Ad
Ad
Ad
Pages: « 1 [2] 3 4 »   Bottom of Page
Print
Author Topic: Why full frame for landscapes at low ISO?  (Read 16118 times)
MarkL
Sr. Member
****
Offline Offline

Posts: 330


« Reply #20 on: October 21, 2009, 05:19:59 PM »
ReplyReply

Better DR
More wide lens options
Lower noise (very marginal though)

Plus I don't use my camera only for landscapes.
Logged
Ray
Sr. Member
****
Offline Offline

Posts: 8849


« Reply #21 on: October 21, 2009, 06:49:02 PM »
ReplyReply

Quote from: MarkL
Better DR
More wide lens options
Lower noise (very marginal though)

Plus I don't use my camera only for landscapes.

That about sums it up. There's a good case to be made for buying a camera so you can use a particular lens. If my Canon EF-S 10-22 for cropped format, or Sigma 15-30 for Full Frame, had been better quality lenses, I would probably not have bought a D700. My Canon 50D with its 15mp sensor should be at least the equal of the D700 resolution-wise. I haven't done any thorough comparisons between the 50D with EF-S 10-22 and the D700 with Nikkor 14-24 because the differences seem so obvious. I can't waste my time confirming the obvious.

The resolution of the Sigma was reasonably close to the Nikkor in the centre of the image but much worse than the Nikkor at the edges and in the corners. The Sigma 15-30 is much more suitable for a cropped format where it becomes a 24-48 zoom. Even the EF-S 10-22 is no wider than 16mm, full frame equivalent. The difference between the 14mm of the Nikkor and the 16mm equivalent of the EF-S 10-22 is quite substantial in terms of both FoV and over all image quality.
Logged
Daniel Browning
Full Member
***
Offline Offline

Posts: 142


« Reply #22 on: October 21, 2009, 08:04:47 PM »
ReplyReply

Quote from: HSakols
So what is the improvement one gains by going full frame if you are still shooting at a low ISO?

As far as image quality goes, for cameras that have similar performance levels (per area), the principle gains are:

  • Dynamic Range
  • Color Depth
  • Lens selection

Color depth is the SNR over the part of the dynamic range that you care to use. Even when a large sensor has *less* dynamic range than a small sensor (because of lower technology or whatever), it still has the advantage of higher color depth. For example, the D3X has 13.5 stops DR compared to 12.5 stops in the newest P65+. But if you only look at the top 8 stops and forget about the bottom 5.5, the MFDB will have higher SNR, which translates to color depth (AKA tonal gradations). For a more extreme example, compare the LX3, which as 10.6 stops of DR, vs. the 5D2, which has less than 10.6 stops of DR (if you consider pattern noise a disqualification as I do).  Assigning an importance to color depth is an exercise for the reader.

The importance of lens selection varies by taste and circumstance. There are many circumstances where the MTF of an APS-C lens is inferior to an equivalent (and perhaps more expensive) FF lens. That is because APS-C lenses must provide the same MTF at a 1.6X higher spatial frequency, though only over a smaller image circle. They also have the disadvantage of higher ratio between backfocus distance and image circle size (although the EF-S mount helped this by a few millimeters). Such inequalities may or may not be adequately addressed by additional post processing on the APS-C image depending on your taste and preferences for post.
Logged

--Daniel
stever
Sr. Member
****
Offline Offline

Posts: 1062


« Reply #23 on: October 21, 2009, 08:27:42 PM »
ReplyReply

it seems that a lot of responders here have not used full frame and crop frame side by side and or have modest print size requirements.

without stitching, 13x19 is pretty much the absolute limit for a 40D (and i believe the 50D as well which is very marginally better -- hope to be pleasantly surprised by the 7D).  The 5D is noticeably sharper than the 40D (and 50D) and can make larger prints.  The 5DII is in another class for larger prints.

if you're making smaller prints or stitching a crop frame camera gives good results - but it takes 3 stitched 40D images to equal one 5DII image - and i've found plenty of situations where movement in the subject area makes stitching impractical
Logged
juana
Newbie
*
Offline Offline

Posts: 1


« Reply #24 on: October 21, 2009, 10:56:49 PM »
ReplyReply

Quote from: Panopeeper
I do measure not negligable differences. However, I can not show how much it means in real-word images, for I don't have either camera.

The point is, that many photographers do not utilize the entire DR of the camera (most of them don't even understand, that their Black=5 cuts off a lot of the DR); however, just landscape shots do need the highest possible DR in order to have clean shadows.


That's a different issue. It depends on the budget and on the utilization of the high DR. I am landscaper but hobbyist, therefor it is too expensive for me.


I have to pass here. My remark war rather generic; I did not know, that Nikon offers such a large selection of DX WA lenses.


Well this has been most interesting reading.

Phew!

Expose to the right, compose, focus, lock up  your mirror and release your shutter.


Logged
Janifero
Newbie
*
Offline Offline

Posts: 2


« Reply #25 on: October 21, 2009, 11:13:23 PM »
ReplyReply

I agree with the implictions Stever is making.

The size of your prints and your printing technique are key measures in determining if you need a FF or not.

When I had a D200 and mid-range lenses, the upward limit of prints  I considered good was 13x19.  When I switched to the 14-24mm I was quite satisfied with 22x17 prints. So the issue of FF or no-FF was irrelevant.

On occasion, some of the D200/14-24mm  files I printed over 5ft..  I saw, or perhaps imagined seeing, less quality than I would like, but I've never once had a client complain.

I now use an infrared adapted D50 and a D3. I print on an Epson 3800.

I find I can get excellent 13x19 prints on my Nikon D50.  A similar sized print coming from my D3 does not show any significant gains that make any difference to my buyers. --At least not enough for them not to pay a pretty penny for each print.

Paper choice makes a difference too.  On 13x19 prints, when using Luster/Baryta papers, and when I really pixel peep, I  appreciate a finer contrast in the FF pictures.  If I make the same print on Fine-Art cotton papers I really cannot see any difference.  I can send out to a lab prints  4ft long coming from the D50, and if coated behind plexi and printed on metallic paper, there is no visible difference from a similar file coming from the D3 (or D200 back in the day).

I do see differences on 22x17 prints on Luster/Baryta paper, and to a lesser degree on cotton papers.  There is an almost intangible "wash-out" effect on pics from the D50, contrast details seem weaker, and the overall intangible feeling is of something softer.  To the layman I doubt anything is really noticeable though.

In summary, before upgrading, I would ask myself:

- Is my shooting technique solid?  ie. adequately weighted tripod, use of RAW, mirror lock/self-timer
- Is my glass the best I can afford?
- Can I benefit efficiently from stitching? Perhaps motion and subject don't allow this, or perhaps you don't want to spend the time
- Is my processing and printing technique solid? ie. monitor/profile calibration, sharpening, curves, calibrated printer, etc.
- Does my media allow the benefits of FF to be apparent?  ie. thick cotton papers and plexicoated can go a long way in forgiving
- Do I print in sufficient size for this to be relevant?  e.g over 13x19

Lastly, and perhaps most importantly:  Will it make a difference in helping sell or making me feel more satisfied?


Logged
Ray
Sr. Member
****
Offline Offline

Posts: 8849


« Reply #26 on: October 22, 2009, 12:02:07 AM »
ReplyReply

Quote from: Janifero
I agree with the implictions Stever is making.

The size of your prints and your printing technique are key measures in determining if you need a FF or not.

No, no, no! The OP is asking whether he should upgrade froma 12mp D300 to a 12mp full frame, like the D700. Resolution and print size is determined by pixel count, as far as the camera is concerned.

Lenses are another matter.

The advantages of the larger format of same pixel count are mainly image quality at shallow DoF which is often not a major concern with landscape photography where getting a sufficiently extensive DoF is often of greater concern.

Whilst it's true that the greater pixel density of the smaller sensor is more demanding on the lens, this again is not an issue when extensive DoF is sought. In other words, a lens at F10 is usually sharper than a lens at F16. Likewise, a lens at F5.6 is usually sharper than a lens at F9.

The same is not true at wide apertures where a shallow DoF is deliberately sought. A lens at F2 is most likely not as sharp as a lens at F3.5, yet you would need to use F2 to get a similar DoF on a D300 to that from a D700 at F3.5. In this situation, the D700 image would be clearly sharper. However, the OP is a specialised landscape shooter with little need for shallow DoF.
« Last Edit: October 22, 2009, 12:05:10 AM by Ray » Logged
pegelli
Sr. Member
****
Offline Offline

Posts: 581



WWW
« Reply #27 on: October 22, 2009, 12:23:43 AM »
ReplyReply

Quote from: ArunGaur
In fact there have been hot debates regarding Sony cameras whether Alpha700 is not equivalent or better than Alpha 900. And many concluded A900 is redundant.
Arun Gaur

Visit My Website

Gimme a break, there's more in life than stitching wide FOV's. Don't get me wrong, I love my APS-C and don't need (can't justify) FF for my hobby but saying a same pixel density FF camera with essentially the same sensor characteristic (granted, pixel peepers will always find a difference) is redundant is nonsence.
Logged

pieter, aka pegelli
bjanes
Sr. Member
****
Offline Offline

Posts: 2756



« Reply #28 on: October 22, 2009, 07:12:55 AM »
ReplyReply

Quote from: Ray
Lenses are another matter.

In the case of the D300 and D700, the resolution in terms of pixels/picture height is the same. However, the Nyquist frequency is 59 lp/mm for the D700 and 90 lp/mm for the D300. If you are shooting with a perfect lens (diffraction limited) at f/8, the MTF at Nyquist will be about 52% with the D700 and about 68% with the D700. The full frame camera can operate at a higher MTF. Nyquist is a poor choice for comparing MTF, since Bayer array cameras can not resolve at Nyquist, but the curve is nearly linear, and similar ratios apply at lower frequencies. The graph was adapted from one on Bob Atkin's web site.

This analysis neglects curvature of field, light falloff and depth of field considerations.

[attachment=17406:D300_D700_MTF.gif]
Logged
grepmat
Jr. Member
**
Offline Offline

Posts: 55


« Reply #29 on: October 22, 2009, 10:26:43 AM »
ReplyReply

I tested the D300 against my old beloved D200, carefully and extensively, and decided to skip it. The D300 was just too prone to noise in the skies at its base ISO, which I felt was a fatal flaw for landscapes. Instead, I jumped to a D700.

Friends, though the D700 has roughly the same number of pixels, it's worlds better for landscapes than the D200/D300. The improved dynamic range and concomitant low shadow noise are enormously liberating. With good lenses, e.g. the 24-70 f/2.8, it's virtually immune to chromatic aberration (I love this). The improved noise also reduces the need for tripod use, which keeps the spousal unit / significant other happy when traveling. I still keep a compact APS camera, but I can instantly tell the difference, and now I'm always slightly disappointed in the smaller brother's images. This is typically true even for telephoto use, despite the APS's imputed focal-length advantage. I still prefer APS for macro, however.

I'm in the camp that feels that we are obtaining diminishing returns with regard to pixel count. Instead, my dream is for Nikon to migrate a D700-class sensor into a camera approaching the size and cost of the D60. Meanwhile, there's more to life and photography than more pixels, and the D700 is well worth the jump from a D300-class camera for so many reasons.
Logged
ArunGaur
Jr. Member
**
Offline Offline

Posts: 57


« Reply #30 on: October 22, 2009, 11:52:14 AM »
ReplyReply

Quote from: aaykay
Are these not 2 entirely different formats to be "better" or "worser" than the other ? For one to be "redundant" just because one has the other ? We are not comparing say an A200 with an A700 here, is it, where both are products in the same format (APS-C) with one having a few more bells-and-whistles over the other ?  

The A900, being  a Full-frame, gathers 240% of the light gathered by the A700 in every identical exposure.  DOF and other factors being additional factors.  That may or may not be important for the person taking the picture and if one is happy with one's tool of choice, I think that is all that matters.

That's right. They belong to different categories and should serve the different purposes. Still there is no end to the debates comparing the two on different accounts. Of course, this confounds the matter, sometimes unnecessarily.
Arun Gaur

Visit My Website
Logged
ArunGaur
Jr. Member
**
Offline Offline

Posts: 57


« Reply #31 on: October 22, 2009, 11:57:09 AM »
ReplyReply

Quote from: pegelli
Gimme a break, there's more in life than stitching wide FOV's. Don't get me wrong, I love my APS-C and don't need (can't justify) FF for my hobby but saying a same pixel density FF camera with essentially the same sensor characteristic (granted, pixel peepers will always find a difference) is redundant is nonsence.

I would also tend to agree with you. After all there should be some-some difference. It is very much obvious. However, debates must go on in face of the obvious things.
Arun Gaur

Visit My Website
Logged
ArunGaur
Jr. Member
**
Offline Offline

Posts: 57


« Reply #32 on: October 22, 2009, 12:06:13 PM »
ReplyReply

Quote from: JeffKohn
That's pretty silly, considering the 900 is a 24mp camera. It may be redundant if you only print 8x10, but at larger output sizes there will most definitely be a difference.

To me also it appears to be pretty silly, though the color and charm in life is nothing but an assemblage of silly things. People have come out with what they claim to be conclusive proofs : "Look here these are the prints, not 8X10, but much bigger, and can you spot the difference between the two prints made from the two cameras? You cannot. So I am proved to be right".
Arun Gaur

Visit My Website
Logged
Luis Argerich
Full Member
***
Offline Offline

Posts: 155


Astrolandscaper


WWW
« Reply #33 on: October 22, 2009, 01:28:58 PM »
ReplyReply

Or you could have just read the OP question.

If you compare a 5DII to a 40D you are comparing a 22mpx camera to a 10mpx camera, the fact that it is full-frame is not the key.

Quote from: stever
it seems that a lot of responders here have not used full frame and crop frame side by side and or have modest print size requirements.

without stitching, 13x19 is pretty much the absolute limit for a 40D (and i believe the 50D as well which is very marginally better -- hope to be pleasantly surprised by the 7D).  The 5D is noticeably sharper than the 40D (and 50D) and can make larger prints.  The 5DII is in another class for larger prints.

if you're making smaller prints or stitching a crop frame camera gives good results - but it takes 3 stitched 40D images to equal one 5DII image - and i've found plenty of situations where movement in the subject area makes stitching impractical
Logged

BernardLanguillier
Sr. Member
****
Online Online

Posts: 7748



WWW
« Reply #34 on: October 22, 2009, 03:29:58 PM »
ReplyReply

Quote from: grepmat
I'm in the camp that feels that we are obtaining diminishing returns with regard to pixel count. Instead, my dream is for Nikon to migrate a D700-class sensor into a camera approaching the size and cost of the D60. Meanwhile, there's more to life and photography than more pixels, and the D700 is well worth the jump from a D300-class camera for so many reasons.

The range of apertures available is getting narrow, but the current 20+ megapixels DSLR still offer a very clear advantage over 12MP sensors when everything is done right, wihch is not always easy and requires serious oversizing of some pieces of equipment for consistency accross a variety of conditions.

Cheers,
Bernard
Logged

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

Posts: 7244


WWW
« Reply #35 on: October 22, 2009, 03:47:02 PM »
ReplyReply

Hi,

I guess your mileage may vary. The D700 has lesser pixel density than the D300 so it places less demands on the lenses. Using full frame lenses on the D300 lets you utilize the swet spot of the lens.

Best regards
Erik

Quote from: HSakols
Hi
I presently use a Nikon D300 for many of my static landscape and nature shots where I always use the native ISO (200).  I notice many landscape photographers have switched to full frame (eg D700). So what is the improvement one gains by going full frame if you are still shooting at a low ISO.  Do larger sized prints (eg 16x20) show a significant  improvement just by using a full frame sensor.   Once again I'm not asking about the high ISO advantage of full frame.
Logged

grepmat
Jr. Member
**
Offline Offline

Posts: 55


« Reply #36 on: October 22, 2009, 05:04:33 PM »
ReplyReply

Quote from: BernardLanguillier
The range of apertures available is getting narrow, but the current 20+ megapixels DSLR still offer a very clear advantage over 12MP sensors when everything is done right, wihch is not always easy and requires serious oversizing of some pieces of equipment for consistency accross a variety of conditions.
I'd agree that there is still juice left above 12 MP, but as you say, it's increasingly hard to take advantage of. And 12 MP may not be the sweet-spot for 35mm - the best all-around compromise for every job if you can have only have one camera - but it's in the ball-park.

Quote from: ErikKaffehr
The D700 has lesser pixel density than the D300 so it places less demands on the lenses. Using full frame lenses on the D300 lets you utilize the swet spot of the lens.
This speaks to my thesis (somewhat tongue-in-cheek) that "every camera is the same": If you could some-how weight all photographic trade-offs in an unbiased way, one would find that every good camera (e.g., shot-noise-limited) scores roughly the same. Do you insist on both high resolution and low noise? Well, you will lose depth of field (and have a monster camera). Want awesome depth of field plus high resolution? You will lose on noise (but you get a pocket camera), etc. The image quality would differ in certain characteristics but the net quality in some sort of lumped fashion would be the same.

Cheers.
« Last Edit: October 22, 2009, 10:29:15 PM by grepmat » Logged
ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7244


WWW
« Reply #37 on: October 22, 2009, 05:10:20 PM »
ReplyReply

Hi,

I have both full frame and APS-C, a Sony Alpha 900 and a Sony Alpha 700 sharing a similar sensor. No question that the Alpha 900 resolves higher, but I see little difference in A2 size prints from Lightroom 2.

Best regards
Erik


Quote from: grepmat
I'd agree that there is still juice left above 12 MP, but as you say, it's increasingly hard to take advantage of. And 12 MP may not be the sweet-spot for 35mm - the best all-around compromise for every job if you can have only have one camera - but it's in the ball-park.


This speaks to my thesis (somewhat tongue-in-cheek) that "every camera is the same": If you could some-how weight all photographic trade-offs in an unbiased way, one would find that every good camera (e.g., shot-noise-limited) scores roughly the same. Do you insist on both high resolution and low noise? Well, you will lose depth of field (and have a monster camera). Want awesome depth of field plus high resolution? You will lose on noise (but you get a pocket camera), etc. The image quality would differ in certain characteristics but the net quality in some sort of lumped fashion would be the same.

Cheers.
Logged

Ray
Sr. Member
****
Offline Offline

Posts: 8849


« Reply #38 on: October 22, 2009, 06:11:34 PM »
ReplyReply

Quote from: BernardLanguillier
The range of apertures available is getting narrow, but the current 20+ megapixels DSLR still offer a very clear advantage over 12MP sensors when everything is done right, wihch is not always easy and requires serious oversizing of some pieces of equipment for consistency accross a variety of conditions.

Cheers,
Bernard

I wouldn't quite agree with this, Bernard. I found no restriction in choice of apertures after comparing two very high pixel-density cameras, the Canon 40D and 50D, which on full frame would be equivalent to 26mp and 39mp.

It's true that there's a law of diminishing returns at work regarding increased resolution in the plane of focus, as one stops down towards diffraction-limited f stops. However, for landscape work in circumstances where as much DoF as possible is sought, the higher pixel-density sensor has a clear advantage.

For example, if one doesn't want to compromise resolution too much at the plane of focus, I think most users of the APS-C format would be reluctant to stop down beyond F11, maybe F13 at the most. Many lenses are sharpest around F5.6 so there is often a trade-off between degree of DoF and sharpness at the plane of focus.

This is where the higher pixel density camera has the advantage. You get approximately the same degree of sharpness and detail with a 50D at F16 as you get with a 40D at F11. But there's no doubt whatsoever, that the 50D at F16 produces greater DoF.

Comparing the 50D at F8 with the 40D at F5.6, using a standard 50mm prime (50/1.4), the 50D image not only has an obviously greater DoF, but is also marginally and noticeably sharper at the plane of focus.

The full frame equivalent of the current Canon 7D would be 47mp. I would have no hesitation in buying a future Canon Full Frame DSLR with the pixel density of the 7D, if the price was right   .

The bottom line is, the higher pixel-density camera will never produce worse results at the plane of focus (using the same F stop and lens), but will frequently produce greater DoF with the same degree of sharpness at the plane of focus, when stopped down.

I should add, it is assumed one always compares equal size images on monitor or print. That's only sensible.
« Last Edit: October 22, 2009, 06:14:13 PM by Ray » Logged
coles
Newbie
*
Offline Offline

Posts: 16


« Reply #39 on: October 23, 2009, 05:17:02 PM »
ReplyReply

One of the primary benefits for a larger sensor size (or film size) is that an image degradation due to camera or subject motion is less detectable because the amount of enlargement for any given size of print is less (using a tripod does NOT guarantee 0 camera motion).
Logged
Pages: « 1 [2] 3 4 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad