The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

Sensor Size/Image Quality

Status
Not open for further replies.

johnnygoesdigital

New member
I really think it's about the overall print size that would influence my choice. I've shot magazine covers with 35mm, FF Canon's, and no one knows the difference. For Billboards or airport installations, I choose MFD, where size does matter. Lenses have post production software to make up for any loss in optical quality, (personally, I'd rather have a good lens first, like the S2). I'd love for 35mm DSLR to catch up, it is the most reliable camera system/format i've ever used. For gallery installations it's always MF, film...yes, film!
 

Godfrey

Well-known member
Thank you for the insightful, factual and very technical answer. I understand the relationships between all these factors will make the question almost an apples to oranges comparison, however I am really seeking a more simple answer. Will the image quality of a 36MP FF DSLR sensor be equal to a 40MP MF sensor?
I don't think it's a valuable exercise to try to answer your question directly with facts and figures and photosite sizes, ad nauseam. It comes up to be yet another long gear-headed discussion. Certainly better, more expensive, more precisely manufactured equipment with more more more should have the capability of achieving technically better results.

That's why my answer was rooted in a qualitative judgement of "when is enough enough?"
Once sensitivity, pixel density, acutance, and dynamic range are at a satisfactory quality plateau for any format and specific shooting scenario ...
When excellence in Photography is reduced to a list of technical capabilities and specifications, Photography vanishes and all we have left is equipment and image.
 

Bob

Administrator
Staff member
Really it is all about finished print size and viewing distance.
For web work, my little gh2 does well enough.
BUT when I want to shoot for print or when I might want more crop-ability or maybe a large print might be desired, I haul out the MF back. I have used many a Canon and Nikon but the wedding shooter influence in their design has rendered them too bland for me. So you pays your money and you takes your choice. Since we are at pixel sizes and isos where I can count the number of photons that are needed to get a sensor response in zone 2, we are also treading near the realm of shot noise and other phenomena, such as the stochastic noise produced at the edges of sensitivity where one photon more or less makes a difference in the noise threshold. The biggest potential improvement in sensor technology at this juncture might be buried interconnect to make the effective aperture of each cell larger, or the improvement of the bayer color filters to approach the quantum limits of sensitivity. Folks, we are within a factor of maybe 3-4 of that limit now assuming that you would like to obtain at least 10 bits of good data from the sensor. Some of my colleagues think we are much closer than that. So at least with silicon sensors, we can see the end of the tunnel. Photography means writing with light. That gets harder the less light we use and the smaller the sensor sites get.
-bob
 

Bob

Administrator
Staff member
The handheld macro shooter (in available light) is my favorite. ;)

Nice post, Bob.:thumbup:

I wonder, though the area is small for a FF sensor, if Nikon or Sony or Canon configure it in a way there is efficient heat loss away from the sensor area as in an IQ back. Even if they attempt it, it is unrealistic with x number of frames/s captures.
Thanks, Vivek,
Perhaps a small liquified N2 dewar flask attached to cool the sensor?
The problem is that given that many of the design parameters are interrelated, somebody will be disappointed since the nature of engineering is compromise.
-bob
 
Last edited:

fotografz

Well-known member
Interesting discussion.

I had dinner last night with a Professional Photographer friend of mine that runs/owns a high volume commercial photography studio. Nine Hasselblad H cameras, mostly H3D-II/39s and a few H2/39 Multi-Shots, and a bunch of Canon 5D-IIs ... big beautiful studio with multiple product sets going all day long ... more high-end Macs than the Apple store, and the most Profoto Pro-7A & Pro-8A 2400 boxes and Pro heads I've ever seen in one place ... LOL!

We got into a similar discussion, and his POV was that there is no comparison, and more megs in a 35mm format won't change that, any more than the 21meg 5D-II did compared to the H3D/22 they also once used. Their 5D-IIs are used for low demand projects and the occasional editorial or location people assignments for website applications.

The reason for this studio's raging success while other commercial studios are dying like May Flies, is that they are relentlessly focused on quality of end product to their clients ... meticulous attention to detail in every imaginable way, total control over subtile tonal separation, dynamic range and faithful color rendition in every respect. Anything less and their clients would take notice immediately ... because retouching cost would skyrocket ... which is as expensive, if not more, than the photography itself.

These aspects had less to do with final size of use, and everything to do with all of the above characteristics. However, he did say that the larger file also allowed his shooters to sometimes pull back to gain DOF rather than stopping down and introducing diffraction.

BTW, his opinion was that 40 meg is enough for their applications and quality standards ... in his words, "We're done". What he would prefer to see is increased operational speed, more usable features and accessories (like the HTS/1.5 which they also use), and in terms of the H lenses, a continued revamp like they've done with the recent upgrades of the HC50-II, 120-II and 150N ... namely, in his opinion, the HC35 needs re-done.

Personally, I took note of that relationship between MFD and 35mm years ago ... with a relatively similar meg count used in studio, the MFD won every time, and as things have advanced on both fronts, that relationship has not changed.

I think putting 36 meg into a prosumer body is a marketing ploy to snag bucks from enthusiasts that buy based on numbers. This will not only place a strain on the optics, the users experience will come under attack in terms of technique ... which seems odd from the standpoint of how these types of cameras are actually used. Besides, I could not imagine a prosumer body being used in a working studio like my friends, which is a hint as to who such marketing ploys are aimed at ... it sure isn't him.

I also think Canon has made a brillant move with the upcoming 1DX FF 18 meg. with improved performance specs rather than sheer meg count. Bravo to them :thumbs:

-Marc
 
T

TheReason

Guest
Marc,

Your response was exactly the type I was hoping to receive. Industry related facts and needs discussed. I have worked in the industry (Here in our Detroit area) as an Art Director, Retoucher, and Shooter. The retouching and final image quality of the files are of supreme importance for certain clients, so I am happy to hear that the newer more pixel packed DSLR bodies still have a niche spot. Just as the MF systems have a time and place of their own.

Would LOVE to see your friends studio! Sounds like heaven to me! Thanks for the great response.

-Chris
 
T

TheReason

Guest
Marc,

Your response was exactly the type I was hoping to receive. Industry related facts and needs discussed. I have worked in the industry (Here in our Detroit area) as an Art Director, Retoucher, and Shooter. The retouching and final image quality of the files are of supreme importance for certain clients, so I am happy to hear that the newer more pixel packed DSLR bodies still have a niche spot. Just as the MF systems have a time and place of their own.

Would LOVE to see your friends studio! Sounds like heaven to me! Thanks for the great response.

-Chris
 

fotografz

Well-known member
Marc,

Your response was exactly the type I was hoping to receive. Industry related facts and needs discussed. I have worked in the industry (Here in our Detroit area) as an Art Director, Retoucher, and Shooter. The retouching and final image quality of the files are of supreme importance for certain clients, so I am happy to hear that the newer more pixel packed DSLR bodies still have a niche spot. Just as the MF systems have a time and place of their own.

Would LOVE to see your friends studio! Sounds like heaven to me! Thanks for the great response.

-Chris
His studio is in Royal oak, close to Birmingham Chris.

BTW, I'm in Franklin with a small studio in house doing some product work ... originally came from an Art Director background, recently retired from my Exc. Creative Director career, and now only do photography and some advertising consultation work. My clients are super picky about IQ also ...

We should have coffee sometime and talk shop :)

-Marc
 

JorisV

New member
BTW, his opinion was that 40 meg is enough for their applications and quality standards ... in his words, "We're done". What he would prefer to see is increased operational speed, more usable features and accessories (like the HTS/1.5 which they also use), and in terms of the H lenses, a continued revamp like they've done with the recent upgrades of the HC50-II, 120-II and 150N ... namely, in his opinion, the HC35 needs re-done.
-Marc
+1
 

Shashin

Well-known member
Now matter what you do, light still is a wave with a frequency. You may improve optics, but only to a point. And when you optimize for resolving power, you need to give up contrast/acutance. So the new Nikon may have the same pixels as my Pentax 645D, but my lenses don't have to work harder and can have nice contrast in the details.

And sensor size matter for how the image renders no matter the pixel density. DoF is the obvious quality that is different. When Kodak came out with Tech Pan, medium-format and large-format shooters did not ditch their cameras for 35mm even though Tech Pan was rumored to have 4x5 quality in a 35mm frame.

But likewise, 35mm has a look and feel that is not in medium format.

BTW, an interesting comparison you can make at DPreview is one between Sony's new APS 24MP sensors in the SLT and Nex 7 camera and the Sony 35mm sensor in the a900. And try different ISOs.
 

FredBGG

Not Available
Not so tiny at all !

about the same size as on a 7D, which if can remember right is one of the best Cameras actually for lowlight stuff and image quality in APS-C.
double the chipsize and there you have 36MPix.

And what about the 80 Mpix IQ180 - 5.2 Micron Pixel size and leader in Dynamics and Image quality on DXO Mark ?

So..........

regards
Stefan
Take a closer look at the results in DXO mark. Real ISO is on the IQ180 is way off. At real 100 ISO the 180 does not have that good of a dynamic range.





That said it does not bother me too much. it's nice to have low iso sensor as I like to shoot wide open with f2 lenses on MF cameras.

However personally I think that Phase one and Hasselblad are barking up the wrong tree with 80 MP. it would be far more interesting to see full frame sensors on 6x7 and 6x8 cameras.

A 40MP sensor on a 6x8 camera would look more interesting than an 80MP sensor on a 645.

Some claim that it's a cost problem. That is rubbish, because MF sensors are made (or can be made) of multiple smaller sensors tiled together.. 44MP sensor coulds be made from twice as many tiles as the aptus 22 MP back.

I would love to have a FF sensor behind my 6x8 125mm and 180mm f3.2 lenses that both have full tilt and shift.
 
Last edited:

FredBGG

Not Available
I think putting 36 meg into a prosumer body is a marketing ploy to snag bucks from enthusiasts that buy based on numbers. This will not only place a strain on the optics, the users experience will come under attack in terms of technique ... which seems odd from the standpoint of how these types of cameras are actually used. Besides, I could not imagine a prosumer body being used in a working studio like my friends, which is a hint as to who such marketing ploys are aimed at ... it sure isn't him.

I also think Canon has made a brillant move with the upcoming 1DX FF 18 meg. with improved performance specs rather than sheer meg count. Bravo to them :thumbs:

-Marc
Anything over 40MP on a FF 645 is by the same logic a marketing ploy.
The Full frame 645 is only twice the size of FF 35mm

personally I prefer the shadows in a 22MP back at 50 ISO than a high MP count sensor. Big photosites have their pluses.
 

PeterA

Well-known member
Yes it would be nice to have a much larger chip able to be used on cameras with movement and lenses with larger image circles....
 

Bob

Administrator
Staff member
Take a closer look at the results in DXO mark. Real ISO is on the IQ180 is way off. At real 100 ISO the 180 does not have that good of a dynamic range.





That said it does not bother me too much. it's nice to have low iso sensor as I like to shoot wide open with f2 lenses on MF cameras.

However personally I think that Phase one and Hasselblad are barking up the wrong tree with 80 MP. it would be far more interesting to see full frame sensors on 6x7 and 6x8 cameras.

A 40MP sensor on a 6x8 camera would look more interesting than an 80MP sensor on a 645.

Some claim that it's a cost problem. That is rubbish, because MF sensors are made (or can be made) of multiple smaller sensors tiled together.. 44MP sensor coulds be made from twice as many tiles as the aptus 22 MP back.

I would love to have a FF sensor behind my 6x8 125mm and 180mm f3.2 lenses that both have full tilt and shift.
I did not understand that DXO measurement at all since I measure a good stop difference between 50 and 100.
I also checked it out with he Sekonic test chart. and their meter calibration software and that phenomenon did not appear.
I have much trouble with the DXO methodology since so often it is very different than what I measure with my own instrumentation. This has gone on so long and it is so bad (not just with the ISO180 but with many others cameras) that I discount their data out of hand once the source is known. I decided not to waste my time trying the uphill battle of explaining why their entire site is rubbish with a convincing veneer.
-bob
 

fotografz

Well-known member
Anything over 40MP on a FF 645 is by the same logic a marketing ploy.
The Full frame 645 is only twice the size of FF 35mm

personally I prefer the shadows in a 22MP back at 50 ISO than a high MP count sensor. Big photosites have their pluses.
Yeah, it still haunts me how beautiful images from the 9 micron backs look. I know a few pro shooters that still use a 528 Multi-Shot tethered in studio ... in 22 meg single shot, or 4 shot (mostly) ... and when needed, 16 shot. Pretty versatile back for studio application.

However, I do think that there are those who do utilize 60 and 80 meg backs effectively because they 1) Work on location, like landscape artists, 2) do print very large, 3) also use tech cameras that feature optics that are up to the task 4) have frequent applications that either require 39/50 Multi-Shot or 60/80 single shot to avoid moiré.

The only reason I have a 60 back is for shooting fabric swatches for GM every year, and a few other assignments where clients will be printing very large and the prints will be viewed very close ... like wall sized product shots for trade-show booths. Otherwise, the 37.8 meg Leica S2 with its incredible optics does everything else ... especially location work.

-Marc
 

Stefan Steib

Active member
Fred

I took some research recently and found out that the larger chips are actually not single chips "glued" one to another, they can be made with steppers using the so called stitching method.
The needed resolutions would be "lowtech" and available on the most Chip fabrication plants in Asia, so the production would be comparably cheap, maybe 500 € per wafer with 4 usable 6x7 chips. The only thing that is necessary for this is designing the chip (can be scaled available CMOS designs) and making the masks (some millions).

But then Phase and Hasselblad would need to make new cameras and lenses too - The actual Rodenstock and Schneider lenses would not work in the shorter range , and these chips would be limited to tech cams using 40-50 mm and up, which would be a very small market.

I agree that longer focal lenghts would work beautifully on such a chip.
But I also see that there are alternatives - the new organic CMOS by Fuji will have a better spectral response, thus using more photons per pixel.
I am pretty sure we will not see larger chips than 4,5x6 for standard imaging- the trend will be shrinking not growing sizes.

and using the CCD´s as low iso devices on a tripod is certainly also possible with a 36Mpix CMOS of smaller 24x36mm size

Greetings from Munich
Stefan

Take a closer look at the results in DXO mark. Real ISO is on the IQ180 is way off. At real 100 ISO the 180 does not have that good of a dynamic range.

That said it does not bother me too much. it's nice to have low iso sensor as I like to shoot wide open with f2 lenses on MF cameras.

However personally I think that Phase one and Hasselblad are barking up the wrong tree with 80 MP. it would be far more interesting to see full frame sensors on 6x7 and 6x8 cameras.

A 40MP sensor on a 6x8 camera would look more interesting than an 80MP sensor on a 645.

Some claim that it's a cost problem. That is rubbish, because MF sensors are made (or can be made) of multiple smaller sensors tiled together.. 44MP sensor coulds be made from twice as many tiles as the aptus 22 MP back.

I would love to have a FF sensor behind my 6x8 125mm and 180mm f3.2 lenses that both have full tilt and shift.
 

Shashin

Well-known member
I wish one day the myth that print size is somehow determined by pixel resolution would die the death it deserves. That does not mean I am against large MP backs, but the MP aren't giving you bigger prints.
 

Stefan Steib

Active member
Hi Shashin

this is what I meant with "rich" files. It´s mostly about cleanness, uniformity, signal to noise and straight colormanagement - means NO harsh corrections/Colorspace transformations and camera profiles.
Sharpness and resolution are important but beyond 100 MB image data this is not essential anymore. (viewing distance and eye´s resolution)

Greetings from Munich
Stefan Steib
 

EsbenHR

Member
I took some research recently and found out that the larger chips are actually not single chips "glued" one to another, they can be made with steppers using the so called stitching method.
This is correct - all commercial MF sensors are created on a single substrate and it is very expensive. Big modern chips (CPUs in desktops for example) are manufactured using similar techniques BTW.

The needed resolutions would be "lowtech" and available on the most Chip fabrication plants in Asia, ...
This seems "obviously correct", but there are some unfriendly dragons dwelling on the pastures you need to pass. You want the entire chip to respond the same way to the light, and that is darn difficult to attain in practice. It requires some ridiculous tolerances that are pretty hard to meet in the real world.

... so the production would be comparably cheap, maybe 500 € per wafer with 4 usable 6x7 chips.
It would be nice if it could be made that cheap. Unfortunately the yield would likely be rather low for a chip of that size. Rather than 4 usable chips, you would probably get 4 unusable chips most of the time.


The only thing that is necessary for this is designing the chip (can be scaled available CMOS designs) and making the masks (some millions).
Actually this would be challenging in itself. It is possible. Canon made an experimental chip some time ago to see how hard it would be, but it is not trivial. Depending on how much money you think it takes to make a problem nontrivial of course ;-)

But then Phase and Hasselblad would need to make new cameras and lenses too - The actual Rodenstock and Schneider lenses would not work in the shorter range , and these chips would be limited to tech cams using 40-50 mm and up, which would be a very small market.
If the chip had a resolution comparable to current backs, the pixels could be larger. This would make it easier to make the back respond better at steep angles, at least for a CCD sensor. CMOS lends itself better to small pixels, so they would probably be screwed.

I also think both Rodenstock and Schneider might be able to find some large format designs in their drawers they could build pretty quickly if there was demand.

I agree that longer focal lenghts would work beautifully on such a chip.
But I also see that there are alternatives - the new organic CMOS by Fuji will have a better spectral response, thus using more photons per pixel.
Chip manufacturers tends to think that "better response" = "more sensitive", which you get by making the color filters wider. Trust me on this: that approach sucks, at least if you need colors.

Case in point: the Foveon sensor is a great idea, but the "color filters" are not a good match to the way human perceive colors. Once you try to wrangle out some colors from the RAW data you end up with a lot of noise.
 
Status
Not open for further replies.
Top