D800 and high ISO photos

...if you really want to use a DX lens on an FX camera, it's better to use it with a 1.4x TC, though"
Why? I agree that you can but why is it 'better'?
More resolution.
For instance, look at the Photozone.de test of the 70-200/2.8G
http://www.photozone.de/nikon_ff/457-nikkor_afs_70200_28_ff?start=1

On a D3X it gives 3953 lp/ph, full frame, f/5.6. If we took a 1.4x crop, that would be 2823 lp/ph.
With a 1.4 TC it gives 3161 lp/ph, so the TC wins by 12%.
--
Bob
 
I had not thought about using a 1.4 TC on a Nikon, but that idea raises an interesting question.

I had formerly tried to use a (non-reporting) Kenko 1.4 TC on my Canon 500D and later on my 7D coupled with my Canon 100-400 lens. Because the 1.4 TC causes a 1 f-stop loss I was never able to reliably focus through the viewfinder. I understood why - the effective aperture of the lens was reduced enough to drop below the minimum required so it would occasionally focus, but never reliably. Better in good light, but even then it was so unreliable that I stopped even carrying it around.

Since the Canon 5D3 had a much better AF system I decided to try the same TC/lens combo on it and found, much to my surprise, that the AF system locked on every time. Good light, bad light, it didn't seem to matter.

So the question for Nikon use would be the same. Will the D800 use a 1.4 TC more reliably than its crop (Nikon) cousins? More specifically, since the Canon 100-400 is a 4.5 - 5.6 lens and since the 1.4 TC drops that to 6.3 - 8 (someone correct me if that is wrong) and since the 5D3 will focus using what is essentially a 6.3 - 8.0 lens, will the D800 also focus at an effective aperture of 6.3?

I do not ask because I am that interested in using a TC with a long lens, but rather assume that if the 5D3 can focus using a combo that its crop cousins will not reliably focus at, it is an indication that its AF system is very good. If the D800 would also focus at such effective apertures it is an indication that its AF system is just as sensitive.

Does anyone know?
 
I was in the same boat as you, just the opposite side of it. I'm a Nikon guy who recognizes the amazing power of the 5D3 camera. What shocked me the most was that the 5D3 performs on par with the Nikon D4 with regards to High ISO noise.

When you add in all the concerns about the D800's bugs, I started taking a real good look at that Canon. Plus, the Canon controls are simply superior to Nikon controls.

But as I continued to read the D800 review, I was awestruck by the D800's other comparisons to the 5D3. The dynamic range is 2 stops more than the Canon, and the Nikon's shadow noise is vastly superior to the Canon. Add in the resolution, and the D800 wins back against the high ISO noise and then some. Understand that I'm a Landscape Photographer, so for me low shadow noise and high dynamic range are very important, even moreso than how the high ISO noise looks.

I would say that if people were brand-blind and totally objective, then most people should buy the 5D3. The files are going to be far easier to work with, it's got better controls, better high ISO noise, better FPS, and more. The Nikon files are going to choke just about anyone's computer after a few layers and applying any kind of filters in photoshop, and there's just not much need for that much resolution unless you're printing 48" prints or something. So unless you need it, I'd vote Canon.

Either camera is absolutely amazing. I think though that people are just emotionally vested. Logic would state to keep with what you know when the playing field is fairly even as it is. So unless you're a Landscape Photographer like me, or otherwise need all that resolution; I think that the Canon is a better choice. At any rate, chances are that even if you put yourself on a waiting list right now... July or so.
--

There are 10 types of people in this world, those who understand binary and those who don't.
 
I had not thought about using a 1.4 TC on a Nikon, but that idea raises an interesting question.

I had formerly tried to use a (non-reporting) Kenko 1.4 TC on my Canon 500D and later on my 7D coupled with my Canon 100-400 lens. Because the 1.4 TC causes a 1 f-stop loss I was never able to reliably focus through the viewfinder. I understood why - the effective aperture of the lens was reduced enough to drop below the minimum required so it would occasionally focus, but never reliably. Better in good light, but even then it was so unreliable that I stopped even carrying it around.

Since the Canon 5D3 had a much better AF system I decided to try the same TC/lens combo on it and found, much to my surprise, that the AF system locked on every time. Good light, bad light, it didn't seem to matter.

So the question for Nikon use would be the same. Will the D800 use a 1.4 TC more reliably than its crop (Nikon) cousins? More specifically, since the Canon 100-400 is a 4.5 - 5.6 lens and since the 1.4 TC drops that to 6.3 - 8 (someone correct me if that is wrong) and since the 5D3 will focus using what is essentially a 6.3 - 8.0 lens, will the D800 also focus at an effective aperture of 6.3?
D800 has focus points that go to f/8. 5DIII actually does not, but can cope as ling as the lens doesn't report.
I do not ask because I am that interested in using a TC with a long lens, but rather assume that if the 5D3 can focus using a combo that its crop cousins will not reliably focus at, it is an indication that its AF system is very good. If the D800 would also focus at such effective apertures it is an indication that its AF system is just as sensitive.
I think with the sophistication of these new AF systems, comparisons are invidious. All the tests I've seen raise as many questions about the test methodology as the AF system. The D800 has Nikon's top level AF system, which is improved over their last one, which was always rated one of the best.

--
Bob
 
Understand that I'm a Landscape Photographer, so for me low shadow noise and high dynamic range are very important, even moreso than how the high ISO noise looks.
I said the same thing in a discussion ands was told that no landscapes need good shadow noise and any examples of ones that do just showed my incompetence as a photographer. The truth is that for many the brand is the most important feature, not the performance.
--
Bob
 
Thanks for all of the information. This thread has largely convinced me that the D800 high ISO performance and its AF system is probably just as good as Canon's and that has made me believe that I should consider changing brands.

I am not in a particular hurry since I still have an older Canon crop camera I can use and that will give me time to perhaps rent a D800 (if they are ever available for rent) and see what it will do.

Thank you to everyone who responded. It was refreshing. There were no personal attacks, no accusations of trolling and no nasty language. Just people trying to be helpful. I am grateful.
 
These extra ISO positions only really mean anything to people who use ex-camera JPEGs , if you use a raw workflow, you can use the D800 at the same exposures as would be given by the 5DIII's high ISO settings - they are only the result of in-camera 'pushing' in the 5DIII in any place.
You say that as though you are sure that it is correct. You may well be (I certainly don't know), but why are you so sure? What do you know (about this) that I don't?
You can tell whether ISO settings are the result of analogue gain or digital pushing by the shape of the read noise curve against ISO
D800



5DIII



Other people (Marianne Oelund and Bill Claff) have measured these with more precision, but for everyday use these will do. The result of increasing analog gain is to reduce the read noise measured as here in electrons. As you can see, the D800 doesn't do very much change of analog gain at all (Marianne and Bill's more accurate figures show that it does a little, but not much). The 5DIII stops applying analogue gain above 3200 ISO, thereafter the ISO settings are a digital push. You can do the digital push yourself in processing just as well, if not better, than it is done in the camera. The advantage of leaving the ISO lower and doing the digital push in processing is that you keep more highlight space, which since low light photos often have light sources in shot, can be a big advantage.
In any case I work with RAW and have taken photos with the 5D3 at both 51200 and 102400 and then used several tools for post. I have found that it is possible to reduce the noise enough to make some of the photos usable. The same may very well be possible at 25600 by adjusting the EV values and, if I still had the 5D3, I would try that.
If you're in the mood for trying, then try leaving it set at 3200 and applying the push from there. For that large a push, you need to apply it to a linear converted file, before gamma correction, the apply gamma to the pushed file.
In any case you may well be right. With the 5D3 you can set the camera to auto ISO and expect that the camera will adjust the ISO values up to 25600 to compensate for shutter speed and light. Can you make the same adjustment with the D800 up to 25600? If so, and if what you right is correct, then there essentially should not be any difference in general use between the two cameras at high ISO values. That would give me more flexibility as I decide on what camera (and camera line) to get.
If you're talking auto ISO, the D800 has three important advantages over the 5DIII.
  1. The 5DIII limits the minimum shutter speed to 1/250 second, which was just a brain dead decision by the designers. The D800's limit is 1/4000 sec, much more flexible.
  2. If you set the minimum shutter speed to 'Auto' the D800 lets you choose the FL multiplier it uses to select the shutter speed, unfortunately only in whole stop increments, but for instance it does allow you to choose 2x FL, which is OK if you err on the side of safety. The 5DIII is fixed at 1x.
  3. The D800 allows EC adjustment in M mode with auto ISO, the 5DIII doesn't, so you can't bias the ISO choice to your own preference (not that you have so much need to with the D800).
The D800's options allow an interesting setup for me which is to set the camera to A mode, set the minimum shutter speed to auto 2x (taking into account the preference for little camera shake on the D800) and the max ISO to, say 800. That becomes essentially a set and forget setting for most everyday photography, I just set the f-number for DOF and the camera takes the rest of the decisions pretty much as I would have taken them, which is what I want.
This is all wonderful info, Bob.

However, can you direct me to how to set up for automatic shutter speed selection by the camera for (say) minimum shutter speed of 2x focal length? I can see where you can set up Auto ISO and then actually dial in a minimum shutter speed, but is there really an option for the camera to automatically select the minimum shutter speed based on any lens you put on the camera?
--
Lance B
http://www.pbase.com/lance_b

 
Really dont know how you cant see the blue magenta shift in the 12800 image

Open both and compare the bottom left corner for example
Sorry don't see it. If something is there (which I don't see) it doesn't bother me...
I guess for casual photography, you can just ignore it. But the original post was by someone who really wanted to know. There are the facts for you if you ever need them.
Maybe it's a problem for you personally, but I'm not sure what you see in those high ISO images that is different from high ISO shots taken with any other recent full-frame camera...

There are 2 images taken at ISO 400 and 12800 .







 
These extra ISO positions only really mean anything to people who use ex-camera JPEGs , if you use a raw workflow, you can use the D800 at the same exposures as would be given by the 5DIII's high ISO settings - they are only the result of in-camera 'pushing' in the 5DIII in any place.
You say that as though you are sure that it is correct. You may well be (I certainly don't know), but why are you so sure? What do you know (about this) that I don't?
You can tell whether ISO settings are the result of analogue gain or digital pushing by the shape of the read noise curve against ISO
D800



5DIII



Other people (Marianne Oelund and Bill Claff) have measured these with more precision, but for everyday use these will do. The result of increasing analog gain is to reduce the read noise measured as here in electrons. As you can see, the D800 doesn't do very much change of analog gain at all (Marianne and Bill's more accurate figures show that it does a little, but not much). The 5DIII stops applying analogue gain above 3200 ISO, thereafter the ISO settings are a digital push. You can do the digital push yourself in processing just as well, if not better, than it is done in the camera. The advantage of leaving the ISO lower and doing the digital push in processing is that you keep more highlight space, which since low light photos often have light sources in shot, can be a big advantage.
In any case I work with RAW and have taken photos with the 5D3 at both 51200 and 102400 and then used several tools for post. I have found that it is possible to reduce the noise enough to make some of the photos usable. The same may very well be possible at 25600 by adjusting the EV values and, if I still had the 5D3, I would try that.
If you're in the mood for trying, then try leaving it set at 3200 and applying the push from there. For that large a push, you need to apply it to a linear converted file, before gamma correction, the apply gamma to the pushed file.
In any case you may well be right. With the 5D3 you can set the camera to auto ISO and expect that the camera will adjust the ISO values up to 25600 to compensate for shutter speed and light. Can you make the same adjustment with the D800 up to 25600? If so, and if what you right is correct, then there essentially should not be any difference in general use between the two cameras at high ISO values. That would give me more flexibility as I decide on what camera (and camera line) to get.
If you're talking auto ISO, the D800 has three important advantages over the 5DIII.
  1. The 5DIII limits the minimum shutter speed to 1/250 second, which was just a brain dead decision by the designers. The D800's limit is 1/4000 sec, much more flexible.
  2. If you set the minimum shutter speed to 'Auto' the D800 lets you choose the FL multiplier it uses to select the shutter speed, unfortunately only in whole stop increments, but for instance it does allow you to choose 2x FL, which is OK if you err on the side of safety. The 5DIII is fixed at 1x.
  3. The D800 allows EC adjustment in M mode with auto ISO, the 5DIII doesn't, so you can't bias the ISO choice to your own preference (not that you have so much need to with the D800).
The D800's options allow an interesting setup for me which is to set the camera to A mode, set the minimum shutter speed to auto 2x (taking into account the preference for little camera shake on the D800) and the max ISO to, say 800. That becomes essentially a set and forget setting for most everyday photography, I just set the f-number for DOF and the camera takes the rest of the decisions pretty much as I would have taken them, which is what I want.
This is all wonderful info, Bob.

However, can you direct me to how to set up for automatic shutter speed selection by the camera for (say) minimum shutter speed of 2x focal length? I can see where you can set up Auto ISO and then actually dial in a minimum shutter speed, but is there really an option for the camera to automatically select the minimum shutter speed based on any lens you put on the camera?
Yes, there is. Set the minimum shutter speed to 'Auto' then use the multiselector > key to take you to a submenu which allows you to change the multiplier in 1 stop increments up to +/- 2 stops (5 settings)

--
Bob
 
...if you really want to use a DX lens on an FX camera, it's better to use it with a 1.4x TC, though"
Why? I agree that you can but why is it 'better'?
When you use a 1.4x TC on an FX lens you only use the central portion of the lens's image circle (the reduction in angle of view is the tele-conversion effect).

The DX lens only has that much image circle, so the TC effectively eliminates the vignette of the DX lens.
 
can you direct me to how to set up for automatic shutter speed selection by the camera for (say) minimum shutter speed of 2x focal length? I can see where you can set up Auto ISO and then actually dial in a minimum shutter speed, but is there really an option for the camera to automatically select the minimum shutter speed based on any lens you put on the camera?
I have wondered the same but it seems to me that since the camera knows what FL you are actually at as you zoom a lens (the camera writes it to the EXIF information so it clearly knows) why could it not adjust the shutter speed to the appropriate 1/2*FL or 1/3*FL speed based on only on the lens but also on the current zoom position? That is something that would be truly helpful for me.
 
I had not thought about using a 1.4 TC on a Nikon, but that idea raises an interesting question.

I had formerly tried to use a (non-reporting) Kenko 1.4 TC on my Canon 500D and later on my 7D coupled with my Canon 100-400 lens. Because the 1.4 TC causes a 1 f-stop loss I was never able to reliably focus through the viewfinder. I understood why - the effective aperture of the lens was reduced enough to drop below the minimum required so it would occasionally focus, but never reliably. Better in good light, but even then it was so unreliable that I stopped even carrying it around.

Since the Canon 5D3 had a much better AF system I decided to try the same TC/lens combo on it and found, much to my surprise, that the AF system locked on every time. Good light, bad light, it didn't seem to matter.

So the question for Nikon use would be the same. Will the D800 use a 1.4 TC more reliably than its crop (Nikon) cousins? More specifically, since the Canon 100-400 is a 4.5 - 5.6 lens and since the 1.4 TC drops that to 6.3 - 8 (someone correct me if that is wrong) and since the 5D3 will focus using what is essentially a 6.3 - 8.0 lens, will the D800 also focus at an effective aperture of 6.3?

I do not ask because I am that interested in using a TC with a long lens, but rather assume that if the 5D3 can focus using a combo that its crop cousins will not reliably focus at, it is an indication that its AF system is very good. If the D800 would also focus at such effective apertures it is an indication that its AF system is just as sensitive.

Does anyone know?
The D800 focuses well in low light with a 300mm F4 and 2xTCIII at F8.
Few examples in the Flickr gallery below.
--
Cheers, BB
FlickR site
http://www.flickr.com/photos/brandon_birder/
Flickr D800 gallery
http://www.flickr.com/photos/brandon_birder/sets/72157629726734905/
FlickR Nikon1 V1 gallery
http://www.flickr.com/photos/brandon_birder/sets/72157628774050455/
 
can you direct me to how to set up for automatic shutter speed selection by the camera for (say) minimum shutter speed of 2x focal length? I can see where you can set up Auto ISO and then actually dial in a minimum shutter speed, but is there really an option for the camera to automatically select the minimum shutter speed based on any lens you put on the camera?
I have wondered the same but it seems to me that since the camera knows what FL you are actually at as you zoom a lens (the camera writes it to the EXIF information so it clearly knows) why could it not adjust the shutter speed to the appropriate 1/2*FL or 1/3*FL speed based on only on the lens but also on the current zoom position? That is something that would be truly helpful for me.
That is exactly what it does, so long as it is a CPU lens and know the FL. Sorry I should have made that clear, the shutter speed will change as the lens zooms.
--
Bob
 
Considering that we enjoy the "magic hour" so much; I'd say that they don't know what they are talking about.
--

There are 10 types of people in this world, those who understand binary and those who don't.
 
I also see a magenta color cast in ISO400 image in that area.

I can't believe the level of pixel-peeping on this forum sometimes? This is shadows at 12800...this is hardly the biggest issue with this image?
 
I also see a magenta color cast in ISO400 image in that area.
I think that comes from the CFL lighting, not the camera.
I can't believe the level of pixel-peeping on this forum sometimes? This is shadows at 12800...this is hardly the biggest issue with this image?
The point is that with fairly simple noise reduction, the D800 can handle low light past ISO25600 at 36MP without the colors falling apart. You don't have to be a scientist to appreciate that.
 
The D800's options allow an interesting setup for me which is to set the camera to A mode, set the minimum shutter speed to auto 2x (taking into account the preference for little camera shake on the D800) and the max ISO to, say 800. That becomes essentially a set and forget setting for most everyday photography, I just set the f-number for DOF and the camera takes the rest of the decisions pretty much as I would have taken them, which is what I want.
This is all wonderful info, Bob.

However, can you direct me to how to set up for automatic shutter speed selection by the camera for (say) minimum shutter speed of 2x focal length? I can see where you can set up Auto ISO and then actually dial in a minimum shutter speed, but is there really an option for the camera to automatically select the minimum shutter speed based on any lens you put on the camera?
Yes, there is. Set the minimum shutter speed to 'Auto' then use the multiselector > key to take you to a submenu which allows you to change the multiplier in 1 stop increments up to +/- 2 stops (5 settings)
Thank you very much. Just tried it and it's a brilliant option. :-)
--
Lance B
http://www.pbase.com/lance_b

 

Keyboard shortcuts

Back
Top