why full frame?

If information recently released related to the new APS-C sensor in the Fuji X-Pro1 is viable, the sensor in that camera beats FF sensors from 3-4 years ago (Nikon D700 and Canon 5DMkII).

I appreciate very much your comments about "best compromise".

Well I wont be trading my d700 for any crop camera whatever the specs on paper! Beats them in what way exactly? The only camera I would replace a d700 with would be another FF Nikon.
Rather than compare paper specs, let's wait until we can see some results from the new wonder camera before we start writing off the likes of D3/700 and Canon 5d's!
 
Well I wont be trading my d700 for any crop camera whatever the specs on paper! Beats them in what way exactly? The only camera I would replace a d700 with would be another FF Nikon.
Rather than compare paper specs, let's wait until we can see some results from the new wonder camera before we start writing off the likes of D3/700 and Canon 5d's!

You're absolutely right, inasmuch as the graph I saw only correlated two indices (S/N ratio and resolution), but obviously left out things like dof that have an enormous impact on the final image. The new Fuji sensor has greater resolution and better S/N ratio versus the D700 up to ISO 800. At ISO 1600, the D700's S/N pulls slightly ahead, but resolution remains far behind that of the X-Pro1. Needless to say, this does NOT translate into conclusive evidence for a better or worse looking image, but my guess is that with the powerful NR software we now see, it would be preferable to have higher S/N and higher resolution. This, indeed, is the basic formula of the compromise offered by the M9. But it does suggest that the relation between FF and APS-C is constantly evolving...
 
Rather than compare paper specs, let's wait until we can see some results from the new wonder camera before we start writing off the likes of D3/700 and Canon 5d's!

Agreed, bigger sensors will always have an advantage, even if it only how they render depth... I know a Mamiya 7 lens will outresolve my view camera lenses but the larger format still has a better feel and presence to me.

But the APS sensors are really good nowadays... even the ($800 used) D300 is quite a respectable camera that has a nice advantage in terms of auto focusing because it uses the same sensor array as the D700-D3 cameras - but by virtue of its smaller APS sensor, the array covers most of the frame compared to being concentrated in the center of the full-frame. So for a bird-wildlife shooter with longer lenses, the D300 is a valuable camera that can deliver better results than its more expensive brethren.

No question though, it is nicer to look through a D700 finder. But the D300 has one of the best APS finders.
 
I used a D300 for a couple of years. That camera is an excellent performer. I think the x100 raw files are a bit better... but not by much. The D300 AF is outstanding. With a battey grip the D300 is still a workhorse for action photography. The D300 body was about $1,600 new when I bought mine... I think.
 
lots of good conversation/points/pov/etc...

when all is said and done...i think that full frame will not likely be part of my future...on a rangefinder or dslr.
 
Agree with the above comments however the last two years have changed these tremendously; in just two short years... The parameters defining the ratio of maximum light intensity measurable at the saturation point of pixel to minimum light intensity measurable right above the read-out noise are not related directly to the pixel size anymore. So do not wonder of why (or how) the sensor/A/D converter on the Pentax K-5 delivering higher Dynamic Range than the Nikon D3x or PhaseOne IQ180... Forget about comparing the known merits of the CCD- against the CMOS-sensors...

http://www.dxomark.com/index.php/Cameras/Camera-Sensor-Ratings/(type)/usecase_landscape

IMHO, we are on the brinks of a new era where APS-C size sensor would become the new standard for what Leica had initiated way back in 1925. And regarding to the technology what Dr. Fossum has predicted (in my previous post) I think it is readily available, in Sony's as well as Canon's plants. For example:

The 16MP 1/2.3" Exmor R CMOS sensor of Sony DSC-HX100V measures 4.62 x 6.16 = 28.46sq.mm. The APS-C sensor is 15.7 x 23.7 = 372.09sq.mm, meaning: Assuming the same processes and technology it's possible to squeeze into the APS-C sensor 372.09 / 28.46 x 16 = 209MP!

Let's prepare ourselves also for over 14 Evs dynamic ranges from the APS-C format in this year or latest in 2013.

Regards,

Bob

The Pentax K-5 sensor technology would perform even better if the technology could be scaled to an increased sensor area.

This is probably way too simple, but think about a sensor with a single photosite the size of an APS-C sensor. This sensor could count a fixed number of photons (signal). Now increase the area of the single photosite to match the size of a 135 format sensor. If the photosites technology scales, you will be able to count more photons while the noise associated with the photosite remains constant. The increase in SNR decreases the uncertainty (errors) in the photon count and improved certainty becomes more important as the available signal decreases.

I used to think the ADC was important (of course the ADC has to be good), but I no longer do. When your subject provides limited signal, nothing beats SNR.
 
lots of good conversation/points/pov/etc...

when all is said and done...i think that full frame will not likely be part of my future...on a rangefinder or dslr.

APS-C
36x24mm
6x6, 6x7, 6x8, 6x9, 6x17 ext
4x5, 5x7, 8x10 11x14 20x24 ext

There's no such thing as the best format. So what matter's is the one that work best for you !
 
When the APS-C will cost 5$ and the FF 50$, such conversations would stop taking place. Of course, bigger is always better. Currently I believe the cost of sensors as disclosed by a Canon spokesman is like 100$ vs 1500$.
 
These days movies look much better on my own TV or computer than they do in the pathetic low end digital projection theaters around here.



fine, then if you want to nitpick, it would be like the difference between watching a movie on your iPhone with Apple earbuds vs. watching it on your brand new 70" next gen HD (4K?) TV with a quality sound system.
 
The Pentax K-5 sensor technology would perform even better if the technology could be scaled to an increased sensor area.

This is probably way too simple, but think about a sensor with a single photosite the size of an APS-C sensor. This sensor could count a fixed number of photons (signal). Now increase the area of the single photosite to match the size of a 135 format sensor. If the photosites technology scales, you will be able to count more photons while the noise associated with the photosite remains constant. The increase in SNR decreases the uncertainty (errors) in the photon count and improved certainty becomes more important as the available signal decreases.

I used to think the ADC was important (of course the ADC has to be good), but I no longer do. When your subject provides limited signal, nothing beats SNR.

True.. but think about gaining 3dB on the SNR meaning ISO 8000 instead of ISO 6400; the realistic difference between the APS-C and 135 formats in today's technology (not even 6dB!!) Is it worth to have an array of Canon EF size lenses, a body larger than the M9 because of AF requirement in exchange of 3dB only? As a Leica user I would not pay for a body larger than the M-body or lenses like the AFS Nikkors. If you think the FF format only for the manual type lenses then I wonder if a sales figure of less than 100K would be feasible for Sony, Nikon or Canon. Further why they would invest in a new line of lens production designed specifically for the FF-rangefinder type for such low volumes? Or, skip the the lens requirement then why they would invest in a body if it would be destined to use lenses of others like Leica, Zeiss or CV? Lenses in one way are to secure the sales of bodies... Remember the Epson RD-1.

The m43 has set a trend: Compact size and has been accepted widely. The NEX series and the X100 have "accentuated" it.. Now the X-Pro1 and the Nex-7 are defining the new rangefinder type compacts with futuristic features. Even the size of the X-Pro1 has caused some raised eyebrows - the size limits for future too have been defined, for the body as well as for the lenses.
 
When the APS-C will cost 5$ and the FF 50$, such conversations would stop taking place. Of course, bigger is always better. Currently I believe the cost of sensors as disclosed by a Canon spokesman is like 100$ vs 1500$.

On a standard 12" wafer there will be an approximate theoretical max of 2.3 times as many APS-C sensors as full frame sensors (197 vs 84.) Of course since the wafers are circular the numbers won't be quite that high.

Then there is the matter of yield. A very rough rule of thumb for similar products is the yield goes down by the square of the die size.

An APS-C sensor has an area of ~370 square mm and full frame ~864mm square mm. Using this rough rule of thumb this means the yield of a full frame sensor is about 18% of an APS-C.

Combine these two and a single wafer with 100% yield for APS-C would result in only 15 working full frame sensors. And 100% yield is definitely theoretical. :)

So a full frame sensor would be about 13 times more expensive; this is very close to the $100 vs $1500 ratio in EdwardKaraa's post.

Of course, the cost of a full frame camera is more than just the cost of the sensor...

This is why I don't think we'll see a full frame Fuji, and why I agree that APS-C is the sweet spot of price vs performance.
 
shallow DOF with small lenses

shallow DOF with small lenses

Why full frame?

Greater control of depth of field for a given lens size and focal length.

It's pure physics really. We could look at online DOF calculators, but I would prefer to talk about some practical examples considering specific lens atributes.

e.g. 35/1.2 Nokton or 35/1.4 ASPH give super-fast and wide. Nothing remotely like these exists on a crop camera to give a sharp subject with creamy background context.

Then consider the 50/1.5 sonnars from the first golden age of rangefinder photography - they are in my opinion the perfect balance of a fast but tiny normal lens. They give incredible control of depth of field for their diminutive size and sub 200g weight.

I think this shallow DOF "look" is responsible for driving non-pros into progressively larger format sizes until they reach whatever point they decide not to carry any more heavy gear around, and so drop back down a format size. As many testify, the 35mm is a very nice compromise between DOF control and supremely portable overall system size.

Also, I LOVE genuine wide angle, and all of the 20-28mm effective field of view lenses I used on crop cameras were f/4 or slower, which is too slow for me. I never used a Nikkor 12-24, but would probably opt for that lens if I were still using a crop camera day-in-day-out. It is not cheap or lightweight, but does have the image quality. I now have 21/2.8 Zeiss which is superb and small and offers more creative DOF control on a FF camera.

Given the choice, I would only prefer to use a crop sensor for macro (for enhanced DOF when hand-holding) or telephoto (the crop factor gives fast telephotos that are much smaller and easier to hold).

Image quality itself is probably just fine with the latest generation or two of crop cameras, but my evolving style has driven me to format+aperture combinations that allow subject isolation at short-to-middle distances. I can't do that with slow zoom lenses, and I can't do it as well on crop sensor formats for the wide and normal focal lengths I prefer to use to follow my vision.
 
I'm currently stuck in DX for my digital Nikon, and really can't wait for an affordable full-frame body, for reasons that have already been mentioned, including better quality at the same Mp resolution and the ability to have more Mp, which I do need for some of my work, higher effective "film" speed/quality, less depth of field, but the biggest difference for me is that I live in the W/A-50mm zone, and the 1.5 crop factor steals a lot of that from me. I would really like my 20mm to be a 20, not a 30, so that I wouldn't have to lug around the gigantic, and not too good, 10-20 Sigma to get a real 20mm.

The specific reason I dumped all of my Leica gear and went entirely to Nikon was that when I did it, Leica had just announced their first digital, and I realized it was going to be a while before they made an (unaffordable to me) full frame camera. I thought Nikon would get to that point faster. I certainly was not going to tolerate using my 15mm C/V as a slow 23mm lens for the next ten years until used full frame digital Leicas became affordable. (I did keep my IIIa and a couple of lenses, though :)
 
Comparison

Comparison

I agree with your arguments. I am still sitting on "FF" film RF bodies, because I cannot afford the only FF digital mirrorless on the market.

Here is an arbitrary comparison, but exactly what it means using a 1.5 crop or FF sensor and related lens, regarding DoF.
Fuji X-Pro 1 with 35mm @ f/1.4 vs D700 with 50mm AF-S @ f/1.4 (Courtesy of vladdodan.ro)



Do you see the big difference in the background blur? Exactly the same DoF (or FoV) problem as using M-lenses, which are designed for FF, on a crop sensor...

I was always very interested in affordable digital bodies for M-lenses like R-D1, m4/3, NEX or GXR-M, as well as the new X-Pro1. But looking at this differences and thinking at the wide lenses :confused: Still intrigued and confused at the same time...
 
Last edited:
I reject the notion that FF cameras need to be physically large. The XA and Epic are full frame cameras.
If by FF we refer to digital and not 35mm film, then you won't get a small FF camera soon.

Think about the XA's design needs. The electronics for shutter and meter are simple and small. You need a small amount of room on each side of the shutter for film cassette and take up spool. Behind the shutter, you don't need much room because of the thin film.

With digital, you need more room for the electronics, the sensor and its mount. Room for the connection between sensor and electronics and the filter in front of the sensor. Combine that with the need of larger lenses that allow the light to hit the sensor at the right angle and FF bodies will be large for some time.
 
I am going to go with bigger being better, in general here. In my film work, that gradually took me to 8x10 with all the format pit-stops along the way. If I could have a MF back for my Hassie's at the price of a used D3, I'd do it in a heartbeat. I love the files from the D3 and the M9. I like having the full range of wide-angle lenses available. And I like how a 50 and 35 look on a FF sensor. Folks talk about "crop-factor" like the laws of physics have been repealed, but the relationship between near and far objects, the drop off of your focus zone etc. are primarily determined by the lens. Before we get into a long thread about how DOF actually is film or sensor-size dependent, let me just say, "I don't care." A 35 mm lens on a 1.5 crop sensor just doesn't do it for me, when a 35mm lens is what I want to use.


My experience using a Hasselblad and I love the quality of it . If I want to go in to digital I prefer the FF as according to my new experience with M8 and it crop factor makes me lot of uncomfortable with using my familier focal length lenses. But Quality of M8 is worthy enough for the price . If I can have a digital back on my Hassy it'll be more worthy than a 35mm FF or M9 camera
 
Back
Top Bottom