Thom on D800 lens choice

Started May 14, 2012 | Discussions
Ilkka Nissilä Veteran Member • Posts: 4,107
Re: Thom on D800 lens choice

I thought we were talking about this all the time. A 2x crop from 200/2 corresponds to 400/4 on full frame. Sounds familiar?

Your claim of correspondence assumes a few things.

1. That there is a 200mm f/2 lens which, at f/2, has the MTF (200/2, f/2,2f) = MTF(400/4, f/4, f), for all f up to Nyquist. Of course for any half-decent 400mm lens, there is no corresponding 200/2 that could even remotely approach that kind of performance. To compensate for this one can attempt sharpening which of course brings up the noise in the cropped D800 shot.

2. That there is an ISO setting available on the camera that is used to do the cropped shot that is two stops lower than the ISO setting used in the camera used for the 400mm shot. In Africa, I would guess a lot of the time you can use base ISO (e.g. on a D3s that is ISO 200) so there is no actual ISO setting available on the D800 to go two stops lower than that. Thus the images from the D800+200/2 cropped to FOV of 400mm lens is going to have lower SNR.

3. That the focusing accuracy of the D800+200/2 is twice as good as that of the (let's say D3S+200-400/4) i.e. errors due to focusing of the former setup, if any, are at least 50% smaller than that of the latter setup, again, if such errors should be present. In any wildlife photography situation especially if the lens is wide open, there will be focusing errors which limit the sharpness in many shots to less than lens MTF would suggest. Since the D800 focusing system isnt' that much more advanced than that of e.g. the D3s, the focusing errors in the 2x cropped 200mm shot will be in practice about twice as large in affected shots.

All of these things are well known to actual experienced users of long lenses which is practically all of them recommend that you get the long focal length glass first, and then the high pixel count body if you have money left.

mattr Veteran Member • Posts: 3,531
Re: Thom on D800 lens choice

noirdesir wrote:

Yes, but a TC is even more useful on lower pixel density cameras such as the D700. Higher pixel density can do exactly the same thing as a TC from an imaging point of view.

It somehow sounds to me as if you want to say that because of the higher pixel density (compared to the D700) using a TC on the D800 is not better than cropping.

There are diminishing returns if you use a TC on cameras with very high pixel density. The increase in subject resolution from the TC will be bigger on the D700 than on the D800.

I think mainly the part where you portray cropping as only affecting MP and not also affecting things like resolution in these MP, and SNR, DR and tonal range when you are not shutter speed limited.

See response below.

mattr Veteran Member • Posts: 3,531
Re: Thom on D800 lens choice

Ilkka Nissilä wrote:

Your claim of correspondence assumes a few things.

No, that's why I said that it will depend on the specific cameras and lenses. I only claim that the total light gathered from a scene at a given shutter speed is a major factor in the resulting image quality.

1. That there is a 200mm f/2 lens which, at f/2, has the MTF (200/2, f/2,2f) = MTF(400/4, f/4, f), for all f up to Nyquist. Of course for any half-decent 400mm lens, there is no corresponding 200/2 that could even remotely approach that kind of performance. To compensate for this one can attempt sharpening which of course brings up the noise in the cropped D800 shot.

My example was the Nikon 200/2 versus the 200-400/4 at 400mm. Looking at the Photozone test results on the D7000 I'm not convinced that the cropped 200/2 on the D800 can not "even remotely approach" the zoom on a D700. But yes, of course the larger pixel sensor is easier on the lens.

2. That there is an ISO setting available on the camera that is used to do the cropped shot that is two stops lower than the ISO setting used in the camera used for the 400mm shot. In Africa, I would guess a lot of the time you can use base ISO (e.g. on a D3s that is ISO 200) so there is no actual ISO setting available on the D800 to go two stops lower than that. Thus the images from the D800+200/2 cropped to FOV of 400mm lens is going to have lower SNR.

In Africa the good animal shots are usually early or late in the day and higher ISO is used very commonly.

3. That the focusing accuracy of the D800+200/2 is twice as good as that of the (let's say D3S+200-400/4) i.e. errors due to focusing of the former setup, if any, are at least 50% smaller than that of the latter setup, again, if such errors should be present. In any wildlife photography situation especially if the lens is wide open, there will be focusing errors which limit the sharpness in many shots to less than lens MTF would suggest. Since the D800 focusing system isnt' that much more advanced than that of e.g. the D3s, the focusing errors in the 2x cropped 200mm shot will be in practice about twice as large in affected shots.

No, both systems have the same DoF and there is no difference in the required focusing accuracy. One could even argue that AF will work better with the faster lens. OTOH, the AF sensor placement will be more accurate with the longer lens.

All of these things are well known to actual experienced users of long lenses which is practically all of them recommend that you get the long focal length glass first, and then the high pixel count body if you have money left.

But a lot of people are using long lenses on high pixel density crop sensors, no?

jonrobertp Forum Pro • Posts: 12,875
Re: Thom on D800 lens choice

Then one may just as well use the 24-120 on the D600 or even the D3200 for the gist of what Thom is saying. No surprise either. Though some wud argue.

 jonrobertp's gear list:jonrobertp's gear list
Canon PowerShot G7 X Canon G3 X Panasonic ZS100
noirdesir Forum Pro • Posts: 13,586
Re: Thom on D800 lens choice

mattr wrote:

noirdesir wrote:

Yes, but a TC is even more useful on lower pixel density cameras such as the D700. Higher pixel density can do exactly the same thing as a TC from an imaging point of view.

It somehow sounds to me as if you want to say that because of the higher pixel density (compared to the D700) using a TC on the D800 is not better than cropping.

There are diminishing returns if you use a TC on cameras with very high pixel density. The increase in subject resolution from the TC will be bigger on the D700 than on the D800.

The 200 mm f/2 wide open clocks in at 3726 LP/PH on a D3x, using a 1.4x crop from it would reduce that to 2661 LP/PH. Using a 1.4x TC instead produces 3218 LP/PH. So, you re-gain roughly 50% of resolution you would have lost in a pure crop operation. I would not call this something close to diminishing returns.
http://www.photozone.de/nikon_ff/499-nikkorafs200f2vrff?start=1

noirdesir Forum Pro • Posts: 13,586
Re: Thom on D800 lens choice

mattr wrote:

Ilkka Nissilä wrote:

Your claim of correspondence assumes a few things.

No, that's why I said that it will depend on the specific cameras and lenses. I only claim that the total light gathered from a scene at a given shutter speed is a major factor in the resulting image quality.

As I said in my previous post, I take issue with you portraying cropping as only resulting in less MP. It also results in less resolution in these MP compared the use of a TC that is then downsampled to your MP target number. And if you are not shutter speed limited it also results in a lesser overal IQ (SNR, DR).

My example was the Nikon 200/2 versus the 200-400/4 at 400mm. Looking at the Photozone test results on the D7000 I'm not convinced that the cropped 200/2 on the D800 can not "even remotely approach" the zoom on a D700.

The question is cropping vs. TC or longer lens for a given camera . Why do keep inserting comparisons with a different camera? Why tie the longer lens (200-400 mm on D700) to the lower resolution camera and the shorter lens the higher resolution camera (200 mm on D800). That of course will disadvantage the longer lens in comparison.

mattr Veteran Member • Posts: 3,531
Re: Thom on D800 lens choice

noirdesir wrote:

I would not call this something close to diminishing returns.

You really want to argue that the additional subject resolution from a TC does not diminish when a camera with higher pixel density is used? (You then probably also say that the benefit from a TC is not greater when a sharper lens is used)

This is getting silly and is unrelated to this discussion anyway.

mattr Veteran Member • Posts: 3,531
Re: Thom on D800 lens choice

noirdesir wrote:

mattr wrote:

Ilkka Nissilä wrote:

Your claim of correspondence assumes a few things.

No, that's why I said that it will depend on the specific cameras and lenses. I only claim that the total light gathered from a scene at a given shutter speed is a major factor in the resulting image quality.

As I said in my previous post, I take issue with you portraying cropping as only resulting in less MP. It also results in less resolution in these MP compared the use of a TC that is then downsampled to your MP target number. And if you are not shutter speed limited it also results in a lesser overal IQ (SNR, DR).

My example was the Nikon 200/2 versus the 200-400/4 at 400mm. Looking at the Photozone test results on the D7000 I'm not convinced that the cropped 200/2 on the D800 can not "even remotely approach" the zoom on a D700.

The question is cropping vs. TC or longer lens for a given camera . Why do keep inserting comparisons with a different camera? Why tie the longer lens (200-400 mm on D700) to the lower resolution camera and the shorter lens the higher resolution camera (200 mm on D800). That of course will disadvantage the longer lens in comparison.

Just re-read the thread if you want to understand what I'm talking about.

noirdesir Forum Pro • Posts: 13,586
Re: Thom on D800 lens choice

mattr wrote:

noirdesir wrote:

I would not call this something close to diminishing returns.

You really want to argue that the additional subject resolution from a TC does not diminish when a camera with higher pixel density is used?

It does diminish but not nearly to the negligible levels that you imply with your formulation of 'diminishing returns' (which implies approaching negligible levels).

noirdesir Forum Pro • Posts: 13,586
Re: Thom on D800 lens choice

mattr wrote:

noirdesir wrote:

mattr wrote:

Ilkka Nissilä wrote:

Your claim of correspondence assumes a few things.

No, that's why I said that it will depend on the specific cameras and lenses. I only claim that the total light gathered from a scene at a given shutter speed is a major factor in the resulting image quality.

As I said in my previous post, I take issue with you portraying cropping as only resulting in less MP. It also results in less resolution in these MP compared the use of a TC that is then downsampled to your MP target number. And if you are not shutter speed limited it also results in a lesser overal IQ (SNR, DR).

My example was the Nikon 200/2 versus the 200-400/4 at 400mm. Looking at the Photozone test results on the D7000 I'm not convinced that the cropped 200/2 on the D800 can not "even remotely approach" the zoom on a D700.

The question is cropping vs. TC or longer lens for a given camera . Why do keep inserting comparisons with a different camera? Why tie the longer lens (200-400 mm on D700) to the lower resolution camera and the shorter lens the higher resolution camera (200 mm on D800). That of course will disadvantage the longer lens in comparison.

Just re-read the thread if you want to understand what I'm talking about.

Yeah, you mean you can crop with the D800 where you could not crop with the D700. That is not in dispute. What is in dispute is your portrayal of cropping as only resulting in less MP.

mattr Veteran Member • Posts: 3,531
Re: Thom on D800 lens choice

noirdesir wrote:

mattr wrote:

noirdesir wrote:

I would not call this something close to diminishing returns.

You really want to argue that the additional subject resolution from a TC does not diminish when a camera with higher pixel density is used?

It does diminish but not nearly to the negligible levels that you imply with your formulation of 'diminishing returns' (which implies approaching negligible levels).

I did not intend this implication. I thought diminish means diminish and not "approaching negligible levels".

mattr Veteran Member • Posts: 3,531
Re: Thom on D800 lens choice

noirdesir wrote:

What is in dispute is your portrayal of cropping as only resulting in less MP.

Where did you find this "portrayal"?

Cropping also reduces the effective sensor size for a start. And as I said repeatedly the specific lenses and cameras will have an effect.

noirdesir Forum Pro • Posts: 13,586
Re: Thom on D800 lens choice

mattr wrote:

noirdesir wrote:

What is in dispute is your portrayal of cropping as only resulting in less MP.

Where did you find this "portrayal"?

Whenever you say: "If only 12MP are required" . This implies that the criteria for cropping or not cropping is in how many MP it results in.

Cropping also reduces the effective sensor size for a start.

That was the sentence I had been waiting for.

mattr Veteran Member • Posts: 3,531
Re: Thom on D800 lens choice

noirdesir wrote:

mattr wrote:

noirdesir wrote:

What is in dispute is your portrayal of cropping as only resulting in less MP.

Where did you find this "portrayal"?

Whenever you say: "If only 12MP are required" . This implies that the criteria for cropping or not cropping is in how many MP it results in.

Cropping also reduces the effective sensor size for a start.

That was the sentence I had been waiting for.

That's why I always said for example that a 200/2 on the D800 can substitute for a 200-350/2-3.5 or a 28/1.8 for a 28-48/1.8-3.1, if only 12MP are required. I did not say 200-350/2 or 28-48/1.8. I thought it was understood.

fad
fad Forum Pro • Posts: 17,165
Re: Thom on D800 lens choice

jonrobertp wrote:

Then one may just as well use the 24-120 on the D600 or even the D3200 for the gist of what Thom is saying. No surprise either. Though some wud argue.

I think what he's saying is that the lens you choose has to do with how you plan to use the camera. He plans to use it to get the highest resolution possible. But there are other ways to use it.

When I use it as a cheap substitute for an S2, I will use different lenses than when I might use it as a smaller FX camera than a pro body. Given that the D3S/D4 will have advantage, I won't know how often that is until I get one. But the D800E and the 50/1.8 D is something I will explore for street shooting with smaller gear.

Someone who does not have other FX bodies may sometimes use it for art, product, fashion, landscape, where MF might have been used. They also might use if for family, travel, street, events, documentary where IQ is secondary to getting the shot. If you understand what you are using it for, Thom's article helps to clarify the tradeoffs. If you have a bunch of bodies, or only shoot a with a tripod, it is easier to be a perfectionist and only use it in MF mode.

It's like large aperture primes. Some only use them wide open. Others think they work better stopped down. The lenses are fine used both ways. I have both the 85L and the 85/1.4G, both exceptional lenses. But perhaps I should get an 85/1.8 when I want lighter gear. They also produce excellent results. It's hard to figure out what is important for a given task, as opposed to merely asking "What's best." Especially today, when Nikon and Canon are producing such excellent equipment at different specs, it takes a lot of thought.

-- hide signature --

Frank

All photos shot in downtown Manhattan unless otherwise noted.
Thanks in advance for the kindness of your comments or critiques.

 fad's gear list:fad's gear list
Nikon D3S Nikon D800 Nikon D4S Nikon D810 Nikon D750 +19 more
Keyboard shortcuts:
FForum MMy threads