Noise, resolution and sensitivity
Our standard lens for testing camera resolution is a Sigma 105mm f/2.8 macro. Unfortunately, there is currently a problem with some Sigma lenses, including this one, when used on the Alpha 33. The lens does not focus automatically and aperture is not controlled correctly, resulting in meter readings and exposures varying wildly between shots. The effect is similar to the camera automatically bracketing. Sigma has promised to update any lenses affected by this problem for free.
Instead of using the Sigma 105mm, I performed the test using the Sony DT 18-55mm f/3.5-5.6 SAM lens, set to around 50mm. This is the lens that is supplied as a kit with the Alpha 33, so is a representative test of what a user can expect.
At sensitivities of ISO 100 and 200, the Alpha 33 is capable of resolving to nearly 26 on our test chart, which is on a par with the results we have seen from other cameras of 14-million-pixels-plus. The resolution drops to around 24 when the sensitivity reaches ISO 400. It stays at around this level until around ISO 1600, where the effects of luminance noise and its reduction take the resolution to around 20.
At ISO 6400 and ISO 12,800 the chroma and luminance noise reduction has taken its toll, and although the camera can still resolve around 20 on our chart there is a definite loss in sharpness. This should not be too obtrusive when images are printed below A4 in size.
The new multi-frame NR mode allows JPEG images to be captured at ISO 25,600, by combining multiple images and merging them together. Although the technique reduces the amount of chroma noise, images are severely smudged. That said, it is a useful ‘get-out-of-jail’ option when you simply have to get the shot.
Image: Sony Alpha 33 noise chart.
There is some debate online about a problem that causes specular highlights captured with the Alpha 55 to be ghosted around 10 pixels below their original position in the image. Despite using the Alpha 33 to photograph a rock concert, where images had many specular highlights caused by spotlights and metallic objects, I found no such ghosting effect.
This is not to say that the ghosting effect doesn’t exist, but that it hasn’t occurred in any of our images taken in a situation where there were ideal conditions for it to appear. To this end, I would not be concerned about the issue, but we will continue to look at the ghosting issue when we test the Alpha 55 in the coming months.
Image: These images show 72ppi sections of images of a resolution chart, captured using the Sony 18-55mm lens set to 50mm. We show the section of the resolution chart where the camera starts to fail to reproduce the lines separately. The higher the number visible in these images, the better the camera’s detail resolution is at the specified sensitivity setting.
