D810 handheld shooting? Why is it even an issue?

There's lots of confusion on this thread.

I don't think there's a question that in many circumstances shooting on a tripod would give sharper images than otherwise. However, the same is true of a D750, 5D Mk III etc. All photography handheld is prone to higher low level blurring caused by motion of the photographer. The question is whether the D810 is especially sensitive to being used handheld?

I would say no, it is not, it's perfectly possible to use the D810 hand held and to get acceptably sharp results. It's a lot less challenging to use the D810 handheld compared to the D800E I used to own. That doesn't mean using a tripod is irrelevant, it just means you have no more need to use one than when shooting a D750 or a 5D MK III.

The real mischief around all of this isn't a high MP sensor, it's combining a high MP sensor with more vibration than is necessary through the mirror and/or shutter. My experience of using not only a D800E and D810, but also a Sony A7R which shares the same sensor, would all go to back that up. I can guarantee you that when Canon does release their 50mp camera, they will have been all over this aspect of the design.
Hi,

as has been mentioned by others , to ensure that I get true value for my money I also shoot whenever possible on a tripod or other support when using the D810 even with mid-sized lenses such as the 80-400mm which is no where as heavy as the large Primes. At first when shooting hand held with the D810 & D800 I was noticing images were not pin sharp as expected and realised that I had to shoot with the tripod which I started . Images are now sharp with the details clearly visible such as in the one shot taken below with the original image included. Notice how the spikes on the insect's legs are razor sharp - this would definitely not have been possible without a tripod. I got this blown up to a 84cm x 70cm and block-mounted all thanks to using a tripod.

Make use of a tripod with the D810 when possible.

Thanks,

Adrian
 
There's lots of confusion on this thread.

I don't think there's a question that in many circumstances shooting on a tripod would give sharper images than otherwise. However, the same is true of a D750, 5D Mk III etc. All photography handheld is prone to higher low level blurring caused by motion of the photographer. The question is whether the D810 is especially sensitive to being used handheld?

I would say no, it is not, it's perfectly possible to use the D810 hand held and to get acceptably sharp results. It's a lot less challenging to use the D810 handheld compared to the D800E I used to own. That doesn't mean using a tripod is irrelevant, it just means you have no more need to use one than when shooting a D750 or a 5D MK III.

The real mischief around all of this isn't a high MP sensor, it's combining a high MP sensor with more vibration than is necessary through the mirror and/or shutter. My experience of using not only a D800E and D810, but also a Sony A7R which shares the same sensor, would all go to back that up. I can guarantee you that when Canon does release their 50mp camera, they will have been all over this aspect of the design.
Hi,

as has been mentioned by others , to ensure that I get true value for my money I also shoot whenever possible on a tripod or other support when using the D810 even with mid-sized lenses such as the 80-400mm which is no where as heavy as the large Primes. At first when shooting hand held with the D810 & D800 I was noticing images were not pin sharp as expected and realised that I had to shoot with the tripod which I started . Images are now sharp with the details clearly visible such as in the one shot taken below with the original image included. Notice how the spikes on the insect's legs are razor sharp - this would definitely not have been possible without a tripod. I got this blown up to a 84cm x 70cm and block-mounted all thanks to using a tripod.

Make use of a tripod with the D810 when possible.

Thanks,

Adrian
 
T O Shooter stated:

Used in a nice situation as well. I picked up a decent CF monopod a year or so ago, but wasn't overly crazy about being tied to that either. Put it back in the vehicle on that day, and pulled out the tripod and gimbel. Haven't used it since. Maybe I should try it again. I really need to get a ballhead dedicated to the 'pod I guess.


-- hide signature --

A Canon G5 and a bit of Nikon gear.
The trouble with normal is it always gets worse - Bruce Cockburn

I also have tried monopods with my Nikon 500mm lens for birds and quickly put it back into my closet. I shoot birds and bif almost always with a tripod and gimbal head this works well for me.

To me, nothing wrong with using a tripod for landscapes and is a necessity for close up stuff such as macro photo work or anything I can take my time with. Of course needed for moving water.

Larry
 
Or....is it? I've read several articles talking about how you have to he super careful when hand-holding shots with the D810. Talking about how it's tough to get tack sharp images unless on a tripod. Why is that? I've had one for two weeks but haven't actually gotten out with it to shoot enough to understand these types of comments. Thx.
158042314.jpg


158042310.jpg


158042312.jpg


All hand held. All will print 20x30 no problem.

Ask yourself this, could you handhold a film medium format camera? Or a 4x5 film camera? Especially with a low noise 6400 speed film?

Answer: Press guys were handholding 4x5's with 100 speed film in the 50's and they're laughing now.

It's not about the megapickles, it's about the technique and photographers knowledge about phtography. There's a lot of people out there that can't handhold a point and shoot.

The rules still apply. In daylight at F5.6 and 1/125 second with a 50mm lens, you don't need a tripod. It won't make any difference.

--
Edward
www.youtube.com/photouniverse
www.edwardthomasart.com
www.pbase.com/edwardthomas
 
Last edited:
It can also be proven experimentally that even when the camera is mounted on a sturdy tripod, mirror and shutter actuation can cause vibration that moves the image on the sensor and therefore blurs the image at pixel level.
I'll say. Even using MUP, the first curtain can blur images. And the D810's EFCS makes a difference here:



Jim
 
It can also be proven experimentally that even when the camera is mounted on a sturdy tripod, mirror and shutter actuation can cause vibration that moves the image on the sensor and therefore blurs the image at pixel level.
I'll say. Even using MUP, the first curtain can blur images.
Absolutely.
But there is no law of nature that says that a high Mp camera will have more mirror and/or first curtain slap than a low Mp camera.
Ceteris paribus the high Mp camera will never have worse motion blur at the same output size.

It may be more visible at some point but that is a different matter, more to do with less resolution of the low Mp camera.
 
Last edited:
It can also be proven experimentally that even when the camera is mounted on a sturdy tripod, mirror and shutter actuation can cause vibration that moves the image on the sensor and therefore blurs the image at pixel level.
I'll say. Even using MUP, the first curtain can blur images.
Absolutely.
But there is no law of nature that says that a high Mp camera will have more mirror and/or first curtain slap than a low Mp camera.
Ceteris paribus the high Mp camera will never have worse motion blur at the same output size.

It may be more visible at some point but that is a different matter, more to do with less resolution of the low Mp camera.
If we hold sensor size constant, I don't disagree with anything you just said. Was there something in my post that made you think I did?

If we take into account the fact that larger sensors tend to have more MP, I have a H2D-39 that has monster mirror slap.

Jim
 
It can also be proven experimentally that even when the camera is mounted on a sturdy tripod, mirror and shutter actuation can cause vibration that moves the image on the sensor and therefore blurs the image at pixel level.
I'll say. Even using MUP, the first curtain can blur images.
Absolutely.
But there is no law of nature that says that a high Mp camera will have more mirror and/or first curtain slap than a low Mp camera.
Ceteris paribus the high Mp camera will never have worse motion blur at the same output size.

It may be more visible at some point but that is a different matter, more to do with less resolution of the low Mp camera.
If we hold sensor size constant, I don't disagree with anything you just said. Was there something in my post that made you think I did?
No, it was more directed at those who do (and there are a few in this thread).
My bad.
 
No, it was more directed at those who do (and there are a few in this thread).
My bad.
No problem; I just wanted to know if there was something I should clear up.

Jim
 
An interesting thread to read as I currently using a d700 and was thinking of upgrading in May. It is such a difficult choice deciding which model is the best for me. The d810 seems a great camera but I had been put off reading about how ultra sensitive it is and 90% of my shots are handheld out walking (yes I know I should use a tripod more) Maybe worth reconsidering this camera as an option.
--
http://www.flickr.com/photos/sarahb86/
Don't believe it. I hate tripods. I have them, I carry them with me, I don't use them. The is no ultrasensitivity in a D810. A 90 year old grandmother could shoot one. I shoot 99.5% handheld with D8x0 bodies since April 1, 2012 and except for some screwy AF issues on BIF, as long as I maintained a suitable shutter speed I had no issue. The D810 is noticeable better than the other two.

As Sgoldswo said in this thread regarding the 800/e and needing to use 1/2xFL " I and others put that down to excessive vibration caused by the mirror and shutter." The 810 is not like that.
 
Zero, if your final output doesn't make it visible.

Just as there is no advantage to having a 360hp engine in your car, instead of a 120hp engine, when you only have the throttle at 20%.
But if you downsize the 36 Mp image to 12 Mp you will have colour accuracy and dynamic range way beyond what the 12 Mp camera can do.
Which of course has everything to do with differences in technology of the subject D700 and D810, and nothing to do with their resolution - which was supposed to be the topic here.

If you still insist on comparing color accuracy and DR, how about comparing the D810 to the Sony A7s?
(And at 20% throttle the 360 hp car will most likely need much less fuel than the 120 hp car at full throttle...)
Unlikely that the 360 hp car would produce 120 hp at 20% throttle, but in any case, relative fuel efficiency at a specified power output, is driven primarily by engine displacement - given similar technology. Since no displacements were specified, how did you come to your conclusion?
 
Zero, if your final output doesn't make it visible.

Just as there is no advantage to having a 360hp engine in your car, instead of a 120hp engine, when you only have the throttle at 20%.
But if you downsize the 36 Mp image to 12 Mp you will have colour accuracy and dynamic range way beyond what the 12 Mp camera can do.
Which of course has everything to do with differences in technology of the subject D700 and D810, and nothing to do with their resolution - which was supposed to be the topic here.
It has everything to do with their resolution.
Higher sampling rate give higher accuracy.
 
Which of course has everything to do with differences in technology of the subject D700 and D810, and nothing to do with their resolution - which was supposed to be the topic here.
It has everything to do with their resolution.
Higher sampling rate give higher accuracy.
The sampling rate is involved in color accuracy, but is by no means the long pole in the tent.

The first consideration for color accuracy is the spectral response of the filters in the color filter array (CFA). If they were a 3x3 matrix multiplication away from the response of a standardized human trichromat (beings that see with three classes of color receptor, unlike 8% of the men, who see with fewer, a small percent of women who have similar deficiencies, and some women who see with four sets -- tetrachromats), then they'd be as accurate as we need them to be. But they're not, in any commercially available camera. They're not even close. But we can come up with compromise matrices that sorta kinda get close, or we invent nonlinear mapping techniques to do better. Those things are in raw developers. But the camera sees some colors that you and I see as different as the same color. There's no way a profile or a compromise matrix can sort that out. The camera also sees some spectrally-different colors, that you and see as the same, as different colors. That's not impossible to sort out, merely very difficult. The name for these kind of inaccuracies is capture metameric error.

Then there's noise. Photon noise that's due to the nature of light. Non-uniform response of the pixels to that light, although that's becoming less important with modern cameras. Noise that occurs whether the camera is seeing any real light or not, which can have components that vary from frame to frame whose spectrum is usually close to white, and fixed (pattern) noise, usually with strong low spatial frequency components. All that noise affects color accuracy, and the errors are worst for the dimmest colors. There are also errors introduced by the camera as it turns the charge on the capacitors in the pixels into numbers. They are called quantization errors, or quantizing noise. There are various ways that relate the brightest colors the camera can distinguish (full scale) to the dimmest ones that meet some quality metric, and what they measure is called dynamic range (usually photographic dynamic range or engineering dynamic range).

So we've got a lot of errors already and we haven't yet talked about resolution. Resolution figures into some of the errors above, but, when images are converted in post to the same resolution, only in a weak way.

Where resolution and color accuracy combine is for subjects with high spatial frequency. Higher resolution sensors will in general have fewer errors for these subjects. But the eye's sensitivity to the part of color that's independent of brightness -- chromaticity -- rolls off at a lower spatial frequency than the eye's sensitivity to brightness -- luminance, so, ignoring obnoxious things like false-color aliasing errors, even that is usually not very important.

And, by the way, I'd be careful arguing with Marianne. She is extremely knowledgeable.

Jim

--
http://blog.kasson.com
 
Last edited:
How soon we forget. When the D800 first came out, there were all kinds of problems with it attaining proper focus and one of the suggestions was that since it had such a high MP count, it really should be used on a tripod in order to get the best sharpness out of it. This was basically an excuse to sort of escape the reality that the camera just did not focus very well.
I've have an early D800 and had not had any trouble getting it to focus with both old and new lenses. It focuses better than the D50, D90, and D7000 that it replaced.
 
Which of course has everything to do with differences in technology of the subject D700 and D810, and nothing to do with their resolution - which was supposed to be the topic here.
It has everything to do with their resolution.
Higher sampling rate give higher accuracy.
The term "color accuracy" refers to the accuracy of hues, measured by their RGB or LAB components, for example, over a macroscopic patch or image area where those values are constant, such as within a single tile in a color chart. In making the measurements, borders, edges and transitions are avoided.

It is not related to how well the camera records color transitions or fine color detail, thus has nothing to do with sampling period. If your intent is to refer to color detail, then you are using the wrong terminology.
 
Which of course has everything to do with differences in technology of the subject D700 and D810, and nothing to do with their resolution - which was supposed to be the topic here.
It has everything to do with their resolution.
Higher sampling rate give higher accuracy.
The sampling rate is involved in color accuracy, but is by no means the long pole in the tent.

The first consideration for color accuracy is the spectral response of the filters in the color filter array (CFA). If they were a 3x3 matrix multiplication away from the response of a standardized human trichromat (beings that see with three classes of color receptor, unlike 8% of the men, who see with fewer, a small percent of women who have similar deficiencies, and some women who see with four sets -- tetrachromats), then they'd be as accurate as we need them to be. But they're not, in any commercially available camera. They're not even close. But we can come up with compromise matrices that sorta kinda get close, or we invent nonlinear mapping techniques to do better. Those things are in raw developers. But the camera sees some colors that you and I see as different as the same color. There's no way a profile or a compromise matrix can sort that out. The camera also sees some spectrally-different colors, that you and see as the same, as different colors. That's not impossible to sort out, merely very difficult. The name for these kind of inaccuracies is capture metameric error.

Then there's noise. Photon noise that's due to the nature of light. Non-uniform response of the pixels to that light, although that's becoming less important with modern cameras. Noise that occurs whether the camera is seeing any real light or not, which can have components that vary from frame to frame whose spectrum is usually close to white, and fixed (pattern) noise, usually with strong low spatial frequency components. All that noise affects color accuracy, and the errors are worst for the dimmest colors. There are also errors introduced by the camera as it turns the charge on the capacitors in the pixels into numbers. They are called quantization errors, or quantizing noise. There are various ways that relate the brightest colors the camera can distinguish (full scale) to the dimmest ones that meet some quality metric, and what they measure is called dynamic range (usually photographic dynamic range or engineering dynamic range).

So we've got a lot of errors already and we haven't yet talked about resolution. Resolution figures into some of the errors above, but, when images are converted in post to the same resolution, only in a weak way.

Where resolution and color accuracy combine is for subjects with high spatial frequency. Higher resolution sensors will in general have fewer errors for these subjects. But the eye's sensitivity to the part of color that's independent of brightness -- chromaticity -- rolls off at a lower spatial frequency than the eye's sensitivity to brightness -- luminance, so, ignoring obnoxious things like false-color aliasing errors, even that is usually not very important.
So in essence you agree, the high Mp camera gives higher accuracy.
And, by the way, I'd be careful arguing with Marianne. She is extremely knowledgeable.
Those are the most important people to argue with!

M says that downsampled to the lower MP output there is no advantage for the high Mp camera.
I argue that it is.
 
Or....is it? I've read several articles talking about how you have to he super careful when hand-holding shots with the D810. Talking about how it's tough to get tack sharp images unless on a tripod. Why is that? I've had one for two weeks but haven't actually gotten out with it to shoot enough to understand these types of comments. Thx.
Check this, applies to the D810 as well.

 
Or....is it? I've read several articles talking about how you have to he super careful when hand-holding shots with the D810. Talking about how it's tough to get tack sharp images unless on a tripod. Why is that? I've had one for two weeks but haven't actually gotten out with it to shoot enough to understand these types of comments. Thx.
It's not. A higher MP camera will NEVER be worse for hand holding than a lower MP one. Having said that, if you buy a high MP camera and want to take full advantage of all the MP that you paid for (for example, to make 20x30 inch prints) then you have to exercise extra care.
 
I was out shooting today and came across only the second snowy owl I'd seen in Ontario. He had other thoughts on getting photographed and preferred to fly to a barn a 1/2 km away. However I saw this Bell box and thought about this thread.

It was cold at -4 degrees Celsius and a light wind, a little light snow, and I had jumped out with only a fall jacket on, but here's a 1/FL taken with the D810 and 300 2.8VR and TC1.7II so 500mm FL. Certainly one of the harder handholdable lenses to try 1/FL with but here it is anyway. One of one shot taken.

09ae2bd0ccfb4702b554c0a177db69ab.jpg

d159dcd9b18442acb22d332bb343d18c.jpg

Sign and lettering is weathered - shot at ISO 1800

--
A Canon G5 and a bit of Nikon gear.
The trouble with normal is it always gets worse - Bruce Cockburn
 
Last edited:

Keyboard shortcuts

Back
Top