Thoughts About Full Frame

[Jonathan Wienke]

All else being equal, higher levels of sensor noise and less dynamic range with the smaller sensor. You have to shoot at lower ISO even though you can use a larger aperture, so you don't end up ahead in the long run.

[Madness]

But... a compact G5ish digicam with an APS sized sensor? now that would be tempting

Already waiting in line for one of those

[Jonathan Wienke]

so how do you justify your claim that smaller format and pixels have "higher levels of sensor noise"?

That would be based on the pretty much universally observed phenomenon that larger format sensors have lower ISO noise than smaller ones. Digicams are pretty much useless noise-wise above ISO 400, DSLRs can yield acceptable results up to ISO 1600. It's fairly easy to demonstrate this, just go to your local photo store, shoot some test shots with a variety of cameras, and correlate observed noise levels with the pixel size on the sensor.

[Keith Wong]

Michael, allow me the indulgence of posting an article I had prepared for another forum.

1. You already have a lens system which is optimized for 35mm. Every glass in your range (except (except for newly introduced glass, like the EF-S and DX) delivers an image circle for a 35mm frame. I would argue that EF-S and DX is a kludge, definition of a kludge being "an error introduced to fix another error". If Hasselblad wanted to produce a digital back for its MF camera, would it buy a 6.0MP APS sized sensor from Sony? The answer is "no" for many reasons, but one of the prime reasons is because too much of the imaging circle has been wasted. The argument remains true, in a smaller scale, for FF vs. APS sized sensors.

2. Expanding on my first point, the greatest expense in designing a lens is to ensure the lens performs consistently from corner to corner at wide apertures. This is one of the touted advantages of DX/EF-S/Four-thirds lenses. Paying a small fortune for a 600 F/4 and then throwing away the most expensive part of the image sounds wasteful to me. This would not be so bad if the manufacturer rationalised the entire lens range so that savings could be realized across all the lenses - something that the 4/3 consortium is trying to achieve, something Nikon is halfheartedly doing, and something that Canon will not do.

3. To continue along this line, where are the savings which reduced crop lenses have promised us? First 3 examples, price obtained from B&H, last example, Canon Australia website:

Nikon 10.5mm DX fisheye, USD599.95 - Nikon 16mm fisheye (full frame) USD$564.95
Nikon 17-55/2.8 DX, USD$1349.95 - Nikon 28-70/2.8 (full frame) USD$1329.95
Canon EF-S 10-22/3.5-4.5 USD$599 - Canon 17-40/4L (full frame) USD$674.95 (not strictly comparable, since the 17-40/4L is a constant aperture lens)
Canon EF-S 17-85/4.5-5.6 IS AUD$1099.00 - Canon 28-135/3.5-5.6 IS AUD$999

... I am sure you will agree there is no difference. We are still paying the same price for reduced crop lenses as we were for full frame lenses.

4. The next part of my argument concerns sensor design. Among other things, sensor designers are concerned about resolution, dynamic range, and noise suppression. Pixel pitch (the size of each individual photo pixel sensor, in micrometers) is just one out of many factors which influence these. People like to use the rainwater and bucket analogy which is a good one - a larger bucket allows more precise measurement of rainfall. A small bucket spills water too quickly. If there is very little rainfall, the small bucket may not gather any rainwater at all.

You can design a more efficient water gathering system (microlenses), you can design a better ruler (both tricks employed in the new 20D and 1DMk2) ... but in the end bigger buckets are still better.

A full frame sensor ultimately allows you to fit more, and larger pixels. Eventually you will hit a limit on a smaller sensor before the conflicting demands of resolution and larger sensors force you to look at other means of increasing dynamic range and reducing noise. You just hit the limit later on a FF sensor. If two manufacturers have exactly the same noise reduction and image processing technology, the manufacturer that can produce a larger sensor will have higher resolution, increased dynamic range, and less noise.

5. Viewfinders. Its not strictly necessary to have a larger viewfinder - it's just nice to have.

[didger]

Only people who would buy $8000 Camera would be either professionals or serious/affuluent amateurs which accounts for no more than 10% of total DSLR revenue.  

So what?  What percentage of total film camera revenue is for 8x10?  This format has been around since the very invention of photography and may stay around the longest of any film format, since it will be the last to be made obsolete by digital.

The 35mm film size sensor Camera will only get more rare and hard to find.

The Canon marketing folks don't seem to feel that way and their #1 position makes me trust their judgement more than yours.  I run into countless photographers on my excursions and very few are 4x5 shooters and I've only actually once seen an 8x10 in the field.  However, these large format cameras are NOT hard to find.  Try a Google search.

Here's my prediction:  35mm film size sensor cameras will only get more popular and easier to find and also gradually somewhat cheaper.  Another player entering the arena could drive prices down more and a major technological breakthrough (which no one can predict) could drive prices down a lot.  Even now a lot more people are getting into large sensor because of all the used 1ds's for sale now so much cheaper.

I'm neither affluent (you should see my house and my car) nor very much pro so far, but I have a 1ds and I vacillate about hustling to get the money together for a 1dsMKII.  It's more a matter of priorities than affluence.  Countless ordinary folks have cars worth $30,000+ and houses worth $100,000+.  That's their priorities.  1ds owners tend mostly to be pretty fanatically enthused about this camera and format and this can't help but promote growth.  I "converted" a fellow Sierra backpacking photographer (Eos film camera) to 1ds and Zeiss lenses without even trying and this guy is also neither pro nor affluent; just another quality fanatic.  Stay tuned and watch your convictions erode and stand by to watch the growth of MF digital.  Horseless carriages were a passing fad too.

[EAD]

I believe that the simple fact that Canon an Nikon have started investing big bucks in the manufacture of "digital-lenses" of the likes of the EF-S ones is a clear signal of what is going to happen in the near future.

I believe whe are going to have APS size for quite a while. It suites very well the needs and conveneinces of the majority of the potencial future buyers, and we musnt forget that Canon is in this for the money and not for their love for art...

Just the opinion of a poor mortal camera buyer... :cool:

[didger]

You addressed your question to Michael, but I'm not shy, so I'll go for it also.  Yes, a smaller image circle lens can theoretically be better within that image circle than an otherwise comparable lens that covers a bigger circle.  HOWEVER, the problem is that wide angle lenses are not as wide with a small sensor, so if you want a small sensor camera to be ultrawide then you're faced with the problem of making a sharp lens that's around 10-12mm focal length.  Even for small image circle that wouldn't be easy.  What you gain with a bit of extra resolution and a relatively bigger "sweet spot" in a small image circle lens, you lose with the fact that the total final image area is also smaller and therefore effectively less enlargable than a large sensor image.  The issue is which format will give you the biggest sharp enlargement and not which format has the sharpest lenses.  MF lenses are not as sharp as 35mm, but MF nevertheless give you sharper big prints.  Same thing applies to the difference between large and small sensor 35mm DSLR's.

Maybe Michael has a different take on this and can provide a clearer answer.

[Digi-T]

Please, someone correct me if I'm am wrong here or oversimplifying, but it is my understanding that the main advantage of using a smaller area of a lens is to avoid the outer area that may show more distortions. However, there are some lenses that are relatively distortion-free across the entire image circle and if you are using one of these lenses with a smaller sensor you would not be getting the greatest results from the lens. Lenses have a quality and sharpness factor similar to that of sensors where the bigger you are able to go the better the overall image quality. Of course the other factor that many people consider as an advantage with smaller sensors is the bigger magnification factor which is not an advantage for everyone.

T

[Tony Collins]

Michael made the point about APS sensored cameras not having a size advantage against FF. When APS film slrs were made by Canon and Minolta they were more compact. I think that size may have more to do with perceived value to the buyer who wants their new toy to look like £1500 of metal. I am sure that there is a place in the firmament for a range of DSLRs which take advantage of the smaller and lighter body and lenses that a reduced image circle optical geometry allows. It will take time to get there as manufacturers trim 3mm off the body size year on year to give us an excuse to upgrade.

But... a compact G5ish digicam with an APS sized sensor? now that would be tempting.

Tony Collins; 10D but sitting out this years upgrade. 30D anyone?

[didger]

I don't think anyone would contest the economic and weight benefits of smaller sensor cameras and smaller image circle lenses.  When I think of what carrying about 17 lbs. of photo gear on long backpacking trips does in terms of risk to my back and pain to my old climbing injuries, I think about how much nicer a Digital Rebel and 17-85mm IS lens would be.  Total weight about 1.5 lbs. and no need for a tripod and I'm sure a Digital Rebel is more economical for battery weight and use too.  However, I won't switch to a smaller camera and lenses unless the quality gets at least as good as a 1ds.  The Nikon DX2 (or whatever exactly it's called) has more Mpixels than a 1ds, but I'm so doubtful about the sensor quality (esp. noise) and the quality of those special zoom lenses (especially ultrawide) that I'm not even very seriously inclined to check this camera out.  I'm also a bit put off that a $5000 camera would have such a mickey mouse consumer feature as in camera multiple exposures.  My hopes that Nikon would end up a serious contender in the high end pro DSLR field have pretty much evaporated.  I think there's a much better chance that I'll end up getting a 1ds MKII than getting any sort of small light DSLR any time soon.  Too bad for my back, but better for my images.

[BJL]

Smaller sensors mean smaller focal lengths for the same FOV, and therefore a larger DOF.

Only if you assume the same aperture ratio, but why do that? The long time trend in cameras is that lower minimum aperture ratios are available and usable for shorter focal lengths and smaller formats, and this lives on today with things like f/2-2.4 zoom lenses on some 2/3" format digital cameras.

As soon as a proportionately lower aperture ratio is available on the shorter focal length lens used with the smaller format, you can get the same DOF as in the larger format, but with a "faster" f-stop, so that it is possible to use a higher shutter speed, or a lower ISO setting to reduce noise levels. What is the downside to the smaller format then?

[Edward]

> As soon as a proportionately lower aperture ratio is available on the shorter focal length lens used with the smaller format, you can get the same DOF as in the larger format, but with a "faster" f-stop

That is where the aberration problem creeps in.  As the aperture gets larger relative to the focal length, it gets harder to make the lens.  Look at the ultra wide problem even on 35mm and the great quality from the F1 normal lenses - even the Canon and Leica F1s are terrible compared to the cheapest F1.8s.  I would need a 28 F1.0 on my 10D to match my 50 1.4 on a film camera.  How about a 20 F 1.4 to match my 35 F2?  Probably the only good lens match is the 35 F1.4 on an APS camera - but it is a big and expensive lens.  Things are easier at the long end, but then they always are because you do not have bend the light as much.

[BJL]

so how do you justify your claim that smaller format and pixels have "higher levels of sensor noise"?

That would be based on the pretty much universally observed phenomenon that larger format sensors have lower ISO noise than smaller ones. Digicams are pretty much useless noise-wise above ISO 400 ...

One more time: smaller formats can generally use lower ISO settings, so comparisons at equal ISO are "correct" but almost completely irrelevant. More so since the original topic raised by Mike Johnson is APS vs 35mm formats, not digicam vs DSLR.

That would be based on the pretty much universally observed phenomenon that larger formats generally must be used with higher, slower aperture ratios, due to either depth of field requirements or the practical limits set by cost and weight. Neither Hasslebland nor Rollei offer anything faster than f/2 in medium format.

It would be perhaps accurate but very misleading for me to say that larger formats are slower, because of the larger aperture ratios that they often need to use. I would be telling only one half of the story, you are telling only the other half.


What is true is that at the same ISO, shutter speed and pixel count, a larger format has lower noise levels, but also less DOF.

Surely, in a landscape oriented forum we can acknowledge that depth of field is very often a limiting factor in choice of aperture. Then, as I have explained already, either the larger format must use a higher ISO than the smaller, enough higher to cancel its noise advantage, or the smaller format can use its optimum (minimum?) ISO, in which case noise is not a significant issue. At least not for any DSLR that I know of, and not even for the tiny 2.7 micron pixels of the Sony 8MP 2/3" format sensor according to Michael.

Once DOF is relevant, the worse performance at ISO 400 of digicams compared to DSLR's is irrelevant.

[mikepak]

I agree with your argument about full frame sensor.  However, you are over-looking one important point.  

The COST FACTOR.  

Currently, the cost of produing full frame chip is prohibitively expensive.  And, this will not get any better in short run (please refer to my earlier posting), definitely not in next several years.  

This important COST factor will always drawn line between professional market and consumer market.

[mikepak]

Larger sensor manufactured with same technology will have advantage over smaller sensor in terms of signal to noise ratio.   However, I am not so convinced that pixel in full frame sensor will have significantly better result than pixel in APS size sensor given advanced in-camera image processing and on-chip noise reduction process.  Let's take 1D MKII versus 20D for an example.  Most people would be hard pressed to see any difference in noise performance between two camera upto ISO 1600.

I predict that performance gap between full frame sensor and APS size sensor will get harder to detect at average viewing distance.  It will still be there, it will get obvious only in extreme cases such as sever under-exposure or ISO rating of1600 or greater.

Furthermore, performance of typical 35mm SLR lens declines along the periphery.  Thus, by using the APS sensor, we get to use the sweet spot of a given lens.


As for the 35mm film size sensor, Canon may support them in a $8000.00 Camera, but how many people would buy the $8000.00 Camera?  If they ever produce the full frame Camera under $3000, I will be the first to buy it.  I am predicting that such will never happen within next several years.

Only people who would buy $8000 Camera would be either professionals or serious/affuluent amateurs which accounts for no more than 10% of total DSLR revenue.  

I am predicting that, over next five years, market will get dominated with APS sensor size DSLR.  With Minolta, and Pentax jumping in, this category of DSLR will only get better, price will go down, and ultimately, consumer will be the winner.  The 35mm film size sensor Camera will only get more rare and hard to find.  

[drew]

Didger, you see we can agree. I have already contacted CNN, but I was rejected by Al-Jazeera, because I foretold the demise of Bin Laden in Tora Bora. Sadly, he would appear to still be with us. Alas...
The last bit of your message is a human aspiration we can all relate to.

[didger]

Ooops, Michael's reply and mine "crossed in the mail" and he got there a little sooner.  His reply is substantially the same as mine, only briefer.

[dtrayers]

Great point!  Might not matter for rocks and trees, but if you use selective DOF, it makes a huge difference.

The convergence of aberrations and diffraction as the sensor gets smaller is also an issue, but not yet critical at APS size.

And this is the point I think both Michael and Mike haven't addressed ...

Smaller sensors mean smaller focal lengths for the same FOV, and therefore a larger DOF.

Some feel that's a virtue of the digicams and their small sensors: the tremdous FOV.  You don't have to worry about focus accuracy as much.  In fact, I see a lot of posts by new comers to DSLR's complaining that their cheaper digicam took sharper pictures, not realizing they were seeing the large DOF of the digicam, not a sharper picture.

I see it as a disadvantage.  I like being able to control the DOF.  It seems to me that the traditional 35 mm size is just right for being able to have a nice small DOF wide open and still get a reasonable one when stopped down.  Search for posts from novices having learned about their craft and are trying to get background blurr from the digicam.  Lord knows I tried with my G3 at f/2 and didn't get much.

Sure, someday they'll put a 8MP sensor the size of a pin head in a cell phone and you'll get a clean image with no noise.  Just don't go looking for any bokeh.

[BJL]

All else being equal, higher levels of sensor noise and less dynamic range with the smaller sensor. You have to shoot at lower ISO even though you can use a larger aperture, so you don't end up ahead in the long run.

But nor are you worse off, so how do you justify your claim that smaller format and pixels have "higher levels of sensor noise"?

The physics is this, for the example of doubling linear size.

To get the same field of view, depth of field and pixel count, doubling linear pixel and sensor size requires doubling focal length and doubling aperture ratio, which means pixels have four times the area and 1/4 the intensity of illumination, so the same total rate of illumination per pixel. For the same shutter speed, each pixel receives the same amount of light: same signal. As to noise: photon shot noise will be the same, dark current noise in larger pixels will be somewhat greater (more total length of well edges, along which the thermal electron diffusion happens, which is what causes dark current noise). Amplifier noise can be expected to be the same or slightly higher with the larger components needed to handle the larger maximum output of a larger pixel.

So in this situation, smaller pixels give the same signal and the same or lower noise levels, so the same or better S/N ratio.


I agree that with current technologies, bigger pixels have higher dynamic range, manifested as more highlight headroom due to larger electron well capacities.

However, this is only relevant when there is enough light that the sensor can be used at minimum ISO or close to it. At higher ISO, the "underexposure" mostly eliminates the risk of blown highlights even with smaller pixels.

[didger]

Hey, call the news services.  I find myself in total agreement with a Drew message.  I don't see how things could possibly fail to go just that way.  
I don't care what the bulk of sales are now or will be in the future or about ideal compromises, but I care a great deal about high end digital 35mm becoming as common (and well supported with lenses) as high end 35mm film used to be and I'm looking forward to digital MF becoming as common as film MF.  Film MF has never been cheap, but also not exhorbitant like digital MF so far.

I just hope it all comes together while I still have the health and fitness to get out so much.