Comparing resolutions correctly

[ramarren]

Advertisement (gone after registration)

All this talk of needing 40, 50, even 100 Mpixel and still seeing pixelation and artifacts seems crazy to me. Did you ever see John Isaacs' photography in exhibition? Huge, vast prints .. Bigger than 20x24 inch for sure .. Superb, beautiful detailing on eyes, feathers, everything. Not a pixel in sight. All made with an Olympus E-1 ... a 5 Mpixel camera released in 2003. 

 

Everything beyond that is icing on the cake. 

Edited March 11, 2016 by ramarren

[talt03]

You may need a microscope for that on an A3+ print and may not find a difference.

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[CheshireCat]

24 Mpixel              1.0       (100%)                 square root of 1.0

[..]

50 Mpixel              1.44     (144%)                 square root of 2.08                         plus 44%

So despite all the marketing hype the differences are rather small.

 

Not small at all.

 

Think about it this way:

- Assume 24 MP are enough for you (as for most of deniers in this forum).

- You have a 24-70 zoom.

- The zoom provides 70mm with 24MP.

- The same zoom provides the equivalent of 100mm with 24MP, when used on a 50MP camera.

 

Another example ?

- A good 24mm prime is also a virtual 24-35 zoom on a 50MP camera (24MP at the long end).

 

Of course, when Leica cameras will have 50MP and other competitors will have 100MP, the deniers will say that 50MP are just right for them and they don't need 100MP.

 

...But let's keep talking about print sizes and viewing distance and retina resolution and Leica knows better and Sony is for losers, and all the other nonsense. I find it quite amusing.

Edited March 11, 2016 by CheshireCat

[pico]

Has there ever been a compelling photograph rejected because it did not meet high LP/MM metrics?

.

[ramarren]

Has there ever been a compelling photograph rejected because it did not meet high LP/MM metrics?

.

 

 

Back when 3 to 6 Mpixel were the common norm and professional, $8000 Nikon/Canon things managed to eke out 10-12 Mpixel, a couple of the stock agencies would reject images with less than 10 Mpixel. A couple of simple upscalings and my 5Mpixel images were 10Mpixel so that took care of that—they never figured it out since I intentionally stripped all EXIF data from anything I posted or sold. 

[ramarren]

I feel that as a "happy Leica Forum denier", I'm in an elite club of people who couldn't give a rat's arse about bazillions of megapixels. We're more interested in making great photographs, not analyzing equipment specifications to the point of nausea.

 

I think I'll go fondle an M4-2 and a Leicaflex SL now.    

[talt03]

Advertisement (gone after registration)

I'm the 2nd denier who happens to own a 42MP A7RII and wanted to learn more about great shots than why more pixels is better. Unfortunately we got posts here short of saying '24MP SL is nothing but an underdog 42MP A7RII'. This is what's amusing instead of offering knowledge that will show technique to get the same quality of image produced by 24MP vs a downscaled 50MP at the same focal length.

Edited March 12, 2016 by talt03

[CheshireCat]

[...] instead of offering knowledge that will show technique to get the same quality of image produced by 24MP vs a downscaled 50MP at the same focal length.

Instead of insinuating, just kindly ask and you will be given.

[talt03]

Insinuating? Asking how should be in another post as it will be off topic. But to each his own, I want to know how I could make my SL and Sony output a 50k ISO image that looks like what a Nikon D5 can do.

[talt03]

At 20MP.

[jaapv]

Not small at all.

 

Think about it this way:

- Assume 24 MP are enough for you (as for most of deniers in this forum).

- You have a 24-70 zoom.

- The zoom provides 70mm with 24MP.

- The same zoom provides the equivalent of 100mm with 24MP, when used on a 50MP camera.

 

Another example ?

- A good 24mm prime is also a virtual 24-35 zoom on a 50MP camera (24MP at the long end).

 

Of course, when Leica cameras will have 50MP and other competitors will have 100MP, the deniers will say that 50MP are just right for them and they don't need 100MP.

 

...But let's keep talking about print sizes and viewing distance and retina resolution and Leica knows better and Sony is for losers, and all the other nonsense. I find it quite amusing.

You  are missing the point entirely. The "deniers" are not arguing against the advent of higher MP cameras, in fact they will happily buy them, albeit for other reasons, they are arguing against THEIR need for more megapixels for THEIR photography. Whether you need more megapixels is your business, whether they do not is theirs. It will be up to you to show an improvement in your photography when such cameras arrive.

[CheshireCat]

But to each his own, I want to know how I could make my SL and Sony output a 50k ISO image that looks like what a Nikon D5 can do.

 

 

 I like the "50k ISO" part.

I don't want to insinuate that you are just worried that it would be too easy for me at normal ISO, so I will assume most of your shots are taken at 50k ISO

 

But sure, please provide a download link of the very same image shot with your very own SL, Sony and D5 (because you have all three cameras, right ?) --- and I'll see what I can do.

[CheshireCat]

You  are missing the point entirely. The "deniers" are not arguing against the advent of higher MP cameras, in fact they will happily buy them, albeit for other reasons, they are arguing against THEIR need for more megapixels for THEIR photography. Whether you need more megapixels is your business, whether they do not is theirs. It will be up to you to show an improvement in your photography when such cameras arrive.

 

Like I said in my previous example, if it makes sense to have a N mm lens and a N*1.5 mm lens on a 24MP camera, then it makes sense to have a N mm lens only on a 50 MP camera.

It actually makes more sense, as you do not even need to switch the lens to get N*1.5 mm at 24MP.

 

Yes, the 50k ISO guy may not like the lower low-light performance

And yes, you need a good lens and the N*1.5 equivalent lens aperture is different.

[pgk]

Can we also look at some 'assumptions' please?

 

1. More MPixels + greater image 'quality' (for example greater detail or 'better' tonality) - certainly only potentially true if appropriate lenses are used at appropriate apertures - this may well mean that slower. higher resolution lenses are built for specific purposes such as landscape photography. Other lenses may simply be too stretched, even with software help, to provide more detail and the images that they produced will simply be sampled at a greater rate - is this higher image 'quality' or just oversampling? It will be up to the user to decide and there will no doubt be arguments as ever.

 

2. There are/will be other ways of increasing image detail - such as image stitching (very good on static subject matter already), focus stacking (images with far more data than MPixels alone can provide) and more (such as the potential for multi-shots with marginal repositioning of the camera between them and subsequent image analysis which has the potential to create increased data for example). Equating increased MPixels with greater information is far too simplistic an approach.

 

3. Cropping allows for greater flexibility. Up to a point and then all the problems associated with increased sampling will kick in. In my book cropping all too often means a lazily taken shot but there are of course exceptions (using a camera for an image intended to be used as a square print for example). However for 'quality' purposes its always best to get the composition right in camera and relying on increased MPixels to compensate for a poorly composed image is poor practice in my view, though as I've said, at time inevitable.

 

So I have a problem with the assumption so often peddled, that more MPixels is inevitably better. FWIW some of the 'demo' images I've seen showing the virtues of high MPixel cameras/backs appear to show (from their exif or captions) that they make use of as much optimisation as possible (aperture/tripod/lighting/etc.). This is fine and illustrates their optimal results but isn't 'real world' in as much as we often shoot way outside the optimal settings and consequently those higher MPixels will be 'wasted' as they do little more than oversample the less than ideal image projected onto the sensor. This happens already at lowish MPixels sometimes so Higher MPixel cameras are relevant only if the end output requires them, optimal settings can be used and technique is as precise as possible, and they are used for relevant subject matter and relevant styles of photography.

 

Personally I still consider that cameras of 15~25 MPixels are more than adequate for most photography with hand held portable cameras. Some required tripod mounted heavy medium format film cameras in the past when most cameras were 35mm and the parallel is that this situation exists today except that we now have portable hand held cameras that are being used in ways that heavy medium format cameras were in the past. Users of higher MPixel cameras could do well to determine their needs rather than, as I've said before, their wants. Some will be able to justify high MPixels and associated specialist lenses, for most of us they may still hold a fascination and be desirable, but are they really necessary, well that's up to the individual but trying to justify them by extolling virtues with they either don't have or are irrelevant is a pretty pointless thing to do.

[jaapv]

Like I said in my previous example, if it makes sense to have a N mm lens and a N*1.5 mm lens on a 24MP camera, then it makes sense to have a N mm lens only on a 50 MP camera.

It actually makes more sense, as you do not even need to switch the lens to get N*1.5 mm at 24MP.

 

Yes, the 50k ISO guy may not like the lower low-light performance

And yes, you need a good lens and the N*1.5 equivalent lens aperture is different.

You're still missing the point - but no matter.

 

It is not true that a lens must match the sensor. This is not a weakest link situation, as is often mistakenly assumed. The lens-sensor combination will work as a system to produce an image, both parts contribute.

A higher resolving sensor will still perform better than a lower one on a low-res lens and the other way around.

You will see the difference between the Summicron 50 and the Apo-Summicron on an M8 despite an alleged resolution mismatch. And you will see the difference between a 100  MP sensor and a 10 MP sensor using a Thambar. Provided your pixel-peeping skills are up to it.

In the end, however, the limiting factor will be the result that the printer brings to the paper after the image has run through the whole process of technique and technology. That is what the photographers on this thread are implying.

[CheshireCat]

It is not true that a lens must match the sensor. This is not a weakest link situation, as is often mistakenly assumed. A higher resolving sensor will still perform better than a lower one on a low-res lens and the other way around. You will see the difference between the Summicron 50 and the Apo-Summicron on an M8.

 

 

I mildly agree with you, but it was not my point.

 

All I am saying is that if you take two identical photos as follows:

- Apo 75/2 stopped down to f/2.8 on a 24MP camera

- Apo 50/2 wide open on a 50MP camera and then crop to 75mm equivalent (about 24 MP)

 

It will be very tough for you (and me) to understand which is which, without relying on lens rendering character (which is quite neutral for both cited lenses).

As a consequence: if the price difference between the 50MP camera and the 24MP camera is less then the price of the Apo 75 (which will eventually happen) then you are saving money buying only the 50MP camera and the Apo 50, and still get the same quality at 75mm and better quality at 50mm (even if you don't need it). Therefore more megapixels make you save money.

[jaapv]

That is a new one... It won't be too difficult to tell the difference - by the DOF from the smaller "sensor".

[CheshireCat]

That is a new one... It won't be too difficult to tell the difference - by the DOF from the smaller "sensor".

 

 

Nice try, but won't work. That's why I told you to shoot the 75 at f/2.8

[talt03]

I only have the SL and A7RII. D5 is yet to come. Did Nikon deliver the new D5 to shops in your place? For my purpose, clean high ISO is all that matters at A3+ prints. Soon as it comes, I'll get your expert knowledge to convert 50k iso from my SL and the 12k iso from my sony to match the D5's if there's even truth to prelaunch tests of that camera, well again at 20MP. In case you're not informed, 3M iso is part of the D5 specs sheet. Unfortunately I haven't been converted as a Sony, Nikon or Leica fanboy because of their new cameras. It's GAS that and that search for a clean high ISO 35mm camera that keeps me going.

[Advertisement]

Hallo,

Look here: Leica L-Mount

[pgk]

All I am saying is that if you take two identical photos as follows:

- Apo 75/2 stopped down to f/2.8 on a 24MP camera

- Apo 50/2 wide open on a 50MP camera and then crop to 75mm equivalent (about 24 MP)

 

Therefore more megapixels make you save money.

 

Interesting. I hadn't actually realised that buying an Apo 50/2 was a money saving solution. I'll bounce this reasoning of my wife and see if I can get her o see the logic in the argument.

 

No, wait a minute, I'm not sure I can convince her.....

 

I'll just have to stick to my very cost ineffective 50/1.4 asph and 75/2 - not as economic I know but needs must .