Sigma 150-600 Contemporary OS Anti-Vibration Algorithm Comparison

This article analyzes the anti-vibration (OS) algorithms available using firmware version 1.02 for the Sigma 150-600 mm Contemporary lens. Sigma hasn’t published any information on the relative effectiveness of the three available OS algorithms, so I took it upon myself to see if there is any difference.

The three available OS algorithms are called “Dynamic View Mode”, “Standard”, and “Moderate View Mode”. The default setting, if you don’t program any customization, is “Standard”. Each of these modes is available regardless of selecting “OS1” or “OS2”. The “OS1” is the normal hand-held mode, while the “OS2” mode is used for horizontal panning, such as while mounted on a tripod. All tests reported here are using “OS1”. Sigma "OS" is the same as Nikon "VR".

You must use Sigma’s “USB Dock” with their Optimization Pro software to program any customization into the lens. The dock also allows programming focus fine-tune adjustments (16 settings for 4 focal lengths and 4 distances) and focus limiter modifications.

To perform the tests, I used a tripod to rest my fist on, and then I rested the lens on my fist. This arrangement afforded me some level of aiming control, while still letting the lens “wiggle around” to simulate hand-held. I shot about 10 frames of my “A0” size resolution chart at 55 feet for each lens switch setting. The resolution chart images were analyzed using the MTF Mapper program from Frans van den Burgh (see this link). The resolution measurements require that the chart images be pretty level and perpendicular to the lens axis, which is why I didn’t simply try hand-holding the lens while shooting the chart.

I shot each resolution chart image at 400 mm and f/6.3, and I used back-button focus with “AF-C” continuous auto-focus. These are typical shooting conditions for me, which is why I chose them for the testing. Beyond 400 mm, accurate aiming just gets too difficult for reliable/repeatable testing measurements.

My lens was programmed with the “C1” switch setting having “Fast AF Priority” focus speed and “Dynamic View Mode” for the OS setting. I have already made tests that show this “fast” focus mode is essentially as accurate as the “Standard” (default) focus mode, at least when using firmware 1.02. The “C2” switch was set up with “Standard” AF focus speed and “Moderate View Mode” for the OS setting. If the customization switch is turned off, then you get “Standard” AF focus speed and “Standard” OS as well.

Tests such as these are 'statistical' in nature, since they involve taking measurements with a lens waving around. I have included some data below, to give you an idea of how the measurements vary. Yes, I could have made 1,000 measurements at each setting to raise the confidence level; I leave that as an exercise to the reader.

High Shutter Speed Tests

My previous testing has shown an insignificant difference in resolution when you leave OS active at higher shutter speeds (1/1000 and above). This statement is not valid for all lenses!! Any OS algorithm is equally effective (or ineffective, if you wish) at high shutter speeds (you need firmware 1.01 or newer to get this result, however). I only saw a decrease of about 1.0 lp/mm MTF50 by leaving OS active above 1/500 shutter speed.

OS Setting Screen

C1 Switch Settings That I’m Using Now

Medium Shutter Speed Tests

The following tests were conducted using a shutter speed of 1/250 second. For 400mm using an APS-C sensor (600mm equivalent), I consider this “medium”, and starting to get into the realm of needing anti-vibration. Some people would benefit with stabilization at this speed, and some wouldn’t.

Dynamic View OS, High-speed AF MTF50 Measurements:

36, 40, 38, 30, 36, 40, 42, 42, 40, 36, 40. Average = 38.2 lp/mm

Moderate View OS, Normal (Standard) AF MTF50 Measurements:

42,40,42,40,44,40,45,38,42,42,42,42. Average = 41.5 lp/mm

Standard (default) OS, Normal (Standard) AF MTF50 Measurements:

40,45,42,38,42,42,45,42. Average 42.0 lp/mm

OS Off, Normal (Standard) AF MTF50 Measurements:

44,42,40,42,40,34,38,42,38. Average 40.0 lp/mm

Results here don’t show much difference with OS active or not. The “Standard” OS algorithm got the best results, but not enough to really matter.

Low Shutter Speed Tests

These tests used a shutter speed of 1/60, or a little more than 3 stops beyond the traditional limit of 1/600 for an equivalent of 600 mm (DX frame). This is roughly the rated effectiveness of OS for this lens.

Dynamic View OS, High-speed AF MTF50 Measurements:

23,26,34,30,26,30,22,28,28,28. Average 27.5 lp/mm

Moderate View OS, Normal (Standard) AF MTF50 Measurements:

30,24,28,30,30,28,30,30,30,28. Average 28.8 lp/mm

Standard (default) OS, Normal (Standard) AF MTF50 Measurements:

24,23,24,26,32,22,24,18,28,24. Average 24.2 lp/mm

Results here show that the “Moderate View” is the winner. I didn’t show the “OS Off” here, because the images were mostly blurred beyond recognition.

Bear in mind that the MTF Mapper software is extremely picky, so the numbers here may lead you to believe that OS is not that helpful. Not true. The pictures are enormously helped by the OS system when you shoot at slower speeds, but there’s no substitute for high shutter speeds.

Conclusion

It appears that “Moderate View” wins, although not by a huge margin. Sigma (rather cryptically) describes the effect of how each OS algorithm “looks” through the viewfinder. To me, what counts more is which algorithm provides the best anti-vibration effect in the final picture. It seems to me that there is a different end result in your pictures, depending upon which OS algorithm you pick.

I prefer the “look” of Nikkor VR to Sigma OS when looking through the viewfinder, but both companies seem to provide roughly equivalent results in the final shot. Newer lenses invariably provide better stabilization, though. Sigma has the advantage of future OS algorithm improvements, however, available through a new firmware update.

I saw a definite improvement in the Sigma auto-focus system (speed and accuracy) after loading the firmware versions 1.01 and 1.02. I also saw an improvement in the ability to not “mess up” the shot when forgetting to turn off anti-vibration at high shutter speeds. Don’t be surprised if Sigma has more tricks up their firmware sleeves in the future.

Recently, Tamron finally saw the light, and is now copying Sigma with the ability to reprogram lens firmware (in their new 150-600 mm offering), with an essentially identical set of features as Sigma. Nikon et al. hasn’t yet seen the light.

#review