Big Megapixel Development Announcement in the Fall? [CR2]

[Stu_bert]

Its about time Canon responded to the Nikon D800 and D800E. This fixation with high ISO, low DR, and high noise needs to stop. We need a quality camera to bring back the 1Ds range, a camera that is best in class.

For whatever reason Canon have been asleep at the wheel for a while now and its time they woke up. I have no wish for ISO extremes, nor do I shoot video at all, but I do shoot landscapes, so want a camera that has a minimum of noise and world beating DR. Maybe removal of the anti-aliassing filter?

It's a valid fixation. There are many more sports and wedding photographers than landscape photographers. Hence why Canon has dominated the market.

High ISO if clean is great for landscape shots when movement is not desirable - for instance freezing stars without wishing to get star trails. If you're taking shots from a moving plane, then faster speeds are essential (>1/1000th is ideal). Add in the desire to shoot in the golden hour, and suddenly higher iso is useful. Finally, as has been mentioned, not having to take a tripod everywhere opens up flexibility - although I appreciate that may be negated by the higher resolution.

1Dx bodies are also perhaps better in harsher conditions - be that cold, wet or sand, all often encountered by landscape photographers.

I thought as mentioned in other threads, Canon's latest L glass is not sensor limited. And certainly not by a 40MP sensor.

Finally, as also mentioned elsewhere, higher MP resolves the subject detail better.

Cropping is useful, even for landscapes, where you can't change your position or zoom - for a variety of reasons.

The simple conclusion is of course, everyone has different needs. And sure, eventually, Canon will try and satisfy them all, but they're never gonna keep everyone happy...

But then if they did, these forums would be a lot quieter ;D

That's great! However, it doesn't have anything to do with why Canon has neglected a high MP body. The reason is plain and simple. High ISO/high shutter/super AF goes to sports and wedding photogs. Not landscape photogs. Take all the shooters, especially pros. What would you guess? 98% wedding/sports, 2% other? That's all great that everyone has different needs, but is beside the point.

I don't think that's the whole story. You can say that high MP/DR sensor is also for status symbol, bragging rights. If you really want to be the market leader, you need to prove that you have the best or at least will compete with the best, no matter the arena. Sometimes, it's all about reputation. Yes, for most of us, that's not how we see it since some things are really trivial and there are other things that are more important but on the business side for Canon, there's a reputation they must maintain.

Reputation? You mean the one they have supported by numbers, money, and sales? Oh that one!

"If you really want to be the market leader..." you say. Well, guess what. They are!

You can never rest on your laurels... That's what I'm saying. Clear?

No not really, but ok. If you think producing a super high MP camera that only about 1-2% of the DSLR users will purchase is "stepping it up" then that's cool. However, the majority of camera users are NOT asking for a high MP camera. I think it will be produced in small quantities, be produced for a short time, and won't be updated for an even longer time after that. Also, like the 1Ds Mark III, there won't be many units sold. Look at the sales of 5D Mark II vs. 1Ds Mark III. Even most of the pros I knew bought the 5D2 over the 1Ds 3. Too pricy for not noticeable enough features over the other.

Guess I bucked that trend instead, bought my 1Ds MK III second hand and never looked back in comparison to the 5D II

As for high MP - sure the market for MF style resolution is far lower than for "traditional resolution" dSLR, but again, I'd be interested to know how good / bad the D800 has sold for Nikon as this is perhaps the best barometer...

Given the sensor density for high MP is not significantly different from that of the current APS-C sensors, then frankly, why not? You get better resolving power and in a FF format

Again, everyone's mileage is different....

 

[art_d]

With regards the Bay image, seeing as how you yourself say

"Granted, I don’t run into such issues very often"

and you did successfully make a 20"x30" print to your satisfaction from it, I don't see that the Canon let you down. Was it more work to achieve your intended goal than a Nikon might have been? Yes. But that wasn't my request.

I thought you wanted to see circumstances where limited dynamic range in Canon sensors has a real world impact. That is what I have presented The fact that those issues are infrequent does not mean they are insignificant. If I am on a photoshoot and I need a specific shot, then it doesn't matter how often or not it happens. It is happening now, and I need to deal with it.

Another case. I was shooting an interior design job where the designer decided towards the end of the shoot he wanted to use someone as a model in the scene. The problem was there was a bright window on one side, and a furniture arrangement creating a lot of dark blocky shadows which will make for an ugly scene. Lifting the shadows in post is not an option. Exposure blending is not an option anymore because there's no way you can get a model perfectly still across multiple exposures. So I have to go down to the car, haul up my lighting gear, and fill in the shadows with strobes. There wasn't supposed to be a model at the shoot, it just occured by happenstance, and the only thing that saved me was that I always pack extra lighting gear "just in case." But if I had not, there was no way the Canon could have made that photo.

So it may be not be needed often, but that doesn't mean it wouldn't be hugely beneficial to have more dynamic range.

 

[Sporgon]

@privatebydesign, I can't quote you 'cos I'm on my iPhone, but you mention a serious landscape photographer not using one FF frame even if it is 36mp.

This is quite right, high mp on a smaller format will never equal a larger format in landscape photography because your subject or detail in the picture will only cover a very small areas of the sensor, ie the light image projected onto the sensor will be tiny, because the detail is not close to the camera, whereas on a larger format the actual capture is larger. So MF or LF is always going to beat a smaller format in these circumstances. ( This is why when people test FF against APS and fill the frame with a subject close to the camera they see no difference - because in this scenario there is essentially no difference).

Digital has allowed us to easily stitch frames together to mimic a larger sensor, the advantages you gain are the same because each section has a larger image on the sensor. So a 13 mp 5D for instance, stitched from four vertical sequences, will produce a higher 'IQ' than a single frame 36mp FF, because the 5D has been turned into a larger format, - about 36x 90 in this instance, about the same size as the old 645 film medium format.

For applications where the subject detail is larger in the frame, 18 or 22 mp is going to be enough to make enormous prints, so the 36 or 50 mp FF is a little in no man's land. Just like the D800, a gimmick. And you want the E? just add unsharp mask to the regular one.

I think Canon know this. Canon are quite good at avoiding the fluff on their higher end cameras.

"Does Sir require a little pop up flash to fill in his 36 mp ? May I direct Sir to Nikon"

40 mp belongs on a larger format.

 

[art_d]

What you said was

" I understand now your rationale, and I'm not saying you were trying to be sneaky or duplicitious...but I don't think it is reflective of the difference between screen and print."

The word now implies that previously you didn't, hence my surprise. Now you don't, but previously you did.

I don't agree that an image projected on a screen accurately represents what you see in a print just because it is the same size; and some people could be misled into believing that the 700 px image you posted can be printed 7 inches wide and look that crisp. But I was trying to make clear that I didn't want to insinuate that you were being duplicituous in any way. Either then or now.

And that is why I explained what I did, have offered to post the 1680 pixel crop pre downsample, and have offered to send anybody my full original file and the print file. You all seem to find this so hard to believe, have you never used a well set up Print View in PS? Doesn't your screen give you a pretty accurate WYSIWYG?

Yes, thank you for the clarification. I haven't found anything you say hard to believe. I don't use Print View for judging fine detail in a print, no. But some people feel it is suitable for them, and that is fine.

 

[EvilTed]

Maybe by then, Intel will have a superfast processor and SATA express will be out. Current SSD's are limited by SATA III to about 500mb/sec, and huge image files take a lot of time to process.
At least Canon has options for sraw that you can select when you don't want full blown resolution. With the D800, you get those huge files every time if you want to use raw.
That way a user would have a choice.

Yeah but sRAW isn't really all that RAW and with Nikon you can use crop mode FULL RAW which is great for wildlife since you maintain reach while not wasting storage on all the outer border areas so I'd actually WAY rather they went to the Nikon way of handling it.

I'd rather it be 39MP for perfect video and keeping 6fps than 47 or 50+ and being less than 5fps and having worse video.

Sata 3 has a limit of 600MB/s not 500mb/s.
There is a huge difference!

Megabyte per second
(not to be confused with Mb/s – Mega bits per second) A megabyte per second (MB/s) is a unit of data transfer rate equal to:
8,000,000 bits per second, or
1,000,000 bytes per second, or
1,000 kilobytes per second, or
8 megabits per second.

ET

 

[Sabaki]

Hey guys

Seems that the most desirable quality of the D800 would be the impressive dynamic range it offers.

I'd just like to ask, you folks think cameras will ever reach the DR our eyes and brains could resolve and if so, the impact it could have on photography?

Thanks peeps

 

[KyleSTL]

Seriously, some guy just said "very rarely do people shoot in high iso"....are you kidding me?

To be fair, what he said was above ISO 6400. I suspect he's true, only because until the most recent Canon bodies, shooting above ISO 6400 gave unusable results. That's not true anymore. On my 1D X, ISO 6400 is the new ISO 800.

I am in the minority as I never shoot at ISO 6400 (because that ISO setting does not exist in the 5DC).
I know I can buy a camera today that produces wonderful results (for my purposes) at ISO 6400, and I would be thrilled to own one of them (6D, 5D3, D600, D800). I have no desires to have a camera more advanced or feature-rich than the 5D Mark III. I would also be happy with the 6D. My hobby and my disposable income do not allow me to invest in a 5D3 or 6D, even though I know I would be completely content with either one. In my case my upgrade is not limited by what is offered by camera makers, it is limited by budget.

Also, regarding the form factor debate - I would not want a camera with a built-in vertical grip. Personally, I would not care what size the body is for any future cameras (i.e. 5D-class, 7D-class, 60D-class, Rebel-class, SL1-class, mirrorless) so long as the viewfinder is as big and bright as the 5D and the sensor is full frame. I enjoy the handling of my 5D, but I also think the 60D is the perfect mix of handling size and weight.

 

[LetTheRightLensIn]

LetTheRightLensin,

"The difference in DR a lot more than one stop. You definitely do not get that holy cow crispness from 20MP at 47"."

Anybody that, like LetTheRightLensin, doesn't believe my image, email me, I'll send you the untouched original RAW and my 47" print file. Take a crop, print it, then tell me it isn't crisp.

I'm not saying it's unusable or a blurry mess but it doesn't have that crazy crispness you get if you print 19" or less from 20MP.

 

[sanj]

Please do, I have been dying to see some optimally exposed shots where the DR of a Canon has substantially ruined a shot yet a Nikon capture would have been perfect.

Here is one from a recent shoot of a prison complex:
http://www.arthurdomagala.com/blog/wp-content/uploads/2013/02/cell-block-2nd-level.jpg

The common area on the first floor is illuminated by a skylight. The dark gray cell doors on the second level have no lighting on them at all. Exposing correctly for the highlights in this scene severly underexposes the doors. There is no way to set up any additional lighting. Lifting the shadows on the doors in post leads to very obvious pattern noise on the doors. The eventual solution is blending multiple exposures. If this had been shot with an Exmor sensor simply lifting the shadows in a single exposure would not have been a problem.

Another example:
http://www.arthurdomagala.com/blog/wp-content/uploads/2012/04/IMG_6414-web.jpg

This image was exposed to capture the colors along the horizon. But because of dynamic range limitations, the water that should have been dark blue in the lower left corner was instead black. This requires lifting the shadows again. On the first 20x30 inch print I made, the shadow banding was clearly evident:
http://www.arthurdomagala.com/blog/wp-content/uploads/2012/04/IMG_6414-20x30crop-no-nr.jpg

I had to go back, reprocess the image multiple times, blend exposures, apply noise reduction with debanding, apply a manual blur brush, and apply grain to even things out. Again, with a better sensor, this processing scenario would have been greatly simplified.

(A longer explanation can be found at this link: http://www.arthurdomagala.com/blog/2012/04/dynamic-range-canon-dslrs-and-shadow-noise-dealing-with-it/ )

I will just add one more note: "ruined" is your term, not mine. And it's a loaded term. I don't think Canon images are "ruined" by not having more dynamic range. But there are circumstances where it becomes problematic.

Good examples. Thx.

 

[sanj]

All this negativity about advancement in technology!!! Why??
There WILL be a high mpix camera anyone likes it or not.
There WILL be people who will buy it as it will promise better IQ...

We should not fight this and just wait to see if it suits our needs. If it does we will buy it.

I am looking forward to it in a 1d style body... But that is ME.

Just out of interest, what for?

As I said, I am no King Canute, it will all come whether we want it or not, but what, specifically, do you want 40+ MP in a 135 format camera for?

Good question. Simple answer: Higher MP holds the promise of better IQ. And perhaps if you feel (I guarantee you know more about this than me) that 40+ will not work for 135mm format then perhaps Canon will not make it more than 40mp...

I remember when cameras where 12mp everyone said that was enoughhhh. But who would buy that now?

 

[sanj]

REQUEST:
Can anyone guide me to a study which indicates that high MP results in poorer IQ? PLS.

I so curious about this negativity towards high MP cameras. I want to read up and understand.

A buddy of mine shifted to Nikon for D800 and sings (literally) its praises.

 

[heptagon]

Would you as a landscaper be disappointed if it only had ISO100, no VF and no Phase-AF?

Could you use ISO25 if it gave you more DR?

Is the Live-View-AF of the EOS-M sufficient for your studio shots?


If you are OK with all these, the new Camera would be the ultimate mirrorless resolution and DR monster which doesn't compromise with general usability wich already is provided by the 1Dx.

 

[mb66energy]

[...]

Seems that the most desirable quality of the D800 would be the impressive dynamic range it offers.

I'd just like to ask, you folks think cameras will ever reach the DR our eyes and brains could resolve and if so, the impact it could have on photography?

[...]

In my opinion:

AD 1 (DR of D800): I think DR will allways help to capture reality as close as possible.

AD 2: Very good question because you included the brain: I think a real 20bit DR sensor with a 20bit DR display will convince our eye and brain to see the real thing.
Physically a 20bit DR sensor is easy to built, but technically it is not. Think about an ISO 6 sensitivity and the capacity to collect large amounts of charge per pixel without saturation ... you will be fine: Pedestal noise is the same but saturation is far away and increases the DR. Physically.
Technically: I think you have to go into the 3rd dimension for sensor pixels charge storage and a good 20bit ADC is available but slow (at the time). 3D-Sensor design is IMO the real challenge because you have to create millions of 3D structures with small tolerances between them in the size of some microns. A 12 Mpix full frame sensor with a DR of 20bit would be sufficient (for the beginning) and blow away a lot of other sensors for some photographic fields.

I am shure that a lot of companies do research to built such (sensor) chips. A lot of chip technology has 3Dish structures just now. Hopefully we will see good sensors soon in Canon cameras - I don't have money/time to reinvest in new glass ... and hopefully we will see large OLED displays which provide DRs of 14 or 16 bit easily.

 

[mb66energy]

REQUEST:
Can anyone guide me to a study which indicates that high MP results in poorer IQ? PLS.
[...]

I cannot guide you to such a study and I am shure you will never find such a study because:

If the per pixel quality stays the same, more pixels give always better IQ until other factors like lens IQ are limiting.

Lower pixel counts help to reduce processing time and filesize which may be valuable for some purposes but nothing less/more.

I am physicist and it was never bad to measure values as precise as possible - it was bad to measure with less than required precision! If you don't need the precision you can report rounded measurement values - if a light bulb has a power consumption of 60.22412 Watts or 60 Watts essentially is not important for a bulb user, but perhaps for a lamp design researcher who optimizes the production process.

 

[sanj]

REQUEST:
Can anyone guide me to a study which indicates that high MP results in poorer IQ? PLS.
[...]

I cannot guide you to such a study and I am shure you will never find such a study because:

If the per pixel quality stays the same, more pixels give always better IQ until other factors like lens IQ are limiting.

Lower pixel counts help to reduce processing time and filesize which may be valuable for some purposes but nothing less/more.

I am physicist and it was never bad to measure values as precise as possible - it was bad to measure with less than required precision! If you don't need the precision you can report rounded measurement values - if a light bulb has a power consumption of 60.22412 Watts or 60 Watts essentially is not important for a bulb user, but perhaps for a lamp design researcher who optimizes the production process.

Hmmmm. Then why the resentment to better IQ? That is what I am trying to understand.

 

[Sporgon]

REQUEST:
Can anyone guide me to a study which indicates that high MP results in poorer IQ? PLS.
[...]

I cannot guide you to such a study and I am shure you will never find such a study because:

If the per pixel quality stays the same, more pixels give always better IQ until other factors like lens IQ are limiting.

Lower pixel counts help to reduce processing time and filesize which may be valuable for some purposes but nothing less/more.

I am physicist and it was never bad to measure values as precise as possible - it was bad to measure with less than required precision! If you don't need the precision you can report rounded measurement values - if a light bulb has a power consumption of 60.22412 Watts or 60 Watts essentially is not important for a bulb user, but perhaps for a lamp design researcher who optimizes the production process.

Hmmmm. Then why the resentment to better IQ? That is what I am trying to understand.

I think may people are saying they would like the improvement in 'IQ' with files that are not as large, because they already have more than enough resolution for a highly portable 35mm type camera.

+ I feel that the reality of much higher MP on FF chip ( or APS for that matter) only manifests itself in 'higher IQ' when the photographer is viewing his images at 50 - 100% on his computer screen. Although this may give the photographer immense satisfaction it is not actually producing a better picture for display to anyone else.

 

[mb66energy]

REQUEST:
Can anyone guide me to a study which indicates that high MP results in poorer IQ? PLS.
[...]

I cannot guide you to such a study and I am shure you will never find such a study because:

If the per pixel quality stays the same, more pixels give always better IQ until other factors like lens IQ are limiting.

Lower pixel counts help to reduce processing time and filesize which may be valuable for some purposes but nothing less/more.

I am physicist and it was never bad to measure values as precise as possible - it was bad to measure with less than required precision! If you don't need the precision you can report rounded measurement values - if a light bulb has a power consumption of 60.22412 Watts or 60 Watts essentially is not important for a bulb user, but perhaps for a lamp design researcher who optimizes the production process.

Hmmmm. Then why the resentment to better IQ? That is what I am trying to understand.

I think may people are saying they would like the improvement in 'IQ' with files that are not as large, because they already have more than enough resolution for a highly portable 35mm type camera.

+ I feel that the reality of much higher MP on FF chip ( or APS for that matter) only manifests itself in 'higher IQ' when the photographer is viewing his images at 50 - 100% on his computer screen. Although this may give the photographer immense satisfaction it is not actually producing a better picture for display to anyone else.

I forgotten to refer to TECHNICAL IQ in my lines.

Sporgon, your argument (underlined) is IMO correct. I myself are satisfied with my 40D in terms of technical IQ .... FOR WHAT I DO: Photographing things and landscapes because I like to do it and viewing the best pictures for pleasure as 30x40cm (12x16 inch) prints or via a XGA-Beamer (1024 x 768 Pixels, 0,8 MPix (!!!)). A photograph with valuable content and expression works always on 12x16 inch or e.g. a 2 MPix beamer.

Furthermore I see no dramatical technical IQ gain with the 18MPix sensor of the 600D.

But I would take a 48 MP full frame sensor for one reason: Coulors of e.g. monochromatic light sources would be much cleaner if you downscale it to 12 MPix to have 1 red, 2 green and 1 blue subpixel for one final image pixel. Bayer demosaicing gives strong artefacts for e.g. LED tail lights and red sunsets. Therefore I miss a sRAW mode in my 600D which gives me 10 MPix or so.

sanj: "Then why the resentment to better IQ?"
Alternatively I could say "24 MPix for APS-C is bullshit, 10 MPix is enough". BECAUSE I have only 10 MPix at hand and no money to buy a better resolving camera. I think is more or less a psychologial resentment.

 

[sanj]

So a higher MP camera, case in point D800, will not show any better IQ due to resolution on my 27" Imac?

 

[mb66energy]

So a higher MP camera, case in point D800, will not show any better IQ due to resolution on my 27" Imac?

What ist the resolution of your 27" Imac?

 

[sanj]

I checked, it says 1920x1080 resolution.