EM1 shuter shock measurment

In the last four weeks I did some shutter shock experiments which I want to share with you today, especially because EM1 has an improved firmware now.

I observed the shutter shock issue last year when I bought the 75-300mm mFT lens for my EM5. At 1/200s pictures always had vertical blur. With slower shutter speeds, picture quality of handheld shots often were less blurred. Which puts the IBIS in good light but raises some questions about shutter shock.

Later, when I bought the EM1, I made a short test with my tele lenses and I found the EM1 being a little bit better.

Meanwhile I sold the 75-300mm lens and replaced it by the heavy 50-200SWD lens, which solved the problem for me (at least partially, you will see).

As the discussion recently also included shorter focal lengths, I got curios whether my other lenses have the problem or not.

I already had some posts in the German Olympus forum, where some very clever Olympus users have their discussion. And I met Reinhard Wagner from this forum for exchanging our findings and for discussing our conclusions. He writes very good books about Olympus cameras and he also wrote a short Oly ePaper about his findings regarding shutter shock. Very good reading (it is in German). Simply search and download "oly-e-paper 1/2014"!

But now, to sum up my own findings: I found two ways to make shutter shock visible:

1. A moving bright point
2. MTF measurments with slightly slanted black squares

1. Moving bright point

One very good way to make shutter shock vibrations visible is an oscilloscope. I did not put any signal on the inputs, the scope was simply drawing a straight line. Shutter shock will cause a twisted line if the camera goes up and down while the shutter is open. But the effect is really small, only some pixels. Therefore, I had to step away until the scope screen was only 300 pixels high on the resulting photo.

My settings were like this:

• Scope running free with 20ms/cm (5 lines per second)
• Camera handhold with 1/25s (40ms) and f11
• Autofocus on the knobs right to the screen
• Doing 3 to 6 shots for each setting

100% crop oftypical blur (green LED) and corresponding twisted line

Using good settings (like IBIS=auto), can I reproduce the shutter shock?

I think: Yes. The results were quite easy to reproduce, at least with the "bad" camers/lens combinations. Here is an example of a bad pair: EM5 with 40-150/4-5.6 tele-zoom:

3 consecutive shots with EM5 and 40-150mmlens at 123mm/f11

As I had a method to reproduce it, I was curios about all the options which might influence the effect.

First question: EM1 compared to other cameras:

I did not have to many cameras at hand and I had to borrow a PM2 and a Nikon D700. Here is a quick comparison:

Different camera types

You see that smaller point and shoots (XZ1) or bridge cameras (Minolta A200) have an advantage here because they don't have a focal plane shutter.

Second question: What about different lenses?

I tested with:

• Samyang 7.5mm fisheye
• Olympus 12-40/2.8
• Olympus 45mm/1.8
• Olympus 40-150/4-5.6
• Olympus 50-200/2.8-3.5SWD
• Olympus 75/1.8
• Olympus 75-300/4.8-6.3

You see that even wide angle lenses show some shock. That was a surprise for me. And that EM5 and PM2 have some problems to keep the 75-300mm lens steady enough at 1/25s. That was not a surprise for me.

Third question: How to reduce shutter shock?

I tried different external things to reduce shutter shock. My favorite which I knew that it will work is putting some weight under the camera, inmy case i use my folded tripod.

I tried:

• IBIS = off
• IBIS = on
• Cable release and camera on a soft surface (a pillow)
• Cable release and camera on a tripod
• Weight (1.1kg tripod)
• EM1 HLD7 grip
• And portrait orientation (Ok, the blur is no longer vertical now, instead it is horizontal.)

External or mechanical things to reduce shutter shock

Another surprise: my 40-150mmlens goes mad on my tripod.

No surprise for me: without IBIS, blur is even worse.

Last not least: What do all these nice in-camera options do, including AntiShock=0sec?

Now we have these things like "short delay", Anti-Shock=1/8s and of course very new and very hot: Anti-Shock=0s

I tried these options:

• normal settings
• Shutter delay = short
• Antishock =1/8s
• Shutter delay = short and Antishock =1/8s
• Antishock =0s
• Shutter delay = short and Antishock =0s
• EVF IBIS =off

A bad surprise was the combination of short shutter delay and 1/8s anti-shock.

And the good surprise is the AntiShock=0s option, seems toe work good with short shutter delay. This is my favorite option now!!

This is part one of my findings.

But I have another method: the MTF measurements

1. MTF measurements

Some time ago I did some lens tests with a tool called "MTF_Mapper".

http://sourceforge.net/projects/mtfmapper/

You generate and print some test targets wth it and it will annotate the photos with the measures MTF values. And you can test if your AF is spot on or not with it.

But this time, I did not use the official MTF targets, I made some by my own, specialized to test shutter shock. My idea was to tilt the squares 15 degrees which is the angle of the shock movement. This way two edges of a tilted square are sharp and the other two edges are blurred.

This is the target, you see, no rocket science:

Test target for shutter shock

If I feed a test shot into MTF-mapper, I can use the MTF values on the edges of the inner squares and type them into a Excel table.

Annotated test photo

I made some tests with these settings:

• Lens was 75mm/.9
• Aperture f4
• Camera on pillow wit cable release
• Time from 1/8000s to 1/10s
• Using two FP flashes to get enough light above 1/320s
• Target size was A2 (for sheets A4)
• Distance was 2-3 meters (target fills the frame 95%)

I took 3 photos and extracted 8 MTF values vertical and 8 values horizontal, I calculated the median of these 24 values v and h for each shutter time.

The dotted line is the horizontal blur which is only affected by shutter shock at very slow shutter times (camera sits on a pillow, it is clear that the shutter movement will cause some slow camera movement too). But the most interesting part is the gren solid line: the vertical blur of my EM1.

Shutter Shock with EM1 and EM5 compared

You see that EM1 gets blurry below 1/500s. EM5 is better at 1/200s but at 1/50 both cameras are far from perfect (perfect would be 0.5 cycles per pixel pattern, 1 perfect white and 1 perfect black pixel).

But please note: the MTF values are very sensitive to even the slightest blur. I never noticed any severe shutter shock with my 75/1.8 until I did these tests. It would be hard to find the shutter shock in any real world photo shot at 1/200s with the 75mm lens.

This said, here is the same test with the new firmware, AntiShock=0 enabled and disabled.

AntiShock=0 compared to AntiShock=off

I would say quite a dramatic improvement.

That was my longest thread ever, hope you find it interesting enough to read to the end. After having the new firmware the "shutter shock" poblem is no longer a big issue for me. I don't think I would have spend all this effort if the option would have been available some weeks earlier ...

But anyhow I have learned a lot, and now I know that even my 50-200mm lens is affected by shutter shock. It shows no double lines, but if I enable AntiShock=0, there is much less blur in my photos. Which makes quite a difference between a good and a very good lens!

Best regards

Christof