Originally posted by RioRico ...No, please bore us. Redundancy is the soul of communications.
Well, with an invitation like that, how could I refuse (you do have a talent for the pithy observation though)!
Rather than post a link or two, I'll try to illustrate what's going on SR-wise:
With SR off, camera shake will result in a certain amount of image blur, caused by movement of the image relative to the sensor.
The object of SR is to move the sensor by just the right amount, in order to exactly counter the movement of the image caused by the shake. With the correct FL information, the camera is able to do this (well, nearly).
So, assume a 100mm lens:
SR off, sensor moves 0% of ideal, you get 100% blur.
SR on:
Set FL value to 25mm. Sensor moves 25% of ideal, you get 75% blur.
Set FL value to 50mm. Sensor moves 50% of ideal, you get 50% blur.
Set FL value to 75mm. Sensor moves 75% of ideal, you get 25% blur.
Set FL value to 100mm. Sensor moves 100% of ideal, you get no blur.
Set FL value to 125mm. Sensor moves 125% of ideal, you get -25% blur.
Set FL value to 150mm. Sensor moves 150% of ideal, you get -50% blur.
Set FL value to 175mm. Sensor moves 175% of ideal, you get -75% blur.
Set FL value to 200mm. Sensor moves 200% of ideal, you get -100% blur.
Set FL value to 225mm. Sensor moves 225% of ideal, you get -125% blur.
etc. etc.
A negative sign is used to indicate that the blur is now acting in the opposite direction to the original (not that this matters).
So at 200mm, you get the same blur as with SR off (but in the opposite direction).
Using the above model, you can devise a formula which gives an optimum compromise value for a zoom lens. It turns out this formula is
Best compromise FL value = 2*H*L/(H+L)
Where H is the max FL of the zoom, L the min FL.
Using this for some popular vintage zooms:
28-70: theoretical setting: 40mm, actual setting: 40mm, max blur: 43%
28-80: theoretical setting: 41mm, actual setting: 40mm, max blur: 50%
28-90: theoretical setting: 43mm, actual setting: 40mm, max blur: 56%
28-105: theoretical setting: 44mm, actual setting: 45mm, max blur: 61%
35-70: theoretical setting: 47mm, actual setting: 45mm, max blur: 36%
35-105: theoretical setting: 52.5mm, actual setting: 50mm, max blur: 52%
35-135: theoretical setting: 56mm, actual setting: 55mm, max blur: 60%
70-150: theoretical setting: 95mm, actual setting: 100mm, max blur: 43%
70-150: theoretical setting: 95mm, actual setting: 85mm, max blur: 43%
75-150: theoretical setting: 100mm, actual setting: 100mm, max blur: 33%
70-210: theoretical setting: 105mm, actual setting: 100mm, max blur: 52%
80-200: theoretical setting: 114mm, actual setting: 100mm, max blur: 52%
80-200: theoretical setting: 114mm, actual setting: 120mm, max blur: 50%
80-210: theoretical setting: 116mm, actual setting: 120mm, max blur: 50%
60-300: theoretical setting: 100mm, actual setting: 100mm, max blur: 67%
If you were able to set the exact theoretical setting, you'd get maximum blur at each extreme of the zoom range. Choosing the nearest lower FL value will favour the wider end, and vice versa for the nearest higher value. Obviously, if you have the zoom set to exactly match the value you entered, you'll get full SR.
Anyway, you can see from the above that you can get quite reasonable SR for lenses with modest zoom ranges (and let's face it, you wouldn't want a vintage zoom with a big zoom range for reasons of IQ). 50% blur implies 1 stop of SR - not nearly as good as the 3-4 stops maybe for full SR, but still worthwhile.
Having said all that though, there aren't too many vintage zooms out there which are worth putting on your camera (in my view, anyway)...