Will MF resolution win over 35mm stabilisation?

[eronald]

The first real A7RII images are trickling in on dpreview. They are pretty detailed. I think there is going to be a good market in high-rez lenses. Also, for landscape users without the most stable tripods it's not at all obvious that an MF back will be able to hold more rez than a stabilised 35mm camera of this new generation.

http://www.dpreview.com/forums/thread/3881960

I think MF needs new tech. The gap is narrowing more and more.

Edmund

[Don Libby]

Is this another attempt to say that medium format is dead? Getting rather old.

[ddolde]

I just ordered a Credo 80 back.  If that doesn't kick Sony's ass I don't know what will.

[dchew]

Also, for landscape users without the most stable tripods it's not at all obvious that an MF back will be able to hold more rez than a stabilised 35mm camera of this new generation.

You mean to tell me my $24.99 Ritz Tripod is not going to cut it with my IQ180?

Say it isn't so!!

Dave

[eronald]

You mean to tell me my $24.99 Ritz Tripod is not going to cut it with my IQ180?

Say it isn't so!!

Dave

At the very least I expect no contest for any handheld use.

Edmund

[BernardLanguillier]

I just ordered a Credo 80 back.  If that doesn't kick Sony's ass I don't know what will.

The credo will of course deliver more detailed files when everything is done perfectly.

In real world applications, I'd love to be able to compare the average detail in the files captured. I would think that the superior focusing abilities of the Sony (including eye detection, AF from the sensor,...) and its superior DR may result in better files.

You can try that yourself, buying one is a trivial cost compared to your Credo!

Cheers,
Bernard

[MrImprovement]

I am not convinced that stabilization , if the camera is on a tripod, is the way to go.  My understanding was that if the camera was put on a tripod, then stabilization should be turned off.  Surely with each pixel site being so very very small, almost any vibration is going to cause problems - even the vibration of the anti-shake sensor!

With a $3K body, $5K Otus lens from Zeiss, you are at $8K for a sharp Sony, which is within striking distance of the Pentax 645Z, which has 50 megapixels.  I don't know the current prices of the other stuff.

In my limited experience with even the 30-year old 75mm standard lens on the old, original 645 from Pentax - that ancient non-AF lens could resolve enough to keep the 645Z happy.  Newer lenses, even more so. 

[Bart_van_der_Wolf]

Is this another attempt to say that medium format is dead? Getting rather old.

Hi Don,

While I agree that Edmunds questions do tend to have a similar 'angle of approach', there is of course also something one can agree on. There is a need for new tech, if the MF market wants to survive in the long run. The low cost alternative are getting better, while the developments in MF are mostly stagnant.

However, what most people fail to recognize is that physically large sensors, while relatively expensive due to low yield, small volumes, and required high margins, have a huge technical/physical benefit over any smaller siblings. A large sensor means that a longer focal length is used for the same angle of view, and that combined results in a larger on-sensor magnification factor of the projected image.

As we know, larger feature sizes usually have a higher MTF response. So for an identical field of view (assuming the same aspect ratio) the physically larger sensor will produce a more robust image for post-processing, not only but especially with the micro contrast. That has (almost) nothing to do with dynamic range or other qualities that sensors may have.

Therefore, to maintain that discrimiinating advantage and not let it be (more than) compensated with other benefits of smaller sensors, the MF sensors (and ecosystem) do need to at least keep up with some of the smaller sensors features. A switch to larger MF CMOS sensor versions with high DR may be all that's needed for the near future (I do understand the issues for work that requires WA and/or shifted lenses). That coupled with much better calibration, very high quality lenses and better tethering, maybe even with better Bayer CFA filters, would make a combination that's hard to beat for real professional work.

I think that a larger threat comes from clients who want to lower their cost by squeezing out Photographers, which gets easier when the image quality gap gets narrower (which it doesn't have to). Of course there is more to being a successful photographer than producing high quality images, people management and relation skills and a good network are also very important.

Cheers,
Bart

[tom b]

Image quality from digital cameras has been acceptable for 99% of pro demands since the EOS 1Ds. Best practice, such as using strobes or a tripod probably makes more difference in the final image. To be honest I would say that the determining factor in the future of MF vs 35mm cameras is what camera you were trained with and what camera system you like to use.

Cheers,

[Gel]

Since getting the 5DSr I am using the 645z a lot less....

It's that Sigma 24-35mm F2 y'see.

I don't wish the demise of medium format either. On the subject of blurring I'm running around shooting non stabilised lenses / bodies handheld at the focal length = shutter speed and the images are pin sharp. Must be that smaller, lighter mirror

[Gel]

I think that a larger threat comes from clients who want to lower their cost by squeezing out Photographers, which gets easier when the image quality gap gets narrower (which it doesn't have to). Of course there is more to being a successful photographer that producing high quality images, people management and relation skills and a good network are also very important.

Cheers,
Bart

This.

People need a certain level on image quality. They need the images for promotional material, fashion, sales, whatever. Does it need to be shot on a 25k body anymore when a 135 sensor from Sony is enough. Or enough so that people can't tell the difference and halves their campaign costs.

[JoeKitchen]

Image quality from digital cameras has been acceptable for 99% of pro demands since the EOS 1Ds. Best practice, such as using strobes or a tripod probably makes more difference in the final image. To be honest I would say that the determining factor in the future of MF vs 35mm cameras is what camera you were trained with and what camera system you like to use.

Cheers,

+1

But I also feel there are advantages and disadvantages of each system.  Most younger photographers are now being trained on the 35mm DSLRs, and have little knowledge on the advantages MF offer or why they are advantages.  Primarily these would be use of the tech camera and leaf shutters. 

Since few schools are still teaching the tech camera, most do not understand the advantages it has over the tilt/shift lenses, which there are many.  Same thing with leaf shutters.  Even trying to get people to understand why they are necessary some times is difficult.  (I feel like it is easier to teach algebra to snarky high school students sometimes then camera tech to photographers who are interested in photography.) 

Not sure who to blame here, but much of this weight I place on the schools not living up to what they say they teach. 

In any event, going forward, MF companies should place more of an emphasis on the technical aspects of what the cameras can do and how they differ.  This will give them more of an edge then talking about image quality, especially after you show people why you do not need more than one focus point for almost every genre of photography. 

[Joe Towner]

The markets where MF will rule is shifting with the higher resolution 35mm gear.  The places where MF still excel will not change, the places where folks are willing to pay for it will.

It's unfortunate that 35mm gear is accelerating faster than MF gear.  Imagine how pissed off this entire forum would be were it not for the Sony 50mp chip.

[Jager]

Medium format film always had (still has) dramatically higher image quality than 35mm film.  Tonality, detail, richness, colors, ability to crop, ability to print large... it was all there.

Same thing in the early days of MFD versus 35mm digital.  The much larger sensor area and those big fat pixels translated into exactly the same sorts of advantages.

Not so much anymore, IMHO.  Having stolen a march with CMOS-tech, high-end 35mm digital is rapidly closing the gap on MFD.  Increasingly, it takes a technologist (as opposed to simply a visually literate consumer) to discern the difference between the formats in any kind of conventional use case. 

The MFD manufacturers are at the mercy of the chip makers, of course.  And therein lies the rub.  It would seem that between R&D and yield ratios and unit costs and all the rest, the full-frame 35mm sensor space is the place to live.  MFD is a space to be exploited, of course.  But as a secondary market.  Your best scientists and engineers develop for 35mm.  Your fabrication priorities are for 35mm.  Your budget priority is for 35mm.  You do MFD as an afterthought, or give it to the junior guys.

At least it seems that's the way it is.

The much-lauded (deservedly so) 50mp Sony CMOS sensor which represents today's leading-MFD tech is the case-in-point.  33x44 is simply not that much larger than 24x36.

MFD needs a much bigger chip in order to retain its historical advantage.

[eronald]

Jager,

 I've been told that where the tech effort really goes is cellphone sensors; all the rest is spinoff.

Edmund

It would seem that between R&D and yield ratios and unit costs and all the rest, the full-frame 35mm sensor space is the place to live.  MFD is a space to be exploited, of course.  But as a secondary market.  Your best scientists and engineers develop for 35mm.  Your fabrication priorities are for 35mm.  Your budget priority is for 35mm.  You do MFD as an afterthought, or give it to the junior guys.

At least it seems that's the way it is.

The much-lauded (deservedly so) 50mp Sony CMOS sensor which represents today's leading-MFD tech is the case-in-point.  33x44 is simply not that much larger than 24x36.

MFD needs a much bigger chip in order to retain its historical advantage.

[Bart_van_der_Wolf]

I've been told that where the tech effort really goes is cellphone sensors; all the rest is spinoff.

Nothing wrong with good spinoff, as long as it materializes in a physically large sensor with plenty high MTF pixels as a result.

Cheers,
Bart

[EricWHiss]

I've been told that where the tech effort really goes is cellphone sensors; all the rest is spinoff.

I think that is pretty much true, at least the last time I attended Norman Koren's Imatest training here in the valley,  I was the only one not working with cell phones…     

The multi-aperture stuff might really make a difference.

That said, I do believe the real advantage the Sony has over MFDB is basically better high ISO capability - oh yeah and also really useful autofocus.    Personally I haven't been wowed with the detail on the sample images I've seen like I am still with the MFDB.   

[eronald]

I think that is pretty much true, at least the last time I attended Norman Koren's Imatest training here in the valley,  I was the only one not working with cell phones…     
 

What I noticed when I attended was that nobody was interested in cameras or images.
The imaging device  now has as much to do with pictures as an amplifier in electronics has to do with music.

Edmund

[ErikKaffehr]

Hi,

If shooting on free hand in available light, stabilisation may play a major role. But, I don't think stabilisation is a decisive factor.

Why? To begin with I would guess that MFD shooters in general use the cameras where ultimate image quality is sought. In my view and experience this mostly precludes handheld shooting. Now, I am fully aware that some MFD users shoot technical cameras handheld from helicopters, but I would say in general that MF DSLRs are best used on tripod, with mirror lockup, or with high speed strobes. Under those optimal conditions image stabilisation wouldn't help.

With present day CMOS 35 mm had a significant advantage in readout noise, allowing for extending DR in the deep shadows or as an alternative use high ISOs. Those two things are the opposing sides of the same coin.

But, the latest generation of MFD is based on Sony CMOS technology, like the technology employed in Nikon D800/D810 and the Sony A7/A7r. So both MFD and 35mm uses similar generation CMOS from the same vendor. Some Pentax lenses also have image stabilisation.

The present generation MF CMOS sensors are 44x33 mm, making them 1.7X larger than a 24x36mm sensor. So, they collect 70% more light, giving something like 30% improvement in Signal Noise Ratio (SNR).

Also, linear size is around 1.3 times larger. So if an MF-lens reaches say 80% MTF at 20 lp/mm a 35mm lens would need 80% MTF at 26 lp/mm. Here I would suggest the high end lenses like The Zeiss Otus 85/1.4 or the Batis 85/1.8 may reach and surpass the best Hasselblad V generation lenses in spit of the format advantage. See enclosed MTF for the legendary (*) 100/3.5 Planar for the Hasselblad and the Otus 85/1.4.

With CCD higher resolutions are possible, up to 80 MP recently and the sensor surface is larger so less demand is made on the lens.

So my guess is:

• 50 MP CMOS on MF probably dominates 36-50 MP CMOS on 35mm in most aspects.
• 36-50MP CMOS on 35mm with excellent lenses may match and surpass 50MP MFD CMOS with less excellent lenses
• 80 MP CCD with 1.0 crop factor combined with excellent lenses and excellent lenses will be hard to beat, and plays reasonably well with symmetric ultrawides

This may put 30-50MP CCD in a less enviable position, but the competitive situation may press the prices of those systems in the second hand market, so more photographers can get their feet wet in the low end MFD sector. Keep in mind, those systems are still capable of excellent images.

As a final thought, we also need to consider maintenance costs. According to Lensrentals, DSLR repairs are on the order of 350$. I would guess MFD repairs are ten times more expensive. That may also be a consideration for those who regard green backs a finite asset that should be spent wisely.

Best regards
Erik

(*) The Planar 100/3.5 was intended for tasks like architecture and arial photography. It was probably the sharpest of the classical Hasselblad lenses for the V-series. The last generation of Zeiss lenses for the Hasselblad include some very good designs like the 40/4 CFE IF and  the Tele-Superachromat T* 300/2.8.

[EricWHiss]

What's to say that MFDB won't have IBIS at some point?

All the 35mm and smaller format images look flat to me.  I still prefer MF and LF images often even when they are not tack sharp