[Qwntm]

Yes, exactly, to be honest, the main issue that I have , is not not having pixel shift, or not having a FF camera. My main concern is not to be there at the right place and time... I should carry a camera all the time, because when I'm there at the right moment, I don't have the camera and when I have the camera with me, nature does not show-up...

Boring subjects look great in great light, great subjects look boring in boring light. Always try to shoot great subjects in great light... This is actually the hard part, everything else is controllable and easy.

[Ratcheteer]

Always try to shoot great subjects in great light

and don't forget subject isolation, having a sharp background, or funny bokeh aberration that pixel shift causes can distract from the main subject if its in the foreground which is the number one reason why people want full frame in the first place (less pissing around in post)

[Rondec]

and don't forget subject isolation, having a sharp background, or funny bokeh aberration that pixel shift causes can distract from the main subject if its in the foreground which is the number one reason why people want full frame in the first place (less pissing around in post)

I'm not clear what you are talking about. If you are talking about narrow depth of field, then yes, you wouldn't use pixel shift for it. But my experience is that that is more applicable to portrait photography where you probably would not be printing at such large sizes as Ed is talking about here. In the majority of landscapes, the goal is not to obtain bokeh at all, but rather to have everything in focus -- foreground to background and if pixel shift adds detail to those photos, then it can be beneficial.

I don't think Ed was saying that there isn't a place for full frame cameras. Obviously he owns one. He is just saying that for this specific application pixel shift can give results that compare to a high pixel count full frame camera.

[biz-engineer]

and don't forget subject isolation, having a sharp background, or funny bokeh aberration that pixel shift causes can distract from the main subject if its in the foreground which is the number one reason why people want full frame in the first place (less pissing around in post)

Pixel shift does not change the information incoming to the sensor. Pixel shift equates oversampling. The only situation where pixel shift result in aberration is when the image changes between the 4 samples being captured. If sharp / blurred areas do no change during the pixel-shift sampling process, there is no aberration, the edge between sharp plane and blurred plane also gets more resolution.

[Ratcheteer]

Pixel shift does not change the information incoming to the sensor. Pixel shift equates oversampling. The only situation where pixel shift result in aberration is when the image changes between the 4 samples being captured. If sharp / blurred areas do no change during the pixel-shift sampling process, there is no aberration, the edge between sharp plane and blurred plane also gets more resolution.

I'm talking about the aberration you get from the lens, the problem with bayer filters is you always lose some detail due to diffraction but that also helps hide lens flaws. by the way oversampling is the wrong word to use to describe pixel shift (oversampling would imply mathematical guess based on nearest neighbour) leveraging is a better word to use as it takes 4 sets of data and works out the average for each pixel which in turn improves overall image accuracy

[Qwntm]

and don't forget subject isolation, having a sharp background, or funny bokeh aberration that pixel shift causes can distract from the main subject if its in the foreground which is the number one reason why people want full frame in the first place (less pissing around in post)

Most people who use this as reason to justify full frame really have no idea what the depth of field for a F1.4 50mm or 85mm lens looks like on a full frame camera.


The depth of field is SO NARROW that it is really only usable in VERY and LIMITED specific situations. F2.8 on an APS-C camera is MORE THAN sufficient for isolating a bride for a very close head shot bridal portrait, for example. Her portrait would look ridiculous if her nose were in focus and her eyes weren't or her eyes were in focus but her nose wasn't.


As someone else said, I think Full Frame is great. But only very marginally better than APS-c and the gap is closing.


In practice I have been shooting APS-c weddings since the Canon 20D and NEVER had the reason or desire to shoot FF weddings. The only thing I wanted FF for was resolution and rendering in landscapes, and APS-c has killed that need/desire.


I will probably buy a Pentax FF if it's any good.

[chiane]

That is not a very good answer. There is nothing personal in this. We make images with cameras and lenses and then compare them. It is a fine thing to do, but many factors can influence the end product. Making everything equal is not so simple. If we want to declare a particular camera outperforms another camera, we should explain how we tried to keep everything equal. If we skip that part, serious readers will discount our judgment.

Look dude, it's a personal opinion. I hate how this place goes all ham on anyone trying to give any opinion that isn't formulated out of perfect scientific lab test that need to hold up in Scientific America. I mean really. You need to evaluate your life if you are giving this guy grief for making a good effort at providing info for others.

Reminds me of Nation Lampoons vacation when the grouch old father in law, after looking at Clark's giant holiday light display, says, "Clark, you know the little white lights aren't blinking".

[Qwntm]

I'm talking about the aberration you get from the lens, the problem with bayer filters is you always lose some detail due to diffraction but that also helps hide lens flaws. by the way oversampling is the wrong word to use to describe pixel shift (oversampling would imply mathematical guess based on nearest neighbour) leveraging is a better word to use as it takes 4 sets of data and works out the average for each pixel which in turn improves overall image accuracy

I say shoot both and decide what does what and is useful for what.


There's a lot of things about digital photography that in theory are wonderful and in practice are meaningless or bad. Like AA filters.


Pixel shift makes my K3II return D810 images under certain conditions which tend to be 90% of what I would want a D810 to shoot. End of story for me.

[Rupert]

Very interesting! I'm impressed. When you can show me a huge print with a pixel shifted squirrel...I'm a buyer!

Regards!

[biz-engineer]

Well, buying 3 or 4 cameras model (K-5, K5-IIs, K-3, K-3II, D750, D810, 7DII) and lenses might well represent as much money as buying a new MF system. What I learned with Pentax is that I spent as much money for 3 APSC upgrades than if I had bought a D800E straight away. But, Pentax executives once stated that it was good to offer many different lenses (covering the sames FL ranges) because people like to try them out. I agree with the Pentax exec. If we buy a 645z system right away, we have nothing to buy for the next 10 years, no need to wait for something else, it's boring. It is more exciting to buy a K-3, then a K-3II, and then wait for a FF or another APSC with more BSI pixels, then upgrade to another full frame or whatever.

[Ratcheteer]

Most people who use this as reason to justify full frame really have no idea what the depth of field for a F1.4 50mm or 85mm lens looks like on a full frame camera.

that's a bs generalisation your forgetting some of us have shot with actual 35mm film and larger so we know damn well how shallow the depth of field is and what to expect

There's a lot of things about digital photography that in theory are wonderful and in practice are meaningless or bad. Like AA filters.

aa filters are not a bad thing, the are just not needed today as the latest and greatest camera's have several times more pixels to work with than cameras a decade ago



o

[normhead]

that's a bs generalisation your forgetting some of us have shot with actual 35mm film and larger so we know damn well how shallow the depth of field is and what to expect

I shot FF MF and larger, for many years, yet never bothered to get a 1.4 lens, the fastest I shot was 1.8. How many of us have shot ƒ1.4 on film? I doubt it's very many, you want us to run a poll? I know I never did. If you shot 1.4 on film then you also know, the problem on film always seemed to be getting enough DoF, not getting narrow DoF. I often ended up shooting 200 ISO or 400 ISO because I knew I couldn't get enough D0F with a finer grain film.

[stevebrot]

...really have no idea what the depth of field for a F1.4 50mm or 85mm lens looks like on a full frame camera.

The depth of field is SO NARROW that it is really only usable in VERY and LIMITED specific situations.

So true and consistently finding accurate focus into that range is VERY difficult, is out of the capability of currently available PDAF systems and is virtually impossible using optical manual focus with the stock focus screens on current model cameras. F/1.8 is somewhat more accessible, but just barely.


Steve

(...uses a well-calibrated split image finder...)

[stevebrot]

How many of us have shot ƒ1.4 on film?

I have and for the most part, it ain't fun. Ditto for f/2 and faster at greater than 75mm.


Steve

[stevebrot]

What I learned with Pentax is that I spent as much money for 3 APSC upgrades than if I had bought a D800E straight away.

I feel for ya man...

I have been shooting flagship model Pentax digital for almost eight years and have spent less than $2500 for APS-C gear.


Steve