Best Superzoom 2014: Image Stabilization

Select a topic --------------------------------- 1. Eight competitors, 2.5 clear winners 2. Searching for 70s Guy 3. Ultimate Image Quality 4. Image Stabilization 5. Shooting the Moon 6. True Focal Lengths and Other Observations 7. Conclusions and Winners --------------------------------- Reader Comments on Review --------------------------------- Digital Camera Comparometer (TM) --------------------------------- <-- Digital Cameras (Home) <-- Digital Camera Reviews Index

Handheld IS Test

As mentioned elsewhere, we didn't intend this shootout to be as comprehensive as our full camera reviews; there just wasn't time or resources available to do that. (Way too many other hot cameras waiting for us to test and write up!) We did want to do at least a quick spot-check of their IS performance, though, most especially at maximum telephoto, where it's the most critical.

All of the images below were shot at ISO 100 or 125, maximum tele, and maximum aperture, with shutter speeds between 1/140 and 1/160 second. As you can see, this resulted in about the right exposure for the sunny day we were shooting in, so should be pretty representative of what typical users will encounter under daylight conditions. (The one exception in terms of ISO is the Samsung WB2200F - we blew the original shot, so had to reshoot it later in the day. It was a bit less bright then, so we had to shoot at ISO 200, to maintain the 1/160 target shutter speed.)

Do you really need IS in daylight?

It may be surprising to some that you'd need IS in broad daylight, but honestly, you have no idea just how long these teles are until you hold and try shooting with one of them. Without IS, it can be challenging to even maintain the framing you want at these focal lengths, because the image is dancing around in the viewfinder so much. Back in the film era, the old rule of thumb for getting reasonably sharp shots was to shoot at a speed higher than one over the focal length. By that standard, you'd need to be shooting these 1,200mm equivalent lenses at 1/1200 second to ensure sharpness! So even in broad daylight, IS is essential.

Statistics - it's a numbers game

We captured a number of shots with each camera, although frankly weren't thinking of statistics when we were doing so, so a couple of the cameras only had four shots taken with them; the most was seven. The crops below represent the best results obtained with each camera, but we do also tell you the number of acceptable shots out of what total number each camera produced. The good/bad ratio is important, because as you move to slower and slower shutter speeds, the percentage of good shots will decrease proportionately: It's not like there's some magical shutter speed, above which all the shots are sharp and below which they're all blurry; it's a continuum. So while it's useful to compare best-case results between cameras, it's also good to have some idea of roughly how often those results can be obtained.

As we somewhat expected from our previous shooting with the cameras, none of them were stinkers when it came to IS, and none stood dramatically above the others. The Fuji S1 and Nikon P600 seemed to do a bit better than the others (particularly given that the Nikon was shooting at a longer focal length than any of the others), but we consider a 50-60% success rate at shutter speeds this far below the old one-over-focal-length standard pretty great.

Your mileage may vary

Of course, the other factor at work here is just how steady or unsteady the person holding the camera is. Our senior lens specialist Rob Murray is very steady, from the experience of having handheld hundreds (and hundreds) of shots as part of his IS testing for SLRgear (there, we calculate the number of stops of improvement across a range of shutter speeds, rather than looking at results at a given shutter speed, as we did here). Editor Will Brawley is also pretty steady, likely more so than most people reading this. So we had IR Publisher Dave Etchells do the camera-holding for this IS test, figuring that he's probably somewhere in the middle ground in terms of steadiness; not as good as the best, but not as bad as the worst. So depending on your own camera-holding skills, you may get better or worse results than we did in this range of shutter speeds, but at least conditions were pretty similar between the cameras, so it gives us some relative idea of how well they do compared to each other.

OK, enough talking, on to the results! In each group below, we show 1:1 crops of images shot on a sturdy tripod, and the best handheld shot we got with each camera. There's also a larger image showing the full frame, so you can have another reference on the relative reach of each lens.

Superzoom Shootout: Test Shot #4 - tripod vs handheld [with 1:1 crops]
SX50 - tripod	SX50 - handheld ("Shoot Only" IS Mode) (3 out of 6 shots usable)
Canon SX50 - Tripod 1/143s / f/6.5 / ISO 100 / 215mm (1200mm eq.)
The SX50 performed pretty well in this shot. There is a slight decline in image quality compared to the tripod shot, but it would be pretty negligible looking at a print at a normal viewing distance, unless you'd cropped way in.
SX60 - tripod	SX60 - handheld (3 out of 6 shots usable)
Canon SX60 - Tripod 1/159s / f/6.5 / ISO 100 / 247mm (1365mm eq.*)
In some ways the handheld image above looks slightly crisper than the tripod shot. We saw this same behavior on a second tripod vs. handheld comparison, where the handheld shot looked slightly better on higher contrast detail. Unfortunately, both images here are rough enough that its difficult to make a very valid comparison. There's a lot of mush in both images, making it difficult to judge the differences. The tripod shot looks ever-so-slightly back-focused, though, and we found the same issue on the several other tripod shots we took of this same scene. Our theory is that the center autofocus point covers more area than we expected it to, so it was "seeing" some of the background grass in the shot shown above, and thus focused a bit behind the sign itself.
S1 - tripod	S1 - handheld ("Shoot Only" IS Mode) (6 out of 9 shots usable)
Fujifilm S1 - Tripod 1/200s / f/5.6 / ISO 100 / 215mm (1200mm eq.**)
The S1 displays one of the best image stabilization systems of the whole bunch, as seen above. The handheld image is as good or better than the tripod image in some areas, and only slightly worse in a few. The Fuji also delivered a higher percentage of keepers than most of its competition. Two thumbs up for the Fuji S1 IS!
P600 - tripod	P600 - handheld (3 out of 4 shots usable)
Nikon P600 - Tripod 1/160s / f/6.5 / ISO 100 / 258mm (1440mm eq.)
In keeping with the P600's excellent image quality in other parts of our test, its IS did an excellent job at keeping detail nice and crisp. This even more surprising, given that its the longest-zooming camera in the shootout! In shooting with it, we felt that the P600's IS performance easily bested that of the other cameras in the lineup.
SP-100 - tripod	SP-100 - handheld (2.5 out of 5 shots usable)
Olympus SP-100 - Tripod 1/160s / f/6.5 / ISO 125 / 215mm (1200mm eq.)
The SP-100 image wasn't very good to start with on the tripod side, so it's again a bit difficult to judge the difference vs the handheld image. The handheld does show a bit more degradation in quality, though, primarily seen as noise and splotchiness in a few areas. (Sometimes, if an image becomes slightly more blurry, parts of it will drop below the local-contrast threshold for the noise-reduction processing, and will get smooshed further.) We indicated 2.5 out of 5 shots usable handheld with IS on, as one was right on the borderline; neither bad enough to call completely unacceptable, nor as good as we'd really like to have seen it.
FZ70	FZ70 - handheld (2.5 out of 5 shots usable)
Panasonic FZ70 - Tripod 1/160s / f/5.9 / ISO 100 / 215mm (1200mm eq.)
The tripod image from the FZ70 is not bad in comparison to many here, but the handheld shot is noticeably softer. (As with the SP-100 above, one of the five handheld shots was really right on the borderline for what we'd call acceptable, so we gave the camera a half-point for that one.)
WB2200F - tripod	WB2200F - handheld (4 out of 7 shots usable)
Samsung WB2200F - Tripod 1/206s / f/5.9 / ISO 100 / 215mm (1200mm eq.)
The WB2200F's tripod shot is noticeably riddled with mottling and splotchy areas, and the handheld shot displays just a bit worse. The image stabilization is likely not bad given how similar the images are, but unfortunately neither are good so it's really of little consequence here.
HX400V - tripod (IS enabled)	HX400V - handheld (2 out of 4 shots usable)
Sony HX400V - Tripod 1/160s / f/6.3 / ISO 80 / 214.8mm (~1200mm eq.)
Similar to a few other models, the HX400V's tripod image is so bad that trying to gauge the IS performance here is almost pointless.

As noted above, we didn't find any obvious failures in these cameras' IS functions, but did feel that the Fuji S1 and Nikon P600 did better in this area than the others.

Lab Sample Analysis

As a general rule in digital photography, smaller sensors tend to yield less detail and more noise. The image processing firmware is therefore configured to attempt to counter these unwanted side effects and is altered and tweaked with virtually every new model trying to overcome the obvious limitations. Unfortunately, physical constraints mean that superzoom cameras are going to have tiny sensors, in order to reach such extremely long equivalent focal lengths. Small sensors give a higher multiplier ratio between the actual focal length of the lens and the 35mm equivalent. A larger sensor thus means either shorter telephoto reach, or a much more bulky and expensive lens design.

It's our job to show you what you're buying, and test images shot under carefully controlled laboratory conditions are the single best way to strip away the hype and reveal the truth about the image quality each of these cameras produces. These images are shot at a medium focal length (neither wide nor particularly tele), and thus give us another data point for each camera's performance. They're shot at settings that should be in the sweet spot of each camera's performance parameters (medium focal length, and a pretty wide aperture, to avoid diffraction limiting with such tiny pixels), and the crops below are all at each camera's native (base) ISO.

What we're looking for, especially in sensors this size, is how each company handles the trick of removing noise and sharpening the image while still maintaining detail. Too much sharpening and/or processing usually strips detail, and can render images that look unrealistic (or even cartoon-like) while too little processing often leaves images that appear soft, dull or without "pop". And while we'd love to be able to experiment with different settings within each camera, our time constraint only allows us to test them here at their default JPEG settings, which is how the majority of users will use them anyway.

We've also included a camera with a Four Thirds sensor, the similarly-priced Olympus E-PL7, in order to show you what a larger sensor is capable of at base ISO for comparison purposes.

Superzoom Shootout: Lab Test Still Life (at 1:1 resolution)
Olympus E-PL7 (for reference only)
These E-PL7 images will give you a good idea of what this target actually looks like, including what the tiles in the mosaic around the monk look like, the horizontal stripes in the Pure bottle lettering, the clarity in the smaller lettering and the fairly flat wall behind the target that we use to judge noise in the shadowy areas.
Canon SX50
With a 12.1mp sensor the SX50's images naturally appear smaller in these 1:1 crops. The first two crops are soft compared to a Four Thirds sensor obviously, and there is little in the way of sharp detail, but it's also important to note that there is very little noise in the shadow area; bear this in mind as you scan the others.
Canon SX60
At 16.1 mp we'd hoped for better detail from this newest member of the superzoom fraternity, but the first two crops are fairly soft and lack crispness. There is a hint of noise in the shadows, but at least it's not bad compared to some below.
Fuji S1
Quite the contrast to the SX series cameras, the S1 takes a very heavy-handed approach in its default processing, especially sharpening. Sharpening and noise reduction add pop to the images, but they also rob the mosaic tiles and lettering of some detail, and add some obvious noise and mottling to the shadow crop.
Nikon P600
In what appears the most evenly balanced of these superzoom crops, the P600 images have a decent amount of detail for this sensor size, including the mosaic tiles and the horizontal stripes in the Pure lettering. There is a bit of grain in the shadowy areas, but it doesn't have the mottled appearance of several of the others.
Olympus SP-100
Where the S1 removed some of the detail in the mosaic tiles, the SP-100 removes almost all of it. That crop looks a bit like it was doused with water, and this is at base ISO. The noise levels in the shadowy areas are smooth enough, but the washed out detail in the other images is disconcerting.
Panasonic FZ70
The FZ70 handles the mosaic tile area quite well, showing good detail without any artifacts, and the Pure lettering is not bad although a bit soft. But the noise in the shadow crop is similar to what we saw in some areas of our real world images (blurred backgrounds, for instance), giving an unwanted choppiness and mottled look to many of its images.
Samsung WB2200F
The mosaic area of the WB2200F is not too bad overall, but the bottle lettering shows an odd sharpening pattern that turns the horizontal lines into little cubes in places. And there is a bit too much noise in the shadowy areas as well, rendering the latter two crops unnatural looking.
Sony HX400V
The HX400V is the only camera with a higher resolution of 20.4mp and hence the larger-sized images in these 1:1 crops. The Pure lettering is not too bad, yielding fairly good detail, though with more pumped saturation than from any of the others. The mosaic tile loses some detail, similar to the S1 processing, while the shadow area shows more noise and mottling than in any of the others. This is similar to the unnatural mottling we saw in the background of the first outdoor test shot, and is a big factor in the camera's poor image quality.