• Please use real names.

    Greetings to all who have registered to OPF and those guests taking a look around. Please use real names. Registrations with fictitious names will not be processed. REAL NAMES ONLY will be processed

    Firstname Lastname

    Register

    We are a courteous and supportive community. No need to hide behind an alia. If you have a genuine need for privacy/secrecy then let me know!
  • Welcome to the new site. Here's a thread about the update where you can post your feedback, ask questions or spot those nasty bugs!

Will the Mk3 kill the Backs?

Here we go again, nothing more tasty than speculation threads just before the introduction of a brand new product... difficult to resist, so why not.

I talked today with a friend that is in the top of his field and shoots for the best artists and Galleries in this city. He has a Scanning back Betterlight, but most of his work was captured with the Canon Mk 2.

When we talked today he said that he just got an almost new Leaf 22MP back (Mamiya mount) at an incredible price for the conditions of the machine.

I was surprised because I knew he was thinking about the Mk3 for the last year, but he said something that makes sense: "too many pixels" in too small size sensor. The FX, --to use the format just invented by Nikon that applies exactly to Canon[-- is half the size of a 22MP DB, so the "sensels" have to be half.

It is the same thing that happens with other formats, for example all the point and shoot cameras that have 10MP are not the same as the DSLRs with same 10MP. It is true, the first are very convenient and some can be carried in the shirt pocket.

So, lets see what happens when the images are out...

What will probably will make things more interesting is when the backs begin to get closer to the 8k range of the Canon...

ah, and there is also the problem of resolving limitations of Canon glass...
 
"too many pixels" in too small size sensor."

The real question is whether more is too many ...? It depends on the quality threshold one sets.

So, lets see what happens when the images are out...

Yes, before that (and some fundamental objective testing) it cannot be more than (educated) speculation.

What will probably will make things more interesting is when the backs begin to get closer to the 8k range of the Canon...

Which is unlikely to happen (despite the Mamyia attempts). I is a (relative) niche market, and those physically larger sensor arrays cost exponentially more.

ah, and there is also the problem of resolving limitations of Canon glass...

Sure, good lens quality is always needed for a good MTF response of the total optical system (lens/AA-filter/micro-lenses/sensel shape). However, the lens alone is just one factor. The sensor array's sampling density is still the most determining factor for resolution, and the AA-filter adds some additional loss of modulation. The Bayer CFA also plays its role, since it (in particular) reduces chroma resolution. Dynamic range on the other hand, benefits from larger sensels.

Another important factor to consider in comparisons, is the relatively lower resolution of MF lenses. It's unavoidable due to the need to cover a larger image circle, which causes larger trade-off concessions in the reduction of residual lens aberrations. Its down-sides are somewhat compensated for by a lower need for output magnification, but at a price (literally).

We therefore are confronted with conflicting requirements, and it ultimately boils down to how the different factors that contribute to "image quality" are weighted. In addition, there are of course also other deciding factors like portability, long focal length reach, and sensitivity, that may be more important for certain types of usage.

Bart
 
Medium Format's user base is obviously more reduced than 35mm or smaller, but you have to consider that it is consolidating and absorbing the applications of large format from 4x5 to 8x10. So it is an upper class of photographers that are not going to, suddenly, become part of the middle class of Nikon/Canon carrying masses.

The need for IQ, will always exist.

"deciding factors like portability, long focal length reach, and sensitivity" are real, but in that case the 10MB point and shoot wins hands down: very portable, super zooms and high sensitivity -- at least as advertised-- and... cheap
 

Steve Saunders

New member
That info about smaller pixels meaning more noise is generally right. But Nikon with the D3 have managed something revolutionary with the high-ISO noise performance. Considering there is NO on-chip noise reduction, this makes it all the more remarkeable. I know the D3 only has 12MP, but then the Canon 1DsII only has a bit more at 16MP and the previous Canon FF model had around the same as the D3, but they come nowhere near to delivering the amount of detail and low noise as the new Nikon.
The point I'm making is that all of a sudden the DSLR technology has leapfrogged and the amount of detail that can be had from an FF or FX sensor is just amazingly good compared to what we've seen to date, from any DSLR or MF back. Bear in mind that MF backs tend to evolve more slowly than DSLR's, so I think for the moment FF DSLR's are going to have an edge until MF catches up.
 
That info about smaller pixels meaning more noise is generally right. But Nikon with the D3 have managed something revolutionary with the high-ISO noise performance. Considering there is NO on-chip noise reduction, this makes it all the more remarkeable.

Miracles, and especially the ones that deny the laws of physics, don't happen that often. In this particular case, and after looking at the D3 sample images, noise reduction is what makes the difference. Even with the (more agressive) noise reduction switched off, the Nikon images show a level of Chroma noise reduction that can't be explained by physics (you can't have different chroma and luminance noise performance unless it's postprocessed). It's not really surprising, because Nikon have been doing that ever since the D70 came to market (to the dismay of amateur astronomers) and even in their Raws with noise reduction switched off.

The point I'm making is that all of a sudden the DSLR technology has leapfrogged and the amount of detail that can be had from an FF or FX sensor is just amazingly good compared to what we've seen to date, from any DSLR or MF back.

Although it' s not as sudden as you think, there are a differences between CCD and CMOS technology, and it seems like CMOS technology has evolved much more and faster than the already mature CCD technology. It is easier, and cheaper, to make large quantities of CMOS devices than specialized CCDs (which also help getting research funding!). CMOS devices also generate much less heat, which helps to keep power consumption and noise down. A benefit of CCD based photoreceptors is that they potentially can have a higher Dynamic range because of a larger photoactive surface relative to similar pitch CMOS devices.

Bart
 
Steve, I agree that technology is moving fast and we will probably soon have more than we could want of everything .. cheap.

But the XF format will probably reach a limit of what it can resolve with the physical size of the sensor and image circle. I would be more than sufficient for the big masses of photographers.

But the question is: will the Mk3 kill the backs? I think probably not yet. But the fight will be something worth while to watch...
 
I don't think any 35mm DSLR is ever going to be truly capable of killing off the medium format backs. With the 1DsMkIII and the fabled D3X (if it is real) at around 22MP, we have pretty much reached the practical limit of the quantity of pixels in the 35mm form while retaining the most functions (read as ISO versus noise). The exception being something like the foveon sensor with the same number of photo-sites (20-25Million photosites in a 35mm format).

Meanwhile the medium backs can probably achieve in excess of 100MP before they get close to their practical limits. BUt since some folks are already balking at the file sizes, even this may not be practical. As always the medium backs will continue to exceed 35mm for overall quality, while 35mm will probably continue exceed the MF for utility.

Just my opinion,
 

Asher Kelman

OPF Owner/Editor-in-Chief
If one works as a pro with a good paying client base, a first rate DB pays for itself and will be better than Canon can offer except for fast shooting or low light.

I'm jumping to another way of doing things. LF for detail-rich slow a meditative work and large prints. Otherwise my current cameras and lenses do everything.

If one can manage with the DSLR one has, one can get off the replacement treadmill and just do photography perhaps!

Asher
 
I don't think any 35mm DSLR is ever going to be truly capable of killing off the medium format backs. With the 1DsMkIII and the fabled D3X (if it is real) at around 22MP, we have pretty much reached the practical limit of the quantity of pixels in the 35mm form while retaining the most functions (read as ISO versus noise). The exception being something like the foveon sensor with the same number of photo-sites (20-25Million photosites in a 35mm format).

I think there is still a lot of room for progress, but not in the way we know it today.

Given the trend for miniaturization, we can expect sensor arrays with approx. 1 micron sensel pitch. That will allow to exclude an AA-filter (and thus save a lot of cost), because the optical chain will serve as an effective low-pass filter. That data will be binned in adjustable quantities at the analog signal level (so before the ADC and Raw output) to reduce noise to the required level, or for the required resolution, sensitivity, or file size. Larger sensor arrays will still offer the best noise versus resolution trade-off in that scenario.

Just my opinion,

Bart
 

Tim Gray

New member
The digital SLRs have been chasing the MF format up market for a while. The quesiton was posed when the 1DS came out. Now MF backs are pushing 40mpx so they keep moving as well.

I guess that my point would be that the market's appetite for high end isn't growing anywhere near as fast as capability of the technology. This means that for more and more commercial applications a high end dslr is becoming "good enough", and the segment of the market for which a 22mpx $8k body isn't "good enough" is getting smaller. That means that the epxensive leading edge MF development is going to be in demand from a progressively smaller and smaller market. The economics aren't hard to figure out. FWIW, Clayton Christensen's "Innovator's Dilemma" is an excellent discussion of the process and impact of this kind of "disruptive innovation".
 

Asher Kelman

OPF Owner/Editor-in-Chief
The digital SLRs have been chasing the MF format up market for a while. The quesiton was posed when the 1DS came out. Now MF backs are pushing 40mpx so they keep moving as well.

I guess that my point would be that the market's appetite for high end isn't growing anywhere near as fast as capability of the technology. This means that for more and more commercial applications a high end dslr is becoming "good enough", and the segment of the market for which a 22mpx $8k body isn't "good enough" is getting smaller. That means that the epxensive leading edge MF development is going to be in demand from a progressively smaller and smaller market.

People miss that the digicams have some of the most advanced new technology, as Texas Instruments chips (Digic type) that bring out shadow areas and suppress overblown areas. So Kopdak and HP license use them and their camera sell but no one makes a fuss. Same with new optical plastics.

Now Canon makes it's own electronics, has it's own three layer "Foveon" sensor, is already partnered in a consortium for a competing functioning cheap to produce CMOS chip with each pixel individually addressed as a spearate camera, so binning is straight forward and images can be take thousands of times per second from different pixels!

So Canon could easily move into the MF area if it wished. No doubt they alrready have the chips. They can make a new camera from completed design in 3 months if they wish. However, they have no need to do so. The 1DIII was ready for market a year ago, but didn't release it until now since Nikon had nothing that worried them then.

So there is tremendous potential for improvement in these cameras. A lot will be far beyond pixel number; rather, smart processing according to photographer's choice and vertical market needs.

The pressure will come from below from now on!

Asher
 
Top