Sunday, 30 January 2011

Comparison: Lumix 14mm vs Lumix 20mm pancake lenses

The Lumix G 14mm f/2.5 and Lumix G 20mm f/1.7 pancake lenses are similar, yet still very different. They are of course similar since they share a pancake characteristic: They are both very compact. They are shown below, with "home made" lens hoods:


(From the left: 20mm and 14mm.)

Still, there are many issues which make them different. The Lumix G 14mm f/2.5 has internal focusing, which makes the autofocus very fast and silent.


The Lumix G 20mm f/1.7, on the other hand, has a more traditional focus assembly, which moves all the lens elements back and forth. This is not as fast, and does generate some noise. This is not an issue in daylight, since both are very fast in sufficient light. But in low light, the 20mm lens can take in excess of one second to focus, which can be annoying.

And low light brings us to another area where the lenses are different: The 20mm lens is a true low light lens, with a maximum aperture of f/1.7. The 14mm, still gains about one stop of light gathering compared with the kit lens, but it is not at all a low light lens.

And, of course, the field of view is quite different: The Lumix G 14mm f/2.5 is a wide angle lens, and the Lumix G 20mm f/1.7 is more of a "normal" lens, which can be used for, e.g., environmental portraits in low light without a flash.

In terms of pricing, they tend to sell for approximately the same amount. The 14mm lens was included as a kit lens for the Panasonix GF3 camera, and these kits did not sell well. They were often split, and the lens sold separately off auction sites, which lowered the perceived value of the lens for some time.






Sharpness

What about the sharpness? The general opinion is that the 20mm lens is very sharp in the centre, even wide open at f/1.7. It does require stopping down to at least f/2.8 to get sharp corners, thought. When it comes to the 14mm lens, people generally say that it is not as sharp as the 20mm lens.

I prefer to find out for myself, so I made a field study. I put the Panasonic Lumix GH2 camera on a tripod, set it at base ISO (160), and used 2 second shutter delay to avoid camera shake.

I took the same picture using both lenses, at different apertures. I used the out of camera JPEG images. The shutter speed was always 1/100 second or faster. Below are the full images, scaled down and resharpened. Both images were taken at max aperture.



Lumix G 14mm f/2.5
Lumix G 20mm f/1.7

To better compare the sharpness, I have cut out 100% crops from the images. These crops were not sharpened. Here are some comparisons. These are from the centre of the images:



And here are crops from the top left corner:



Click to see larger versions of the images.

Sharpness Conclusion

First of all, we can conclude that the centre images are boring: They are virtually the same from max aperture down to f/5.6. They show that there is little to gain by stopping down the aperture when it comes to the centre resolution. Perhaps we can say that the Lumix 20mm lens is a tad bit softer at f/1.7 than f/2. But the difference is very subtle.

In the corner, though, there are more issues to comment. The Lumix 14mm lens does sharpen up a bit when stopping down, and appears to reach an optimal aperture around f/4. Stopping down further to f/5.6 does not appear to give better performance.

The Lumix 20mm lens appears to sharpen up quicker when stopping down. f/2.8 appears to give a sufficient sharpness, bit there is a tad bit of improvement also when going to f/4.

What about comparing between the two lenses? In the centre, I would say they are equally sharp. There is little to complain about in terms of sharpness at any of the aperture values.

In the corner, though, it seems that the 20mm lens is a bit sharper overall. Also, the 14mm lens has some purple fringing artifacts around the branches, which cannot be found in the 20mm corner images.

When it comes to vignetting, the 14mm lens again appears to have somewhat more vignetting wide open. You must close down to around f/3.5 to lose the vignetting, but it is not a huge problem even at larger apertures.

Optically, it seems that the 20mm lens still has an edge over the 14mm lens, especially in the corners. However, unless you are very critical, and make huge enlargements, I don't think any of the lenses will disappoint in terms of optical performance.

Chromatic aberrations

When using these lenses on Panasonic Lumix G cameras, the JPEG output images are automatically corrected for some Chromatic Aberration artifacts, like red/green fringing around high contrast areas, especially in the corners of the frame.

As at the current date, Olympus cameras do not employ this CA adjustment.

Based on my examination of the CA artifacts, these lenses do not generate a significant amount of them anyway. So even without the in camera CA correction, CA artifacts are not a significant problem.

Some purple fringing are left after the in camera processing, as we can see in these images as well.

Geometric distortion correction

Perfect rectilinear projection is one of the traditional quality indicators of lenses. If the lens gives a pincushion or barrel distortion, then that is commonly interpreted as a sign that the lens design is bad.

Both of these lenses give a significant barrel distortion without any post processing. Bear in mind, though, that this processing is done automatically in the camera, so that the JPEG images come out looking rectilinear. And when using RAW, most RAW converters will apply the geometric distortion correction seamlessly. So the user might very well never notice that the lenses feature significant distortion.

Using some third party RAW converters, it is possible to look at the images prior to the distortion correction. This reveals the true nature of the distortion properties of the lens. Below are images of a tiled wall. The black lines show how the image looked after the automatic in camera correction, while the red lines illustrate how the camera sensor actually saw the scene:



Lumix G 14mm f/2.5
Lumix G 20mm f/1.7

Read more about this study here. The 14mm lens has somewhat more barrel distortion. To get a rectilinear image, the 14mm lens requires a correction of -16% in the "Lens Distortion" filter in The Gimp, while the 20mm lens requires -11%.

What this means, is that the 20mm lens wastes less pixels in the corners of the image frame, and, potentially, can give slightly better corner sharpness. On the other hand, this effect is rather subtle, and for any real life application, I'd say you can basically ignore it.

Further, the 14mm lens does not correct enough for distortion at close focus distances. Hence, while you get good, rectilinear images at moderate to far focus, you'll get some small amount of barrel distortion at close focus. Again, this is not a problem for most real life usages, but it may be worth to note that this is not a lens for close focus reproduction of art, for example. The 20mm lens, on the other hand, is well corrected for all focus distances. I would guess that this difference is due to the internal focus of the 14mm lens, which is known to produce different geometric distortion properties at different focus lengths.

The lenses are designed to require post processing for a reason: Lens design is a matter of balancing various optical properties against each others. With this choice, the lens designers can focus improving the artifact that cannot be corrected in post processing, while leaving the geometric distortion to be adjusted in post. This can, potentially, lead to smaller lenses with better quality.

Bokeh

I have made a comparison of the out of focus highlights rendering for both lenses. The study shows that neither lens has a "perfect" bokeh. They exhibit various problems, for example non-circular out of focus highlight discs, ringing, dirty and uneven bokeh. See another bokeh comparison here, which has the same conclusion.

However, you must focus quite close in order for these problems to show. When photographing people, you will normally want to keep a distance of one meter or more to avoid perspective distortion, and the bokeh should not be a problem with this distance.

Field of view

Obviously, the 14mm lens has a wider field of view than the 20mm lens. The 14mm lens is a wide angle lens, while the 20mm lens is what people would normally call a "normal" lens. Normal lenses have a focal length which correspond roughly to the diameter of the sensor. The Four Thirds sensor diagonal measures 21.6mm, so the 20mm lens is in fact a slightly wide normal lens.

Based on the field of view difference, which is quite significant, which lens would you want to buy? Experienced photographers will probably not ponder long about this. They are already well aware of the concepts "wide angle" and "normal lens", and know their preferences. What about the rest of us?

If you have used the kit zoom lens for some time, you could take a look at your favourite photos and see what focal length they were taken with. Did you typically use the wide end of the zoom lens? Or the longer end? The answer here might determine your focal length preference.

There is a philosophy which goes like this: You can always get closer to an object, but you cannot always get further away from it. So to be able to photograph what you want, choose the widest lens. In this case, this philosophy dictates that you choose the Lumix G 14mm f/2.5 lens over the Lumix G 20mm f/1.7 lens, since the former is wider.

However, it doesn't take much thinking to see that the premises are not always right. Let's say you want to photograph people. Then, you should not get closer to them than around 1 meter. Going closer will give you perspective distortion, which can make the photo unflattering.

Hence, if you intend to photograph a person, and want to have their face as the main part of the image, you will want to choose the longer lens. At a 1 meter distance, their face will be just a small spot in the frame with the 14mm wide angle lens. Even the 20mm lens is not long enough to be a portrait lens, but it is still the better choice. For a portrait headshot, you will generally want a focal length of around 40mm or higher. But the 20mm lens can be used to take an environmental portrait.

On the other hand, if you intend to photograph a group of people, you will want to choose the wide angle lens. You cannot always back up more, so the widest lens is best to cover a group of people.

Aperture range

We have already discussed the different maximum apeture. The 20mm f/1.7 lens has the larger maximum aperture, obviously. However, the 20mm lens also has the larger minimum aperture. Here are the ranges.

14mm: f/2.5 - f/22
20mm: f/1.7 - f/16

The smaller possible minimum aperture for the 14mm lens is an advantage when shooting video. Generally, one would not want to have too fast shutter speed when recording a video.

For motion pictures, a 180° shutter is commonly used. This means that the shutter is open half the time. If you have 30 frames per second, this means that the shutter speed should be 1/60 second.

When recording a video outdoors on a sunny day, you may need to close down the aperture a lot to achieve 1/60 second shutter speed. In that case, the f/22 option comes handy with the 14mm lens. Otherwise, you may need to use an ND filter to get the right shutter speed.

Common knowledge says that you should avoid using small apertures, due to diffraction. Diffraction is known to blur the image at pixel level when using very small apertures. However, when shooting video, the resolution used is smaller, so I don't think diffraction is any problem. In fact, one could say that diffraction acts as an extra anti aliasing filter, which could actually improve the image quality during video capture.

Video use

Both lenses work perfectly fine with video. However, the quick and virtually noiseless autofocus of the Lumix G 14mm f/2.5 pancake lens makes it preferable for general video use.

In low light situations, you could find that when using the Lumix G 20mm f/1.7 you can lose focus for some seconds when there is movement in the scene. This is not so likely to happen with the 14mm lens, in my experience.

Apart from the autofocus differences, the choice between the two lenses largely comes down to the same issues whether you intend to use them for video or photo: The field of view and the maximum aperture. So the considerations in the rest of the article apply just as well for video use.

Here are some example videos.

Low light video with some action using the Lumix G 14mm f/2.5 pancake lens on a GH2:



More information about the video parameters used in the movie above. You'll notice that the audio quality is poor in the video. However, this is due to the sound system, which clips the sound at high levels.


This video showing the ice breaking up in Stockholm was recorded using a Lumix G 14mm f/2.5 pancake lens on a GH2, at 1080p, 24fps:




A low light concert movie using the Lumix G 20mm f/1.7 pancake lens on a GH2:



More information about the video parameters used.

The following video was recorded outdoors using the Lumix 20mm f/1.7 lens on a GH1. You'll see that the camera loses focus now and then, which is a bit annoying. Both the lens and the camera have had firmware updates since this video was recorded, and the autofocus performance during video has improved.



More information about the video parameters used.

Compared with the Sigma 19mm f/2.8 EX DN

In 2012, Sigma released their first Micro Four Thirds lenses, the Sigma 19mm f/2.8 EX DN and the Sigma 30mm f/2.8 EX DN. The 19mm lens is quite similar with the Lumix G 20mm f/1.7 pancake lens, so it makes sense to compare them.

See my main comparison article here. A brief summary: The Sigma lens focuses much quicker, and more silently. It also has a short startup delay, just like the Sigma 30mm lens, and it rattles when not in use. The rattle is no problem, it can just be a bit annoying.

The Sigma 19mm lens has the most pleasing bokeh. It is also cheaper.

In terms of image quality, I think it is clear that the Lumix G 20mm f/1.7 lens is the better. The 20mm lens also has a larger maximum aperture, and a smaller size.

I think that reasons for buying the Sigma 19mm lens over the Lumix G 20mm lens could be to save money, and to get better autofocus performance, especially during video recording.

After just a year, Sigma discontinued the Sigma 19mm f/2.8 EX DN lens, but introduced a new version at the same time. The new version has a different exterior design, but other than that employs the same optical layout, and, hence, the same image quality. The new lens retails for a bit more than the old one did before being discontinued. I guess that Sigma thinks the new metal exterior appears more desirable, and allows them to charge a premium price:



New version of the 20mm lens

In the summer 2013, this lens was discontinued, and a new version of the lens, Lumix G 20mm f/1.7 II (H-HS020A) was released. The new version has the same basic specifications, and has the same optical design. The exterior design is new, though, with a black or silver metal finish.

As the new lens has the same optical design, it still has the old style focus assembly which moves all the lenses back and forth. Reports indicate that the focus speed is the same as the first one, i.e., not very impressive. Even with the new design of the lens, the autofocus is still the slowest among the Micro Four Thirds lenses.

So the only reason to buy the new version of the lens would be if you prefer the new design to the old one.

The new designs of the Panasonic Lumix G 20mm f/1.7 II:


Conclusion

It seems to me that the Lumix G 20mm f/1.7 is valuable as a sharp, low light lens, while the main benefits of the Lumix G 14mm f/2.5 are the very compact size and the fast autofocus. Both lenses are optically very good, but the 14mm lens, lacking the true low light capability, is not as interesting. From my perspective, anyway.

This note is written six months later: After using the Lumix G 14mm f/2.5 lens a lot, I have come to like it more and more. Now, I use it more than the Lumix G 20mm f/1.7 lens.

The reasons for liking it more are the same as I have written above: Fast and silent autofocus, small size, very good optical qualities. It's also very good for video, due to the AF performance and silence. Besides, the field of view is generally quite useful when photographing and videographing people.






Appendix

To make the comparison of the images easier, I have applied auto levels to each row. That way, the exposures are more comparable.

The centre of the images:



The corner of the images:

Monday, 24 January 2011

Wide angle: 8mm vs 9mm

The Micro Four Thirds format is blessed with a number of compatible very wide angle lenses. There are:

Lumix G 8mm f/3.5 Fisheye

Lumix G 7-14mm f/4

Olympus M.Zuiko 9-18mm f/4-5.6

The Olympus Zuiko 9-18mm f/4-5.6 Four Thirds lens can also be used on Micro Four Thirds cameras, given that you have the appropriate adapter, e.g., Panasonic DMW-MA1, Olympus MMF1 or Olympus MMF2. These are all functionally similar. This lens will autofocus on Micro Four Thirds cameras, but the focus can be a bit slow.

In this article, I am comparing the first and the last on this list. Here's a picture of them both:


The Olympus Zuiko 9-18mm f/4-5.6 (left) is shown without the appropriate adapter. Mounting the adapter will add 18.67mm length, since that is the difference between the register distance of the two formats.

The Panasonic Lumix G 8mm f/3.5 fisheye is a truly compact lens. Olympus has a similar lens on their 2011 roadmap, and it remains to see how compact it will become.

Field of view

At 9mm focal length, the Olympus lens has a diagonal field of view of 100º. The 8mm fisheye lens, on the other hand, has a diagonal field of view of a whopping 180º! How can one mm difference in focal length make up such a massive difference in field of view?


The answer is that the projection is different in the two lenses. Projection in this case refers to the mapping of the real world objects in three dimensional space, down to the image sensor and two dimensions.

Most photographic lenses feature a rectilinear projection. This is what we have become used to. A rectilinear lens will produce an image where straight lines in the real world object are straight also in the resulting image.

Fisheye lenses are fundamentally different. With a fisheye lens, only straight lines the pass through the image centre are straight. All other lines will be bent. There is a significant amount of barrel distortion.

Within fisheye lenses, there can also be variations. Circular fisheye lenses will give a disc of exposure. A 180º view in all directions is mapped into a single disc, and the rest of the sensor frame is left black.


Full frame fisheyes are perhaps more common. They feature a 180º only in the diagonal, and otherwise fill out the entire sensor area. The Lumix G 8mm f/3.5 is a full frame fisheye lens.  To most users, these are more useful, as they give a rectangular image, as we are used to.

Example images

To further illustrate the difference between a rectilinear wide angle lens and a fisheye lens, let's look at an example. The images below were taken at base ISO, and on a tripod.



Olympus 9-18mm @ 9mm f/4
Lumix G 8mm Fisheye @ f/3.5


Olympus 9-18mm @ 9mm f/8
Lumix G 8mm Fisheye @ f/8

As you can see, the fisheye image is wider, and also features significant barrel distortion. Straight lines in the real objects are bent in the depiction.

This is not really the right type of image to evaluate the vignetting, but it seems that the 8mm fisheye lens vignettes a bit more at f/3.5. However, with such a wide angle of view, it is unlikely that you have the same tone across the field anyway, so I cannot see that vignetting will be a significant issue with this lens.

Note that the light source, the setting sun, is in the middle of the frame.  Both lenses handle this fairly well. There is not a big amount of flare or lack of contrast caused by the light source in the centre of the frame.

Sharpness

To evaluate the sharpness, let's look at 100% crops from various parts of the image. These images have not been sharpened. Click for a larger version.

Here are crops from the centre of the frame:


We see quite clearly that the Lumix 8mm Fisheye is the sharpest lens, straight from wide open at f/3.5.

Corners and Chromatic Aberration

From these border crops, we see basically the same thing. We can see some softness in the 8mm Fisheye image at f/3.5, but it sharpens up well at f/5.6.


Also, we see some Chromatic Aberration lens distortion artifacts in the Olympus images.  There's the red and green fringing off high contrast areas.  These artifacts typically appear near the borders, and become more significant the further away from the image centre you get.  This can be corrected pretty well by software, so it's not a big issue.

Panasonic lenses are automatically corrected for Chromatic Aberration (CA) distortion during the in-camera image processing, when using Panasonic cameras. So the JPEG out of camera images I have used could have been corrected for these effects, which may be why we don't see any CA in the Fisheye images.

To examine the effects of the automatic CA correction in the 8mm fisheye lens, let's look at one example. This picture of the Morris Jumel Mansion was taken with the GH1 and the Lumix G 8mm f/3.5 Fisheye:


Here are 100% crops from the extreme top right corner, and from the middle right border:

We see that the corrected image still shows some colour fringing artifacts in the extreme corner, but they are mostly gone in the border crop. From the rest of the frame, i.e., not the extreme borders, you will be hard pressed to find any CA artifacts in the out of camera JPEG image.

From the RAW images which have not been corrected for CA distortion, we see that there are some CA artifacts. These fringes are about 1-2 pixels wide in the extreme corner, which is not very significant. For comparison, the colour fringes are about 2-3 pixels wide in the images from the Olympus Zuiko 9-18mm f/4-5.6 wide zoom lens.

It's also fair to comment that the image from the Lumix 8mm fisheye is remarkably sharp in the extreme corner. Keep in mind that the corner is at a 90º angle from the optical axis.

Defishing

It is possible to convert the fisheye image to a normal image. This process is usually refered to as defishing the image.

Many different programs allow this kind of transformations. I have used a program called Hugin for this purpose.

Below is a comparison of the original images taken with the rectilinear lens at 9mm, the fisheye, and, in the bottom row, the fisheye image converted to rectilinear.



Olympus 9-18mm @ 9mm f/8
Lumix G 8mm Fisheye @ f/8


Fisheye image converted to rectilinear
Fisheye image converted to rectilinear and cropped

The original fisheye image can be stretched to look fairly similar to the rectilinear image. But it has an even wider field of view. I would say the difference in field of view is significant.

This defishing process is hardly optimal, though. The corners have been stretched, and hence lack some resolution compared with the centre of the image. Also, it is difficult to frame the image correctly if you intend to defish it later. But having the option to defish the image makes the fisheye lens more useful.

Here is another example of defishing. The original image is from the Apple Center in New York, Manhattan:


After defishing, it looks like this. There is still a bit of barrel distortion, which could have been removed with some tweaking of the parameters. You can see that the image is less sharp in the corners, due to the stretching needed in the defishing process.


Another note is that the aspect ratio changes when defishing the image. The original fisheye images were taken with a 4:3 aspect ratio, while the defished image has a much wider aspect ratio, closer to 16:9.

On first inspection, this might look like a mistake. However, it does actually make sense. A fisheye lens creates an image where the field of view is not constant across the frame. What I mean is that the field of view is more compressed in the corners than in the centre. Hence, the ratio of horizontal to vertical field of view becomes larger than that of the original image.

When using the 4:3 aspect ratio, the output image of course has a 4:3 ratio in terms of pixels.  However, due to the compressed field of view in the corners, the horizontal to vertical field of view ratio is 124º:92º. 

Conclusion

In concluding, it is clear that the Lumix G 8mm f/3.5 Fisheye lens is better optically than the Olympus 9-18mm zoom. The Lumix 8mm fisheye appears to vignette a bit with wide open aperture, but I can't see that being a big problem.

On the other hand, the 9-18mm zoom is more versatile. In the longer end, it gives a pretty normal field of view, and can be used for general photography. In the wide end, it is an extreme wide angle lens. It can be used to make stunning and interesting wide angle images, as well as pictures with a more normal perspective.

The fisheye lens is an exotic lens, and is not always easy to use. When you nail an interesting image with the fisheye lens, it can be very rewarding. But many pictures end up looking just hideous, or like clichés. It is a lens with a required taste. Given the high price, I would not recommend buying it unless you know what you are doing.

Finally, the 8mm fisheye lens focuses much faster and more silently on Micro Four Thirds cameras.

Of course, nobody with a Micro Four Thirds camera should buy the Olympus Zuiko 9-18mm f/4-5.6 Four Thirds lens, since it requires and adapter, and is much larger than the M.Zuiko Micro Four Thirds version of the lens.  From what I have read, the M4/3 version has comparable optical qualities, and focuses much faster.  It is also collapsible, and much lighter and more compact.

Sunday, 23 January 2011

GH2: Face recognition works with toys

Like many other CDAF cameras, the Panasonic GH2 has face recognition. This means that it will find out where faces are in the image frame, and will prioritize focusing on them. One fun fact is that it also works with toy faces.



In the video, I have put the Leica Lumix 45mm f/2.8 macro lens on the GH2 camera. This is the only macro lens (at this time), which can autofocus reasonably fast with the GH2. The Olympus Zuiko Digital 50mm f/2 macro also autofocuses on the GH2, however, the focus is very slow.

You'll see that in the video, the camera starts focusing straight away. This is because I set the Pre-AF mode to AFQ. In this mode, the camera will start focusing when the image is stable, assuming that you have framed your subject and are ready to release the shutter. Used this way, the Pre-AF mode let's you get your picture faster, since the camera has already focused.

Note that the Pre-AF mode only works when there is a generous amount of light. In dim indoor lightning, it does nothing at all. You will have to half press the shutter release as usual for focus when the lightning is dim.

Friday, 21 January 2011

Moon photography and GH2 ETC

The Panasonic GH2 introduced a new concept: Extended Tele Conversion (ETC). This has been much hyped, and perhaps a bit misunderstood.

Will ETC give you a 2.6 tele conversion for any lens? So the Lumix G 20mm f/1.7 becomes a 52mm f/1.7 lens? The much anticipated portrait prime lens? The short answer is: Sort of, but only for video.

What ETC does, is simply to record the pixels in the centre, rather than the whole sensor area. It is, quite simply, a digital zoom. Digital zooms have been around for ages.

So why is this big news? Because it also works for video. If you record a video at full 1920x1080 resolution with the GH2, it is sampling the pixels from the whole 4976x2800 sensor area, and downscaling it to 1920x1080. In ETC mode, it simply samples the 1920x1080 pixel area in the centre of the sensor, and does no downscaling at all. So you still get full resolution, but with a 2.6 crop factor. The crop factor is calculated by dividing 4976 by 1920. Here's an illustration:


This means that when recording movies at full HD resolution, you can use ETC to convert the Lumix G 45-200mm f/4-5.6 zoom lens into a 117-520mm f/4-5.6 lens. In terms of a traditional 35mm film camera, that is equivalent to 234-1040mm range, since the Four Thirds format has a 2x crop factor itself.

My answer above was "sort of". The reason for that is that digital noise will be somewhat larger in ETC mode, especially with higher ISO values. This is because there is no downscaling. When using the whole image sensor, the camera can use more pixels to compose the video output, to avoid noise. However, when using only the centre pixels, downscaling is not possible. So the noise level is typically higher. I have made a study which shows that there is indeed more noise, and some less sharpness and contrast when using the ETC mode.

Here is a video recording of the full moon, using the Lumix G 45-200mm f/4-5.6 zoom lens at 200mm f/7.1. The camera was put on a tripod, and I used spot autofocus. I used ISO 250, 1/400s exposure. I used the ETC mode.



You can see that the moon moves in the frame. This is simply due to the rotation of the earth.

To see the effect of not using the ETC mode, here is a still image taken with the same image parameters, but without ETC:


Without the digital zoom from the ETC mode, the moon is just a small disk in the centre of the image, even when using a 200mm tele lens.

To capture an image in which the moon makes up a significant part of the frame, you'll need a very long lens. Let's say you want the moon diagonal to take up half of the Four Thirds sensor diagonal. In that case, you'll need a lens which is 600mm. Such a lens is currently not available for the Micro Four Thirds format. Panasonic and Olympus both produce tele zooms that extend to 300mm, but that's still only half way.

I found the exposure parameters by using spot exposure metering off the moon itself. On the other hand, when I used the normal exposure metering mode, I got this image:


In this case, the moon is just an overexposed blob with no features at all. We also see why there is some lack of sharpness in the moon images: There was in fact a cloud cover. But the cloud was too dark to be seen in the image when metering off the moon.

The difference in the exposure is quite huge, around 8 stops:

Spot metering off the moon: f/7.1, ISO 250, 1/400s (13EV)
Center Weighted: f/7.1, ISO 3200, 1/30s (5EV)

One interesting fact is that the moon spot metering exposure is virtually identical to ISO 100, f/16, 1/60. And why is this significant? Because ISO 100, f/16, 1/60 is known as the Sunny 16 rule. Basically, it says that when exposing objects lit directly by the sun, you can use that exposure as a rule of thumb. And the moon is also an object lit by the sun. It is just a bit farther away from us than what we are used to photographing.

So, the Sunny 16 rule also works for the full moon. But you may still want to expose a little bit more. After all, you don't want the moon to be average gray, you want it to be a bit bright. So you could adapt a similar Loony 11 rule.

Pentax lens design patents related to M4/3?

In 2008, Pentax filed patents for some lens designs that have an image circle similar to that of the Four Thirds format. There has been some speculation as to what these lenses might be used for.

Some have speculated that Pentax might revive the 110 format, but for digital mirrorless interchangeable lens cameras. This format used negatives of the size 17x13mm (21.0mm diagonal), while the Four Thirds format has an active sensor area of 17.3x13.0mm (21.6mm diagonal). So they are virtually identical. Here is an illustration:


The lens designs Pentax patented were: 17mm f/2.8, 17mm f/2 and 14mm f/2.8. From a Micro Four Thirds perspective, these sound like sensible wide angle lenses.

Some have speculated that these designs were licensed to Panasonic and/or Olympus for use in their Micro Four Thirds lens lineup. To evaluate this, we can take a look at the latter: It has some resemblance to the Panasonic Lumix G 14mm f/2.5 pancake lens. We remember that when Panasonic first announced this lens, it had the specifications 14mm f/2.8. It was later changed to f/2.5.

Comparing the lens design layouts, we see the following:


Are these lens designs similar? I don't really have the right competence to answer this myself, but a simple inspection reveals that they are in fact quite similar. They have basically the same lens elements and groups. So it is not inconceivable that the Lumix G 14mm f/2.5 is based on the Pentax patent. But this is far from concluding evidence.

What about the Olympus M.Zuiko 17mm f/2.8 lens? Could it be based on the Pentax patent for a 17mm f/2.8 lens? Not really, is my guess. Their lens designs differ quite a bit.

Sunday, 16 January 2011

GH2: Magnified focus with legacy lenses

The Panasonic Lumix GH2 represents an evolution from the original GH1, which not only includes better video quality, but also better ergonomics. One example is the possibility to use the click wheel to select magnified focus assist mode when using legacy lenses. With the GH1, this mode was available through two button presses: Left arrow key followed by the down arrow key.

On the GH2, it is much simpler. Normally the rear control wheel controls the aperture (in A mode), and the exposure compensation. Switching between these two is done by pressing the click wheel.


When using a legacy lens, however, the camera cannot control the aperture anyway, so the default function of the control wheel is to adjust the exposure compensation. However, Panasonic has added a new function on the GH2: Pressing the control wheel once brings up the magnified focus assist view. You can exit the magnified view by pressing the control wheel again, or by half pressing the shutter release button.

I will demonstrate how to do this.

To use a legacy lens, you will first need to make sure the camera operates when a non-compatible lens is attached. This is done by setting the "Shoot without lens" option. You'll find this option under the custom options (with the "C" and wrench symbol). Here is the location in the menu:


As an example, I am using the manual focus Nikkor 50mm f/1.8 AIS lens. This is a classic, cheap and light lens. It has a pancake characteristic, however, with the adapter it is still not very compact.

To attach such a lens, you will also need a Nikon to M4/3 adapter. Here's how to attach the adapter and the lens to the GH2 camera:



In the next demonstration video, you can see that I use the magnified focus assist view. In this mode, rotating the control wheel changes between 5x and 10x magnification. You can also use the arrow keys to move the enlarged area.



The magnified view can also be brought up by pressing the touch screen. However, since I usually have the thumb near the rear click wheel, I find it much easier to use this function for focus assist.

When focusing, you will normally want to use a large aperture. E.g., f/1.8 for the Nikkor 50mm f/1.8 AIS in my example. It is easiest to verify that the focus is correct when the aperture is large, since the depth of focus is the smallest.

On the other hand, when actually taking the photo, you'll often want to close down the aperture for extra sharpness. Hence, you end up adjusting the aperture ring on the lens back and forth all the time.

Virtually all cameras do this process automatically. But when using legacy lenses on a Micro Four Thirds camera, you must set the largest aperture manually when focusing, and then stop down the aperture before taking the picture.

Sunday, 9 January 2011

Concert video with GH2 and Lumix 20mm

Here is an example concert video recorded using the Panasonic Lumix GH2, and the Lumix G 20mm f/1.7 lens. I recorded the movie in 1080i, 50fps. However, it was scaled down to 720p, 25fps before uploading to YouTube.

I recorded by pressing the red video button on the top camera plate. This gives automatic handling of most imaging parameters. Sadly, these parameters are not visible during video recording, so it is not possible to see what aperture, ISO, and shutter speed the camera has chosen. However, since this was recorded during dim light, it is natural to guess that the aperture was the maximum, f/1.7. The shutter was probably around 1/60 second, and I guess the ISO was around 1600.



One difference between the GH1 and the GH2 is that the latter supports manually adjusting the microphone recording levels. In this video, I let the camera set the levels automatically, which worked well. I think the sound quality is good.

I used the free software Openshot to edit the video. The editing in this case was limited to cutting off the start and end, as well as the downscaling to 720p.¨

The camera was handheld. To get some more stability, I had the strap around my neck, and rested my hands on the camera. By using the tiltable LCD display, I can record the video without getting much attention at all.

Compared with a similar recording using the GH1 and the Lumix 20mm lens outdoors, this camera manages to keep the focus pretty well throughout the video recording.

The band is the Frank Znort Quartet, playing "Mink Schmink".

Sunday, 2 January 2011

AF comparison: Lumix 20mm vs Sigma 30mm

I previously compared the Lumix 20mm f/1.7 pancake lens with the Sigma 30mm f/1.4. The Sigma lens was put on a Pentax K10D. While the comparison is not entirely fair, since the K10D is an older camera, my conclusion was that the Lumix 20mm lens was a better performer.

In this article, I am comparing the lenses again. I set the Lumix 20mm pancake lens on a Panasonic GH2 camera, and the Sigma 30mm f/1.4 lens goes on a Pentax K10D camera.

The Sigma 30mm f/1.4 (left), and the Lumix G 20mm f/1.7

Here you can see both cameras set on tripods, pointing towards a LEGO figure on a table. Both cameras are set to f/2.8, and base ISO. I raised the built in flash on both cameras, and set the initial focus to infinity. The focus mode was centre point.

Here's a video showing both cameras' shutter releases being pushed fully at the same time:

video

Not surprisingly, the Panasonic setup fires off first. It has the fastest autofocus.

With the Panasonic camera, focus was reached after 0.4 seconds, and the pre-flash was fired after an additional 0.17 seconds. There is a pre-flash to measure the required intensity of the main flash, since the TTL flash metering does not work in the same way as with older SLR film cameras. The main flash was fired after a total of 0.77 seconds since the shutter release button was pressed the first time.

With the Pentax camera, I don't have the measurement of the time until focus was reached, since this is not displayed. But the main flash went off after 1.10 seconds. The pre-flash was fired around 0.1 seconds before the main flash.

Now, let me be the first to say that this comparison is unfair. The Pentax K10D camera is from 2006, while the Panasonic GH2 was launched in 2010. Also, I'm using a third party lens on the Pentax camera, which is known to not always give the best AF performance.

The Sigma lens is operating closer to it's minimum focus distance, which is 40cm. The Lumix 20mm lens, on the other hand, has a more generous minimum focus distance of 20cm.

Let's look at the images. Here are the full images, scaled down and resharpened a bit:

Lumix 20mm, f/2.8, base ISOSigma 30mm, f/2.8, base ISO

To better compare the images, here are 100% views from the centre, directly from the out of camera JPEG files. These are unsharpened. Click for a larger view.


It is apparent that the the Lumix 20mm lens is much sharper. Also, the Panasonic exposure is better. This is a tricky situation for a DSLR to expose, anyway. The camera has no way of knowing that the white surface is supposed to be white, and not grey. So in this case, I should have helped the camera by saying that it should over expose. The Panasonic camera sees the whole picture prior to making the exposure, and the automatic exposure is clever enough to see that the surface should in fact be white.

When looking at the 100% crops from the two different cameras, one could conclude that the Pentax/Sigma image is unsharp due to front-focusing or back-focusing. I.e., that the AF sensor or lens is not calibrated correctly. Since I am stopping down to f/2.8, though, I doubt that poorly calibrated focus is the reason for the unsharp image. But perhaps it is, and this kind of uncertainty is common when using DSLRs. Thankfully you don't need to worry about it with Micro Four Thirds.

Micro Four Thirds cameras use Contrast Detection Autofocus (CDAF) to verify the correctness of the focus, and it is generally more accurate than Phase Detection Autofocus (PDAF), used on DSLR cameras.

I am confident that the Pentax/Sigma combination focused on the LEGO figure, and not on the background. If I set the camera with the figure slightly off-centre, it would not focus at all, since the white table is too even and glossy.

Conclusions

This study reveals several interesting issues. First of all, we see again that the the Lumix G 20mm f/1.7 pancake lens is very sharp. It's autofocus is not quite as fast as other Micro Four Thirds lenses, however, it is still fast enough for most uses. All in all, it is a very good lens, especially for environmental portraits.

Again, I see that the Sigma 30mm f/1.4 is not very sharp. This disappoints me, since it is otherwise a very interesting combination of focal length and aperture for use on an APS-C DSLR. Some might suggest that my camera/lens combination requires calibration. Perhaps they do, but again, this just illustrates one of the drawbacks with DSLR cameras and PDAF technology.

Some would say that the Sigma 30mm f/1.4 lens is not intended to be very sharp anyway. Rather, it is designed for available light photography of people, to be used with apertures around f/1.4-f/2, say. In these settings, the quality of the bokeh (out of focus rendering) is also important. One could say that the Lumix 20mm lens is more of a generalist lens. Still, I think it's bokeh is usually adequate.

The autofocus performance of the Pentax/Sigma combo was better than I had expected. On the other hand, it is also significantly more noisy. Keep in mind that the Lumix 20mm lens is one of the most noisy Micro Four Thirds lenses I have tried, but the Pentax/Sigma still makes much more noise.

The GH2 camera has a flash lag of about 1/4 to 1/3 second. This is due to the fact that the TTL flash metering does not work during the main flash, like with older SLR film cameras. So to figure out how large the main flash exposure needs to be, the camera does a pre-flash. After having evaluated the pre-flash exposure, the main flash is fired.

1/3 second flash delay is in fact significant, and this is one of the drawbacks of Micro Four Thirds. Note that this is not just related to the built in flash. An external TTL flash, like the Panasonic FL360 also has the flash delay in TTL mode. To avoid the flash delay, you will need a flash that can do traditional flash auto mode. This is a bit more cumbersome to use, and may require some calibration from situation to situation. But an advantage is that you can use just about any flash for this purpose, even if it is an older flash from a different system. Just be sure the flash trigger voltage is not so high that it fries your camera.

Note that the flash delay is not exclusive to Micro Four Thirds. DSLRs also have it, like the Pentax K10D in my example above. However, since DSLRs will use the exposure sensor for the flash evaluation, rather than the main imaging sensor, the pre flash delay will usually be quite short.

Saturday, 1 January 2011

New year's eve fireworks with GH2, 8mm Fisheye

Using the Panasonic GH2 and the Lumix G 8mm f/3.5 fisheye lens, I made a video recording of the new year's eve fireworks.

Since the light was very dim, I set the ISO to 3200, and used the lens wide open at f/3.5. Setting the mode dial to "creative movie mode", and using manual exposure and manual focus, it is possible to set the shutter speed to slower than 1/30 second, which is the usual limit when recording videos. To get a sufficient exposure, I set the shutter speed to 1/4 second. Of course, this means that the camera only records about four frames per second.

The camera was handheld. With such a wide field of view, it is fairly easy to handhold the camera stably.

To get more "action" in the clip, I set the speed to 300%, meaning that it is three times faster than normal.



The Panasonic GH1 can also record videos with a slow shutter speed. The process for using this feature is not well documented, but you can follow the instructions in the linked article.