Canon EOS sensors, and technology

[Phil Indeblanc]

What I find odd in a number of sensor discussions and comparisons of 35 to MF, The size of the sensor takes the main topic of focus while other aspects are sidelined and one of those is more simple and an integral reasons why MF is better quality.

I read a number of posts saying that its only for enthusiasts and such, which is not true. There are legitimate fields of photography beyond landscape that the MF is crucial, such as high end product, and many other macro related shooting that needs details, also some medical segments. Just because many things get dummied down to the web doesn't mean that's it. There are fine print publications and most importantly, trade shows that surge our economy and industries world wide... they need large clear prints.

So the main feature MF renders better IQ is simple. Its the AA filter I have been asking Canon to make optional for a decade now.

There's a reason why Nikon with D800E, and Sony A7R have product lines without the AA filter.
There's a reason why you read in PRO and top forums the bridge from MF and 35 is getting tiny due to the D800E, and A7R!
There's a reason why you have some shooters waiting for Canon, while others are switching to Sony or Nikon for the need.(I'm holding out as much as I can, and at the threshold).

The other reason used to be the photosite size. MF used to be 11-12, 9 microns. Even Canon 1Ds had large ones.
But the tech is getting cleaner, and sophisticated with lenes and such..as we see MF dB's even scale down the microns to around 5-6 to get higher mpixel count.
Anyway...

Bottom line is that the AA free images have a 3D quality mainly due to the image not getting filtered. I remember shooting with the old Kodak DSLRc. The images were very comparable to the PhaseOne P25 files IQ(never mind the color issues it had). Except one was 14~mp vs 22~mp. Like the MF.

It would be my dream for canon to have its own Foveon-like sensor in a 30-40mp of the highest IQ.

No DOUBT Foveon has something special. No doubt they are being protective about it too long for their own good as other technologies develop. License the thing, or use a Canikon mount and let developers find solutions for the megapixel limit. I wonder if Sony has their own version? I think Fuji does with the "Honeycomb" pattern.

CANON!!! Please come up with a 28+mp AA FREE CAMERA!!! How long do we have to wait? If its so important for a few scientists that have little real world expereince, how hard is it to give the option of putting some filter on the sensor, on the glass, or just a model option for those rare niche shooters who do close-up textile work, or those that worry about moire. It rarely happens and is easy to correct. Don't look at it as a flaw. Or buy Foveon from Sigma and make a real IQ based camera. Blow the doors off others and take that segment too!!! DO IT NOW!!(in my best Arnold voice).

 

[mackguyver]

Have you seen the comparisons between AA filtered and non-AA filter cameras? The differences are miniscule, even at 100% and only then using the best lenses at their sharpest (usually f/5.6) apertures. To me, it's more of a marketing gimmick, but Canon has reduced the intensity of their AA filters over the years. If you really want sharp landscapes, buy the TS-E 24 II or 17 lenses. These lenses and their Tilt will give you way sharper and way more in-focus images than you'll see by removing the AA filter.

 

[Phil Indeblanc]

Have you seen the comparisons between AA filtered and non-AA filter cameras? The differences are miniscule, even at 100% and only then using the best lenses at their sharpest (usually f/5.6) apertures. To me, it's more of a marketing gimmick, but Canon has reduced the intensity of their AA filters over the years. If you really want sharp landscapes, buy the TS-E 24 II or 17 lenses. These lenses and their Tilt will give you way sharper and way more in-focus images than you'll see by removing the AA filter.

To you, YES. NO to all other counts relative to IQ.
Maybe miniscule in people, landscapes and the like. For many product macro shoots, it is a very different story. It's obvious you haven't needed to photograph macro products, or even needed the use of multiple systems and testing. If you did, you would never say this. I also know of some medical use that a AA free sensor will gain from. So for your needs, this conversation doesn't really make a difference.

For someone who has actually used the Canon's or Nik since/before the 10D, 1Ds, 5Dm2... against OLDER(than the 1Ds/5D) sensors like a Kodak Pro SLRc, it is easy to see that the non AA sensor has the IQ, and 3D like sharpness and look of the MF( CCD also plays a role). Why do you think Foveon users are such lovers of their cameras, (until they hit the mpixel limit)? Its not debatable. Look at DXo. Look at the D800E, the A7R. Then tell me it is a gimmick. Have you used these cameras? The difference is a good chunk of a bite they gain. I opt for Medium Format because what I shoot demands it. My customers demand it. Life would be MUCH easier if I didn't have to bust out a 4x5/MF dB for a shoot. Work would be much smoother.

 

[Hannes]

and not to forget about moire either, now that is something that will actually ruin images whereas an AA filter will make then slightly less crisp.

 

[Phil Indeblanc]

and not to forget about moire either, now that is something that will actually ruin images whereas an AA filter will make then slightly less crisp.

How long have you been shooting for?

I keep hearing this, and from a scientific standpoint, it maybe something to consider. Do you think Nikon AND Sony would launch it if it were as big an issue as you think?

Do you think ALL MF camera users would use the optional AA filter if they had to? We don't, unless we shoot some close up textile stuff. It just makes it easier for post.

Otherwise, with about 400K images in my database, I think I came across this happening once, and all it took was a click in editing. DONE! Sorry if I come off strong on this, but it is sad to hear this song played everytime the AA conversation comes up. In real world experience it is hardly an issue. I think an EASY choice to not have it for image quality.
Yes some sensor specs and sizes make it more atuned to getting moire, yet I don't see anyone care about it except in the lab.

 

[unfocused]

CANON!!! Please come up with a 28+mp AA FREE CAMERA!!! How long do we have to wait? ...DO IT NOW!!(in my best Arnold voice).

Okay. Here we go again.

People get upset when told this, but the truth is, it all boils down to economics.

You believe there is a market for the product you want. But, is that market large enough and sufficiently competitive to justify Canon to take the action you request? The only one who knows for sure would be Canon and they are not talking.

No company can survive going after 100% of potential customers. There are always customers that have to be left on the table because it just isn't possible to serve them and make sufficient profit – the people who want to buy a 5DIII at $1,200; those who want a Medium Format body at half the price of current competitors; etc. etc.

Companies have limited resources and a responsibility to put those resources where they will offer the best return. Canon has emphasized cinema in recent years. They obviously have determined that the market justifies their investment. That may be why they haven't released a high megapixel camera yet...the market may not justify the investment.

The hard truth is this: What we as individuals want is irrelevant. What we in the aggregate, comprising tens of thousands or even millions of like-minded consumers want is all-important.

Killing off the AA filter just isn't important enough to enough consumers to justify it at this point, regardless of whether or not it might improve image quality.

 

[Don Haines]

Let's start with a bit of logic.....

A medium format sensor is typically in the 50mm x 40mm size range..... about 2000 square mm's

A FF sensor is typically 36mm x 24mm, or about 860 square mm's

A crop sensor is typically 22.4mm x 14.8mm, or about 330 square mm's

The technology is the same for all of them.... yes, the number of pixels and the size of the pixels varies, but the underlying technology is the same.... light shines onto a sensor where the photoreceptors are arranged in a simple geometric pattern. the only significant difference between them is that with more surface area, a larger sensor can have larger pixels and that means better low light performance. With the ratio in size, and given the same level of technology and number of pixels, that give FF about 1.5 stops better performance than APS-C and MF about 1.5 stops better performance than FF.... The FF/MF debate is the same as the APS-C/FF debate...

If the image that you are trying to capture also has a geometric pattern, there will be an interference pattern created between the two geometric patterns. The closer these two are in size, the more pronounced the pattern is.

Some people claim to never have problems with moire... and they are right! For the images they shoot they don't have the problem of geometric pattern interference.

Some people claim that moire is a big problem.... and they are right! For the images they shoot, geometric pattern interference is a problem.

The AA filter is a half-way measure.... it does not eliminate all moire, but it gets rid of a lot... it is a balancing act between artifacts and sharpness.... it was chosen as the best GENERAL PURPOSE solution because there is no way that Canon is going to come out with ten cameras with AA filters ranging from strong to none.

There is no almighty "one solution for all"... but there is one general consideration here... As focusing gets better and as the quality of your lenses gets better, and as your technique gets better.... pictures become sharper and moire is more noticeable.

We should all remember that just because we do not have a problem, it does not mean the problem does not exist. And we should also remember that because we do have a problem, not everyone is affected.

 

[Phil Indeblanc]

There is no almighty "one solution for all"... but there is one general consideration here... As focusing gets better and as the quality of your lenses gets better, and as your technique gets better.... pictures become sharper and moire is more noticeable.
We should all remember that just because we do not have a problem, it does not mean the problem does not exist. And we should also remember that because we do have a problem, not everyone is affected.

Unfocused...What does it mean "OK, here we go again" ? Is it maybe that this voice is getting louder? Maybe?

One thing is certain. If Canon DOESN'T do anything about the still image evolution, and squarely concentrates on Motion-video, they will drop a large segment of users. A very small insignificant amount have already adapted to the A7R for lens options, crisp clear sharp images, portability, mpixel, very well suited for landscape, architecture, automotive, product, and many other things. For a numberi of users it has been the alternate for MF.

I'm certainly not saying for Canon to jump into MF market. I am saying to up the pixel count and give the OPTION of having no AA filter. (maybe a clever way of a insert between the lens and sensor. That's what the top engineers are for).

Don your response is most logical and agreeable.

 

[Don Haines]

There is no almighty "one solution for all"... but there is one general consideration here... As focusing gets better and as the quality of your lenses gets better, and as your technique gets better.... pictures become sharper and moire is more noticeable.
We should all remember that just because we do not have a problem, it does not mean the problem does not exist. And we should also remember that because we do have a problem, not everyone is affected.

Unfocused...What does it mean "OK, here we go again" ? Is it maybe that this voice is getting louder? Maybe?

One thing is certain. If Canon DOESN'T do anything about the still image evolution, and squarely concentrates on Motion-video, they will drop a large segment of users. A very small insignificant amount have already adapted to the A7R for lens options, crisp clear sharp images, portability, mpixel, very well suited for landscape, architecture, automotive, product, and many other things. For a numberi of users it has been the alternate for MF.

I'm certainly not saying for Canon to jump into MF market. I am saying to up the pixel count and give the OPTION of having no AA filter. (maybe a clever way of a insert between the lens and sensor. That's what the top engineers are for).

Don your response is most logical and agreeable.

It is hard to guess the future....

So what happens if Canon goes mirrorless on their DSLR's.... If they keep the existing lenses you now have all this space where the mirror used to be.... perhaps they could have an AA filter that slips in and out... and perhaps you can have the option of replacing the AA filter like we replaced focus screens back in the good old days...

There is always a way...

P.S. If there are a bunch of Canon engineers reading this and saying "Brilliant! why didn't we think of that"... feel free to ship me a free 600F4

 

[Loswr]

Its not debatable. Look at DXo. Look at the D800E, the A7R. Then tell me it is a gimmick. Have you used these cameras? The difference is a good chunk of a bite they gain.

You are arguing that lack of an AA filter results in a sharper image, and citing DxO as evidence? Please point to a DxO reference where they state that they test sensor sharpness, or point to examples of the same lens tested on a D800E or a7R compared to that lens tested on a D800.

Comparing the D800E to the D800, images from the latter are much more amenable to sharpening in post, and when properly sharpened, images from the two are not significantly different.

One thing is certain. If Canon DOESN'T do anything about the still image evolution, and squarely concentrates on Motion-video, they will drop a large segment of users. A very small insignificant amount have already adapted to the A7R for lens options, crisp clear sharp images, portability, mpixel, very well suited for landscape, architecture, automotive, product, and many other things. For a numberi of users it has been the alternate for MF.

I'm certainly not saying for Canon to jump into MF market. I am saying to up the pixel count and give the OPTION of having no AA filter. (maybe a clever way of a insert between the lens and sensor. That's what the top engineers are for).

Nikon released the D800 and the D800E, and expected to sell fewer of the D800E. The 5DIII, with its AA filter and lower MP count, has outsold both. Where's that 'large segment of users' that Canon has dropped?

I'll echo 'here we go again' – another prediction of doom for Canon if they don't make the product you want. This is a common refrain from people who don't seem to realize or cannot accept that they are in the minority.

 

[Phil Indeblanc]

You are arguing that lack of an AA filter results in a sharper image, and citing DxO as evidence? Please point to a DxO reference where they state that they test sensor sharpness, or point to examples of the same lens tested on a D800E or a7R compared to that lens tested on a D800.

I'm not arguing the point, it is a fact.

Comparing the D800E to the D800, images from the latter are much more amenable to sharpening in post, and when properly sharpened, images from the two are not significantly different.

No such thing as creating real information in post. Refer to answer one.

Nikon released the D800 and the D800E, and expected to sell fewer of the D800E. The 5DIII, with its AA filter and lower MP count, has outsold both. Where's that 'large segment of users' that Canon has dropped?

Didn't the D800E outsell the D800?!

OK, I shouldn't say large segment. But "ground breaking" changes or improvements are what big tech companies need to keep the spirit of innovation alive, don't you think? People who echo these sometimes small, and sometimes game changing innovations are what can snowball the market direction.

I'll echo 'here we go again' – another prediction of doom for Canon if they don't make the product you want. This is a common refrain from people who don't seem to realize or cannot accept that they are in the minority.

Doom for Canon? No.
I am certainly aware that I'm in the minority.
What you might be saying is that minorities don't have a voice in large organizations that only tailor product to the general standard/mediocre requirements for the simple equasion of maximum profit gain for the minimal output? Are you saying Canon has no more canon-balls? If other companies are using such innovations to gain market share, doesn't Canon have the right to protect its integrety and say, "Ya, we know all about that, and yes that is cool, But, BAM!, try this on for size!" And keep the band marching? I certainly don't expect for Canon to be quiet and say, "Hmmm, that sucks, they did it first, so we will ignore it and just work on more video related stuff...If we ignore it, it will go away".

 

[Loswr]

You are arguing that lack of an AA filter results in a sharper image, and citing DxO as evidence? Please point to a DxO reference where they state that they test sensor sharpness, or point to examples of the same lens tested on a D800E or a7R compared to that lens tested on a D800.

I'm not arguing the point, it is a fact.

Yes, it is a fact. The magnitude of that difference is not so obvious. Detail that is present but blurred in a predictable manner can be brought out in post. (Side note - optical microscopes can now resolve beyond the Abbé diffraction limit, and one way of achieving that uses post-processing analysis of moiré resulting from patterned illumination, i.e. since the pattern is predictable, detail not present in the image can be extrapolated mathematically.)

Also, depending on the lens much of that extra detail may not be there to begin with, which is why the D800 with a Nikon 24-70/2.8 barely outresolves a 5DIII with a Canon 24-70/2.8 II, despite the 60% higher MP count of the D800's sensor.

Regardless, your statement, "Look at DxO," has absolutely no bearing on the fact at hand. Why invoke a completely irrelevant source of information to support your contention? It certainly does not benefit your credibility.

OK, I shouldn't say large segment. But "ground breaking" changes or improvements are what big tech companies need to keep the spirit of innovation alive, don't you think? People who echo these sometimes small, and sometimes game changing innovations are what can snowball the market direction.

Dual pixel AF could be considered a 'groundbreaking' innovation. Canon can certainly innovate, however they tend to bring forth innovations that impact large market segments rather than niche market segments.

Doom for Canon? No.
I am certainly aware that I'm in the minority.
What you might be saying is that minorities don't have a voice in large organizations that only tailor product to the general standard/mediocre requirements for the simple equasion of maximum profit gain for the minimal output? Are you saying Canon has no more canon-balls? If other companies are using such innovations to gain market share, doesn't Canon have the right to protect its integrety and say, "Ya, we know all about that, and yes that is cool, But, BAM!, try this on for size!"

The question is, are those other companies gaining market share? In particular, Nikon cited slower than expected sales of high end dSLRs (i.e., the D800) as a reason for the YTD loss they posted last quarter. It's also relative - a significant gain for Sony (not that there's evidence of that) would be a minor loss for Canon.

 

[mackguyver]

Have you seen the comparisons between AA filtered and non-AA filter cameras? The differences are miniscule, even at 100% and only then using the best lenses at their sharpest (usually f/5.6) apertures. To me, it's more of a marketing gimmick, but Canon has reduced the intensity of their AA filters over the years. If you really want sharp landscapes, buy the TS-E 24 II or 17 lenses. These lenses and their Tilt will give you way sharper and way more in-focus images than you'll see by removing the AA filter.

To you, YES. NO to all other counts relative to IQ.
Maybe miniscule in people, landscapes and the like. For many product macro shoots, it is a very different story. It's obvious you haven't needed to photograph macro products, or even needed the use of multiple systems and testing. If you did, you would never say this. I also know of some medical use that a AA free sensor will gain from. So for your needs, this conversation doesn't really make a difference.

For someone who has actually used the Canon's or Nik since/before the 10D, 1Ds, 5Dm2... against OLDER(than the 1Ds/5D) sensors like a Kodak Pro SLRc, it is easy to see that the non AA sensor has the IQ, and 3D like sharpness and look of the MF( CCD also plays a role). Why do you think Foveon users are such lovers of their cameras, (until they hit the mpixel limit)? Its not debatable. Look at DXo. Look at the D800E, the A7R. Then tell me it is a gimmick. Have you used these cameras? The difference is a good chunk of a bite they gain. I opt for Medium Format because what I shoot demands it. My customers demand it. Life would be MUCH easier if I didn't have to bust out a 4x5/MF dB for a shoot. Work would be much smoother.

LOL, you're in your own world, and obviously there's no arguing with you.

 

[Don Haines]

OK, I shouldn't say large segment. But "ground breaking" changes or improvements are what big tech companies need to keep the spirit of innovation alive, don't you think? People who echo these sometimes small, and sometimes game changing innovations are what can snowball the market direction.

I think we are on the edge of a shift in digital cameras.

We need to step back and ask "why mirrors". In the days of film, you needed the mirror and optical viewfinder to know what you were looking at and we needed focusing screens to know if we were in focus.... Then came digital sensors and we treated them like film... because that is what we were used to.

A digital sensor is NOT film. It has different strengths and different weakness.... and the mirror is no longer the only way to see through the lens. A decent mirrorless camera (and there are several on them out there) will be designed to the strengths of digital technology. They already do many things better than DSLRs, but a great mirrorless camera will have to do everything better. Right now, the two big stumbling blocks are focusing and viewfinders.

Dual pixel technology may well be the end of the focusing dilemma... and as it matures we should be able to have far more capable focusing systems on mirrorless cameras than with DSLRs... the point I keep bringing up is that we should be able to recognize a bird and track it as it flies through the air, even though the operator is not steady. We already have P/S cameras that recognize individual faces and can even tag them for use on social media and I have a waterproof P/S that has "cat mode" and "dog mode" and when you put it in "cat mode" it tracks the face of the cat and not the dog so please don't tell me this is a far-fetched idea... It's not coming, it's already here!

The second stumbling block is viewfinders. Right now, optical viewfinders are better than EVFs.. A few years ago EVFs were garbage... there are some real nice ones now.... who knows what the future will bring? At some point, people will stop trying to design an EVF to be like an optical viewfinder and design them to the strengths of digital... perhaps they will get a bit bigger... perhaps you will have a little window open up on it to check focus at 10X... or exposure preview.... or whatever... but until they stop pretending it is optical they will be inferior. I am sure that what is currently in the labs is good enough for the real world.... we are that close.

I can see Canon coming out with a new camera that shakes things up. I would love to see a quad-pixel technology 7D mirrorless camera where you could address the sub-pixels individually for a 24megapixel image with similar ISO and noise to the 70D, or bin them together for a 6megapixel low-light camera that had better low light performance than a 1DX....

Canon has a HUGE R+D department.... they are not all siting on their rear ends playing solitaire... something is coming and the delays to the successor to the 7D may just mean that the change is big.

 

[Don Haines]

I opt for Medium Format because what I shoot demands it. My customers demand it. Life would be MUCH easier if I didn't have to bust out a 4x5/MF dB for a shoot. Work would be much smoother.

Are you using a digital back on your 4X5? If so, which one and do you like it?

 

[unfocused]

Unfocused...What does it mean "OK, here we go again" ? Is it maybe that this voice is getting louder? Maybe?

No. It means exactly what I wrote in my post. Someone convinces themselves that there is some feature that they absolutely have to have and then they make the leap that if they don't get it, Canon is making a huge and costly mistake. So they start a thread on this forum to whine about it.

You are fixated on the AA filter.

This same thread gets repeated over and over again with the only difference being the obscure feature that the individual has fixated on -- dynamic range, shadow banding, sensor size, number of megapixels, the list goes on and on.

The only voices getting louder are the imaginary ones in people's heads.

Canon is a business. A very successful business. There isn't anything anyone on this forum can come with that Canon has not considered and researched in far greater detail.

If it comes to a point where a business case can be made for changing the AA filter, they will change it. But, starting a forum thread and shouting "Hurry up Canon" isn't going to change anything.

 

[jrista]

OK, I shouldn't say large segment. But "ground breaking" changes or improvements are what big tech companies need to keep the spirit of innovation alive, don't you think? People who echo these sometimes small, and sometimes game changing innovations are what can snowball the market direction.

I think we are on the edge of a shift in digital cameras.

We need to step back and ask "why mirrors". In the days of film, you needed the mirror and optical viewfinder to know what you were looking at and we needed focusing screens to know if we were in focus.... Then came digital sensors and we treated them like film... because that is what we were used to.

A digital sensor is NOT film. It has different strengths and different weakness.... and the mirror is no longer the only way to see through the lens. A decent mirrorless camera (and there are several on them out there) will be designed to the strengths of digital technology. They already do many things better than DSLRs, but a great mirrorless camera will have to do everything better. Right now, the two big stumbling blocks are focusing and viewfinders.

Dual pixel technology may well be the end of the focusing dilemma... and as it matures we should be able to have far more capable focusing systems on mirrorless cameras than with DSLRs... the point I keep bringing up is that we should be able to recognize a bird and track it as it flies through the air, even though the operator is not steady. We already have P/S cameras that recognize individual faces and can even tag them for use on social media and I have a waterproof P/S that has "cat mode" and "dog mode" and when you put it in "cat mode" it tracks the face of the cat and not the dog so please don't tell me this is a far-fetched idea... It's not coming, it's already here!

The second stumbling block is viewfinders. Right now, optical viewfinders are better than EVFs.. A few years ago EVFs were garbage... there are some real nice ones now.... who knows what the future will bring? At some point, people will stop trying to design an EVF to be like an optical viewfinder and design them to the strengths of digital... perhaps they will get a bit bigger... perhaps you will have a little window open up on it to check focus at 10X... or exposure preview.... or whatever... but until they stop pretending it is optical they will be inferior. I am sure that what is currently in the labs is good enough for the real world.... we are that close.

I can see Canon coming out with a new camera that shakes things up. I would love to see a quad-pixel technology 7D mirrorless camera where you could address the sub-pixels individually for a 24megapixel image with similar ISO and noise to the 70D, or bin them together for a 6megapixel low-light camera that had better low light performance than a 1DX....

Canon has a HUGE R+D department.... they are not all siting on their rear ends playing solitaire... something is coming and the delays to the successor to the 7D may just mean that the change is big.

Very well put. Especially that last bit...the change would really have to be big like that, for it to be justifiable. Otherwise, it's just a demonstration of a major Canon blunder, if the 7D II comes out and is a mediocre improvement over the 7D, and not much in terms of competition against counterpart offerings from other brands.

Regarding the two points about viewfinders and focusing. I'm not sure were "nearing the end" of the issues. I think DPAF marks the beginning of finally moving down the right path, however I think there is a lot of innovation along that path that needs to take place before you start seeing action photographers seriously think about dumping their dedicated AF sensors and familiar AF points for a mirrorless image-sensor-based AF system. DPAF should at least become QPAF, so we can detect phase in at least two directions. I think we may ultimately need to see one further innovation, dual-direction QPAF, where you have horizontal and vertical with one half of the sensor's pixels, as well as phase detected diagonally in two perpendicular directions with the other half of the sensor's pixels. Only then would you be technologically similar to how dedicated PDAF sensors are designed, and only then could you really start building advanced firmware to really produce high rate, high accuracy AF without a dedicated AF sensor.

There is still another problem, however, that mirrorless AF systems will need to overcome before they can really achieve parity with their dedicated AF system counterparts: Low Light Sensitivity. Modern dedicated AF systems are sensitive to light down to the -2 to -3 EV range. Not only that, each dedicated PDAF point receives a tiny fraction of the total light entering the lens (thanks to passing through a half-silvered mirror and an AF unit splitting lens), and each line sensor that comprises an AF point recieves at most half of that tiny fraction of total light. All that, down to at least f/5.6, and in "pro" grade cameras, down to f/8. Dedicated PDAF sensors are ludicrously sensitive to the smallest amount of light...and largely thanks to the fact that they can be fabricated independently of the image sensor, so they can be explicitly designed with huge photodiodes in each line sensor that have massive SNR. I'm not sure how camera manufacturers will overcome this issue, as even at very high ISO settings, image sensors are nowhere near as sensitive as the photodiodes in PDAF sensors. I'm sure one of the big manufacturers will figure out something brilliant to solve this problem...but I think it is definitely something that needs to be dealt with.

(BTW, I am aware that Canon's current DPAF supports live view focusing up to f/11, however the speed of that focusing is nowhere even remotely close to as fast as a dedicated PDAF unit. The slower speed gives DPAF a bit of an advantage in that area...similar to the advantage Canon creates when they force a slower AF rate when attaching one of their teleconverters to a lens.)

As for EVFs, I can only hope they get significantly better. I'm very curious to see what Canon does with their Hybrid VF...I wonder how that will ultimately work, and whether it will be as flexible and user configurable/selectable as it really needs to be to be a success. I suspect it will be rather inflexible, and only activate the EVF under very specific circumstances (such as recording video).

 

[tolusina]

........What does it mean "OK, here we go again" ? …...

It means that temple guy has been here already, now he's gone. Are you he?

..... sensor discussions and comparisons of 35 to MF …...

Stop. Stop right there, please? Just stop.




.

 

[verysimplejason]

Hmmm... Just buy a Sony A7R + EF adapter. Since you're doing food photography and other macros, you don't need that "better" AF. ;D You don't need to stay with one brand (and I know you don't). Give some love to other brands also.

 

[Phil Indeblanc]

Yes, it is a fact. The magnitude of that difference is not so obvious. Detail that is present but blurred in a predictable manner can be brought out in post. (Side note - optical microscopes can now resolve beyond the Abbé diffraction limit, and one way of achieving that uses post-processing analysis of moiré resulting from patterned illumination, i.e. since the pattern is predictable, detail not present in the image can be extrapolated mathematically.)

Also, depending on the lens much of that extra detail may not be there to begin with, which is why the D800 with a Nikon 24-70/2.8 barely outresolves a 5DIII with a Canon 24-70/2.8 II, despite the 60% higher MP count of the D800's sensor.

I thought DXO showed the D800E as having the highest IQ if not almost as high as the Phase One IQ280 ? I could be wrong as I didn't study it, it was something I read in discussion of MF DB's. If that is not the case, please do list the top 3-5 markers of IQ in the DXO testing.

Do all the calculations you want. I am shooting for over 20 years now.
I have done my side by side apples to apples test using a Leica R Macro 60 lens mounted on each on a studio stand with controlled lighting. If you shoot jewelry, you will know the difference without a second thought.

Yes Don, I use a P25/P45 and have used, Kodak pro back, Blad CF39, Sinar evol75 on a Sinar 4x5. I have used Nikon before, but focus was horrible about 10+ years ago. Switched to Canon and have loved it for the product and service. I think they can do better in specialized features.

I use it in the studio on still subjects. For people I often use it for portraits that are slow moving. Otherwise I use the 5D mark2.

There is a significant difference in the lowest end which is the P25(22MP) DB vs a 5D mark2. Regardless of all the numbers these guys want to crunch. A good portion and I don't hesitate to say it is due to the AA filter.