[1r0nh31d3]

I must be the only person who likes to take photos at night.
I was out with a friend taking pics, we both had 50mm 1.4 but I have a pentax kr and he has a 5d mark ii. I was amazed at how much more light his camera was taking in. That was the point where I started to ask questions about what I could do differently if I could put my pentax lense on a full frame camera.

[ElJamoquio]

Not sure where this hypothetical is headed, but one aperture per focal length would have a near linear relationship (with a levelling off at the wider angles) in order to maximise acceptable DoF and sharpness. 28mm f/2.8, 40mm f/4, 70mm f/5.6, etc. but why limit FLs to a single aperture anyway? How does it relate to the debate here? If it's about the extra stop of DoF the larger format offers, then that's all good. Not everyone needs this whilst it adds to the bulk of the gear.

Not sure if I was clear enough - I meant one 'lens' per focal length (that could be your answer, just checking). Just curious where your tradeoff of cost/weight/sharpness/etc is.

[andreybs]

I think that FF will not die because growing megapixel density has its reasonable cause. Modern sensors technology does not allow making photos without noise issue at all. Every new sensor of the same format can have increased megapixel density but its effective ISO level will be the same (or may be worse). So there is always a compromise between ISO setting and noise level while shooting in low-light conditions. And it could be easily solved if your sensor has enough megapixels to spare for image resizing (lowering resolution) in order to eliminate almost all noises received even on high ISO levels. Its all about the light... If you have FF-sensor you can increase megapixel density without decreasing pixel surface saving the light for every single pixel. That is why 16Mp APS-C sensor is good, but 32Mp FF-sensor is twice better. And It will be great if Pentax will make FF-camera with 30-40Mp sensor like Nikon D800. I think it will definitely find its popularity among Pentax users.

[ElJamoquio]

FF will not die out anytime soon.

In the past SNR improved dramatically with every generation, though. It still can, it has a ways to go before it reaches theoretical limits.

Nevertheless actual technology has stagnated - over the past 3.25 years the best SNR has remained roughly constant. Are we at a practical limit? I don't think so, but we'll see.

[Ash]

I wonder what the theoretical limits are, since as it is, we have 24Mp APS-C with excellent SNR to at least ISO 1600, and acceptable noise to ISO 12,800. Of course, there will be more head room with SNR with larger sensors, and that advantage is up to each photographer to decide whether it is a worthwhile venture or not. With the current lenses Pentax have made, I can say I am satisfied photographically with the results I have so far on the K-5 IIs. If a K-3 stumbled into my hands, I'd keep it, but at the moment, I'm enjoying shooting with the K-5 (when I have time to do so).

As for the hypothetical one lens per FL, I almost go by this already, as my go-to lenses are all primes, the exception being the DA 12-24, which I have found just that little bit more versatile than the DA 15 Ltd in WA photography. My tradeoff (personally) for cost/weight/sharpness is prioritising sharpness = size/weight > cost. So, as much as I'd be keen on the added benefit of 36Mp in a FF format, I do not really want the added weight and size of the camera and its lenses, and personally cannot justify the added cost anyway, since the IQ I am currently garnering from the K-5 IIs is adequate for my 'needs'.

[ElJamoquio]

Honestly I don't agree with your assessment of ISO 1600.

[wildman]

I wonder what the theoretical limits are, since as it is, we have 24Mp APS-C with excellent SNR to at least ISO 1600, and acceptable noise to ISO 12,800

A hypothetical so far as ISO is concerned:

If we did have a camera who's base ISO was say 12 or 24 K the ISO setting would become a non-issue.
Everything would be taken at say 24k and the only issue is to turn down sensor sensitivity electronically to avoid over exposure.
For my bird photography with slow long lens' that would be photographic nirvana.

[Ash]

If we did have a camera who's base ISO was say 12 or 24 K the ISO setting would become a non-issue

Whilst this does sound feasible, pragmatically it is probably a challenge to the engineers to create such a sensor and still produce acceptable images at ISO 100.

Honestly I don't agree with your assessment of ISO 1600

Depends on the exposure of the image taken. Underexposed, and obviously there will be noise creeping in, particularly in the shadows, when processing the image. Anyway, for arguments sake, there is an ever advancing capability these sensors have, and it may be that there is a practical limit to APS-C, but unless we know this then we can't say for sure FF has an economical advantage for SNR benefit over APS-C. i.e. the point where there is a diverging difference in SNR benefit between APS-C and FF notwithstanding resolution differences, and thus the law of diminishing returns no longer applies. That's when FF will come out as a more desirable camera to the greater population (but it still has to be small enough and light enough with its FF lenses to carry around...)

[Rondec]

A hypothetical so far as ISO is concerned:

If we did have a camera who's base ISO was say 12 or 24 K the ISO setting would become a non-issue.
Everything would be taken at say 24k and the only issue is to turn down sensor sensitivity electronically to avoid over exposure.
For my bird photography with slow long lens' that would be photographic nirvana.

Funny. I would much prefer a sensor with a base iso of 25.

[wildman]

it is probably a challenge to the engineers to create such a sensor and still produce acceptable images at ISO 100.

If the the base ISO is, say, 12K, why would we need to mimic an ISO of 80?
By "base" iso all I mean is If the base ISO is now, 100 and the base ISO was moved up to, say 1600 that would mean the performance you get at 100 you now get at 1600.
At least that's all I mean for the purposes of this post and I am speaking only hypothetically which is taking a lot for granted I'll admit.

But sometimes it's just fun to speculate.

[ElJamoquio]

A hypothetical so far as ISO is concerned:

If we did have a camera who's base ISO was say 12 or 24 K the ISO setting would become a non-issue.
Everything would be taken at say 24k and the only issue is to turn down sensor sensitivity electronically to avoid over exposure.
For my bird photography with slow long lens' that would be photographic nirvana.

The premise isn't really valid, though. Even at quantum efficiency the signal on 24k ISO isn't very high.

[ElJamoquio]

Depends on the exposure of the image taken. Underexposed, and obviously there will be noise creeping in, particularly in the shadows, when processing the image.

The colors at ISO 1600 are usually poor. The noise is there.

I'm a little picky but not SUPER picky. That said when I take out the noise the colors are smeary.

1600 isn't terrible, don't get me wrong, but it's marginal in terms of being able to produce a 'great' picture IMO. I like to keep at at ISO 400 and below, and honestly most of the time I work to keep at at 80-100.

We can ignore our differences in opinion though and I acknowledge that at 'some' ISO, FF will be marginal and APS-C unacceptable, etc., etc.

In my experience the improvement over APS-C isn't really noise at high-ISO (I always just think in equivalent terms)... it's that the equivalent ISO goes down (on FF), to, say, ISO 40 or so. I love that, personally, but alone it probably wouldn't be worth it to purchase a FF camera.

but it still has to be small enough and light enough with its FF lenses to carry around...)

That's the thing though... my D600 kit is lighter than my K-5 kit. It depends on what you're comparing of course. If I had just the K-5 and a 40mm it'd be lighter. If I had the K-5 and a few primes, it's heavier. It's a shame because I love primes, but it's tougher for me to justify them with good quality zoom lenses on FF.

I'm sure Pentax is working on a great FF, though, so I can't wait to see it come out...

[ChristianRock]

Those who think Canikon are "imposing" FF are being irrational.

The fact is, the market dictates those things. If FF is increasing market share, has it occurred to the "persecuted mentality" people that it just might be that this is what people want to buy?

I've been budgeting a used Canon FF+3 primes setup... still not where I can swing it price-wise, but there's no denying the FF appeal to some of us. Other than my kit lens, everything Pentax I've ever bought covers the FF circle.

[jsherman999]

A hypothetical so far as ISO is concerned:

If we did have a camera who's base ISO was say 12 or 24 K the ISO setting would become a non-issue.
Everything would be taken at say 24k and the only issue is to turn down sensor sensitivity electronically to avoid over exposure.
For my bird photography with slow long lens' that would be photographic nirvana.

What you're describing sounds like what the guys at places like SensorsGen call an 'ISO-less camera'.

You have the ISO base backwards, but your idea is effectively the same - the way an ISO less camera would work is that every image is actually captured at base ISO, which would be 80, 100, whatever (no matter what ISO you have dialed in,) - and then gain is applied in-camera only to the image that appears on the LCD, so you can see the image as you wanted to capture it, get the proper feedback in the LCD. Then, the image would be saved to the card at base ISO with some metadata written to the Exif so that your raw converter (ACR, whatever) would know how much brightness to apply to the image to simulate that ISO setting. Basically brightness is being added in post to a base-ISO image vs. baking it into the image before writing it to card.

The sensor for such a camera would have to have a very flat read-noise curve for that to work. Right now, your camera has higher read-noise (different than shot noise) at base ISO than at higher ISOs, so if you try to simulate this by shooting a low light shot at base ISO and then push it in post, the higher read noise becomes visible and you see a noisier image than if you had just shot at the 'proper' higher ISO in the first place.

ISO-less would allow you to change ISO after the fact, during PP, with a slider - just like you can change things like 'white balance' or shot temperature. And it would be much harder to blow highlights.

As soon as we have a truly flat-curve sensor (and a camera body that supports it and updates to ACR to read the new Exif metadata,) we will be able to truly adjust 'ISO' in post. It's also interesting that the K5 and D7000 already has a very flat read-noise curve, and any improvement on that would come close to being a true ISO-less camera.

.

[Rondec]

Those who think Canikon are "imposing" FF are being irrational.

The fact is, the market dictates those things. If FF is increasing market share, has it occurred to the "persecuted mentality" people that it just might be that this is what people want to buy?

I've been budgeting a used Canon FF+3 primes setup... still not where I can swing it price-wise, but there's no denying the FF appeal to some of us. Other than my kit lens, everything Pentax I've ever bought covers the FF circle.

I don't think there is any "persecuted mentality." I just don't personally think that most people wake up in the morning wondering when full frame will be cheap enough to buy it. Yes, there are people for whom that is the driving focus of their life, but most people could care less about the size of a sensor in their camera.

The issue is when we forumites project our desires onto the population as a whole and assume that they are tracking the same direction that we are.

I do want full frame, but I understand as well, that it is not likely to dramatically change my photography ability. If a family member asks for a camera recommendation, I would still recommend a mid-level APS-C camera, as I feel like these will meet their needs nicely with a big bang for a relatively small buck.