1.3.3 Three Different Low-Light Modes

These three features are designed for when you’re shooting in low light and don’t have a tripod handy. (And you just hate noisy pictures at high ISO.) For years, professional photographers (astrophotographers in particular) had a trick up their sleeves when it came to reducing noise in static images – they would take several different shots in succession, and then merge them all in Photoshop. The underlying principle was that each frame had the same subject but completely random noise, and by combining the images the noise would just get “averaged” away, while the subject, which appeared consistently in each shot, would be reinforced.

Using the same intelligence found in the panorama stitching algorithms, your camera can use this very same technique. Using either Handheld Twilight mode or the Multi-Frame Noise-Reduction function, the camera will take several handheld shots in rapid succession, line them all up (in case your hand wasn’t perfectly steady), merge them all together, and produce one high-resolution, low-noise, low-light image – all in-camera!

These all can get kind of confusing since these three features are all advertised to do the same thing. The features are:

Function	How it's Invoked	What it Does
Multi-Frame Noise Reduction (MFNR)	Fn. --> ISO --> top-most option (Section 6.21.2)	Averages four or 12 sequential shots together. Gives you complete control over white balance, exposure compensation, creative styles, picture effects, ISO, etc. You can’t turn it on while in RAW or RAW+JPG mode.
Anti-Motion Blur (AMB)	Scene mode (Section 6.38.2)	Identical to MFNR above, except 1) you have no control over image settings like ISO, white balance, etc. 2) the camera analyzes the scene and doesn’t merge parts of images that contain blurs. As a result, areas around things that move may appear noisier than surrounding areas. It also shoots at a faster shutter speed - if the camera decides that the shutter speed should normally be 1/10th of a second, the HHT function will take four images of 1/40th of a second and then add them all together, whereas HandHeld Twilight will take four pictures at 1/10th second each.
HandHeld Twilight (HHT)	Scene mode (Section 6.38.7)	This is like AMB above except there's no intelligent removal of moving artifacts, and it uses the slower shutter speed. Still no control over image parameters.

 

 

Figure 1-7: Handheld twilight allows you to take low-noise, low-light images without a tripod. This was taken from the inside of Los Angeles International Airport’s iconic “Theme Building”.

These are very innovative features and you can see an example of Handheld Twilight in Figure 1-7.

 

TIP 1: To invoke the Scene (SCN) mode, set the Exposure mode dial to SCN; rotate rear wheel until the setting you want is selected. If you want to see what you're selecting using words instead of icons, though, use the MENU --> 7 --> Scene Selection --> [Choose one].

TIP 2: There’s a fourth feature called “Night Scene” (Section 6.38.6) which is supposed to be for nighttime shots. But as far as I can see doesn’t do anything different than AUTO mode with the flash disabled. You can safely disregard this feature.

TIP 3: Handheld Twilight (HHT) was Sony's first attempt to do multi-frame image merging. Since then they came out with AMB and MFNR, both of which do a better job. So in my mind there’s no reason to use HHT. (I know, too many acronyms…)

 

Three Low-Light Modes Compared

This seems like as good a place as any to compare the multi-frame low-light modes (MFNR, HHT, AMB). (Figure 1-9). (Mind you, none of these fancy modes are better than using a tripod at low ISO.)

I have to tell you that I’ve done this test many times on books with different cameras, and in the past the visible differences between these modes have been striking and apparent. In more recent cameras the differences are so subtle that I really have to pixel peep in order to see them. (And I believe that if you have to pixel peep that much to see a difference, then the differences are not at all meaningful.)

Figure 1-8: A random equipment drawer. Figure 1-9 shows comparison close-ups of the yellow square.

Bottom Line: While nothing beats a tripod, of all the low-light modes offered, Multi-Frame Noise Reduction really does the best job of reducing noise without degrading the detail (more info in Section 6.21.2). Next runner-up is good old Program mode :-) , followed by Anti-Motion blur, which works well with moving objects (and is described more in Section 6.38.2). Handheld Twilight (which I feel is now obsolete and doesn’t really offer any advantages) is discussed further starting in Section 6.38.7.

Figure 1-9: Pixel-peeping the yellow rectangle in the previous figure, comparing the three low-light exposure modes. The best image to my eye is Multi-Frame Noise Reduction, the only low-light mode which allows you some control over your shooting variables.

Another test showing the real advantage of Anti-Motion Blur can be seen in Figure 1-10. Pretty significant.

Figure 1-10: The ultimate challenge through an airplane window: A low-light shot with a moving subject. While Multi-Frame Noise Reduction will do a good job reducing the noise (left), Anti-Motion Blur pulls off a miracle by severely reducing the blur (right).

1.3.4 Handheld High Dynamic Range (HDR)

In Appendix A, I talk about the limited dynamic range of the digital sensor, and how our eyes can see a significantly greater range of light (bright to dark) than what the camera can see. There have been lots of attempts to correct this intentional artifact of photographic representation of real light, the most popular of which is known as High Dynamic Range (HDR) photography.

The time-honored way to create an HDR image is to put the camera on a tripod and take 3 (sometimes more) pictures of the same scene, each at different exposures – some darker, some lighter. Then, you merge them all in your computer so it sort of looks like the way you saw it in real life. An example of HDR photography appears in Figure 1-11.

5 years ago, HDR photography was labor-intensive and unintuitive. But Sony's HDR feature makes it simple – without needing a tripod, the camera will take 3 pictures of different exposures, line them up (in case your hand was moving), and merge them so the brightest and darkest parts of all 3 come through.

Figure 1-11: In-camera High Dynamic Range can turn difficult light with blown-out highlights and too-dark shadows (left) into something a little closer to the way you remember seeing it (right).

Figure 1-12: Peaking Level and Peaking Color work together to make manual focusing easy again. Things that are sharp (= high contrast) are highlighted in the color of your choice.

I should warn you, though, that this feature only covers the first step in traditional HDR photography, and skips entirely a 2^nd important step called “Tone Mapping”. Therefore, you might look at your HDR images and conclude they appear kind of flat. This is normal, and in fact it’s the reason your sensor has reduced dynamic range to begin with. Do NOT expect this feature to produce these kinds of uber-extreme over-the-top HDR effects like those at http://tinyurl.com/7sjnd5n. You’ll be disappointed. This feature is discussed more in Section 14.2.

1.3.5 Peaking Color

This camera gives you TWO great manual focusing aids (compared to zero with most DSLRs). The first is the Focus Magnifier, which shows you a magnified area of the image so you can fine-tune your manual focusing. The other is a very useful feature called “Peaking Level” (and its counterpart, “Peaking Color”). As you turn the manual focusing ring around your lens, areas that have high contrast (which equates to sharp focus) will be highlighted in the color of your choice. Faster than using the ground-glass focusing screens! (Figure 1-12.)

This feature is insanely useful even if you don't focus manually. When combined with the DMF (Direct Manual Focus) feature, the peaking color can show you with greater clarity what the camera has focused on, so you can shoot with autofocus with greater confidence. I talk about that configuration at the end of Section 2.2, and about the Peaking Level in greater detail in Section 7.10.

TIP: Peaking Color is now available when shooting movies! See Section 7.10.

1.3.6 Face Detection and Improved Eye AF

Figure 1-13: Face recognition finds the subject quickly even though other objects may be closer. Much faster than switching to spot focusing and then using the focus-recompose-shoot method.

If you’re used to using a DSLR, and you shoot people a lot, then you’re really going to like these two features.

Your camera employs face detection to determine the subject far faster than was previously possible. Have a look at the example in Figure 1-13.

Keep in mind that I was always a “focus-lock-recompose-shoot” kind of guy, which is just completely incompatible with shooting kids. Using Face Detection changed all that -- it will find the face, focus on it, bias the exposure for the face, and then it’s ready – all faster than my experienced skills could do. It’s also much more accurate than the pure “wide-area AF” method, since in the old days there was no intelligent scene analysis going on – the camera would usually focus on whatever was closest.

You can also register a face so it can have focusing priority when it’s amongst a group of other faces. Is it useful? Find out more about my experiences with this feature in Section 7.29.

The A6300 also has an improved Eye AF feature. Once it’s been assigned to a button (I've assigned it to the "AEL" button since it's easy to hold down while you're shooting), pressing that button focuses on the closest eye. (The eye is the first thing YOUR eye goes to when looking at a portrait. It’s human nature. Furthermore, the eye is the first thing you pixel peep on to ascertain the picture’s sharpness. It’s a good feature and works well when you have a native E-mount lens attached, and the feature now works in Continuous AF (AF-C) mode too.

Much has been made of the new “Eye AF” feature online, but its implementation has me scratching my head. The camera has face detection, doesn’t it? If the camera can recognize a face, can't it then zero in on the eye on its own? Why is this even a feature that needs a button press?

Worse, if you use it often your brain now has one additional mental hoop to jump through when picking up your camera: You now have to think to yourself, “Am I shooting a person whose face is close enough for Eye AF to work?” If so, you can use Eye AF function to focus. If not, you'll have to press the shutter release button halfway. If you used Eye AF when there's no Eye to find, the camera will not find anything to focus on and (depending upon configuration) will refuse to take the shot. Yes, it’s a very small thing to complain about in such an otherwise awesome camera, but as a person who used to shoot in high pressure situations (like weddings), anything that slows you down is a detriment. In my opinion this should have been implemented automatically and invisibly.

Anyway, Eye AF can now track an eye in AF-C focusing mode after focus has been acquired should the eye move across the screen.

TIP: Eye-AF works with native E-mount lenses only. (Not even A-mount lenses with an LA-EA1 or LA-EA3 adapters will work.)