Want lower noise? Bump the ISO!

It sounds like nonsense, but it is not.
What I am talking about is action shots in low-light conditions, where the subject must be in focus and low shutter speeds are just not an option. Wildlife photography is a good example.

We all know that the higher the ISO, the more noisy the picture. This is why bumping the ISO for no good reason is a mistake, as it will produce an image that could have been much cleaner if captured at lower ISO.

That being said, sometimes there are very good reasons to bump the ISO well above our comfort zone
Some photographers freak out at the idea of shooting at ISO 400. For many, shooting at ISO 800 strikes fear of noise into their hearts. Certainly, most photographers (though not all) avoid ISO 1600 or higher unless they are shooting the Milky Way.

As for me, in low-light conditions, I am very happy to push the ISO to whatever value it takes to avoid underexposing the shadows.

Look at the picture below:

Sunset on mid-Atlantic crossroad
1/1250s, F/4.5, ISO 1000

Fin whale heading offshore, while a Cory's shearwater flies back to its nest on Pico Island
 

I shot this scene in shutter speed priority (1/1250) with my Sony A77 Mark 2, an APS-C camera that features a translucent mirror (its full-frame sister is the Sony A99 Mark 2). The mirror is fixed and reflects some light up to a phase-detection autofocus sensor, thus reducing the amount the light reaching the camera sensor by about 1/2 stop.

To make things apparently more challenging, my lens was a Tamron 70-300mm F4.5-5.6 Di VC USD.
With a maximum variable aperture of 4.5-5.6, we can all agree it is not a fast lens.

(as an aside, if you fear that relatively inexpensive gear will never take you very far in photography, you may find it encouraging that I used exactly this kit to photograph my Portuguese Man o' War)

Now the key part: I decided to let my camera choose the ISO in the range 100-1600 and that resulted in a value of 1000. Do you find my image grainy? Do you see any significant loss of details due to noise reduction applied in post? Honestly, I don't.

Here is the trick: I knew that, in such light conditions and with the settings indicated above, my camera was going to expose to the right - i.e. it would expose the highlights correctly.
As you probably know (but just in case you did not), the highlights populate the right side of the histogram, hence the expression exposing to the right ;-)

I knew I was exposing to the right because I checked the histogram before the Fin whale surfaced next to the boat. From the histogram, I also knew in advance that no part of the image was going to be underexposed.
If your camera features an electronic viewfinder, like mine does, things are even easier because what you see through the viewfinder is basically what you get.

At lower ISO, the image would have been significantly darker. And you know what? If you drastically open the shadows of a dark picture in post, the amount of noise introduced will often exceed the noise produced by shooting the same scene at a higher ISO value.

Below is an image I took at ISO 1600 in Postojna Cave, Slovenia. There is no action going on here but, since tripods are not allowed in the cave, I handheld my camera and managed to get a pretty sharp image at 1/20s (Light Bless In-Camera Stabilisation!).

Faces everywhere (Postojna Cave, Slovenia)
1/20s, F/3.2, ISO 1600, handheld

 While the trick works in general very well, sometimes conditions are just too extreme and it becomes impossible to produce a usable shot (although how usable a shot is depends on what we want to do with it!). 

But you know what? When that happens to me, at least I know I would not have got a better exposure shooting at lower ISO.

If you enjoyed reading, please pick a time to like and share.

Cheers,
Enrico

Why I do NOT need a full frame camera

Some people raise their eyebrows when I tell them I am an APS-C shooter, which means the crop sensor inside my camera is significantly smaller than the 35mm equivalent sensor of full-frame cameras.
Truth be told, I never owned a full-frame camera and, at the moment, I am not planning to upgrade to it.
Full-frame cameras and the lenses designed for them are more expensive than their crop sensor equivalent, but it is not just about the money
The main reason for me is that I doubt it would be a real upgrade in the first place!


Editor's pick on GuruShots.com, Prime Collection on 500px

I believe there are only two fields of photography that really benefit from a full-frame sensor:

  1. Portrait photography (or, in general, whatever field of photography where it is desirable to have a subject in focus and everything else nicely blurred).
  2. Low light (high ISO) photography.

In the above scenarios, the best results are achieved by using fast lenses and shooting them wide open (i.e. at low F-stops, such as 1.2, 1.4, 1.8, etc.).
Now the point: full-frame cameras can take full advantage of the widest aperture indicated on the lens, whereas APS-C cameras cannot.

This is because on crop sensor cameras, you must multiply both the focal length and the aperture indicated on the lens by the crop factor to obtain the values that correctly describe the behaviour of your lens

On APS-C cameras the crop factor is 1.5, the only exception being Canon (1.6).

Example: a 50mm F1.8 lens only behaves as such on a full-frame camera. On an APS-C camera with a 1.5x crop factor such as mine, the same lens behaves like a 75mm F2.7. Definitely a much narrower field of view and over a stop darker in terms of aperture!

FOCAL LENGTH:   50mm x 1.5 = 75mm 
 APERTURE:   1.8 x 1.5 = 2.7

 


Most portrait photographers use full-frame cameras to take full advantage of their fast lenses (maximum aperture such as 1.2, 1.4 or 1.8), thus being able to produce a very pleasant, defocused background known as bokeh.

Fast lenses at low F-stops are also helpful if you shoot in low light conditions, so that the sensor will gather the maximum amount of light allowed by your lens. Once again, full-frame cameras perform better than APS-C cameras in these conditions.
Of course, you could still bump the ISO to compensate for the smaller aperture on APS-C, thus making the image as bright but... the higher the ISO, the more grainy the image

By now you probably see why I am not upgrading to full-frame. Since I am a landscape and wildlife photographer, the limitations above do not really impact my work.

As a landscape photographer, I very often want to have everything in focus back to front and such an effect is more easily achieved on APS-C.
As a wildlife photographer, my main struggle is getting close enough to my subjects to reasonably fill the frame. Since a 300mm zoom becomes a 450mm on my APS-C camera (300mm x 1.5), this is an advantage for me rather than a limitation!

I also love shooting interiors and exteriors and, to have everything in focus, I almost always set my aperture to either 7.1 or 8. This means I do not need to buy an expensive fast lens, because I would rarely (if ever) shoot it wide open anyway!

I want to conclude this post with a sort of disclaimer, just in case one day you see me with a full-frame camera in my hands.
My Sony Alpha 77 Mark 2 has a 24MP, APS-C sensor. That is already quite a lot of megapixels for a crop sensor. On the other hand, I would love to have a 36MP or even a 42MP sensor. 
If one day I ever feel the need to have bigger files for my prints, or to be able to crop more into my images while retaining a reasonably high resolution, that day I will definitely upgrade to full frame.

I think a 42MP crop sensor camera would pay a significant toll in terms of image quality in low light conditions.
Why? Consider a full-frame sensor (a) and an APS-C sensor (b) with exactly the same resolution (e.g. 42MP).
The amount of light hitting the sensor per unit time and unit area will be the same for (a) and (b). But since the sensors have the same resolution and (a) is bigger than (b), the amount of light gathered by each pixel will be less on the APS-C sensor (b), because there will be more pixels per unit area sharing the light hitting that area.
In other words, since the sensors have the same resolution (the same number of pixels), the pixels will need to be more densely packed on the crop sensor (b) than on the full frame sensor (a) because the crop sensor is smaller.

I hope it makes sense. A surprisingly high number of photographers ignore these facts ;-)

Cheers,
Enrico