canon rumors FORUM

Rumors => EOS Bodies => Topic started by: marekjoz on April 21, 2012, 08:55:05 AM

Title: Next Canon step in the matter of their sensors
Post by: marekjoz on April 21, 2012, 08:55:05 AM
So the battle for 5d3 image quality will continue for some time. No matter for a while whether DXO results are fine or not - the numbers are presented to the public and caused some mess anyway. People in Canon probably are not happy looking at comparisons knowing the public look at them as well and buys equipment following those summaries. So what Canon will do next do defend it's positions?
Title: Re: Next Canon step in the matter of their sensors
Post by: Astro on April 21, 2012, 09:03:29 AM
they will care a S____ and listen to pros who actually use cameras out in the field.
Title: Re: Next Canon step in the matter of their sensors
Post by: lol on April 21, 2012, 09:17:52 AM
My impression is they're spending all their time and effort in video, and have neglected any major development on the stills performance side of the sensors.
Title: Re: Next Canon step in the matter of their sensors
Post by: altenae on April 21, 2012, 09:20:28 AM
And still it's a superb tool which is capable of more then most of us will ever need or use..

And I think this is a fact..
Title: Re: Next Canon step in the matter of their sensors
Post by: Astro on April 21, 2012, 09:31:50 AM
or they will just post a link to the techradar test with a big question mark ?

and ask how two DxO test can have different results.

http://www.canonrumors.com/forum/index.php?topic=5868.0 (http://www.canonrumors.com/forum/index.php?topic=5868.0)
Title: Re: Next Canon step in the matter of their sensors
Post by: TrumpetPower! on April 21, 2012, 09:34:18 AM
Here's something I'd generally be interested in -- but, please, only reply if you have real-world experience, and not if you're only projecting your expectations from interpretations of an unfocused high-ISO macro shot of the inside of soWmebody's lenscap you saw posted on the 'Net.

In what real-world situation does the 5DIII have insufficient image quality for the job but the D800 does have sufficient IQ? That is, when would you put down the 5DIII and pick up the D800 instead of a medium format kit?

I personally can't think of any.

Cheers,

b&
Title: Re: Next Canon step in the matter of their sensors
Post by: gene_can_sing on April 21, 2012, 09:50:22 AM
My impression is they're spending all their time and effort in video, and have neglected any major development on the stills performance side of the sensors.

That's a huge misstatement. Their current video products are consider so overpriced and underspec'd that they're getting destroyed on the video forums. Nikon D800 video is even shining over the 5D3.

Canon has just slacked too hard the last few years. They lost the hunger after the success of the 5D2 on all fronts.
Title: Re: Next Canon step in the matter of their sensors
Post by: lessmore on April 21, 2012, 09:51:06 AM
Here's something I'd generally be interested in -- but, please, only reply if you have real-world experience, and not if you're only projecting your expectations from interpretations of an unfocused high-ISO macro shot of the inside of soWmebody's lenscap you saw posted on the 'Net.

In what real-world situation does the 5DIII have insufficient image quality for the job but the D800 does have sufficient IQ? That is, when would you put down the 5DIII and pick up the D800 instead of a medium format kit?

I personally can't think of any.

Cheers,

b&

Ask the same question about the 5D Mark II (or any number of other cameras) and you may get the same answer.

Whether the camera does everything you need or not, Canon also has to live up to their customers' expectations. Those expectation are shaped partially by what the competition is doing.

How many people who own a Porsche have ever driven it at full speed around a corner? Does that mean that Porsche should make cars that go slower because for most people "that's fast enough"? Sure, they could. I wonder how sales would look if they did?
Title: Re: Next Canon step in the matter of their sensors
Post by: Astro on April 21, 2012, 09:53:28 AM
How many people who own a Porsche have ever driven it at full speed around a corner? Does that mean that Porsche should make cars that go slower because for most people "that's fast enough"? Sure, they could. I wonder how sales would look if they did?

exactly.... 95% of all porsche owners don´t need a porsche or could even use it´s full potential. it´s a status symbol.
Title: Re: Next Canon step in the matter of their sensors
Post by: lol on April 21, 2012, 10:40:57 AM
In what real-world situation does the 5DIII have insufficient image quality for the job but the D800 does have sufficient IQ? That is, when would you put down the 5DIII and pick up the D800 instead of a medium format kit?
Currently I primarily use a 7D plus assorted telezooms for wildlife. Lack of reach at a reasonable quality is a limiting factor, but at the same time, I need to be ready to adapt to changing situations. Typically I'd go out with two bodies, one tele zoom, one wide zoom. I would love a high MP density FF body to use. 36MP full frame would be an acceptable minimum, although a Canon parallel would be 46MP (make the 18MP APS-C sensor bigger). Basically this would give me the same effective quality and reach as the 7D by cropping on the long end of zoom. When I don't need to crop, I can make use of the increased resolution. Also, by being a bigger sensor, the zoom just got extended 1.6x on the wide end. This would increase my flexibility significantly as I look towards longer focal length solutions. Yes, I'd trade off the fps for the increased resolution. For obvious reasons medium format is the wrong tool for this job.

e.g. using the 100-400L as an example.
On 7D: 160-640mm equivalent at 18MP
On hypothetical 46MP full frame sensor: 100-400mm at 46MP, 400-640mm effective at >18MP by cropping.

Increased resolution over most of the range, while maintaining the total reach, and more effective zoom range compared to 7D.
Title: Re: Next Canon step in the matter of their sensors
Post by: altenae on April 21, 2012, 10:52:03 AM
More and more and more superb reviews about the 5D mark iii are becoming available.
Very very good user experiences also.

But these are all users who own the camera and use the camera.

I am really sick and tired over these DXO numbers, price to high , sensor not good , Canon is a shame, etc.

Buy the 5D or not.
I am really interested in the portfolio of all the complainers , so that I can see why the 5D mark iii is not good enough and why Canon is a shame.

There is so much more to an improvement over the mark II then only the sensor and the BS DXO numbers.
-----
www.wildlife-photos.net (http://www.wildlife-photos.net)
www.scramble.nl (http://www.scramble.nl)
www.planepix.nl (http://www.planepix.nl)
Title: Re: Next Canon step in the matter of their sensors
Post by: sach100 on April 21, 2012, 11:15:39 AM
More and more and more superb reviews about the 5D mark iii are becoming available.
Very very good user experiences also.

But these are all users who own the camera and use the camera.

I am really sick and tired over these DXO numbers, price to high , sensor not good , Canon is a shame, etc.

Buy the 5D or not.
I am really interested in the portfolio of all the complainers , so that I can see why the 5D mark iii is not good enough and why Canon is a shame.

There is so much more to an improvement over the mark II then only the sensor and the BS DXO numbers.


+1 Amen!

We need not go any further than CR forum. People are quietly posting away great pictures (and video) while some are going out of their way in proving 5d3 (sensor etc) is crap. btw, if guys are churning out great photos with their 'crap' 5d3 then they must be brilliant photogs and they can do magic with ANY camera.
There are issues with 5d3, agreed (like every other camera that came before it), but i am more than happy with mine! yeah baby!!
Title: Re: Next Canon step in the matter of their sensors
Post by: jrista on April 21, 2012, 11:25:02 AM
Here's something I'd generally be interested in -- but, please, only reply if you have real-world experience, and not if you're only projecting your expectations from interpretations of an unfocused high-ISO macro shot of the inside of soWmebody's lenscap you saw posted on the 'Net.

In what real-world situation does the 5DIII have insufficient image quality for the job but the D800 does have sufficient IQ? That is, when would you put down the 5DIII and pick up the D800 instead of a medium format kit?

I personally can't think of any.

Cheers,

b&

Landscapes...in a heartbeat! The average sunset landscape can have a DR well above 12 stops, even above 14 stops. The more DR the better for landscape shots. Not that you will necessarily use all of it in the final product, but that you can always use the leeway when it comes to landscape shots. Being able to do even a touch of shadow recovery where necessary without encountering the color-blotched noise-infested pattern-ridden shadows that Canon is well-known for would be a godsend. Not to mention the improved sharpness on the D800e.
Title: Re: Next Canon step in the matter of their sensors
Post by: jrista on April 21, 2012, 11:28:53 AM
e.g. using the 100-400L as an example.
On 7D: 160-640mm equivalent at 18MP
On hypothetical 46MP full frame sensor: 100-400mm at 46MP, 400-640mm effective at >18MP by cropping.

Um...what? On a 47.6mp FF sensor (which would be the equivalent pixel density of an 18mp APS-C), if you crop the 18mp APS-C sized center of the frame...you would get the exact same effective focal length...160-640. Are you thinking about the 200-400mm with 1.4x TC? If so, that would be 320-640 effective @ APS-C crop. I'm curious where you got 400-640mm effective...
Title: Re: Next Canon step in the matter of their sensors
Post by: itsnotmeyouknow on April 21, 2012, 12:00:30 PM
In what real-world situation does the 5DIII have insufficient image quality for the job but the D800 does have sufficient IQ? That is, when would you put down the 5DIII and pick up the D800 instead of a medium format kit?


e.g. using the 100-400L as an example.
On 7D: 160-640mm equivalent at 18MP
On hypothetical 46MP full frame sensor: 100-400mm at 46MP, 400-640mm effective at >18MP by cropping.

Increased resolution over most of the range, while maintaining the total reach, and more effective zoom range compared to 7D.

So why do people bother paying thousands for EF 400 L's and above if they could simply buy 1 camera and crop in? Because of quality.  I would rather have Canon get the quality right first at 22.3. I have a 41mp medium format if I want bigger, but then medium format also has a different quality to it that the D800 can't fully replicate.  If you need to get closer, GET closer.  Stop decreasing you IQ by cropping in in post.
Title: Re: Next Canon step in the matter of their sensors
Post by: lol on April 21, 2012, 12:23:38 PM
Um...what? On a 47.6mp FF sensor (which would be the equivalent pixel density of an 18mp APS-C), if you crop the 18mp APS-C sized center of the frame...you would get the exact same effective focal length...160-640. Are you thinking about the 200-400mm with 1.4x TC? If so, that would be 320-640 effective @ APS-C crop. I'm curious where you got 400-640mm effective...
I was saying you'd get 46MP at up to 400mm, but to effectively go beyond 400mm you can crop. Setting the floor at 18MP, you reach the same 18MP at 640mm effective as with 7D.

So why do people bother paying thousands for EF 400 L's and above if they could simply buy 1 camera and crop in? Because of quality.  I would rather have Canon get the quality right first at 22.3. I have a 41mp medium format if I want bigger, but then medium format also has a different quality to it that the D800 can't fully replicate.  If you need to get closer, GET closer.  Stop decreasing you IQ by cropping in in post.
If I could get closer, I would be able to walk on water and fly. As far as I know, I am not related to Superman, so that is not an option. Sure, I'd love the massive telephotos too. If I had the cash, I'd get them, as well as an entourage to help carry them around. Medium format is about the worst camera for my needs. For less demanding uses, I have the 5D2 at half the price of the 5D3, but neither come close to the 7D in this application. And that is the big point here. I don't care if the kit does better in another application. I'm concerned with what it does for me. Do not try to impose a different set of irrelevant criteria to my needs.

Besides, I'm not decreasing my IQ from my benchmark, which is the 7D. Having a high MP will allow cropping to that when necessary, and when not, a significant resolution increase. Put simply, if Canon had released a 5D3 with a sensor like the D800 in it, I'd have one in my kit bag right now.
Title: Re: Next Canon step in the matter of their sensors
Post by: altenae on April 21, 2012, 12:32:49 PM
Um...what? On a 47.6mp FF sensor (which would be the equivalent pixel density of an 18mp APS-C), if you crop the 18mp APS-C sized center of the frame...you would get the exact same effective focal length...160-640. Are you thinking about the 200-400mm with 1.4x TC? If so, that would be 320-640 effective @ APS-C crop. I'm curious where you got 400-640mm effective...
I was saying you'd get 46MP at up to 400mm, but to effectively go beyond 400mm you can crop. Setting the floor at 18MP, you reach the same 18MP at 640mm effective as with 7D.

So why do people bother paying thousands for EF 400 L's and above if they could simply buy 1 camera and crop in? Because of quality.  I would rather have Canon get the quality right first at 22.3. I have a 41mp medium format if I want bigger, but then medium format also has a different quality to it that the D800 can't fully replicate.  If you need to get closer, GET closer.  Stop decreasing you IQ by cropping in in post.
If I could get closer, I would be able to walk on water and fly. As far as I know, I am not related to Superman, so that is not an option. Sure, I'd love the massive telephotos too. If I had the cash, I'd get them, as well as an entourage to help carry them around. Medium format is about the worst camera for my needs. For less demanding uses, I have the 5D2 at half the price of the 5D3, but neither come close to the 7D in this application. And that is the big point here. I don't care if the kit does better in another application. I'm concerned with what it does for me. Do not try to impose a different set of irrelevant criteria to my needs.

Besides, I'm not decreasing my IQ from my benchmark, which is the 7D. Having a high MP will allow cropping to that when necessary, and when not, a significant resolution increase. Put simply, if Canon had released a 5D3 with a sensor like the D800 in it, I'd have one in my kit bag right now.

Getting the reach with good lenses is better then crop in most circumstances.
Title: Re: Next Canon step in the matter of their sensors
Post by: jrista on April 21, 2012, 01:58:09 PM
Um...what? On a 47.6mp FF sensor (which would be the equivalent pixel density of an 18mp APS-C), if you crop the 18mp APS-C sized center of the frame...you would get the exact same effective focal length...160-640. Are you thinking about the 200-400mm with 1.4x TC? If so, that would be 320-640 effective @ APS-C crop. I'm curious where you got 400-640mm effective...
I was saying you'd get 46MP at up to 400mm, but to effectively go beyond 400mm you can crop. Setting the floor at 18MP, you reach the same 18MP at 640mm effective as with 7D.

So why do people bother paying thousands for EF 400 L's and above if they could simply buy 1 camera and crop in? Because of quality.  I would rather have Canon get the quality right first at 22.3. I have a 41mp medium format if I want bigger, but then medium format also has a different quality to it that the D800 can't fully replicate.  If you need to get closer, GET closer.  Stop decreasing you IQ by cropping in in post.
If I could get closer, I would be able to walk on water and fly. As far as I know, I am not related to Superman, so that is not an option. Sure, I'd love the massive telephotos too. If I had the cash, I'd get them, as well as an entourage to help carry them around. Medium format is about the worst camera for my needs. For less demanding uses, I have the 5D2 at half the price of the 5D3, but neither come close to the 7D in this application. And that is the big point here. I don't care if the kit does better in another application. I'm concerned with what it does for me. Do not try to impose a different set of irrelevant criteria to my needs.

Besides, I'm not decreasing my IQ from my benchmark, which is the 7D. Having a high MP will allow cropping to that when necessary, and when not, a significant resolution increase. Put simply, if Canon had released a 5D3 with a sensor like the D800 in it, I'd have one in my kit bag right now.

Getting the reach with good lenses is better then crop in most circumstances.

+1 Definitely.
Title: Re: Next Canon step in the matter of their sensors
Post by: marcust on April 21, 2012, 02:43:53 PM
they will just quit makeing still cameras, going to video only,,,,,,,,,
Title: Re: Next Canon step in the matter of their sensors
Post by: lol on April 21, 2012, 02:59:02 PM
Getting the reach with good lenses is better then crop in most circumstances.
To the point when they either become too heavy or too expensive they are no longer an option. Optimise the whole system.
Title: Re: Next Canon step in the matter of their sensors
Post by: briansquibb on April 21, 2012, 03:10:05 PM
In what real-world situation does the 5DIII have insufficient image quality for the job but the D800 does have sufficient IQ? That is, when would you put down the 5DIII and pick up the D800 instead of a medium format kit?


e.g. using the 100-400L as an example.
On 7D: 160-640mm equivalent at 18MP
On hypothetical 46MP full frame sensor: 100-400mm at 46MP, 400-640mm effective at >18MP by cropping.

Increased resolution over most of the range, while maintaining the total reach, and more effective zoom range compared to 7D.

So why do people bother paying thousands for EF 400 L's and above if they could simply buy 1 camera and crop in? Because of quality.  I would rather have Canon get the quality right first at 22.3. I have a 41mp medium format if I want bigger, but then medium format also has a different quality to it that the D800 can't fully replicate.  If you need to get closer, GET closer.  Stop decreasing you IQ by cropping in in post.

That is why I got the 600 as well. The 400 f/2.8 makes a good portrait lens on a ff too :D
Title: Re: Next Canon step in the matter of their sensors
Post by: JR on April 21, 2012, 03:32:02 PM
They will never admit their new sensor is not good.  Especially with another new sensor coming into the 1dx and 1dc!!!!

If this ultimately hurt their sale and they start loosing ground to nikon and sony, they will react, if not, they wont care!
Title: Re: Next Canon step in the matter of their sensors
Post by: LetTheRightLensIn on April 21, 2012, 04:46:42 PM
or they will just post a link to the techradar test with a big question mark ?

and ask how two DxO test can have different results.

http://www.canonrumors.com/forum/index.php?topic=5868.0 (http://www.canonrumors.com/forum/index.php?topic=5868.0)

none of the techradar results make much sense at all
Title: Re: Next Canon step in the matter of their sensors
Post by: LetTheRightLensIn on April 21, 2012, 04:49:32 PM
Here's something I'd generally be interested in -- but, please, only reply if you have real-world experience, and not if you're only projecting your expectations from interpretations of an unfocused high-ISO macro shot of the inside of soWmebody's lenscap you saw posted on the 'Net.

In what real-world situation does the 5DIII have insufficient image quality for the job but the D800 does have sufficient IQ? That is, when would you put down the 5DIII and pick up the D800 instead of a medium format kit?

I personally can't think of any.

Cheers,

b&

shooting distant birds (high density sensor puts more pixels per duck)

for making 13x19" or larger prints

shooting scenes with lots of dynamic range (IMO a huge difference here, this is the biggest difference)
Title: Re: Next Canon step in the matter of their sensors
Post by: LetTheRightLensIn on April 21, 2012, 04:52:02 PM


There is so much more to an improvement over the mark II then only the sensor and the BS DXO numbers.


http://www.fredmiranda.com/5DIII-D800/index_controlled-tests.html (http://www.fredmiranda.com/5DIII-D800/index_controlled-tests.html)
DxO is BS and has no bearing on reality??


that said of course there are huge improvements over the mark2 other than the sensor, yes
Title: Re: Next Canon step in the matter of their sensors
Post by: Heidrun on April 21, 2012, 05:02:10 PM
How about a 36x36 sensor.
I think that this is the ideal sensor size
Title: Re: Next Canon step in the matter of their sensors
Post by: marekjoz on April 21, 2012, 05:12:00 PM
They will never admit their new sensor is not good.  Especially with another new sensor coming into the 1dx and 1dc!!!!

If this ultimately hurt their sale and they start loosing ground to nikon and sony, they will react, if not, they wont care!

Sales numbers seem to be indeed the best indicator of marketing efficiency and technical effort.
Thanks all for the participation in this survey as far.

I'm really interested in your opinion, not only what you think about 5d3, but also and even more, what do you think, that Canon really thinks about it. Critics suggested, that Canon has hurried up with this release and it was marketing pressure. Enthusiasts emphasize (and here is more of them), that this is a good product fullfilling their expectations.

If Canon decided to show up this release under pressure only, we could maybe sooner than later see it's successor or rather something between 5d3 and 1dx like Craig suggested. If Canon is happy with 5d3 then most probably things would go as usual (I'm convinced that it's too early to speculate before 1dx was sent to public and tested in the field).
Title: Re: Next Canon step in the matter of their sensors
Post by: well_dunno on April 21, 2012, 06:47:59 PM
They will never admit their new sensor is not good.  Especially with another new sensor coming into the 1dx and 1dc!!!!

If this ultimately hurt their sale and they start loosing ground to nikon and sony, they will react, if not, they wont care!

I could not agree more!..
Title: Re: Next Canon step in the matter of their sensors
Post by: Aglet on April 21, 2012, 07:24:14 PM
Canon is likely quite happy with the performance of the 5D3; it's selling like crack, it actually does do a lot of things considerably better than the 5D2, at least for most of the crowd who asked for those improvements.

I sure hope they are listening to all the forum griping tho, as well as those of us who've also bothered to let them know directly, that we expected better overall sensor performance from them, especially at the low ISO end.

I'm very curious to see how the 1DX performs, at BOTH ends of the ISO range.  There'd better be some seriously clean low ISO performance with sensor read noise as low as what the competitors are achieving.
If they can't pull it off, it will hurt their marketing in the longer term, especially if they can't compete with the likes of the D800/D4/D7000/D5100/D3200/K5.
After decades of being an all-Canon gig I've started to explore what the competition has to offer, and some of it's mighty tantalizing. So much so I've started to collect some Nikon gear and even the Pentax K5 is of some interest for a few of its unique features.

I really think Canon needs to look at what the competition has brought to the market and pay attention to its loyal customers who'd like some of the same without having to buy it from the competition.

There's one way to help them do that.... TELL THEM.
Title: Re: Next Canon step in the matter of their sensors
Post by: TrumpetPower! on April 21, 2012, 07:32:20 PM
Here's something I'd generally be interested in -- but, please, only reply if you have real-world experience, and not if you're only projecting your expectations from interpretations of an unfocused high-ISO macro shot of the inside of soWmebody's lenscap you saw posted on the 'Net.

In what real-world situation does the 5DIII have insufficient image quality for the job but the D800 does have sufficient IQ? That is, when would you put down the 5DIII and pick up the D800 instead of a medium format kit?

I personally can't think of any.

Cheers,

b&

Landscapes...in a heartbeat! The average sunset landscape can have a DR well above 12 stops, even above 14 stops. The more DR the better for landscape shots. Not that you will necessarily use all of it in the final product, but that you can always use the leeway when it comes to landscape shots. Being able to do even a touch of shadow recovery where necessary without encountering the color-blotched noise-infested pattern-ridden shadows that Canon is well-known for would be a godsend. Not to mention the improved sharpness on the D800e.

I'm not suggesting that improvements are unwelcome or claiming that the D800 doesn't hold a slight edge in certain areas.

But are you really shooting landscapes in situations where you'd be able to tell the difference in the final print between the 5DIII and the D800, short of examining them with a loupe?

And if your landscapes won't sell with the 5DIII, would you really be able to sell them with the D800 instead?

That's my point. If the 5DIII isn't good enough, the marginal improvement the D800 might be able to offer really isn't going to do the trick, either, even if it's better. If you can't cut the branch with a hunting knife, a Bowie knife isn't going to cut it either; you need a saw or a hatchet.

shooting distant birds (high density sensor puts more pixels per duck)

Does the extra density, combined with noise characteristics and AF performance / frames per second of the D800 really result in a better print once both have been cropped, even if you've got more PPD from the D800?

Distance-limited telephoto (or macro) shooting is about the only scenario I can think of where the difference might be discernible, but I'm still skeptical about whether or not there'll be a real-world difference in prints based on sensor characteristics alone. I'm even more skeptical that the non-sensor bits of the D800 will let it make as good a use of its sensor as the 5DIII will of its.

Quote
for making 13x19" or larger prints

Eh, 13" x 19" is puny. If you're buying these cameras to print at 13" x 19", you've wasted your money.

I'll buy a cup of coffee for anybody who can show me side-by-side 24" x 36" prints made from a 5DIII and a D800 that, when made from the same tripod in the same studio and independently processed by the same (competent) person to maximize final print quality (not just the same recipe applied to both) for each which can be readily distinguished at standard viewing distances.

I can maybe see the D800 pulling away at 36" x 54", but still not at standard viewing distances. And, once again, if your bread and butter is 36" x 54" prints (or bigger) and the 5DIII format won't get you the gigs, you need medium format, not the D800.

Quote
shooting scenes with lots of dynamic range (IMO a huge difference here, this is the biggest difference)

In what real-world situation does the 5DIII have insufficient DR but the D800 does where you wouldn't be doing multiple exposures for HDR in the first place? And let's not forget that the 5DIII can do 6 FPS fire-and-forget several-shot brackets with a just couple button presses.

Once again, as I see it, if the 5DIII is inadequate, the answer isn't the D800 but something else entirely (in this case, a different technique).

Cheers,

b&
Title: Re: Next Canon step in the matter of their sensors
Post by: Aglet on April 21, 2012, 07:49:19 PM
In what real-world situation does the 5DIII have insufficient DR but the D800 does where you wouldn't be doing multiple exposures for HDR in the first place? And let's not forget that the 5DIII can do 6 FPS fire-and-forget several-shot brackets with a just couple button presses.

You sound like you never shoot storms or other action where HDR bracketing is impractical.
HDR bracketing is great for limited DR jpegs, they should not be necessary with a properly exposed raw file.

Real world example is right here.

www.fredmiranda.com/5DIII-D800/index_controlled-tests.html (http://www.fredmiranda.com/5DIII-D800/index_controlled-tests.html)

Just imagine it as a landscape.

here's another, less extreme:

www.a2bart.com/tech/darknoise.htm (http://www.a2bart.com/tech/darknoise.htm)

Then again, if you and your customers are happy with snaps out of the camera that you don't care to process to look better that's fine.  Why disparage those whose artistic work is benefited by a camera that can provide a better base file?
Title: Re: Next Canon step in the matter of their sensors
Post by: TrumpetPower! on April 21, 2012, 08:24:40 PM
You sound like you never shoot storms or other action where HDR bracketing is impractical.

I'll grant you that I haven't shot storms. I am, however, having a hard time imagining even a fast-moving storm that would have sufficient DR that you could use a shutter speed that won't move / blur in a single D800 exposure but that would be a problem with a single pair of 6FPS 5DIII exposures (with nearly twice the resultant DR of the single D800 exposure).

Quote
HDR bracketing is great for limited DR jpegs, they should not be necessary with a properly exposed raw file.

Um...no. Not even close. The attached HDR I did with three 5D (classic) RAW exposures manually composited. And not even the D800, I'm sure, would have had a hope of a chance at doing this with a single exposure. Notice the shadows under the fern in the lower left? That's at the bottom of Muir Woods. And notice the blue sky peeking through the top center? It's still blue.

Quote
Real world example is right here.

www.fredmiranda.com/5DIII-D800/index_controlled-tests.html (http://www.fredmiranda.com/5DIII-D800/index_controlled-tests.html)

Just imagine it as a landscape.

I'm sure it wouldn't occur to Fred to try to make a print of either with the shadows pushed all the way to the midtones. And, if that's really the vision he had in mind, he'd have taken a second exposure for the shadows and composited it in -- just as I did in the attached shot.

You'll also notice that dynamic range was the only advantage Fred found that the Nikon had, and it had some other very crippling focus-related problems (and, frankly, laughably inferior glass) that kept him from using it for anything real.

In other words, it's the Canon, inferior DR and all, that's putting bread on his table and the Nikon that's good only for those JPEG snapshots you're referring to. And, as I've been pointing out everywhere, there's no visible resolution differences at 24" x 36".

Cheers,

b&
Title: Re: Next Canon step in the matter of their sensors
Post by: briansquibb on April 22, 2012, 01:28:17 AM

shooting distant birds (high density sensor puts more pixels per duck)

for making 13x19" or larger prints

shooting scenes with lots of dynamic range (IMO a huge difference here, this is the biggest difference)

I think you will find

- 22mp will be near enough native printing for A3 - 18x11 ish
- a dr of 12 will print the same as a dr of 14

I think you will find that 12mp images were printing very well on 30 x 24 using software to expand the image

If you are not close enough to birds with a large white then it is your field skills that need improving, not the camera. If you are a serious BIF shooter then the 5DIII is the wrong camera for you - and the D800 would be even worse.
Title: Re: Next Canon step in the matter of their sensors
Post by: TrumpetPower! on April 22, 2012, 01:38:12 AM
If you are a serious BIF shooter then the 5DIII is the wrong camera for you - and the D800 would be even worse.

Now you've got me wondering...I'd think the 5DIII would be second only to the not-yet-for-sale 1DX for BIF. I'd expect the 5DIII's new AF to Me_Me_Me even the 1DIV's framerate -- and, as you note, with a Big White, cropping shouldn't be a big deal.

On the other hand, I've not done any serious BIF (yet). So what's better than the 5DIII, presumably mounted to a 600 f/4, with or without the 1.4x, and why?

b&
Title: Re: Next Canon step in the matter of their sensors
Post by: stevenrrmanir on April 22, 2012, 01:58:07 AM
Here's something I'd generally be interested in -- but, please, only reply if you have real-world experience, and not if you're only projecting your expectations from interpretations of an unfocused high-ISO macro shot of the inside of soWmebody's lenscap you saw posted on the 'Net.

In what real-world situation does the 5DIII have insufficient image quality for the job but the D800 does have sufficient IQ? That is, when would you put down the 5DIII and pick up the D800 instead of a medium format kit?

I personally can't think of any.

Cheers,

b&

Landscapes...in a heartbeat! The average sunset landscape can have a DR well above 12 stops, even above 14 stops. The more DR the better for landscape shots. Not that you will necessarily use all of it in the final product, but that you can always use the leeway when it comes to landscape shots. Being able to do even a touch of shadow recovery where necessary without encountering the color-blotched noise-infested pattern-ridden shadows that Canon is well-known for would be a godsend. Not to mention the improved sharpness on the D800e.

great pictures can be taken with ANY camera - I know people with cheap cameras that get amazing pictures without blowing thousands of dollars on bodies alone! they actually enjoy photography and don't waste time arguing over the interwebs over their toys!
Title: Re: Next Canon step in the matter of their sensors
Post by: Aglet on April 22, 2012, 02:12:16 AM
I'll grant you that I haven't shot storms. I am, however, having a hard time imagining even a fast-moving storm that would have sufficient DR that you could use a shutter speed that won't move / blur in a single D800 exposure but that would be a problem with a single pair of 6FPS 5DIII exposures (with nearly twice the resultant DR of the single D800 exposure).

have a closer look at the example from a2bart from the other link.
1/1000s, f/5, ISO100.
that held the cloud hilites and motion in check. A push in post brought up the rest of the landscape and then other localized contrast and toning was used to embellish it.
That is a real-world example of a wide DR scene nicely squeezed into the DR of a print or display. An even wider DR could be accommodated using the same methods but only if the shadow noise level of the camera was low enough to not make a mess of it - as in the Fred Miranda example.  You don't have to bring shadows up to midtones to see the noise from a camera like the 7D, you can leave them 3 stops below mid and there's still obvious banding.

When lively storms have clouds moving 100+ mph, trees thrashing around and all the foliage on the move, no camera's fast enough to capture 2 or more frames without serious movement artifacts to deal with in conventional multi-shot HDR.  I've wasted plenty of time trying to bracket such scenes and put them together in post; results were not acceptable even in slower moving weather conditions.  Not to mention re-curving a multishot HDR to try and get a natural feeling image is a hassle.  IMO, you can get a much nicer result with a simple fill-light or shadows adjustment using Adobe's products.  LR4 does this with minimal artifacting or halos and it takes about 15 seconds to do it.  More complex toning control is available if needed.

No more multi-shot HDR silliness for me unless I need to capture more than 9 or 10 stops of DR and present them all.  Even then, I'll likely get a better result from just one ETTR raw file from one of the quieter new Nikons (or older Canons) where the hilites aren't clipped.

Um...no. Not even close. The attached HDR I did with three 5D (classic) RAW exposures manually composited. And not even the D800, I'm sure, would have had a hope of a chance at doing this with a single exposure. Notice the shadows under the fern in the lower left? That's at the bottom of Muir Woods. And notice the blue sky peeking through the top center? It's still blue.

Yes, not bad if you like that sort of look but again, multishot HDR has a hard time maintaining natural looking color and contrast. The shadow levels of that one could have been left a little lower too.
And I'd love to see what I can do with 1 shot from the D800 against a 3-shot HDR on the same scene.  A nicer job in much less time, I'm thinking. :)  Multishot HDR has its uses, but it's not needed if the camera can handle the scene without it.  And that would be any scene where the camera's noise-free DR is at least one or 2 stops greater than the scene being shot. That means the D800 and D5100 are gonna give me 2 more stops to work with, in ONE shot, than anything from Canon at this time.

You'll also notice that dynamic range was the only advantage Fred found that the Nikon had, and it had some other very crippling focus-related problems (and, frankly, laughably inferior glass) that kept him from using it for anything real.

yes, the nikon was hindered on Fred's shoot by inferior glass and a less then adequate live view mode, pity that latter especially as it's super useful on the Canons.
But to say the the better DR was the ONLY advantage the D800 had is like saying the only advantage one of these 2 motorized vehicles has is wings.  Wings are pretty damn important if you intend to fly.

Glass is one thing causing me some consternation because Canon's latest TSE 17 and 24mm are on my wish list.  Really wish there was something comparable for the D800 so I hope Samyang pulls a supermodel out of a cake with the 24mm F-mount version they're working on.  Or maybe (chuckle) the 1DX will have a much lower noise pattern than the 5Ds...

In other words, it's the Canon, inferior DR and all, that's putting bread on his table and the Nikon that's good only for those JPEG snapshots you're referring to. And, as I've been pointing out everywhere, there's no visible resolution differences at 24" x 36".

I'd be plenty happy with the rez from my 5D2 if I could push it more. The difference in rez is secondary for this purpose. I've printed 24x36" from my 40D that looks just fine with appropriate sharpening in post, and that's at nose-to-paper distances and ~100ppi.  Getting 180+ppi at the same size would only look better at much closer to normal viewing distance. I'll take it if if I got it. :) No complaints about file size either. (NEF compressed raw isn't all that bad for size)

Title: Re: Next Canon step in the matter of their sensors
Post by: briansquibb on April 22, 2012, 02:24:18 AM
If you are a serious BIF shooter then the 5DIII is the wrong camera for you - and the D800 would be even worse.

Now you've got me wondering...I'd think the 5DIII would be second only to the not-yet-for-sale 1DX for BIF. I'd expect the 5DIII's new AF to Me_Me_Me even the 1DIV's framerate -- and, as you note, with a Big White, cropping shouldn't be a big deal.

On the other hand, I've not done any serious BIF (yet). So what's better than the 5DIII, presumably mounted to a 600 f/4, with or without the 1.4x, and why?

b&

The simple answer is the 1D4 - if pushed it can also cope with the 600 + 2x - and of course the 1.3 crop helps too.

The 1D4 rolls along at 10fps -the AF doesn't come into it. The 5DIII strolls along at 6fps (my low continuous speed on the 1d4)

The beauty of the 1D4 (and other series 1) is that wherever the AF point is is the place whee the metering is done, the 5DIII meters at the centre point. So with the 5DIII unless you are using centre point focussing you have to choose between correct metering or AF.

The 1D4 also has the option to manually set the shutter speed and the aperture with auto iso and also be able to set exposure compensation to get further refinements on metering. This isn't possible with the 5DIII

I think that you will find that there is so little in the AF speeds (40ms) that you wont notice the difference - this is one of those headline moments that means nothing in real life "5DIII AF faster than the mighty 1D4". Likewise with the D800 "you dont need long lens because you can get it from cropping" - providing that at 4fps you can capture the moment.

The 1D4 high speed burst is 28 images, the 5DIII is 18 images

I am talking here about the advantages of the 1D4 for sporting/birds/wildlife - not for the traditional 5D strengths of weddings/studio/landscape
Title: Re: Next Canon step in the matter of their sensors
Post by: marekjoz on April 23, 2012, 05:24:08 AM
88 people voted - quite a nice feedback. Thank you.

Summarizing results:
1. They will work harder to get better results and follow the path they do now
2. They don't care and think 5d3 sensor is fine
3. They will run some marketing campaign to defend the results achieved with current 5d3 sensor
4. They will care a S___ and listen to pros who actually use cameras out in the field.

So we here rather think, that Canon finds 5d3 a good product and doesn't see a need to buy sensors form third party company but rather work harder.
What I find very important for us is also (no matter what we think about current gear prices) help Canon not to decrease income because of those comparisons. If sales decreases because of public not buying gear scared with tests, numbers and comparisons then we also will notice it in less money spent on R&D. Canon might not care for numbers and tests and listen to pros but the question is whether the public buying equipment will do it as well?
Title: Re: Next Canon step in the matter of their sensors
Post by: TrumpetPower! on April 23, 2012, 12:23:38 PM
If you are a serious BIF shooter then the 5DIII is the wrong camera for you - and the D800 would be even worse.
(W)hat's better than the 5DIII (for BIF), presumably mounted to a 600 f/4, with or without the 1.4x, and why?
The simple answer is the 1D4 - if pushed it can also cope with the 600 + 2x - and of course the 1.3 crop helps too.

The 1D4 rolls along at 10fps -the AF doesn't come into it. The 5DIII strolls along at 6fps (my low continuous speed on the 1d4)

I can see how a faster framerate would be welcome for flying birds, to get the wings positioned exactly how you envision. But I'd have thought the repetitive motion of the wings alone would mean a sufficient number of shots with a good composition...and that you'd want the better autofocus to make sure that you have more in-focus shots to choose from.

Quote
The beauty of the 1D4 (and other series 1) is that wherever the AF point is is the place whee the metering is done, the 5DIII meters at the centre point. So with the 5DIII unless you are using centre point focussing you have to choose between correct metering or AF.

Really? Autoexposure with BIF? I'd have thought the way to go is with an incident meter. This is outdoors, right? Except for fast-moving partial cloud cover, the light's not changing. I mean, if these birds were in the trees I could understand...but "flying" implies "out in the open," right? You want great metering for weddings where you're shooting in all directions and every subject is in different light. But for shooting in a single direction from a fixed location outdoors? I don't think it'd even occur to me to check what the camera thought the exposure should be. Even if you don't have a meter, that's what your gray card is for.

Quote
The 1D4 high speed burst is 28 images, the 5DIII is 18 images

Yes, but both represent half a minute of continuous shooting. Are you really going to be machine-gunning for thirty seconds straight without letting your finger off the trigger even once? This is still photography we're talking about, not the movies.

Quote
I am talking here about the advantages of the 1D4 for sporting/birds/wildlife - not for the traditional 5D strengths of weddings/studio/landscape

Me, too. And, frankly, the only real advantage I can personally think of for the 1D4 is for its framerate...and that again can go away if the 5DIII gets more frames-in-focus per second even if the 1D4 has some additional frames-not-in-focus thrown in the mix for a higher total frames-per-second. Obviously, I haven't compared the two and I don't recall seeing anybody else doing a side-by-side comparison.

Cheers,

b&
Title: Re: Next Canon step in the matter of their sensors
Post by: jrista on April 23, 2012, 12:57:34 PM
Quote
I am talking here about the advantages of the 1D4 for sporting/birds/wildlife - not for the traditional 5D strengths of weddings/studio/landscape

Me, too. And, frankly, the only real advantage I can personally think of for the 1D4 is for its framerate...and that again can go away if the 5DIII gets more frames-in-focus per second even if the 1D4 has some additional frames-not-in-focus thrown in the mix for a higher total frames-per-second. Obviously, I haven't compared the two and I don't recall seeing anybody else doing a side-by-side comparison.

Well, there is also the extra reach of the 1D4 with its APS-H sensor. That is something people keep seeming to forget about with all the new FF cameras flying about, but high density cropped sensors really do have something extra to offer that FF can't touch. Reach is everything when it comes to birds/BIF/wildlife, and you get a hell of a lot more bang for your buck with a cropped sensor and a middle-grade lens.
Title: Re: Next Canon step in the matter of their sensors
Post by: LetTheRightLensIn on April 23, 2012, 04:17:54 PM

shooting distant birds (high density sensor puts more pixels per duck)

for making 13x19" or larger prints

shooting scenes with lots of dynamic range (IMO a huge difference here, this is the biggest difference)

I think you will find

- 22mp will be near enough native printing for A3 - 18x11 ish
- a dr of 12 will print the same as a dr of 14

I think you will find that 12mp images were printing very well on 30 x 24 using software to expand the image

If you are not close enough to birds with a large white then it is your field skills that need improving, not the camera. If you are a serious BIF shooter then the 5DIII is the wrong camera for you - and the D800 would be even worse.

Why would a D800 be a worse birding camera? 16MP 6fps with higher density beats 22MP 6fps with lower density
And not everyone can afford 800mm or 1200mm lens. And even dragging a 300 2.8+2x TC around is a pain enough. Why do you think so many birders have 7D and/or 1D4???

yes 22MP is fine for 13x19, but you can see the difference between that and 36MP at that size if you use all of the highest print settings, it's not critical, but it does look crazy sharp when go to say 540ppi vs 300 ppi.

The DR difference is a good 3 stops, not 2, and yes it does make a difference.

Also, 4k displays are coming soon.
Title: Re: Next Canon step in the matter of their sensors
Post by: jrista on April 23, 2012, 05:28:48 PM
yes 22MP is fine for 13x19, but you can see the difference between that and 36MP at that size if you use all of the highest print settings, it's not critical, but it does look crazy sharp when go to say 540ppi vs 300 ppi.

That assumes your 36mp sensor is capturing enough detail to start for a 540ppi print (which would probably be more like a 720ppi, 600ppi, or 360ppi print, depending on the printer, unless you want uneven and potentially unsightly interpolation.) You need the very best glass, at an ideal aperture, with absolutely zero camera shake, and a low enough ISO to capture a photo that will actually fully and perfectly utilize all the spatial resolution the D800 sensor is capable of. Thats a lot of factors, and outside of stable tripod-based landscape photography in zero wind, its going to be a tough thing to achieve most of the time.

Its pretty tough to capture a perfectly sharp, maximum detail image with the 7D's 18mp sensor, which offers slightly more spatial resolution than the D800. When it comes to BIF, its simply expected that your not going to capture that much detail in perfect clarity unless your about 2 feet from the bird or have ungodly reach, and that is kind of the holy grail of BIF. Most of the time, your well outside of ideal conditions. Reach becomes ever more important than resolution at that point, as the more you can fill the frame and utilize more of sensor area available, the better the quality of each pixel will be, offering more for your final output. Far more important than sensor, in that respect, is the glass and the reach of your glass. I would happily lug around a 300/2.8 with a 2x TC, or even a 500/4 or 600/4 with a 1.4x TC, if it will help me fill the frame.

When it comes to printing, the native print size between the 1D IV, 5D III and D800 is enough to print at any common paper size up to A2 with a bit of post-process sharpening for the lower res images, with very little to no visible difference between them all. The linear difference in megapixels translates to a far smaller area difference those megapixels can print at. Dynamic range in print is far lower than it is directly out of the camera, and unless your working with a true 10-bit display, even your monitor is going to be displaying less than 12 stops of DR. The native DR you get out of the camera has to be considerably compressed if your final destination os print. Gamut is a far more important factor for print, and both DR and gamut for print are things primarily based on software, several degrees removed from the camera itself. Your probably printing about 6-7 stops of DR most of the time, and potentially much less than that if you opt for a matte paper over an idealistic high dMax gloss or metallic paper, with a gamut around about as large as AdobeRGB (and a bit larger in either the greens or blues/violets...depending on the printer and inks.)

If we are going to talk about the benefits of any given camera for print, I'd say any of them will do far more than superbly. Printing is so many degrees removed from how good the camera hardware is its rather ridiculous and largely a moot point. Reach, be it via sensor crop factor or, more idealistically, longer glass, is a far more critical factor of capturing and printing BIF shots.
Title: Re: Next Canon step in the matter of their sensors
Post by: briansquibb on April 23, 2012, 07:02:43 PM


Why would a D800 be a worse birding camera? 16MP 6fps with higher density beats 22MP 6fps with lower density
And not everyone can afford 800mm or 1200mm lens. And even dragging a 300 2.8+2x TC around is a pain enough. Why do you think so many birders have 7D and/or 1D4???

yes 22MP is fine for 13x19, but you can see the difference between that and 36MP at that size if you use all of the highest print settings, it's not critical, but it does look crazy sharp when go to say 540ppi vs 300 ppi.

The DR difference is a good 3 stops, not 2, and yes it does make a difference.

Also, 4k displays are coming soon.

I would guess that the vast majority of BIF shooters have a 7D or a 1D3/4. Go onto a BIF shooter forum and ask.

I would guess the majority of BIF have lens 400mm+ with, I expect, the 500mm being the most common.

I would expect a 7D with 8fps and a 1D4 with 10fps to be significantly better at BIF than the D800. Surprisingly DR is not a critical factor for BIF shooters - key factors are

- AF
- reach
- metering

The issue is not density but filling the frame, issues only start when cropping start. Putting all 16mp/18mp on the subject will always give top IQ

There is no way the D800 can compete with the 1D4 for BIF.
 
Title: Re: Next Canon step in the matter of their sensors
Post by: jrista on April 23, 2012, 07:13:26 PM


Why would a D800 be a worse birding camera? 16MP 6fps with higher density beats 22MP 6fps with lower density
And not everyone can afford 800mm or 1200mm lens. And even dragging a 300 2.8+2x TC around is a pain enough. Why do you think so many birders have 7D and/or 1D4???

yes 22MP is fine for 13x19, but you can see the difference between that and 36MP at that size if you use all of the highest print settings, it's not critical, but it does look crazy sharp when go to say 540ppi vs 300 ppi.

The DR difference is a good 3 stops, not 2, and yes it does make a difference.

Also, 4k displays are coming soon.

I would guess that the vast majority of BIF shooters have a 7D or a 1D3/4. Go onto a BIF shooter forum and ask.

I would guess the majority of BIF have lens 400mm+ with, I expect, the 500mm being the most common.

I would expect a 7D with 8fps and a 1D4 with 10fps to be significantly better at BIF than the D800. Surprisingly DR is not a critical factor for BIF shooters - key factors are

- AF
- reach
- metering

The issue is not density but filling the frame, issues only start when cropping start. Putting all 16mp/18mp on the subject will always give top IQ

There is no way the D800 can compete with the 1D4 for BIF.

As a BIF shooter, I pretty much agree with all of that. Particularly the most important camera features: AF, Reach, Metering.
Title: Re: Next Canon step in the matter of their sensors
Post by: V8Beast on April 23, 2012, 09:54:16 PM
Landscapes...in a heartbeat! The average sunset landscape can have a DR well above 12 stops, even above 14 stops. The more DR the better for landscape shots. Not that you will necessarily use all of it in the final product, but that you can always use the leeway when it comes to landscape shots. Being able to do even a touch of shadow recovery where necessary without encountering the color-blotched noise-infested pattern-ridden shadows that Canon is well-known for would be a godsend. Not to mention the improved sharpness on the D800e.

That's a fair answer, but based on all the b!tching, you'd think everyone is a landscape photographer all of a sudden. It's either that, or people who may have been in the market for a 5DIII don't like the idea that Nikon makes a better camera for landscapes, even though they don't shoot landscapes. Or maybe landscape photographers are an awfully vocal bunch :)
Title: Re: Next Canon step in the matter of their sensors
Post by: dr croubie on April 23, 2012, 10:19:00 PM
Quote
They don't care and think 5d3 sensor is fine
This.

Quote
They will run some marketing campaign admitting to mistake and offer some solution for future.
They will try to get next sensors from Sony as Nikon did
They will try to get sensors from another company than Sony because of some Nikon exclusive agreements
Not going to happen.
They didn't make a mistake, they knew exactly what they were doing.
They will never buy from Sony or Nikon, that's the worst thing for you, for me, for competition, and for the industry in general. Rewind a few years, and imagine that Canon bought Sony sensors too. The 5D2 wouldn't exist. The 1Ds3 wouldn't exist. The 7D/60D/550D/600D wouldn't exist. We'd all be stuck using the same 12MP D300/D700 sensors. The new 24MP NEX-7/D3200 sensor wouldn't exist either, without the competition from the 5D3 and 7D sensor, there's no incentive to go any better.

Quote
They will care a S___ and listen to pros who actually use cameras out in the field.
They do.
They did.
5D3 was what happened when they listened to the Pros.
The Pros they listened to were event photographers, wedding photographers, people whose livelihoods depend on being able to focus accurately in low light, being able to take the least noisy shots possible in low-light, silently from the back of the church or on a dimly lit stage.

Quote
They will run some marketing campaign to defend the results achieved with current 5d3 sensor
They don't have to.
As said, it works perfectly well for its intended audience.

Quote
They will work harder to get better results and follow the path they do now
Debatable.
The 1DX and 5D3 were both releases aimed exactly at target markets. The 1DX target market is for photojournalists who *must* get a shot, and *must* get it to the publishers (paper or 'net) ASAP, or they're out of a job. Hence the best Canon AF ever, the 12/14 fps, the ethernet port.
5D3 was aimed at the wedding photogs who can't afford the 1DX, who need silent shutter, who have to carry the damn thing all day and don't want a gripped body or need the best sealing.

The Pros they haven't listened to lately are the wildlife/birders, who are almost always focal-length limited, need more pixels-on-target, need f/8 focussing. The 1Dmk4 is still the camera for them. Will the 1D4 be replaced soon, in a 3D or 7D2 with more MP-on-target, and f/8 focussing?

They also haven't listened to Landscapers lately. They are the ones most disappointed by the 5D3 and the ones most likely to switch to the D800. They need the Low-ISO Dynamic Range that canon just isn't giving them. They need the Megapixels. They don't need ISO6400 performance.
Will they introduce a 30-40MP+ 1Ds/2/3/4D body to make landscapers happy?

With both of those segments, there's a *lot* less market share than for the photojournalist and wedding/event photographers that were targetted by the 1DX and 5D3 bodies. So maybe Canon will just ignore birders and landscapers forever, and continue on their path of most-market-share? I hope not, but they're going to be fine and making enough money from the 1DX and 5D3 bodies to not care.
Title: Re: Next Canon step in the matter of their sensors
Post by: LetTheRightLensIn on April 23, 2012, 10:25:06 PM
Quote
They don't care and think 5d3 sensor is fine
This.

Quote
They will run some marketing campaign admitting to mistake and offer some solution for future.
They will try to get next sensors from Sony as Nikon did
They will try to get sensors from another company than Sony because of some Nikon exclusive agreements
Not going to happen.
They didn't make a mistake, they knew exactly what they were doing.
They will never buy from Sony or Nikon, that's the worst thing for you, for me, for competition, and for the industry in general. Rewind a few years, and imagine that Canon bought Sony sensors too. The 5D2 wouldn't exist. The 1Ds3 wouldn't exist. The 7D/60D/550D/600D wouldn't exist. We'd all be stuck using the same 12MP D300/D700 sensors. The new 24MP NEX-7/D3200 sensor wouldn't exist either, without the competition from the 5D3 and 7D sensor, there's no incentive to go any better.

Quote
They will care a S___ and listen to pros who actually use cameras out in the field.
They do.
They did.
5D3 was what happened when they listened to the Pros.
The Pros they listened to were event photographers, wedding photographers, people whose livelihoods depend on being able to focus accurately in low light, being able to take the least noisy shots possible in low-light, silently from the back of the church or on a dimly lit stage.

Quote
They will run some marketing campaign to defend the results achieved with current 5d3 sensor
They don't have to.
As said, it works perfectly well for its intended audience.

Quote
They will work harder to get better results and follow the path they do now
Debatable.
The 1DX and 5D3 were both releases aimed exactly at target markets. The 1DX target market is for photojournalists who *must* get a shot, and *must* get it to the publishers (paper or 'net) ASAP, or they're out of a job. Hence the best Canon AF ever, the 12/14 fps, the ethernet port.
5D3 was aimed at the wedding photogs who can't afford the 1DX, who need silent shutter, who have to carry the damn thing all day and don't want a gripped body or need the best sealing.

The Pros they haven't listened to lately are the wildlife/birders, who are almost always focal-length limited, need more pixels-on-target, need f/8 focussing. The 1Dmk4 is still the camera for them. Will the 1D4 be replaced soon, in a 3D or 7D2 with more MP-on-target, and f/8 focussing?

They also haven't listened to Landscapers lately. They are the ones most disappointed by the 5D3 and the ones most likely to switch to the D800. They need the Low-ISO Dynamic Range that canon just isn't giving them. They need the Megapixels. They don't need ISO6400 performance.
Will they introduce a 30-40MP+ 1Ds/2/3/4D body to make landscapers happy?

With both of those segments, there's a *lot* less market share than for the photojournalist and wedding/event photographers that were targetted by the 1DX and 5D3 bodies. So maybe Canon will just ignore birders and landscapers forever, and continue on their path of most-market-share? I hope not, but they're going to be fine and making enough money from the 1DX and 5D3 bodies to not care.

1. More DR actually can be useful for wedding shooters.

2. It can really be useful for landscape shooters and the 5D3 IS also their top landscape cam so no I don't think they listened to the pros too well about it all or to the wrong set.
Title: Re: Next Canon step in the matter of their sensors
Post by: dr croubie on April 23, 2012, 10:47:41 PM
2. It can really be useful for landscape shooters and the 5D3 IS also their top landscape cam so no I don't think they listened to the pros too well about it all or to the wrong set.

Well yes, by default 5D3 is their best landscape camera.
It's also their best wildlife camera of their current models, tied with the 1Dx and 7D (all 3 have varying degrees of MP-On-Target, AF, and FPS), none of those 3 are as good as the 1D4 imho.

Maybe canon are planning a *true* landscapers camera, to compete with the D800?
Maybe they're planning a *true* replacement to the 1D4?
(Maybe they can combine the two together, high-MP FF sensor with f/8 AF, which will be a real D800-competition, but with 8+ fps?)

Maybe they're planning an MF camera system to compete with Hassy/Pentax/Phasemiya? (Because if your income depends on getting the best landscape shot with the most MP to print at the biggest sizes, then you're pretty much stuck with MF, at least you were before the D800)
Title: Re: Next Canon step in the matter of their sensors
Post by: jrista on April 23, 2012, 11:31:13 PM
Landscapes...in a heartbeat! The average sunset landscape can have a DR well above 12 stops, even above 14 stops. The more DR the better for landscape shots. Not that you will necessarily use all of it in the final product, but that you can always use the leeway when it comes to landscape shots. Being able to do even a touch of shadow recovery where necessary without encountering the color-blotched noise-infested pattern-ridden shadows that Canon is well-known for would be a godsend. Not to mention the improved sharpness on the D800e.

That's a fair answer, but based on all the b!tching, you'd think everyone is a landscape photographer all of a sudden. It's either that, or people who may have been in the market for a 5DIII don't like the idea that Nikon makes a better camera for landscapes, even though they don't shoot landscapes. Or maybe landscape photographers are an awfully vocal bunch :)

Hah, very true. Ironically, I think more people use higher ISO's than low, particularly ISO 100. If I leave my ISO on auto, even when I'm not shooting BIF under dark clouds, it only rarely reaches 100. It is usually at 160 or 200 at least, and often around 400 or so even in pretty darn good light. The only times my ISO is actually 100 is when I'm using full-manual mode for landscapes, particularly when I want long exposures with polarizing, ND and GND filtration. ISO 100 is pretty slow and frequently mandates extremely wide apertures or long shutter speeds, neither of which is ideal for a whole lot of photography.

I think it really boils down to the simple technicals. People here are largely Canon users, and Canon has fallen behind. No one likes their future upgrade track to offer less IQ than the competition for more money. Sadly, thats exactly the position Canon has left many of their customers in (even if its only perceptually, as there is more to IQ than just what the sensor is capable of.)
Title: Re: Next Canon step in the matter of their sensors
Post by: V8Beast on April 24, 2012, 01:02:00 AM
Hah, very true. Ironically, I think more people use higher ISO's than low, particularly ISO 100. If I leave my ISO on auto, even when I'm not shooting BIF under dark clouds, it only rarely reaches 100. It is usually at 160 or 200 at least, and often around 400 or so even in pretty darn good light. The only times my ISO is actually 100 is when I'm using full-manual mode for landscapes, particularly when I want long exposures with polarizing, ND and GND filtration. ISO 100 is pretty slow and frequently mandates extremely wide apertures or long shutter speeds, neither of which is ideal for a whole lot of photography.

I still think the 5DIII's files look better than the D800's, but subjective impressions aside, the Nikon is the better camera in terms of DR. Sure low ISO DR is useful, and I do plenty of ISO 100-400 shooting but I also crank the ISO up to 800-3200 very frequently for various reasons, and it's not always due to low light. I often turn up the ISO so I can stop down a lens for either maximum sharpness or deep DOF. Smaller apertures also come in handy for something like pan blurs as they yield a deeper focal plane and a bigger margin for focusing error. Someones maximum DR is important in these situations, but often times it isn't. Plus, the 5DIII has the edge in high ISO DR over the D800, but no one seems to care. DR is important, but it's just one of multiple factors that come into play when setting up a shot.

Quote
I think it really boils down to the simple technicals. People here are largely Canon users, and Canon has fallen behind. No one likes their future upgrade track to offer less IQ than the competition for more money. Sadly, thats exactly the position Canon has left many of their customers in (even if its only perceptually, as there is more to IQ than just what the sensor is capable of.)

I understand the money issue, and if I could have purchased the 5DIII for $500 less, I certainly wouldn't complain. Even so, if you really need the features the 5DIII has to offer, an extra $500 isn't going to be a deal breaker. If the extra $500 puts it out of the hands of hobbyists, that's a different story entirely. The message Canon seems to be sending is that they did their research, determined that the 5DIII's target audience are pros, packed it with features that pros wanted, and priced it accordingly. The funny thing is I don't see a $3,000 body as a "hobbyist friendly" price point, either.

If it's a money issue, I understand, but to harp on one metric of performance (DR) and determine that the 5DIII and the entire Canon system is junk because you can't afford one is a bit extreme. Many DR fanatics post links to blogs and sample images of other photographers as examples of situations where more low ISO DR is beneficial, rather than posting their own sample images. So, perhaps they're more infatuated with the idea of having more DR should they need it opposed to having personally experienced shooting situations where more DR would have been beneficial. 

It's like a 5'6" dude with a Napoleon complex that buys an SUV because of its off-road prowess, yet never actually goes off roading. He likes the idea of what the truck can do in theory even though he never uses its steep approach angle, locking differential, generous ground clearance, and big mud tires.

Photographers have been taking stunning images with cameras that have far less DR than the current generation of Canon's DSLRs. If someone's going to complain about it incessantly, then surely they must have images in which their gear - and not their technique - has let them down instead of having to rely on sample images posted from other photographers to prove their point :o
Title: Re: Next Canon step in the matter of their sensors
Post by: torger on April 24, 2012, 04:23:53 AM
Canon sensor is surely "good enough". It's easy to forget that the 5Dmk3 sensor is not bad, it is the Sony Exmor sensor in the D800 that is exceptionally good, in a class of its own. No sensor except Sony Exmor show these properties, and it is not too many cameras that have it.

However, once you've seen that it is possible to have cleaner dark colors than we were used to, it is hard not to desire it.

It is not about the Canon sensor being so bad it ruins pictures. It doesn't. The older version I lenses did not ruin pictures either, but still sharper version II are introduced to set a new quality standard. Sony has now raised the bar concerning sensor image quality, and Canon is currently lagging behind and shows no sign of having the technology to respond. Sure thing I like to whine a bit about that :-).
Title: Re: Next Canon step in the matter of their sensors
Post by: BillyBean on April 24, 2012, 04:37:38 AM
...

The Pros they haven't listened to lately are the wildlife/birders, who are almost always focal-length limited, need more pixels-on-target, need f/8 focussing. The 1Dmk4 is still the camera for them. Will the 1D4 be replaced soon, in a 3D or 7D2 with more MP-on-target, and f/8 focussing?

They also haven't listened to Landscapers lately. They are the ones most disappointed by the 5D3 and the ones most likely to switch to the D800. They need the Low-ISO Dynamic Range that canon just isn't giving them. They need the Megapixels. They don't need ISO6400 performance.
Will they introduce a 30-40MP+ 1Ds/2/3/4D body to make landscapers happy?

With both of those segments, there's a *lot* less market share than for the photojournalist and wedding/event photographers that were targetted by the 1DX and 5D3 bodies. So maybe Canon will just ignore birders and landscapers forever, and continue on their path of most-market-share? I hope not, but they're going to be fine and making enough money from the 1DX and 5D3 bodies to not care.

WELL SAID! I think a great many folks on this forum are enthusiasts (me too) but who seem to forget that Canon are a BUSINESS not a CHARITY designed to help photographers. Ultimately, they don't give a stuff about current or prospective customers *except* in so far as they can make a few bucks from them, now and in future. Lose sight of that truth at your peril.

This is the joy of capitalism - love it or loath it. Personally, I think it works rather well.

If Canon can make a camera that suits birders and landscape well, without destroying other market segments in the process, and make money off it, then they undoubtedly will. But reality bites: they will not destroy other market segments, so expect AF to be worse, or other stuff that maybe landscape doesn't really need. So my hunch - expect a 5D2 with an uprated sensor, but still rubbish AF.

Title: Re: Next Canon step in the matter of their sensors
Post by: KeithR on April 24, 2012, 07:48:04 AM
1. More DR actually can be useful for wedding shooters.
I see very little evidence of that - most wedding photography has nothing to do with image quality per se, and more to do with dicking about with cross processing and other PP gimmicks that will work just fine regardless of the source of the Raws.

In fact, I'll very happily argue that bird photographers have far more need of great IQ than wedding photographers, and the 5D Mk III will make an excellent birding body, assuming enough focal length...

Quote
2. It can really be useful for landscape shooters
A tiny subset of the potential customer base for the 5D Mk III. For everyone else, it's a bloody superlative bit of kit - the first FF camera I've ever been actively interested in, and yes, I'm a bird 'tog.

Quote
and the 5D3 IS also their top landscape cam so no I don't think they listened to the pros too well about it all or to the wrong set.
Far more likely, the 5D Mk III isn't intended as a landscape camera.

This whole thing of DxO making such a big deal of the low ISO DR from the D800 to the extent that this characteristic alone has effectively raised it above all other DSLRs, is just another example of DxO bias: the weight they've given this one D800 metric - which it just happens to beat the 5D Mk III in - makes no sense whatsoever except in the context of "massaging" the final mark.

For many potential users, "amazing" base ISO DR is an irrelevance, and there are areas where the 5D Mk III's files are better than the D800's; and in most others, particularly once we're off base ISO, the files are essentially equal in terms of the end result.

Yet the D800 is the star simply because of its low ISO DR performance?

It's a bloody fix, pure and simple: give me excellent all-round file performance (and anyone who says that the 5D Mk III doesn't do that in spades, is a liar, an idiot, or both), stellar AF and high FPS any day. 

Despite what DxO and the habitual whiners and axe-grinders on here have to say about it.
Title: Re: Next Canon step in the matter of their sensors
Post by: 3kramd5 on April 24, 2012, 09:45:31 AM
Man, I'm getting really tired of "100mm = 160mm effective" and all that noise.

Those numbers are only useful to someone who on a regular basis shoots simultaneously with an APS-C body and a FF body and wants lenses which will frame the same.

Beyond that, it's useless. 100mm is 100mm.

It's especially absurd in the world of DSLRs when more people than ever are using SLR cameras and most have never so much as picked up a full frame body. If anything, they should go the other way (crop equivalent framing, not the other way around).

BRB, building an adapter for my SD960 so I can turn my 300mm into 1700mm.
Title: Re: Next Canon step in the matter of their sensors
Post by: Ivar on April 24, 2012, 11:11:46 AM
+1

Canon sensor is surely "good enough". It's easy to forget that the 5Dmk3 sensor is not bad, it is the Sony Exmor sensor in the D800 that is exceptionally good, in a class of its own. No sensor except Sony Exmor show these properties, and it is not too many cameras that have it.

However, once you've seen that it is possible to have cleaner dark colors than we were used to, it is hard not to desire it.

It is not about the Canon sensor being so bad it ruins pictures. It doesn't. The older version I lenses did not ruin pictures either, but still sharper version II are introduced to set a new quality standard. Sony has now raised the bar concerning sensor image quality, and Canon is currently lagging behind and shows no sign of having the technology to respond. Sure thing I like to whine a bit about that :-).
Title: Re: Next Canon step in the matter of their sensors
Post by: jrista on April 24, 2012, 11:35:39 AM
Man, I'm getting really tired of "100mm = 160mm effective" and all that noise.

Those numbers are only useful to someone who on a regular basis shoots simultaneously with an APS-C body and a FF body and wants lenses which will frame the same.

Beyond that, it's useless. 100mm is 100mm.

It's especially absurd in the world of DSLRs when more people than ever are using SLR cameras and most have never so much as picked up a full frame body. If anything, they should go the other way (crop equivalent framing, not the other way around).

BRB, building an adapter for my SD960 so I can turn my 300mm into 1700mm.

Its not that simple. Bird photographers like myself, for example, who explicitly appreciate cropped frame sensors DO like to know EXACTLY what the benefit of that cropping is. Knowing that, say, a 400mm telephoto lens produces an effective field of view similar to a 640mm lens on a full-frame due to a 1.6x crop or a 520mm due to a 1.3x crop is very valuable. Focal length is not just a physical measure when crop factor and effective FoV are taken into account for subjects like BIF.
Title: Re: Next Canon step in the matter of their sensors
Post by: Eengineer on April 24, 2012, 11:53:11 AM
They will use their existing research as they see fit. 
http://www.canon.com/news/2010/aug31e.html (http://www.canon.com/news/2010/aug31e.html) or this sensor for pixel peeping for APC-H sensor
http://www.canon.com/news/2010/aug24e.html (http://www.canon.com/news/2010/aug24e.html)
Title: Re: Next Canon step in the matter of their sensors
Post by: 3kramd5 on April 24, 2012, 12:27:31 PM
Knowing that, say, a 400mm telephoto lens produces an effective field of view similar to a 640mm lens on a full-frame due to a 1.6x crop or a 520mm due to a 1.3x crop is very valuable

How is that valuable? Do you care how a 100mm lens frames on medium format?

If you routinely use multiple formats and want to equip to frame similarly, great. If you don't, you're in the majority, and it doesn't buy anything.
Title: Re: Next Canon step in the matter of their sensors
Post by: NickCat on April 24, 2012, 01:01:59 PM
or they will just post a link to the techradar test with a big question mark ?

and ask how two DxO test can have different results.

http://www.canonrumors.com/forum/index.php?topic=5868.0 (http://www.canonrumors.com/forum/index.php?topic=5868.0)

none of the techradar results make much sense at all

Why don't TechRadar's results make sense? 

The one thing I walk away from those graphs with is the fact that Canon is obviously doing something pretty nasty during the in-body JPG conversion to have that large a discrepancy in the results. 

Maybe it's because I want to believe those results, but they just feel more feasible.  Especially with the D4 having higher scores then the 800.
Title: Re: Next Canon step in the matter of their sensors
Post by: marekjoz on April 24, 2012, 02:59:51 PM
Quote
They don't care and think 5d3 sensor is fine
This.

Quote
They will run some marketing campaign admitting to mistake and offer some solution for future.
They will try to get next sensors from Sony as Nikon did
They will try to get sensors from another company than Sony because of some Nikon exclusive agreements
Not going to happen.
They didn't make a mistake, they knew exactly what they were doing.
They will never buy from Sony or Nikon, that's the worst thing for you, for me, for competition, and for the industry in general. Rewind a few years, and imagine that Canon bought Sony sensors too. The 5D2 wouldn't exist. The 1Ds3 wouldn't exist. The 7D/60D/550D/600D wouldn't exist. We'd all be stuck using the same 12MP D300/D700 sensors. The new 24MP NEX-7/D3200 sensor wouldn't exist either, without the competition from the 5D3 and 7D sensor, there's no incentive to go any better.

Quote
They will care a S___ and listen to pros who actually use cameras out in the field.
They do.
They did.
5D3 was what happened when they listened to the Pros.
The Pros they listened to were event photographers, wedding photographers, people whose livelihoods depend on being able to focus accurately in low light, being able to take the least noisy shots possible in low-light, silently from the back of the church or on a dimly lit stage.

Quote
They will run some marketing campaign to defend the results achieved with current 5d3 sensor
They don't have to.
As said, it works perfectly well for its intended audience.

Quote
They will work harder to get better results and follow the path they do now
Debatable.
The 1DX and 5D3 were both releases aimed exactly at target markets. The 1DX target market is for photojournalists who *must* get a shot, and *must* get it to the publishers (paper or 'net) ASAP, or they're out of a job. Hence the best Canon AF ever, the 12/14 fps, the ethernet port.
5D3 was aimed at the wedding photogs who can't afford the 1DX, who need silent shutter, who have to carry the damn thing all day and don't want a gripped body or need the best sealing.

The Pros they haven't listened to lately are the wildlife/birders, who are almost always focal-length limited, need more pixels-on-target, need f/8 focussing. The 1Dmk4 is still the camera for them. Will the 1D4 be replaced soon, in a 3D or 7D2 with more MP-on-target, and f/8 focussing?

They also haven't listened to Landscapers lately. They are the ones most disappointed by the 5D3 and the ones most likely to switch to the D800. They need the Low-ISO Dynamic Range that canon just isn't giving them. They need the Megapixels. They don't need ISO6400 performance.
Will they introduce a 30-40MP+ 1Ds/2/3/4D body to make landscapers happy?

With both of those segments, there's a *lot* less market share than for the photojournalist and wedding/event photographers that were targetted by the 1DX and 5D3 bodies. So maybe Canon will just ignore birders and landscapers forever, and continue on their path of most-market-share? I hope not, but they're going to be fine and making enough money from the 1DX and 5D3 bodies to not care.

I appreciate your comments and feedback. It's good written.
Title: Re: Next Canon step in the matter of their sensors
Post by: dswatson83 on April 24, 2012, 03:30:07 PM
You are assuming that the 5D mark III has a crappy sensor. Other than the DxO test, I have not seen a single report complaining about the sensor. Most reviews have highly praised the mark III in every way. Though I do think that the MIII was more about perfecting the sum of the parts than working on a specific area. Same with the 1DX which uses even a smaller 18MP and not offering a larger MP body. The Mark III is a clear improvement in every area, including the sensor. And if you believed that DxO was the benchmark for sensor technology, you would sell your D4 for a D7000.
Title: Re: Next Canon step in the matter of their sensors
Post by: marekjoz on April 24, 2012, 04:04:58 PM
You are assuming that the 5D mark III has a crappy sensor. Other than the DxO test, I have not seen a single report complaining about the sensor. Most reviews have highly praised the mark III in every way. Though I do think that the MIII was more about perfecting the sum of the parts than working on a specific area. Same with the 1DX which uses even a smaller 18MP and not offering a larger MP body. The Mark III is a clear improvement in every area, including the sensor. And if you believed that DxO was the benchmark for sensor technology, you would sell your D4 for a D7000.

@drwatson83 - If you're referring to me, then no - I don't think 5d3 has a crappy sensor. I don't own one, so really cannot say anything but only write anything what I've read, so really nothing interesting in this area :). I've read some comments (also from people in the field) and benchmarks and fairly speaking I think I know what kind of animal this camera is. Having 7d and 5d2 I can imagine how it behaves and what IQ it has. What I've read made me rather think that publicity for 5d3 is not quite good comparing of course to d800. So this was rather a reason for this survey as well as interest in your opinion for Canon's reaction for all these benchmarks, comparisons and "vs'es"

What I've read as far I think drcrouble has really summarized it quite well. Also neuro has proved that dxo numbers has nothing to deal with the sales results and canon was always far behind nikon in dxo ratings. Now it's a bit different than before but there are always another circumstances, so I's rather not really worth mentioning.

BTW - The reason I don't own 5d3 is not my doubt in this camera, but rather the fact that I didn't need it so much and when I was decided to get one, it stopped to be available where I live.
Title: Re: Next Canon step in the matter of their sensors
Post by: LetTheRightLensIn on April 24, 2012, 07:45:22 PM
or they will just post a link to the techradar test with a big question mark ?

and ask how two DxO test can have different results.

http://www.canonrumors.com/forum/index.php?topic=5868.0 (http://www.canonrumors.com/forum/index.php?topic=5868.0)

none of the techradar results make much sense at all

Why don't TechRadar's results make sense? 

The one thing I walk away from those graphs with is the fact that Canon is obviously doing something pretty nasty during the in-body JPG conversion to have that large a discrepancy in the results. 

Maybe it's because I want to believe those results, but they just feel more feasible.  Especially with the D4 having higher scores then the 800.

because the DR improvements go flat on the low end to a degree that nobody else has found to be true and the relative differences between cameras don't remotely match what others have found

and, if you take shots, and play around with them, the relative performance between cameras still doesn't match what they got