Ad
Ad
Ad
Pages: « 1 [2] 3 »   Bottom of Page
Print
Author Topic: Why shoot with a high resolution camera?  (Read 4289 times)
KevinA
Sr. Member
****
Offline Offline

Posts: 898


WWW
« Reply #20 on: February 20, 2013, 06:48:15 AM »
ReplyReply

Not just why shoot high res, why bother with the latest and greatest. I seriously think I could shoot 90% of my jobs with the Sony RX100 and would never hear a complaint. I'm working on pimping it out with a eye level screen viewer and a bellows lens shade.
I'm going to mix it in jobs with all the Canon X shots and see if I get any feedback.
Logged

Kevin.
arlon
Full Member
***
Offline Offline

Posts: 126



WWW
« Reply #21 on: February 20, 2013, 09:08:15 AM »
ReplyReply

I am a rank amateur and need all the help I can get. With more pixels I find I can make more mistakes that are more easily fixed in PP. If I were a pro and shot perfectly composed photos of perfect subjects with perfect lenses under perfect light conditions and could wait all day to get the shot, I could probably get away with a "brownie camera". I'm not, I have to shoot what I can when I can and 36mp is a nice crutch.
Logged

Honey, did you bring an extra battery?
PhotoEcosse
Sr. Member
****
Offline Offline

Posts: 581



« Reply #22 on: February 20, 2013, 09:20:34 AM »
ReplyReply

As the owner of both a D800 and a D800E, I am staggered that so much of the above discussion relates image resolution to print size.

Sure - good resolution is handy if you want to make massive prints (although do not over-state this as, the bigger the print, the farther the distance at which it should be viewed).

But, of infinitely greater importance in my opinion is the fact that high-res sensors, such as those in the D800 or modern MF digitals, give you the huge advantage of much larger Raw files. The D800 lossless-compressed files are 45Mb. As modern photo-processing is, in fact, data processing pure and simple, the more data you have available, the greater are your processing options.

The Raw files from my Nikon D3s were only about 16Mb. With the D800 I have three times as much data to play with. That is an immense advantage.
Logged

************************************
"Reality is an illusion caused by lack of alcohol."
Alternatively, "Life begins at the far end of your comfort zone."
jonathanlung
Jr. Member
**
Offline Offline

Posts: 70


« Reply #23 on: February 20, 2013, 11:19:47 AM »
ReplyReply

PhotoEcosse brings up an interesting point: even without bringing up humungous prints or cropping, low-level manipulations are far more forgiving with more resolution. I once worked on a publication that used a lot of masking of photos to do cut-outs. We generally had a 1MP rectangle containing a person that we'd have to mask. Masking was laborious because of the precision required at our (very modest) print sizes. With more resolution, despite the computer being less responsive, the workflow would have certainly been much faster. Oh, how output from a Canon 10D would have been welcomed!
Logged
Josh-H
Sr. Member
****
Offline Offline

Posts: 1905



WWW
« Reply #24 on: February 21, 2013, 02:53:39 AM »
ReplyReply

Quote
36mp is nice, 300 is even better. I tend to agree that the more the better, even in relatively small prints it adds something to the reality of the print, a somehow hard to describe sense of infinite depth, almost fractal in nature. But I am getting carried away.

Bernard you are in danger of sounding like one of the MFDB pundits you have so often bemoaned for claiming their images have a 'quality' not found in lesser pixel devices.... Just sayin... Grin
Logged

BernardLanguillier
Sr. Member
****
Offline Offline

Posts: 7855



WWW
« Reply #25 on: February 21, 2013, 03:15:14 AM »
ReplyReply

Bernard you are in danger of sounding like one of the MFDB pundits you have so often bemoaned for claiming their images have a 'quality' not found in lesser pixel devices.... Just sayin... Grin

Yes, I know I am on a very dangerous slope here.  Wink

Cheers,
Bernard
Logged

A few images online here!
hjulenissen
Sr. Member
****
Offline Offline

Posts: 1675


« Reply #26 on: February 21, 2013, 03:35:27 AM »
ReplyReply

Sure - good resolution is handy if you want to make massive prints ...But, of infinitely greater importance in my opinion is the fact that high-res sensors, such as those in the D800 or modern MF digitals, give you the huge advantage of much larger Raw files. ...
(I edited out a couple of sentences).

If it is not the resolution, then what is the advantage of bigger files? If you re-save the (partially developed) raw file as a full-resolution 16-bit tiff, you will have many times as much data to work with. Does it make the end-result any better?

-h
Logged
ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7308


WWW
« Reply #27 on: February 21, 2013, 04:06:57 AM »
ReplyReply

Hi,

There is nothing wrong having a bit more resolution. On the other hand I don't know if all the pixels are needed.

Very clearly, an 80 MP back with the best lenses can produce better detail than Bernard's D800 (without stitching). Bernard's D800 will produce better detail than my Sony Alpha 99 and that camera gives much more detail then my Sony Alpha 700. Some of my best images were done using that Alpha 700 (12 MP) and they make perfect prints in A2 size.

Same goes for DR, my Sony Alpha 99 is purported to have very good DR (almost like D800) but I took me long until I found a situation that really indicated the DR advantage over lesser Alphas.

A compelling image is compelling whatever MP the camera has and a boring image is a boring image whatever the resolution of the sensor or lens.

Best regards
Erik

Bernard you are in danger of sounding like one of the MFDB pundits you have so often bemoaned for claiming their images have a 'quality' not found in lesser pixel devices.... Just sayin... Grin
Logged

BartvanderWolf
Sr. Member
****
Offline Offline

Posts: 3553


« Reply #28 on: February 21, 2013, 05:16:07 AM »
ReplyReply

If it is not the resolution, then what is the advantage of bigger files?

Hi,

Microcontrast, and the ability to manipulate that accurately without introducing visibly distracting artifacts.

Cheers,
Bart
Logged
BartvanderWolf
Sr. Member
****
Offline Offline

Posts: 3553


« Reply #29 on: February 21, 2013, 05:16:35 AM »
ReplyReply

Bernard you are in danger of sounding like one of the MFDB pundits you have so often bemoaned for claiming their images have a 'quality' not found in lesser pixel devices.... Just sayin... Grin

I agree with Josh, but I understand Benhard's point of having subject surface/material structure becoming almost tangible. One can e.g. see the difference between cardboard and leather. However, there is not much (some still helps, but not much) to be gained by having more pixels than the output modality needs. Enough is enough. When we have enough pixels to satisfy a 720 PPI output without upsampling, there is not much to be gained by having more (other than room to go larger and some postprocessing benefits).

Since microcontrast is lost at the verge of limiting resolution, it helps (some types of photography) to stay away a bit from that absolute black hole which has no life in it anymore. The question then becomes, how far do we have to keep away from that dead-end by having some excess resolution?

In theory, in a perfect world, we could quantify that need as follows:
  • It is generally accepted (e.g. by the ISO organisation) that a 10% modulation of the MTF curve, corresponds quite well with the perceived human visual limiting resolution. That would also leave some room to tweak that remaining low level of microcontrast in postprocessing. So everybody should be happy when we can maintain that level before we reach the Nyquist limit (with associated aliasing issues beyond that), so preferrably with some sort of low pass filter in place.
  • A very good lens, at its optimum aperture, in combination with a sensor element (sensel) that takes an average aperture sample of the projected image, e.g. with a microlens, will produce a result in the optimal focus plane that resembles a Gaussian shaped Point Spread Function (PSF) with a radius of sigma=0.7. Proper deconvolution sharpening can boost that to perfect pixel resolution.
  • Output resolution of 8 lp/mm (406 PPI) is generally considered to be very high quality (5 lp/mm is already good), but because of vernier acuity of human vision, we can still appreciate somewhat higher resolution as an improvement. Hence 600 or 720 PPI (=11.81 or 14.17 lp/mm) without interpolation is a practical goal and upper limit.

If we can accept those criteria, then a bit of math (you'll have to trust me on the validity of the derived formula) will tell us the following:
MTF10 is achieved at cycles/mm = 1.072983 / (pi x senselpitch x sigma),
thus 14.17 = 1.072983 / (pi x senselpitch x 0.7), or senselpitch = 0.0344 millimetre without magnification. When we want to magnify our sensor size 10x for output (e.g. 36mm to 360mm), we then divide the 0.0344 mm, or 34.4 micron, sensel pitch by 10 (so a D800 with a 4.88 micron sensel pitch would not suffice). Or if we have a given sensel pitch, e.g. 4.88 micron, then we know we can magnify 99.98 / 14.17 = 7.06x (36mm to 254mm) without compromise.

When we need larger output with the same reading distance quality, we need more pixels (larger sensor, or by stitching). Of course we can reduce our pixel peeping requirement for more distant viewing conditions, unless we want to allow uncompromised close inspection.

This of course also assumes that one requires the MTF10 at this maximum resolution, which is not required for many subjects, but it is nice to know where to draw the line before being accused of obsessing too much ...

Cheers,
Bart


P.S. OOPS, I had made an error in simplifying the formula (a Log() vs Ln() conversion), I have corrected the relevant bits above.
« Last Edit: February 22, 2013, 05:03:04 AM by BartvanderWolf » Logged
xpatUSA
Sr. Member
****
Offline Offline

Posts: 298



WWW
« Reply #30 on: February 21, 2013, 06:57:29 AM »
ReplyReply

Interesting discussion. There's similar ones raging elsewhere, specifically re: the D800 - which many people are declaring to be the best thing since sliced bread:

http://forums.dpreview.com/forums/thread/3385756

Also a more general discussion here:

http://forums.dpreview.com/forums/thread/3386452

However, many posters over there have a tendency to ignore the laws of physics, unlike this learned forum. So it is quite refreshing to read posts with actual formulae and solid technical facts. Keep up the good work, Bart!

For the adventurous, I've put a spreadsheet here:

http://kronometric.org/phot/mtf/lens%20MTF%20vs%20Pitch.xls

You can put in light wavelength, aperture, sensor pixel pitch, and lens "quality" to get final MTF (as a fraction, not %) at the image plane. Totally un-commented, the yellow cells are inputs. Hopefully it will amuse somebody ;-)

Ted

« Last Edit: February 21, 2013, 07:24:31 AM by xpatUSA » Logged

best regards,

Ted
hjulenissen
Sr. Member
****
Offline Offline

Posts: 1675


« Reply #31 on: February 21, 2013, 07:16:29 AM »
ReplyReply

Hi,

Microcontrast, and the ability to manipulate that accurately without introducing visibly distracting artifacts.

Cheers,
Bart
If it is the filesize in itself that is an advantage, resaving the file as PSD should give the same advantage?

If it is not the filesize in itself (lets hope not), then what is it if not the spatial density with which the sensor plane have been sampled?

-h
Logged
BartvanderWolf
Sr. Member
****
Offline Offline

Posts: 3553


« Reply #32 on: February 21, 2013, 09:37:38 AM »
ReplyReply

If it is the filesize in itself that is an advantage, resaving the file as PSD should give the same advantage?

Obviously, that's why it's not about file size, although it's easier for making small scale edits.

Quote
If it is not the filesize in itself (lets hope not), then what is it if not the spatial density with which the sensor plane have been sampled?

It is the spatial sampling density which will allow a better MTF response than is possible nearer to the Nyquist frequency limit. That means that even low contrast input signals will be preserved for (micro-)contrast manipulation (e.g. Capture sharpening). When we shoot and aim for the lowest significant difference we tend to perceive, a 1% contrast difference would be needed to remain after shooting the original scene. When we aim for an MTF10, that means that original subject contrast at a certain spatial frequency cannot drop below 10%. At higher spatial frequencies we will get lower MTF modulation, and we'll still be able to use some of that.

However, while more subtle original subject contrast may be salvageable with processing, it becomes harder and harder as we lower our MTF response cut-off. When the original scene contrast is down to 1% (thus on the verge of being significant for human vision in real life), the MTF10 will reduce that to 0.1% response, which is only just encodeable in an 8-bit output value. Seems like a good target to aim for to me.

On the other hand, we gain little by going the other direction. If we already have a 10% modulation transfer, we will not improve the image quality very much (some, but not very much) by having a larger modulation for that level of detail, because we can do that with postprocessing, without needing to push very much (which could lead to artifacts).

That's why I suggest that the practical upper limit of significant image quality is reached by the three criteria I mentioned earlier (MTF10, 0.7 sigma PSF, 720 PPI output). We can always lower the bar for image quality for convenience, or file size, or storage capacity, or when the subject doesn't need it, etc., but raising the bar doesn't add significant quality (AKA diminishing returns). IMHO, of course.

Cheers,
Bart
« Last Edit: February 21, 2013, 09:52:17 AM by BartvanderWolf » Logged
arlon
Full Member
***
Offline Offline

Posts: 126



WWW
« Reply #33 on: February 21, 2013, 10:16:37 AM »
ReplyReply

 Forget the graphs and charts and just do some heads up comparing with the kind of pictures you like to take. Take your oldest camera body with fewer pixels and do some comparison shooting with your newest camera (assuming you've upgraded at some point).

You already know the answer or you'd still be using that old camera with less pixels. More pixels density is better or you'd still be using your Sony Mavica.
« Last Edit: February 21, 2013, 10:22:34 AM by arlon » Logged

Honey, did you bring an extra battery?
hjulenissen
Sr. Member
****
Offline Offline

Posts: 1675


« Reply #34 on: February 21, 2013, 11:32:41 AM »
ReplyReply

Obviously, that's why it's not about file size, although it's easier for making small scale edits.
Well, the quote that I responded to said quite clearly: "the huge advantage of much larger Raw files".
Quote
It is the spatial sampling density which will allow a better MTF response than is possible nearer to the Nyquist frequency limit.
There is no evidence to suggest that e.g. my 18MP 7D has a better response close to its Nyquist frequency than my old 350D had at its Nyquist frequency. On the contrary, the system probably has worse response since the Nyquist frequency has been shifted upwards, while some other bandwidth-limiting factors have stayed more or less constant.
Quote
That means that even low contrast input signals will be preserved for (micro-)contrast manipulation (e.g. Capture sharpening). When we shoot and aim for the lowest significant difference we tend to perceive, a 1% contrast difference would be needed to remain after shooting the original scene. When we aim for an MTF10, that means that original subject contrast at a certain spatial frequency cannot drop below 10%. At higher spatial frequencies we will get lower MTF modulation, and we'll still be able to use some of that.
I dont get this line of reasoning. It is what it is. We can use umpteen parameters to describe what is fundamentally a frequency response, but what you (best case) get when sampling at a higher rate is more of everything. The 1% contrast would be shifted upwards, so would the 10% contrast and the 50% contrast. So the question is (in my simple mind) do we need to shift the spatial frequency response upwards, do we need more resolution?

-h
Logged
Hulyss
Sr. Member
****
Offline Offline

Posts: 491



WWW
« Reply #35 on: February 21, 2013, 11:51:49 AM »
ReplyReply

Meanwhile, with a Voightlander Bessa 667 on a 5mm mark at 8 meters :

Logged

Kind Regards - www.hulyssbowman.com
KevinA
Sr. Member
****
Offline Offline

Posts: 898


WWW
« Reply #36 on: February 21, 2013, 12:13:24 PM »
ReplyReply

Forget the graphs and charts and just do some heads up comparing with the kind of pictures you like to take. Take your oldest camera body with fewer pixels and do some comparison shooting with your newest camera (assuming you've upgraded at some point).

You already know the answer or you'd still be using that old camera with less pixels. More pixels density is better or you'd still be using your Sony Mavica.
Well I've dropped the amount of pixels I shoot, gone from 21mp to 18mp, not much I know. But honestly all this talk about more or less pixels making this or that difference, it's the sort of thing that could disappear yourself up your backside. No doubt if I'd gone for a D4 my photography would of suffered even more.
Logged

Kevin.
BartvanderWolf
Sr. Member
****
Offline Offline

Posts: 3553


« Reply #37 on: February 21, 2013, 12:56:49 PM »
ReplyReply

I dont get this line of reasoning. It is what it is. We can use umpteen parameters to describe what is fundamentally a frequency response, but what you (best case) get when sampling at a higher rate is more of everything.

Hi,

Well, when it's more modulation, the local detail contrast gets better. When a given subject detail is sampled with a higher sampling density, it's contrast is retained more accurately, because it is relatively lower on the scale towards Nyquist. That can make a difference (see attachment) when the original contrast was low to begin with (e.g. <=1% original subject contrast), such as in subtle material surface structure.

What you are comparing is not the same level of detail, but coarser (near lower Nyquist frequency) with finer detail (near higher Nyquist frequency). That's a bonus, but will possibly be of too low contrast to make a difference in final output, because some of that the contrast will be lost. It will help in processing some of the other data but it is in the excess detail category where one eventually lands without significant benefits (like using a D800 for web publishing, unless one crops a lot).

Cheers,
Bart
« Last Edit: February 21, 2013, 01:29:47 PM by BartvanderWolf » Logged
ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7308


WWW
« Reply #38 on: February 21, 2013, 01:06:16 PM »
ReplyReply

Hi,

A long time ago C'tein published an article on the issue, it may be worth reading:

http://theonlinephotographer.typepad.com/the_online_photographer/2009/02/why-80-megapixels-just-wont-be-enough.html

Best regards
Erik


Another discussion on hyperfocal distance raised the question-
--
It makes sense only if one doesn't print large output. That would raise a question as to why use a camera with high resolution to begin with, but that's a different subject ...
--

So why does we shoot with high res. cameras?

How big is large print output?

Dave S
Logged

David S
Full Member
***
Offline Offline

Posts: 126


« Reply #39 on: February 21, 2013, 01:47:30 PM »
ReplyReply

Hi,

A long time ago C'tein published an article on the issue, it may be worth reading:

http://theonlinephotographer.typepad.com/the_online_photographer/2009/02/why-80-megapixels-just-wont-be-enough.html

Best regards
Erik



Thanks for the reference. Helps me have better thoughts to think about the whole issue even if I cannot afford the bigger pixel cameras.

Dave S
Logged
Pages: « 1 [2] 3 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad