Test of Minolta A1 anti-shake

Minolta claims on the Dimaga A1 web site that one can shoot up to 3 shutter speeds (factor of 8) slower at full tele when anti-shake (AS) is activated than when it is turned off. In the first part of this web page I investigate if this is true. Further down I investigate how the AS performs when a 1.7x teleconvereter is attached to the camera.

Update May-2004: Firmware updates are now available for the A1/A2 that allow the camera to be customized for use with converter lenses. The test on this page was made before this firmware became available. A test that demonstrates great improvements of the anti-shake performance in combination with the tcon-17 1.6x teleconverter when the new firmware is installed can be found here .

AS at full tele: (no teleconverter)

A number of handheld shots were made at full tele of a test motiv. One picture is shown below (downsized):


Original size: 2560x1920 pixels
Full tele. Shooting distance: 7.2 m

144 pictures were taken using 10 different camera/teleconverter settings:
  - 4 shutter speeds with AS off (no teleconverter)
  - 6 shutter speeds with AS on (no teleconverter)
  - 6 shutter speeds with AS on and a 1.7x teleconverter.
Aperture, ISO setting, and illumination were varied to maintain correct exposure.

I will first discuss the pictures taken without teleconverter. A 70x60 pixel crop containing a circle and two dots (the red rectangle in the picture above) has been taken from each picture. The crops are displayed below arranged in collumns according to the camera settings used:

The magnitude and speed of my handshake motion might change during the shooting session. I therefore started out by shooting only three pictures with every camera setting, making up the first three rows above. Then I completed the pictures for the next three rows, and finally the pictures for the last three rows. This should provide a reasonable statistical material for judging how well the AS improves my shoting results.

We see from the crops above that the AS does not allow the shutter speed to be reduced by 3 stops (a factor 8) without causing extra bluring, since:

  - 1/25 sec shutter speed with AS on results in more frequent blurring than 1/200 sec and AS off
  - 1/13 sec shutter speed with AS on results in more frequent blurring than 1/100 sec and AS off
  - 1/6   sec shutter speed with AS on results in more frequent blurring than 1/50   sec and AS off
  - 1/3   sec shutter speed with AS on results in more frequent blurring than 1/25   sec and AS off.

However, if the speed is reduced by only 2 stops (a factor 4) the bluring decreases:

  - 1/50 sec shutter speed with AS on results in less frequent blurring than 1/200 sec and AS off
  - 1/25 sec shutter speed with AS on results in less frequent blurring than 1/100 sec and AS off
  - 1/13 sec shutter speed with AS on results in less frequent blurring than 1/50   sec and AS off
  - 1/6   sec shutter speed with AS on results in less frequent blurring than 1/25   sec and AS off.

In conclusion I estimate that the AS allows the shutter speed to be reduced with up to 2.3 stops (a factor 5) at full tele without increasing the handshake blurring. This is not quite as good as the "up to 3 stops" claimed by Minolta. It may, however, be that 3 stops can be fully compensated for with different (quicker, slower, etc.) handshake movements than those I happened to produced in this test. Note that effective compensation for 3 stops may still occur for other types of handshake movements (quicker, slower, etc.) than those I happened to produced in this test.

AS performance with teleconverter

I also wanted to see how well the AS works when a teleconverter add-on lens is mounted in the camera lens. A Tiffen Megaplus 2x converter was used. The actual magnification factor is 1.75, i.e. comparable to the Olympus TCON-17 and other add-on lenses that are popular with the A1. I originally bought this lens for use with my old Canon G1 camera, for which it works very well. With the A1 it produces some vignetting and the sharpness at full aperture could been better. However, the lens is good enough for testing the anti-shake system.

A downsized version of one of the teleconverter shots is shown below:


Original size: 2560x1920 pixels
Full tele with teleconverter.
Shooting distance: 7.2 m

The pictures with teleconverter were taken at the same distance as the previous ones, and the blurring of the object for a given handshake movement will therefore be the same in the two cases when AS turned off. I therefore use the same pictures as above, taken without teleconverter and with AS off, as a reference.

Below you see the crops from the reference pictures wit AS off to the left and crops from the pictures with teleconverter and AS on to the right. The crops to the right have been downsized by the teleconverter magnification ratio of 1.75 to make the comparison easier. Click here if you want to see the original sized crops.

Note that a slight reduction in contrast with teleconverter at 1/200, 1/100, and 1/50 sec can be attributed to the low f-stop number used at these shutter speeds combined with converter artifacts. Taking these contrast reduction artifacts into account, one may conclude that the handshake blurring is only slightly reduced by the AS when the converter is used. The improvement corresponds maybe to increasing the shutter speed by 0.5 stops (by a factor 1.4).

Actually, this result is not very surprising, since the A1 anti-shake system uses only mechanical measurements of the camera rotation movements to decide how to move the CCD sensor to compensate. Click here to see a fancy illustration of this. When a converter lens is used, the movements of the the image on the CCD caused by handshake movements is magnified. Since the camera doesn't know about this magnification it doesn't move the CCD enough to compensate properly.

One way to avoid this problem would be to allow the user to enter the magnification factor of the add-on lens used into the camera. The camera must necessarily adjust the magnitude of the correction according to the zoom setting already, so this should be fairly easy to implement. I would not be surprised if Minolta includes some option to tell the camera that a conversion lens is added in a firmware update in conjunction with a release of their own converter lens (which rumors say will come soon). Let us hope then, that they will allow magnification ratios from all kinds of converters (not only the ratio of the Minolta converter) to be entered into the camera.

Update May-2004: The new firmware is now available, and it only allows the camera to be customized for the two Konica-Minolta converters ACT-100 and ACW-100. The firmware works well with converters that are close in magnification factor to these converters, though, like the Olympus tcon-17 or the Raynox 1.54x converter. See here for more details.

In conclusion the AS system doesn't seem to be of very much help when using a teleconverter. The shutter speed can maybe be reduced by 0.5 stops without increasing the blurring. This is a pity, since shooting with a teleconverter lens may be one of the situations when a good AS system is most needed! Let us hope that Minolta soon come out with a firmware update that fixes the problem by allowing the user to tell the camera that an add-on lens is used, and that they make the update such that any magnification factor can be entered.

 

Erlend Rønnekleiv, www.eronn.net
erlend.web&eronn.net [replace & with @]