost might determine whether major market cameras use the RGB version.

I would not presume 16-bit ACD means anything about DR. They could very well be sampling the same range with better gradations.
 
Last edited:
  • Like
Reactions: Michael Clark

dtaylor

Canon 5Ds
Jul 26, 2011
1,416
880
2ev is a lot, I often use 2 stop grad ND filter and it makes a huge difference.
A difference of 2 in a DxO or PtoP score is not literally the same as a ND grad filter. This is the problem with reporting the test results in stops. It implies information is there on one sensor but not on another...as if a ND filter were used...when in fact the score is set by tripping an arbitrary noise threshold.

I simply showed the statement that Sony hasn't improved the DR in years is false. Now you're trying to play it down
Why not? Every Canon improvement is played down.

and call the spirit of Nikon. What does Nikon have to do with the statement about DR improvements in Sony cameras?
If you had read the thread you would know that Nikon set the bar in 2012. So if a Canon sensor is "old" for not having the best DR then so are Sony's sensors.
 

3kramd5

EOS 5D MK IV
Mar 2, 2012
3,083
404
...when in fact the score is set by tripping an arbitrary noise threshold.
That’s not entirely true. The measurement is bounded by an arbitrary noise threshold on the low end, but is unbounded at the high end.
The Nikon D810 seems to have improved well capacity.
 
Last edited:

Quarkcharmed

EOS 5DMkIV
Feb 14, 2018
565
465
Australia
www.michaelborisenko.com
The advantage is so huge that you need to randomly shoot hundreds of images over the course of the year, side-by-side, hoping to find one that shows the advantage?
I never said it was a huge advantage. But you asked me to show you a pair of images taken with Canon and Sony - I don't have them.

This just tells me that you do not understand how the DR scores at various sites translate into real world performance. You just like a higher score and you're going to tout it as a big advantage.
Nope, maybe you remember from another old thread, I'm comparing old Canon cameras to 5DIV, the ones I was actually shooting with. And what's shown as 1.5-2 stop difference in DR (say between 70D and 5DIV) translates into a very noticeable significant difference in IQ.

Read page 4 of this thread and tell me again how much a difference of 3.5 in the DR score of a testing site means.
As above, my practical experience with what's shown as 2 stop difference tells me it's a huge gain.

I worked with older Canon cameras and with raw files from other models, but used 70D and 5DIV the most.
 

Quarkcharmed

EOS 5DMkIV
Feb 14, 2018
565
465
Australia
www.michaelborisenko.com
As I illustrated earlier it's not a huge gain at 3.5.
As far as I remember, your illustration was a single shot from 7D. I have literally thousands images from 70D and 5DIV. Never shoot with them side by side, but you didn't show us a side-by-side comparison either, and at the same time I have lots of shots taken in similar conditions. The images from 5DIV are much cleaner across shadows, midtones and highlights, give much more freedom for editing etc.And again that's 2-stop difference according to http://www.photonstophotos.net/Charts/PDR.htm#Canon EOS 5D Mark IV,Canon EOS 70D. That's my experience but that's what matters to me when I think about spending money on the next camera.

I've seen RAWs from 200D, same sensor as 80D, they're also way better than 70D and can serve as a backup landscape camera, but still lag behind 5DIV and it's not just the numbers, it's the personal experience.

Because I know how my experience translates to the numbers from photonstophotos, I can do an educated guess what I get if a new camera is 1 stop better than 5DIV.
 

dtaylor

Canon 5Ds
Jul 26, 2011
1,416
880
As far as I remember, your illustration was a single shot from 7D.
And? That's the level of shadow recovery you can expect. Nothing can make that less. If someone gets less that's not on the sensor.

The images from 5DIV are much cleaner across shadows, midtones and highlights,
The 5D4 is a full frame camera. How clean a bright blue sky or a mid tone green dress appears is dominated by the format difference and the impact that has on Poisson noise. A 6D2 will also show a cleaner mid tone or highlight.

How hard you can push the shadows is dominated by read noise. The DR tests are tests of read noise.

The bitness of an ADC imposes an upper limit to DR, a 14-bit ADC cannot have more than 14 stops of DR etc.
The ADC conversion is not perfectly linear.
 

Quarkcharmed

EOS 5DMkIV
Feb 14, 2018
565
465
Australia
www.michaelborisenko.com
And? That's the level of shadow recovery you can expect. Nothing can make that less. If someone gets less that's not on the sensor.
Ok. Let's look at your sample (https://www.canonrumors.com/forum/index.php?threads/canon-executives-say-a-lot-more-coming-in-2019.36985/post-771327) closer.
First of all, 5DIV is able to recover totally (visually) dark shadows, in your sample all shadows have detail before recovery. Second, why do you think you've recovered 3.5 stops? I've loaded your sample into PS and tried to match just the exposure of the shadows (see the attachment). I applied CameraRaw EV+2.55 to the left part:
183868


As you can see, the shadows are roughly matched and it's just 2.55 stops.
 

dtaylor

Canon 5Ds
Jul 26, 2011
1,416
880
Second, why do you think you've recovered 3.5 stops?
Where did I say that I recovered 3.5 stops?

I posted that image to make a point to the DRones on this board who miss no opportunity to say that the 5D4 and R sensors are old tech that are so far behind Sony on DR that we're going to switch brands any day now. I replied to a posting of a Panasonic S1 shadow recovery example by someone who kept talking about how great it was to have a "new competitive sensor" implying that the 5D4 sensor is old and uncompetitive.

A D850 / A73 / A7r3 / S1 score roughly 1ev higher than a 5D4 (give or take a fraction of a stop). The S1 scores 3.5ev higher than a 7D. With that kind of score and all the DR rhetoric that appears on this forum you would assume a vast visual difference in shadow recovery between a S1 and 7D. Yet the visual difference is relatively small. If that's what a 3.5 score difference equals in the real world, then a difference of 1 can't be seen outside of pixel peeping conditions.

If I had immediate access to a 5D4 at the moment I would fire off some shots tomorrow against a D800E to put an exclamation point on this. I can borrow the D800E any time I want but can't easily borrow a 5D4 for a few hours (at the moment).

Suffice it to say I've done these tests before between various camera pairs and have a very good idea of how those scores translate to real world prints and the sensors in the 5D4 and R are not behind. They are dual pixel architecture which costs them a point at DxO or PtoP but that doesn't translate to anything substantial in the real world. And if you really want to quibble about it, shoot DP RAWs and process them to add roughly 1ev on the highlight side putting those cameras on par with anything else out there, literally and to the numerical point score. Better even because most photographers would take 1-2 stops from the shadow side and put it on the highlight side if they could. It's the one lingering complaint from the film days since film by nature handles highlight roll off very well.

The "Canon sensor old" meme is old and boring.

BTW: the shadow push was roughly 2.5ev but there was also 0.5ev of highlight recovery. That was why I winked in the original post and said 'but I held the sky.'
 

3kramd5

EOS 5D MK IV
Mar 2, 2012
3,083
404
The bitness of an ADC imposes an upper limit to DR, a 14-bit ADC cannot have more than 14 stops of DR etc.
Let me rephrase:
I would not presume the use of a 16-bit ADC means the sensor has either higher well capacity or lower noise than some earlier sensor with a 14-bit ADC.
 

transpo1

EOS 7D MK II
Jan 12, 2011
750
97
Half the truth is often a whole lie, isn't it?
Half-truth you said was: "Netflix requires 4K for all new programming".
The other half that you forgot to mention is: "Minimum data-rate of Bitrate of 240 Mbps at 23.98 fps" (REF)
What is the max bitrate of Sony MILC cameras (that you said earlier your pro crew use for film making): 100mbps in 4K.
And interestingly, on the same reference, I don't see any MILC of any brand listed as approved camera by Netflix for content creation and delivery!
So who are the pros who use MILC for film making anyway? and what they do with their contents created? Watch it in basement with their grandma?
Now who doesn’t understand about film making in general?
Edit: Same goes with your market analysis statements!:unsure:
Well, denigrating all those who don't have a show on Netflix as "in the basement with grandma" might not win you any friends in the professional community, so I think we can tell who does not hang out with professional crews.

The full truth is that Netflix approved formats only consist of 4K formats: "Camera must have a true 4K UHD sensor (equal to or greater than 3840 photosites wide)." So that point is not able to be disputed.

However, it is important to note that while no DSLR or MILC makes the approved cameras list, Netflix does allow up to 20% for unscripted or 10% of unscripted to be shot on unapproved cameras (B cams, crash cams, etc.), which would include such cameras (if they like your project enough).

Full disclosure: This may also include non-4K cameras converted to 4K for delivery, but I'd wager that the Netflix executives, who have ultimate discretion, would be more likely to accept an non-approved camera if it were shooting 4K, and even more so if it were using the full width of the UHD (4K) sensor. Which is not something the EOS-R does because of its crop in 4K. And this is where material shot on a Sony, a Fuji, or a GH5 might have the advantage.

Also, here's a definition for you because some people obviously aren't intelligent enough to understand the concept: "Professional: A person engaged in a specified activity, especially a sport or branch of the performing arts, as a main paid occupation rather than as a pastime."

There are all sorts of professional filmmakers. Some use Alexas. Some use a Sony A7SII to film a documentary. And some use a GH4 to shoot segments of Planet Earth II.

And, since my point was that quality 4K is relevant and necessary in today's world, congratulations on providing further research to prove that. Please take note, however, that although anyone can Google format specifications, it actually takes a filmmaker's perspective and a brain to analyze them. And I didn't see that anywhere in your post.