Chris_Malcolm
Balmoral Clock Mirror Slap Test: 500mm @ 1/80th sec X8 centre crop
This is a series of photographs exploring the problem of mirror slap image smear with a Sony Alpha 350 with a 500mm reflex lens. The clock tower of the Balmoral Hotel is photographed from the back garden of Inverleith House in the Botanic Garden of Edinburgh. The distance is about 2km.
The camera is mounted on a Benbo tripod with column extended. The column had to be extended to get to head height to clear the obscuring shrubbery. That makes it a heavy tripod, but one prone to vibrate when struck, i.e. not as rigid or as well damped as a non-extended carbon fibre tripod. The lens is Sony's 500mm reflex, a very short light 500mm lens which therefore has no tripod mounting bolt of its own. The lighter weight and shorter length also give it less inertia, and much less angular inertia, with which to counter any camera motion resulting from mirror slap.
Focusing was by autofocus because previous testing has established that autofocus with this very critical lens gives me much more accurate and consistent focusing than I can ever achieve manually by eye, no matter what method or how much care I take over focusing. (That wasn't true at first with this lens -- there was consistent backfocusing in the camera's AF which I had to trim out. Follow this link to see before and after focus trimming shots of this same clock.)
To avoid camera vibration from touching, the shutter was fired by radio remote trigger after a long pause after last touching the camera.
The lens has a fixed aperture of f8, so the shutter speed was changed by changing ISO. The RAW output file was translated to a large high quality JPEG file by Sony's IDC RAW converter. Noise reduction was avoided in the IDC and applied afterwards to the JPEG by Neat Image. The RAW file and Neat Image processing were optimised only for maximum detail from the image. So because it was a dim misty day producing very low contrast images the contrast has been boosted a lot, which increased the noise.
Because the free demo version of Neat Image used here clears the JPEG EXIF data, the camera settings etc. are appended to these descriptions.
The first image shows the entire uncropped image at native unreduced size. The subsequent images crop out the cental clock face and magnify it 8 times by simple pixel duplication, so the pixels are in effect simply enlarged. That makes it easy to judge the detail resolution captured, and to measure any mirror slap smearing in terms of pixels.
These images show clearly that there is no mirror slap smearing at 1/320th sec, some nearly vertical smearing at 1/160th (about five pixels), and a lot more at 1/80th (about ten pixels). I haven't tested slower shutter speeds.
This 1/80th sec image has the worst mirror slap smearing, a smear of about 10 pixels in length at about an angle of 10:30 on the clock face.
The full size ISO 400 image at 1/320th sec with no mirror slap smearing can be seen here. The magnified central crop from that, which is directly comparable with this magnified central crop, can be seen here.
There is no mirror lock up on the Sony A350, so this raises the interesting question of whether more rigid mounting of the camera and lens can prevent this. Initial experiments suggest that it won't be difficult to cut it down by about 1/2, possibly even to 1/4, but that it will be difficult to eliminate. A clock face like this is a very good target for this kind of test, but is not really suitable because over this kind of distance over a city there is rarely little enough thermal turbulence in the air to get shots as sharp as this. Dim misty days with no wind like this one are the best, but while the cloud cover and mist reduce the sun's heat, they also reduce its light, and so it's not possible to use the much sharper ISO 100 for this kind of testing.
So further testing will await my development of an interior target with bright and controllable lighting.
Original DSC06605RW_ntE8X
[Sony A350, SAL 500mm reflex, ISO 100, f8, 1/80th sec.]
Balmoral Clock Mirror Slap Test: 500mm @ 1/80th sec X8 centre crop
This is a series of photographs exploring the problem of mirror slap image smear with a Sony Alpha 350 with a 500mm reflex lens. The clock tower of the Balmoral Hotel is photographed from the back garden of Inverleith House in the Botanic Garden of Edinburgh. The distance is about 2km.
The camera is mounted on a Benbo tripod with column extended. The column had to be extended to get to head height to clear the obscuring shrubbery. That makes it a heavy tripod, but one prone to vibrate when struck, i.e. not as rigid or as well damped as a non-extended carbon fibre tripod. The lens is Sony's 500mm reflex, a very short light 500mm lens which therefore has no tripod mounting bolt of its own. The lighter weight and shorter length also give it less inertia, and much less angular inertia, with which to counter any camera motion resulting from mirror slap.
Focusing was by autofocus because previous testing has established that autofocus with this very critical lens gives me much more accurate and consistent focusing than I can ever achieve manually by eye, no matter what method or how much care I take over focusing. (That wasn't true at first with this lens -- there was consistent backfocusing in the camera's AF which I had to trim out. Follow this link to see before and after focus trimming shots of this same clock.)
To avoid camera vibration from touching, the shutter was fired by radio remote trigger after a long pause after last touching the camera.
The lens has a fixed aperture of f8, so the shutter speed was changed by changing ISO. The RAW output file was translated to a large high quality JPEG file by Sony's IDC RAW converter. Noise reduction was avoided in the IDC and applied afterwards to the JPEG by Neat Image. The RAW file and Neat Image processing were optimised only for maximum detail from the image. So because it was a dim misty day producing very low contrast images the contrast has been boosted a lot, which increased the noise.
Because the free demo version of Neat Image used here clears the JPEG EXIF data, the camera settings etc. are appended to these descriptions.
The first image shows the entire uncropped image at native unreduced size. The subsequent images crop out the cental clock face and magnify it 8 times by simple pixel duplication, so the pixels are in effect simply enlarged. That makes it easy to judge the detail resolution captured, and to measure any mirror slap smearing in terms of pixels.
These images show clearly that there is no mirror slap smearing at 1/320th sec, some nearly vertical smearing at 1/160th (about five pixels), and a lot more at 1/80th (about ten pixels). I haven't tested slower shutter speeds.
This 1/80th sec image has the worst mirror slap smearing, a smear of about 10 pixels in length at about an angle of 10:30 on the clock face.
The full size ISO 400 image at 1/320th sec with no mirror slap smearing can be seen here. The magnified central crop from that, which is directly comparable with this magnified central crop, can be seen here.
There is no mirror lock up on the Sony A350, so this raises the interesting question of whether more rigid mounting of the camera and lens can prevent this. Initial experiments suggest that it won't be difficult to cut it down by about 1/2, possibly even to 1/4, but that it will be difficult to eliminate. A clock face like this is a very good target for this kind of test, but is not really suitable because over this kind of distance over a city there is rarely little enough thermal turbulence in the air to get shots as sharp as this. Dim misty days with no wind like this one are the best, but while the cloud cover and mist reduce the sun's heat, they also reduce its light, and so it's not possible to use the much sharper ISO 100 for this kind of testing.
So further testing will await my development of an interior target with bright and controllable lighting.
Original DSC06605RW_ntE8X
[Sony A350, SAL 500mm reflex, ISO 100, f8, 1/80th sec.]