Monday, 27 February 2012

Samyang 7.5mm f/3.5 fisheye lens review

The Samyang 7.5mm f/3.5 fisheye lens, also marketed as Rokinon and Bower, is special in that it is one of the first third party lens designs made specially for Micro Four Thirds. Some other manual focus lenses for Micro Four Thirds are older designs with a new mount.

The Samyang fisheye lens is an alternative to the native Panasonic Lumix G 8mm f/3.5 fisheye lens, so it makes sense to compare them. Here they are, with lens caps:

They have different type caps. The Samyang, on the left, has a cap held in place with clips, operated by pressing the side tabs. The Lumix's cap is held in place with friction only. I prefer the latter, since the Samyang cap must be inserted correctly rotated, which is somewhat more awkward.

Removing the caps, reveals that the Lumix has a significantly larger front lens element. Both have the same type of integrated hood.

Looking at the rear of the lenses, we see that also the exit pupils have different sizes. The Samyang lens has the smallest exit pupil I have ever seen on a Micro Four Thirds lens. A large exit pupil is usually associated with a high quality lens, although there is of course no definite connection.

Yet another option is the Olympus 9mm f/8 body cap fisheye lens (my review). Both lenses are seen below:

The Olympus 9mm f/8 fisheye is not as wide, covering only 140° diagonally. Further, it is not nearly as fast, with a much smaller aperture of f/8. And it does not have any aperture mechanism, so you cannot stop it down at all. Finally, it does not perform nearly as well as the Samyang lens. I would say: Get the Olympus lens only if you would like a very compact lens at a low price.


Samyang 7.5mm f/3.5Lumix G 8mm f/3.5Olympus 9mm f/8
Aperture rangef/3.5-f/22f/3.5-f/22f/8
Number of aperture blades67None
Lens elements9105
Lens group794
Minimum focus distance0.09m0.10m0.20m
Diagonal field of view180°180°140°

Physical appearance

Even though the Samyang lens is slightly smaller, and has smaller lens elements, it weights more. This is due to the lens being made from metal materials, in contrast to the Lumix lens, which has a plastic exterior.

The Samyang lens has a traditional mechanical focus ring, which moves the whole lens assembly back and forth. The focus ring is well dampened, and feels high quality. There is a distance scale, but sadly no depth of focus (DoF) scale.

Regarding the focus scale, I find it to be a bit lacking. It has no markings between infinity and 0.25m, as seen here:

There is some sense to this practice, of course: The focus is the most critical at shorter distances. Still, I'm guessing that quite many see the need to focus the lens to around 1-2 meter distance more often. So a distance marker at around 1-2 meters would be good. Of course, this lens has a quite deep Depth of Focus (DoF), so the focus is not much of an issue unless you are printing large copies of your pictures.

I solved this by adding a distance marker at 1m myself. It doesn't look very good, but it makes the distance scale much easier to use:

I found the infinity mark to be somewhat inaccurate. But I guess that is not uncommon.

The aperture ring also feels well made, and has clicks at full and half stops.

No electrical contacts

The Samyang lens is a purely mechanical lens, and has no electrical contacts. This means that the camera does not recognize that the lens is mounted at all, and certainly cannot read off information about the focal length or aperture.

It also means that you need to set the "SHOOT W/O LENS" option. Otherwise, the camera will not operate, as it believes that no lens is mounted at all. Here is the "SHOOT W/O LENS" menu option from the GH2 camera:

The camera will still store EXIF data in your image files, however, the information about focal length and aperture is never passed to the camera, and will be missing when using this lens.


Being a fisheye lens, of course you are going to get a lot of distortion. Still, there have been some rumors online that this lens has a stereographic projection, rather than a spherical projection, the norm for fisheye lenses. The stereographic projection is less distorted.

Comparing the distortion of the two fisheye lenses reveals that they are in fact quite similar. Here is a test image showing how a rectangular grid is projected to the sensor with the two lenses:

Read more about the study here. My conclusion is that the Samyang lens in fact features somewhat less distortion, but the difference is rather subtle.

When converting a fisheye image to rectilinear, a process called defishing, it is generally easier to defish the images coming from the Samyang 7.5mm f/3.5 than the Lumix G 8mm f/3.5 lens. The reason being that the Samyang lens has a more common projection model for fisheye lenses.

The Panasonic Lumix G 8mm f/3.5 fisheye lens, on the other hand, has more distortion that what is common for fisheye lenses, making it more difficult to use under certain conditions. Most of the time, this is a non-issue, but if you specifically intend to defish the images, I would go for the Samyang lens.


I have taken the same picture with both lenses, to examine their optical properties. Here they are, both at f/3.5, taken on a tripod. I used the base sensitivity of the GH2, ISO 160.

Samyang 7.5mm f/3.5Lumix G 8mm f/3.5

To better examine the sharpness, here are some 100% crops from the upper right hand corner (click for a larger view):

It is quite apparent that the Samyang/Rokinon lens is surprisingly sharp in the corner. There are virtually no chromatic aberration (CA) artifacts. On the other hand, the Lumix fisheye lens has some residual CAs, even after the in camera CA adjustments.

Here is another sharpness comparison with the Lumix G 8mm f/3.5 fisheye lens. In this test, the lenses come out more alike in terms of sharpness.

The fisheye lenses cover a 180° diagonal field of view. So it can be quite difficult to avoid having the sun, or some other bright light source in the frame. Hence, the handling of flare is quite important. Does a strong, visible light source ruin the image?

Again, here are some comparison images, both taken at the same spot:

Samyang 7.5mm f/3.5Lumix G 8mm f/3.5
Samyang 7.5mm f/11Lumix G 8mm f/11

The Lumix lens gets the green ghosting in the opposite corner of the sun. The Samyang lens avoids this. Overall, I'd say the Samyang lens handles the flare better.

The Samyang lens has six aperture diaphragm blades, and hence, there are six rays around strong flares when stopping down. You can see this in the lower left example image, and even better in an article about the significance of odd and even aperture diaphragm blade numbers.

In use

Here is a discussion about the possible use of a fisheye lens.

This lens can be used on any Micro Four Thirds camera, given that you enable the "SHOOT W/O LENS" option, as described above.

When used on the Panasonic GH2, it is possible to use the rear click wheel to enable the focus assist mode. Pressing the wheel once enables the magnified view. It can then be magnified even more by scrolling the wheel. To disable the magnified mode, half press the shutter, or press the click wheel once more. This GH2 feature is described here. All Micro Four Thirds camera have the magnified focus assist mode, but it may be more involved to enable it on some cameras.

To get the best focus accuracy, you should ideally focus manually using the largest aperture (f/3.5). When finding the focus range, you can then stop down the aperture to the desired size, and take the picture. Most system cameras handle this process automatically, however, since there is no signal connection between the lens and camera, the camera cannot help you with this process.

You can use the camera's automatic exposure with this lens. Just set the aperture you want on the lens aperture ring, set the ISO you want (or leave the ISO on auto), and the camera will set the appropriate shutter speed. This works well, in, e.g., the P and A modes.

The exposure is generally good in low contrast situations, e.g., in daylight. With high contrast scenes, though, the GH2 tends to underexpose significantly. This is strange to me, since using the Lumix G 8mm fisheye lens in the same way yields a good exposure.

At night, I must dial in approximately +1.3 EV compensation on the Samyang lens to get the same exposure as the Lumix G lens would have given. This is at full aperture, f/3.5. At f/5.6, the gap increases, and I must dial in +2.0 EV.

This behavior puzzles me. As the camera sees roughly the same image through both lenses, I would have guessed that it also exposes similarly. But it appears that using a non-electric legacy lens triggers a different exposure mode, which preserves highlights to a larger degree. Or, and I am just speculating here, since the Samyang lens features less flare, perhaps highlights are more defined with the Samyang lens, and hence the camera does more to preserve them.

The need for extra compensation with the Samyang lens can be illustrated with this example. The scene was shot with both lenses at f/5.6.

First, the Samyang lens without compensation:

The Samyang lens with +2.0 EV compensation:

And finally the Lumix lens with no compensation:

In my daytime outdoor sharpness example images above, I used no exposure compensation at all, and they all came out ok. So this is only a problem when the contrast is high, for example at night with artificial light sources.

Since the camera does not know that this is an extremely wide angle lens, it will assume that it has a focal length of around 50mm, and use a shutter speed of around 1/125s. This causes the camera to use too high ISO in some cases, as this lens can be handheld at 1/15s with few problems. Set the ISO lower manually to avoid this.

Example photo

This image of Sergels Torgs was taken at f/5.6, ISO 160. Notice that straight lines which pass through the centre of the frame, are reproduced as straight in the image. The further from the centre a straight line passes, the more curved it is rendered. This is an important property of a fisheye lens.

Here are some 100% crops from the image, showing that the sharpness is very good throughout the frame (click to enlarge):

Another example, this one is from the Stockholm Old City:

The picture was taken by pointing the camera directly upwards inside a circular open yard.

And an example picture from the Rockefeller Center in New York:

Example video

This video was recorded with the Samyang lens at f/3.5.

The GH2 tends to underexpose when using this lens in high contrast situations, e.g., at night. Hence, I had to dial in +1.3 stops of exposure compensation. When using the Panasonic Lumix G 8mm lens, the exposure compensation is not needed, so it is clear that the camera treats them differently.

The first 20 seconds of the video below were filmed with the Samyang 7.5mm f/3.5 fisheye lens, using the Panasonic GH3 camera:

In this example, I filmed a firework display using the fisheye lens.


The process of making rectilinear, or at least near-rectilinear transformations from a fisheye image is commonly called defishing. I have written about this process here.

I have found that when defishing an image from the Samyang fisheye lens at 4:3 aspect ratio, you get a very wide rectilinear image. In my experience, it has a vertical field of view which is a little bit wider than the Lumix G 7-14mm f/4 lens at 7mm, and a much wider horizontal field of view.

Here is a real world comparison with the Lumix G 7-14mm f/4 ultra wide angle lens at 7mm.


Samyang also produce fisheye lenses for DSLR cameras, with the mounts of major camera manufacturers. You could conceivably buy e.g. the Nikon version of this lens, specified as 8mm f/3.5, and mount it to a Micro Four Thirds camera using an adapter for Nikon F on M4/3. This lens is also marketed as Rokinon, Bower and Vivitar.

However, I would not recommend doing so. The DSLR version of the lens is made for a larger sensor, and hence, when using it on Micro Four Thirds, you don't get 180° coverage in the diagonal. Also, due to the longer register distance of the DSLR systems, the 8mm fisheye lens made for these formats is much larger in size and weight. So if you can, go for the lens designed for Micro Four Thirds in the first place.


The Samyang/Rokinon 7.5mm f/3.5 fisheye lens has got impressively good optical properties. This makes it a good alternative to the Panasonic Lumix G 8mm f/3.5 fisheye lens, at a lower cost. You'll have to focus manually, though, which is quite easy for objects far away. When photographing close objects, though, it takes somewhat more effort to use the Samyang lens.

With the GH2 camera, I need to make exposure compensation during night photos with the Samyang lens, to get a sensible exposure. This is not needed with the Lumix fisheye lens. I'm sure this is a camera issue, though, and nothing I should blame the lens for.

I think the Lumix lens generally gives better colours. But this is a subjective matter, of course. As the Lumix lens is quite expensive, the Samyang lens is a good alternative, if you can live with using manual focus. The optical properties are certainly very good.

All in all, the Samyang/Rokinon 7.5mm f/3.5 fisheye lens is a very good lens, and provides good value for money. If you want to try a fisheye lens, it is a good choice.

Further reading

Sharpness comparison with the Lumix G 8mm f/3.5 fisheye.

Comparison with the GoPro Hero 3+ Black.

Comparison with the Lumix G 7-14mm ultra wide angle lens at 7mm.

Comparison with the Olympus 9mm f/8 body cap fisheye lens.

Recording fireworks with the fisheye lens.

Samyang 7.5mm fisheye used on a Sony NEX E-mount camera.

Tutorial on defishing a fisheye lens.

Sunday, 26 February 2012

Reverser ring for macro

There are many ways to achieve macro possibilities. The simplest is of course to buy a dedicated macro lens, like the Leica-Lumix 45mm f/2.8 1:1 macro. This lens does macro very good, and is easy and fun to use. The only negative side is that it is rather expensive, probably in part because of the premium Leica branding.

Another way is to get a cheap legacy lens, and some macro extension rings. If the lens has a mechanical aperture ring and focus ring, then it is quite easy to use with macro. Of course, you must control the focus and aperture manually on the ring, and there is no EXIF information to the camera. For even better control, you could buy some macro bellows, for stepless extension distance.

A third method would be to buy a macro lens to put into the filter threads of the lens. There are a number of third party macro filters to buy, and Panasonic has even launched their own, which can be used with the Lumix G 14mm f/2.5 pancake, and the Lumix G X PZ 14-42mm powerzoom pancake kit lens. I have not tried any of these.

A fourth method is to buy a macro reverser ring. A reverser ring is quite simply an attachment which allows mounting a lens reversed on the camera. I have tried one such ring, easily available on various auction sites for around US$10. It is a simple thing. On one side, it has a Micro Four Thirds lens mount. The red dot makes mounting easy:

On the other side, it has a 52mm lens thread:

The 52mm threads is a good choice, since it fits a number of lenses.

Here is the adapter mounted to the Panasonic GH2. It will of course fit any Micro Four Thirds camera.

Mounting a Nikkor 24mm f/2 AIS lens

I have got an old, rather banged up Nikkor 24mm f/2 AIS lens. It has got a 52mm front lens thread, and hence fits well onto the adapter. Here is a video showing how to mount it:

After mounting the lens, you can still control the aperture with the mechanical aperture ring, which is a good thing. When used like this, the lens has a very short focus distance, good for macro, but it also has a very thin depth of focus. Hence, you will normally want to stop down the aperture to at least f/5.6 to get more in focus. Normally, you will focus at f/2 (for the best focus accuracy), and stop down to f/5.6 (or more) before taking the picture.

I took this test picture of a ruler (in millimeter) to test the magnification:

The test picture shows that the magnification is 17.3:13 (with 17.3mm being the horizontal width of the Four Thirds sensor, and 13mm being the width of the object depicted). This corresponds to 1:0.75 magnification, also written as 1.33x. This is slightly more magnification than what is possible with the Leica-Lumix 45mm f/2.8 1:1 macro. On the other hand, no focus is possible at all with this solution. You can only take pictures at 1.33x magnification.

Mounting a Lumix G 14-42mm f/3.5-5.6 kit zoom lens

The Lumix G 14-42mm kit zoom lens also has a 52mm front thread. Hence, you can mount it to this adapter. Here is a video showing how to mount it:

Using this is quite different from the manual Nikkor 24mm f/2. First of all, the negative sides: The Lumix G 14-42mm lens does not have an aperture ring. Hence, you cannot easily adjust the aperture. This makes macro use difficult, since you will normally need to stop down the aperture for more depth of focus (DOF).

The positive side is that you can use the zoom ring for changing the magnification rate. The focus distance is also changed at the same time, but at this enlargement, it makes more sense to use the word "magnification" than "focus". At 14mm, you get the maximum magnification, and the minimum at 42mm. Let's take a look at this.


Again, I photograph the ruler to find the magnification:

The ruler shows 7mm, hence the magnification is 1:0.4, or 2.5x:

The working distance is 2cm, which is quite short. I try to photograph a LEGO figure here, and as you can see, the subject is almost touching the rear end of the lens:

Here is the result, at 14mm f/3.5. I focused on the eye, but due to the very thin DOF, only one eye is in focus:


Here is the photo of the ruler, showing a 1:0.87 or 1.15x magnification ratio:

At 42mm, the working distance is 4cm:

And here is the result, at 42mm f/5.6:

Example image

This is a picture a tooth brush, and was taken at 14mm, f/14. I could probably have stopped down the lens even more for more depth of focus.

Also note the dots throughout the image. They are probably dust particles on the sensor, and come into focus when using extremely small apertures, like in this case. Otherwise, they don't cause any negative effects.

Mounting a Lumix G 45-200mm f/4-5.6 tele zoom lens

The Lumix G 45-200mm tele zoom lens has a 52mm front lens filter thread, and hence mounts directly to the adapter.

However, when mounted reversed, it focuses near infinity at 45mm, and nowhere at longer focal lengths. Hence, it is useless for reverse macro use.

Mounting other lenses

Both these lenses have a front thread of 52mm, making them easy to mount. But let's say you have the Lumix G 20mm f/1.7, with a front thread of 46mm. What to do?

You can get a step up ring. Mount a 46mm-52mm step up ring to the front of the lens, and then mount the front of the step up ring to the adapter. A step up ring is very cheap to get on various auction sites.

Take care that mounting a wide attachment to the Lumix G 20mm f/1.7 lens can cause the focus mechanism to jam. So don't power the lens with the step up ring attached.


Using the reverser ring adapter, the Lumix G 14-42mm lens can be mounted reversed, and the zoom ring can be used for focusing. The magnification rate is 1.15x-2.5x, which is not a very large range. For comparison, Canon has a specialized macro lens called Canon 65mm MP-E, with a macro focus range of 1x-5x.

Changing the aperture is not very easy with this lens, but it is needed. You cannot photograph macro images using the lens wide open, since the DOF is too thin. To stop down the lens, you can mount it to a camera, and start a long exposure at f/8, 4s, for example. As the camera is exposing, remove the lens. That way, the lens is stopped down to f/8 when you remove it. Do this at your own risk.

This is a cheap way to achieve macro possibilities with the kit lens, a lens that quite many users probably have already. But is is not very easy to use. First of all, the working distance is very short, so you cannot photograph live insects. They will be scared off.

Second, changing the aperture is awkward. And third, the macro range is limited, at 1.15x to 2.5x.

But if you want to experiment with macro images of static objects, than this is a very inexpensive way to do so.

Thursday, 23 February 2012

Fisheye lenses, different projections?

The Samyang 7.5mm f/3.5 fisheye lens is an interesting addition to the Micro Four Thirds lineup. While most third party Micro Four Thirds lenses so far have been existing manual lens designs given a new mount, this lens is designed for the Micro Four Thirds format from ground up.

How can I tell? A full frame fisheye lens has 180° diagonal field of view. Hence, the lens must match the sensor size exactly. If the lens was designed for APS-C, a larger format, then the corners would not correspond to 180° angle of view. A fisheye lens which does not project to 180° in the corners is pretty much useless. Then it is just a wide angle lens with a lot of geometric distortion.

Compared with the existing Lumix G 8mm f/3.5 fisheye lens, the Samyang lens is a much more traditional design, with a manual focus ring and aperture ring.

Some say the Samyang lens features a different projection type, the Stereographic projection. This is supposed to be less distorted than the Spherical projection traditionally associated with fisheye lenses. Let's look into how their projections differ. Here is a picture taken with both lenses, and also using the Olympus Zuiko Digital 9-18mm Four Thirds wide angle zoom at 9mm:

Samyang 7.5mm
Lumix G 8mm
Olympus 9-18mm @ 9mm
By superimposing both the fisheye images in one image, and doing edge detection, we can see how their projections compare:
And let's look at another example:

Samyang 7.5mm
Lumix G 8mm
Olympus 9-18mm @ 9mm
Superimposing the two gives:

(Click for larger images.)


So, what is the conclusion of all this? First of all, from the second example, we observe that both lenses have pretty much the same diagonal field of view. This experiment does not verify that the diagonal field of view is exactly 180°, as it should be, but at least both lenses have around the same maximum diagonal angle. To be more precise, the Samyang lens appears to render a slightly wider diagonal field of view. This might be due to the shorter focal length, 7.5mm versus 8mm.

Regarding the distortion, we can see that the Samyang lens renders objects which are inside the border of the image a little bit smaller. This means that it distorts the images somewhat less. So the rumor is true: The Samyang lens does give less "fisheye distortion".

But surely, the differences are pretty marginal. You're not likely to notice much difference, unless comparing head to head, as I do in this article. So if you're looking at the Samyang 7.5mm lens to avoid the fisheye distortion, you are going to be disappointed.

One thing to note is that if you plan to convert to rectilinear images in post processing, the Samyang lens has the potential for giving you the widest possible rectilinear images of the two fisheye lenses. This process is called defishing, and you can read about the topic here. It is probably not true that the Samyang lens features stereographic projection. It still has fisheye projection, but with slightly less distortion than the Lumix G 8mm f/3.5 fisheye lens.

Saturday, 11 February 2012

What can we expect from the GH3?

The Panasonic GH series has comprised the premium camera models for Micro Four Thirds. The Panasonic GH1 and GH2 have had the best video quality among mirrorless cameras at the same time, and they are also considered to have the best sensor for photos. They are not intended to be volume models, but rather to sell at a premium price for those who want the best camera.

Panasonic GH1 (left) and GH2 (right)

The Panasonic GH2 was announced in September 2010. I bought it the very first day it was available in my market, which was in December 2010. Hence, this camera is starting to get old. In this article, I would like to speculate a bit about what we can expect from the GH3.

Timing of announcement

To speculate about when the GH3 will be announced, it is always useful to look back in time, and see when previous camera models were announced.

In the timeline above, the important tradeshows (PMA and Fotokina) are shown as vertical lines. We see that some camera models were announced in connection with these shows, e.g., the G1, GH1, G2 and GH2. However, after 2011, it seems that the announcements are less connected with the tradeshows. The PMA 2011 was cancelled, and the Panasonic G3 would probably have been announced at this event, if it wasn't cancelled.

Some have speculated that the Panasonic GH3 will not be announced until Fotokina 2012, in September. However, I think that is too late. Panasonic cannot have a top model which is two years old, I think. So the GH3 must be announced before September 2012, I think. The distance between the GH1 and GH2 announcements was one and a half year, hence, I think the GH3 will be announced around March 2012. (After writing this, it has since become clear that the GH3 was not announced around March 2012.  So now I am expecting an announcement at Fotokina 2012, late September.)


And what can we expect from the GH3? The GH series will aim to be seen as the king of the hill in terms of mirrorless cameras. That was fairly easy for the GH1 in March 2009, since the competition did not have any cameras out. Now, however, there is a lot of competition.


The GH2 increased the effective megapixels from 12 to 16. Will the GH3 top this off with even more megapixels? Probably, yes. In an interview, some Panasonic staff were quoted to say that the GH2 got 16 megapixels for marketing reasons, mostly. And the market is back to a megapixel war, sadly. The Sony NEX 7, announced in August 2011, has a staggering 24 megapixels. I don't think the GH3 will top this, but it probably needs to close some of the gap to be seen as a serious competitor in the market. So I would guess the GH3 will be announced with around 20 megapixels.

If true, this is mostly bad news. I don't think many need more than 12 megapixels anyway, and increasing the megapixel count even more probably means that other aspects of the sensor cannot be optimal, e.g., noise characteristics and dynamic range.  But we see a trend towards more megapixels across many segments now.  The new Nikon D800 will have a staggering 36 MP on a full frame sensor.  The Nikon 1 system is the only good news in this respect.  They "only" give 10 MP, which is probably a sensible figure given their smaller sensor area.

Video modes

Both the GH1 and GH2 supplied full HD video modes with 1080 lines. The GH1 could only do this in interlaced mode, and the GH2 improved upon this by adding progressive modes.

The next big thing in video is 4K. Not very well defined, but 4K means that the horizontal resolution is around 4000 pixels. With the 1080p modes of the GH2, the horizontal resolution is only 1920 pixels. So 4K is a big step up from the GH2.

Personally, I don't think the new camera will provide 4K. It is simply a too big leap at this point in time. The new Nikon D4 flagship model does not provide 4K, and still has 1080p as the largest video mode. Rather, I think the GH3 will further refine the video at 1080p, and possibly add higher frames per second modes.


All Micro Four Thirds cameras so far have used the Contrast Detection Auto Focus (CDAF) method. This involves jogging the lens focus back and forth until the contrast is deemed high enough to assume that the focus is optimal. With the current cameras and lenses, CDAF have given fast enough autofocus for still images, even in fairly dim lightning.

This is different from DSLR cameras, which use Phase Detection Auto Focus (PDAF). PDAF is better at following moving targets, which CDAF currently does not do very well.

The Nikon 1 system combines the two focus methods, and employs a hybrid CDAF + PDAF method. In theory, this should be able to give better focus for moving subjects, and also to keep the focus better during video recording. Canon's new mirrorless system, the Canon EOS M, also mixes CDAF and PDAF.

The GH2, while currently probably still the best Micro Four Thirds camera for video, does not do moving targets very well, neither in photo or video mode. So how can the GH3 improve upon this? My guess is that the improvements will be incremental. I think that adding PDAF, like the Nikon 1 system does, is a too large step at this time. There may also be patent issues hindering Panasonic from implementing this. So my guess is that the GH3 will employ faster and better image processing to improve on the autofocus during video recording, but it will not employ any dramatic change to the focus technology.

Global shutter

All Micro Four Thirds cameras so far use a "rolling shutter". This means reading off the sensor values sequentially vertically. For still images, this is just fine, but if the subject moves during exposure, it could give some strange artifacts. We can see this as buildings "leaning" to one side if you pan horizontally during video capture, or rotating propellers warp in a strange way. I have compared the rolling shutter artifacts of the GH1 and GH2 here.

To fix this problem, Panasonic could implement the "Global Shutter". This would involve reading off all sensor values at the same time. Implementing this would mean that they could drop the mechanical shutter. It exists mostly to avoid the rolling shutter artifacts for still images. Hence, it would make the camera less noisy, less expensive to manufacture, and less prone to mechanical failure. Global Shutter would be a huge improvement to the camera.

So can Panasonic implement Global Shutter for the GH3? In an interview before the GH2 was launched, a Panasonic employee was quoted to say that Global Shutter could not be implemented until at earliest in the GH3.

My guess is that a true Global Shutter can not be expected with the GH3. But perhaps they can make the rolling shutter quicker, and avoid most of the artifacts so that dropping the mechanical shutter becomes possible.

For comparison, the Nikon J1 mirrorless only has an electronic shutter, while the Nikon V1 has a mechanical shutter which can be used at the operator's discretion. Using the electronic shutter is silent, but can yield rolling shutter artifacts. The mechanical shutter can be used for faster flash sync speeds.


The GH2 improved upon the ergonomics of the GH1 quite a bit. For example, it moved the control wheel from the front to the rear, and added a focus lever on the left shoulder. I would not expect any significant changes to the ergonomics with the GH3 model. There are not many major ergonomic issues with the GH2 anyway.

Body materials

Both the GH1 and GH2 employ a plastic shell over a stainless steel frame. Currently, there is a trend towards using more metal as the shell material. For example, the recently announced Olympus OM-D E-M5 has a metal exterior, as do most of the competitor system's cameras. Personally, I think that a metal exterior for such a small camera only serves to give a "premium feeling", and not provide any real value beyond that.

So I would hope to see that the GH3 retains the plastic bodies of the predecessors. But just as the megapixel count, the body material might be important from a market perspective. Perhaps it becomes impossible to sell a plastic body camera, in which case Panasonic must change their strategy.

Weather sealing

Pro spec cameras commonly have some kind of environmental sealing, to keep out water splashes and dust. I have used my GH1 and GH2 in light rain, and not noticed any problems. But to use them in challenging environmental situations is probably not a good idea.

If Panasonic add real weather sealing to the GH3, then that is a good way to make it stand out in the crowd. That would be a step towards a more pro camera.

Other functions

The LCD display of the GH2 has 460.000 dots, which is not very impressive. The predecessor GH1 has the same LCD resolution. I think Panasonic need to up this resolution to be competitive. On the other hand, all the more recent Panasonic models have featured the same resolution, GF3, G3 and GX1. So perhaps they think it is still relevant. Using the EVF gives you better resolution, so that can be done for more critical focus control with manual lenses, for example.

The Panasonic GH2 has horribly slow buffer clearing speed when photographing RAW images. To be seen as a serious contender, I think they must improve the write speed for RAW images.

Due to the short register distance, mirrorless cameras are popular for using older SLR lenses in manual focus mode. For easier manual focusing, some have requested a "focus peaking" functionality, which would highlight parts of the image where the contrast is large, and hence, the image is in focus. If the GH3 adds this feature, then it would be easier to use with older manual focus lenses.


This article has been mostly speculations, of course. My guess is that the GH3 will present some evolutionary changes to the GH2, just as the GH2 did when it was announced in September 2010.

On the other hand, if the GH3 does not offer 4K video, no major improvements to autofocus during video, no global shutter, then how can Panasonic still claim that it is a premium video enabled model? It could be that they need to step up and give at least one of these features, to differentiate against the other M4/3 camera, and, not least, the competition.

Adding weather sealing could be another way to signify that this is the top model. Since Olympus has announced the OM-D E-M5 with weather sealing, Panasonic cannot claim that the GH3 is a pemium model without this feature.


Since writing this article, it has become clear that the camera was not announced during spring 2012. So now, I'm expecting an announcement at the Fotokina 2012, late September.

Wednesday, 8 February 2012

Canon G1X, sensor size the same as GH1

The Canon PowerShot G1 X was announced recently, and took a lot of people by surprise. While Nikon have launched their mirrorless, large sensor cameras in the Nikon 1 series, not many rumors existed about Canons mirrorless competitor. It turns out that their answer was a large sensor compact camera, without interchangeable lenses.

The sensor size was also a bit surprising: Larger than Four Thirds, but smaller than APS-C. This sounds like an odd sensor format. But is the sensor format really that odd? Let's try to have a closer look at it.

We know that the Canon G1X sensor is reported to be 18.7mm x 14mm. The Four Thirds sensor size, on the other hand, is somewhat smaller at 17.3mm x 13mm.

But the Panasonic GH1 and GH2 have employed oversized Four Thirds sensors. This design choice was implemented to get the same diagonal field of view in the 16:9 and 4:3 modes, utilizing the full image circle also for video. The Panasonic GH1 has 4000x3000 pixels in 4:3 mode, and is 4352 pixels wide in 16:9 mode. This means that the width of the sensor must be 17.3mm x 4352 / 4000 = 18.8mm. Which is rather close to the reported width of the Canon G1X sensor.  The differences in reported size might be due to rounding off differently.

What about the vertical size? The GH1 sensor is reported to have 14 megapixels in total, which corresponds to a vertical resolution of 3217 pixels. Hence, the vertical size must be 13mm x 3217 / 3000 = 13.9mm. Again this is very close to the reported height of the Canon G1X sensor.

And the total resolution of the Canon sensor? 4352x3264. Again, this is very close to the GH1 horizontal resolution times the estimated vertical resolution based on the total 14 megapixel resolution. This can be illustrated like this:

Both sensors are of the CMOS type, and both have a base ISO of 100.

My conclusion is that the specifications of the Canon PowerShot G1 X sensor are remarkably similar to the Panasonic GH1 sensor. This doesn't mean that they are identical, though. But given that there are not that many sensor manufacturers out there, and certainly not for the Four Thirds sensor size, I would say this is a good indication that Panasonic do in fact produce the sensors for the Canons G1X. As you understand, this is of course purely speculation.

A fundamental difference between the two cameras is that the Panasonic GH1 only gives you a maximum resolution of 12 megapixels for one single exposure, with the option of changing the aspect ratio with the same image circle. 

The Canon G1X, on the other hand, gives the full sensor resolution output for a single exposure.  However, when using video (16:9 aspect ratio), only a smaller image circle is used, meaning that the effective focal length changes.