Skip to content

A couple of days ago I acquired an old Canon A1 SLR camera. Someone was selling off old photo equipment on Ebay and so with the camera came a small collection of four lenses. One of them is a Tokina 135mm f2.8 telephoto. It's quite an impressive item, looking much sturdier mechanically than its Canon FD counterpart. It's made of all aluminium, even the aperture ring is made of metal. It feels quite heavy in your hand and appears to have been well treated since there is no apparent sign of wear. Since I also own the Canon lens, I started wondering which of the two to sell. Casual shooting with both lenses on the G1 didn't reveal major differences in image quality so I did a series of shots under more challenging conditions.

The following photos show performance of both lenses while taking images of a church in my home town against the bright sky before sunset. The sun is hidden behind a cloud layer just outside to the right of the frame on the G1, on a 35mm film camera it would still be inside the frame.

The photos taken with the Tokina are always on the left side. They start at f8 and go up to f2.8, the maximum aperture for both. The lenses feature a built-in hood, it was used while taking the photos but of course for a factor-two crop sensor it's not deep enough. The images were taken as RAW files and then developed without any sharpening or other corrections applied. They were downscaled for presentation and 100% crops of two interesting areas patched on.

Tokina f8 Canon f8

As you can see, at f8 there is not much difference between both lenses. If you look at the 100% crops, there is a slight advantage for the Canon, but it's not obvious. There is a difference in color tone, but I cannot say if its a difference of the lens or just subtle changes in the lighting conditions.

Tokina f5.6 Canon f5.6

At f5.6 the difference is more noticeable. While the Canon lens shows almost no degradation in sharpness and contrast, the photo taken with the Tokina is becoming a little soft with a very slight amount of CA, but it's still pretty good.

Tokina f4 Canon f4

The trend continues at f4, the gap between Tokina and Canon is getting wider. The Canon now starts to show a little softness and a bit of CA, but its overall still pretty good.

Tokina f2.8 Canon f2.8

At f2.8, the Tokina shows a significant loss of contrast and increased softness. Color seams appear around back-lit areas. The Canon has also lost some of the sharpness and contrast it displays at smaller apertures, but its a lot better than its Tokina counterpart.

Summing up, I found the Tokina lens to be performing quite well under most conditions. A challenging environment like strong light sources just outside of the frame requires stopping down to f4 or f5.6 for good image quality. The Canon 135mm however proves to be an excellent lens with good results even wide open.

I remember I didn't write anything about my Easter holidays yet. I went skiing in the Italian Dolomites, to a nice little town in the province of Trento named Moena. Its in the middle of Fassa Valley, a valley in the region of Trentino-Alto Adige where people speak an ancient Rhaeto-Romance language called Ladin. I know this region since I was a kid, we used to go skiing there with all of our family, quite regularly for more than ten years. Moena lies at the foot of the Cantinaccio mountain range, home to the legendary dwarven king Laurin, who was said to having kept a garden of beautiful roses high up near the mountain top. Stories tell he fell in love with a beautiful woman and abducted her. He was defeated by the womans brother and a band of brave knights and taken away into custody, but as a last action he put a ban on his rose garden turning it invisible day and night. But he forgot about dusk, which is not day or night and thus you can sometimes see the mountains glow red of roses in the last rays of daylight when the time is right. So they say.

Easter was late this year but reportedly it had been snowing heavily throughout most of March and April and the conditions were said to be quite good so I took the chance to go there and spend a week skiing and trying my new camera equipment. I had recently bought a Lowe Pro Slingshot 200 back pack that is big enough to hold all my gear, the G1, 14-45, 45-200 and the Canon FD 50 1.4 lens, together with a small tripod and cables, filters, charger, etc. I didn't take the Metz flashgun with me, I have found that it's a part of my equipment I almost never use and since it adds considerable weight I left it at home. I didn't miss it.

I wasn't going there alone. We were a group of around 15 people and we had rented a house for the whole group. A couple of them I knew already from earlier trips to Moena, some of them even from the time I went there together with my parents and brother. The area of Trevalli (Three Valleys) offers quite a number of skiing opportunities. If you have a car, the selection of resorts within 20 minutes travel distance is quite impressive. The image to the left was taken from the top of Col Margherita, which is part of the San Pellegrino skiing resort. Looking at the pictures its hard to believe it was already beginning of April. Not often have I seen this amount of snow that late. However, the warmth of the approaching Spring time was busy melting the snow and even though there was lots of it, its quality degraded massively during the day. So we got up early every day and tried to be on the ski piste soon. That worked rather well, by the time the snow turned soft around 3pm we usually had enough for the day and enough time left to chill, drinking marvelous Italian coffee and eating ice cream.

All the time I carried the Slingshot pack on my back. It was working really well on the piste, I almost didn't feel the weight of the gear and when it was time to enter the lift I could just swing it around from back to front and sit down comfortably. My sun glasses turned out to be a bigger problem, I had to take them off for taking pictures together with my gloves. The viewfinder worked pretty well under the bright sun, I just had to shield it a bit with my hand. The LCD turned out to be mostly useless. I tried to take photos with all the lenses I had, most of them with the 14-45, but some with the 45-200 as well. Sometimes I used a polarizer filter to improve the sky colors and cut down the haze a bit, but still I had to post process most of the images to correct white balance and levels. I used only digikam for raw development and corrections and it handled the task pretty well. I had also taken the MSI U100 netbook with me which allowed me to download the images every evening and inspect them.

I've uploaded a selection of images taken during my stay to Flickr, there's a set named Dolomites. All in all I'm quite satisfied with how the G1 performed. The only thing I'd have wanted is a bit more of dynamic range. One stop more might have made a difference in some situations.


Old Gear, too
Originally uploaded by thinkfat

Today a big jumble sale of photo equipment was held at a university facility, it takes place only once a year and it was the first time I went there. It was quite amazing, the mass of gear for sale was huge. Wide selection of stuff from all old and recent camera makers. The selection of Canon FD lenses was not too rich, though.

Still, I was quite successful and got a 28mm f/2.8 for very little money and a 50mm f/3.5 macro lens in very good condition for considerably less than the cheapest ebay offer.

I was tempted to buy a 24mm f/2 and had another good offer for a 24mm f/2.8, but the difference in speed and field of view was too little against the 28mm I already had in my bag and I declined. Instead I took the 50mm f/3.5 macro, I think it will be nice for flower shots and might work as a portrait lens as well.

The 28mm is a particularly nice lens. It's very light and the effective focal length of 56mm is close to a typical film standard prime. I think I'll like it. The 50mm macro is a bit long physically, but also quite light. Not something to have on the camera all the time, but still something to have in your photo bag.

Some first images with both lenses can be found in my flickr photo stream.

Looks like there is some movement on the Z21 front again. Matthew Garret posted a patch to the linux-acpi mailing list with an rfkill based implementation of dealing with wireless devices. Interestingly, his patch also enables all the hotkey events so now I even get an ACPI event when I flip the speed/stamina switch. But still no idea about how to enable speed mode safely.

Eva made a new release with the rfkill changes integrated, which means you can now control all your wireless devices for even more power saving. Grab it from the usual place.

8

I mentioned in an earlier post that I was intending to go to the bottom of certain rumors about Panasonic including lens defect correction data in the EXIF information of RW2 files. A number of commercial photo software programs are known to manipulate the RAW image data to remove geometry distortion, chromatic aberration and vignetting silently, which lead to enthusiastic reviews posted on certain digital photography sites about the quality of the Lumix G1 kit lens. However, I know from developing my RAW files with ufraw and digikam that the kit lense has a significant barrel distortion at short focal lengths.

One way to automatically correct those distortions is collecting correction data in a database like PTLens and Lensfun. The photo software knows what corrections to apply to images taken by a certain camera with a certain lens at a certain focal length. That approach however requires significant effort for maintaining the database plus you will always be playing catch up with the camera manufacturer developing new lenses. Not an issue for independent open source efforts, we've been running after information from vendors ever since, but not a good situation for the camera vendor. Unless he runs after the software companies to update their databases whenever he releases a new lens, images developed from his cameras RAW files will be not optimal. I would not want that, especially not if I wanted to reduce development cost and retail price for new products.

Panasonic most likely went the second way, which is having software makers implement the same correction algorithms they use in their cameras and embedding the necessary correction parameters inside the photos meta data. Doing it that way allows for lesser synchronization effort between Panasonic and software vendors and faster time to market. With some effort it might be possible to use this correction data also in open source software.

That being said, there's now a couple of tasks to be done until the correction data can be used:

  1. find out what information is stored, and where
  2. reverse engineer the format of the correction parameters
  3. understand how it's applied to image data
  4. proof of concept implementation

I believe I've made some progress today on 1.), finding out what information is stored and where. I inspected some of my RW2 files to see what Exif tags they contain and found a couple of tags that are not decoded by exiftool or exiv2. They just contain some data bytes in unspecified format. I noticed that they seem to vary from file to file and decided to find out if I could reproduce the values.

Reverse engineering is a black box approach. In our case the camera is the black box, inputs are the controls like aperture, focal length, exposure program etc, output are the generated RW2 files. You cannot observe the inner workings of the black box other than by observing differences in the output when making modifications to the input. It definitely helps if you can make an educated guess about the inner workings of the black box. For geometry correction, I know a bit about how its corrected, so that is my first target.

I know that geometry distortion depends on the focal length, so I made a couple of shots with the kit lens at 14mm and 45mm, two at each. Then I looked at the Exif data and found one tag, "Exif.PanasonicRaw.0x0119", where the data did not change for images taken at 14mm, but changed significantly when going to 45mm.

14mm:

84 236 201 47 38 0 0 0 43 1 0 0 141 1 1 0 208 14 238 1 86 2 2 252 196 9 226 3 228 72 36 134

45mm:

152 94 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 86 2 0 0 196 9 0 0 35 234 24 155

Looking at the data, you see something interesting - a lot of Zeros appearing at 45mm. Nothing changes from 45mm to 35mm, which looks like this:

35mm:

161 54 43 194 51 0 0 0 0 0 0 0 0 0 1 0 42 0 1 0 86 2 1 0 196 9 245 1 176 16 17 65

Still a lot of Zeros, but lesser. What does this tell us? Now, I know that the geometry of the lens is almost perfect at 45mm, so I expect that there is less correction necessary. It's reasonable to assume that some parameters of the correction formula will become very small, or zero. The data in the Exif tag apparently follows our expectations. I also assume that the data is not encrypted.

The next task is now to find out how the data is formatted. Most likely these parameters are rational numbers. What I don't know yet is the representation. Is it 16, 32 or 64 bit per coefficient? Am I looking at fixed or floating point numbers, little or big endian? Does this tag only contain correction parameters for geometry, or also for chromatic aberration? The tag contains 32 bytes, so it's either 16, 8, or 4 coefficients. I'll have to do some research on geometry correction, maybe Panasonic uses something that is public domain.

It looks like a prominent method of correcting lens geometry is based on Zernike polynomials. The PanoTools lens model uses polynomials as well. One property of polynomials is that you can use as many parameters as you like, it just depends on the order of the polynomial. It's merely a question of computing power available. I could start with a 2nd or 4th order polynomial like the PanoTools model and try to match the tag information with coefficients computed by Hugin.


Ready!
Originally uploaded by thinkfat

Since we got our new digital camera I have been using it quite regularly and I got very much interested again in digital photography and everything around it. I used to shoot lots of photos when I inherited my grandfathers Cosina SLR, despite the tedious act of having the film developed and paying for prints just to throw lots of them away.

I don't quite remember why I never picked up a digital SLR camera when they became fashionable, but they were quite expensive back then and anyway I spent most of my money (and most of my time) on amateur radio equipment. Today the situation has changed. DSLR cameras today are mass market products, not just for photography enthusiasts.

So, I familiarized with the Lumix G1 quickly and today I carry it with me almost constantly. The photo in this blog post links back to my flickr account, you're welcome to have a look at some results of my early ventures into digital photography.

Interestingly, this camera got me back to KDE programming again. Of course I started using various 'K' programs like KPhotoAlbum or DigiKam to keep track of my growing collection. And sure enough, once you start using a 'K' program, you find lots of small annoyances that demand fixing. So I hacked kflickr to upload resized pictures with an adjustable quality instead of some default value, digikam to recognize the lens description of the Lumix G1, the flickr export plugin to allow uploading to the photostream only instead of requiring a set and I'm now trying to automatize lens defect correction using information embedded in the RW2 files this camera generates when shooting 'RAW'.


p1020151.jpeg
Originally uploaded by thinkfat

If the motorbike virus has you, somehow it's not a question of convenience any more to go for a quick ride or not. So right after lunch today we took out our bikes for the first time this year. It was sure a bit chilly still at 5°C but with the right clothing it was bearable. We drove up to Neunkirchen, had a coffee and apple pie and went back home again.

I also took the opportunity to see if the camera setup I envisioned for our summer vacations would work out, and I'm happy to report that the Lumix G1 fits nicely into the tank pack as planned and is easy enough to operate while on the bike.

I also experimented with the geotagging feature of digikam, which worked quite nicely. I downloaded the trip log from my GPS after the ride and told digikam to correlate the images I had taken. It added a set of tags to the EXIF information that show the location where the photo was taken - well, to be precise, the location where the GPS receiver was when the photo was taken, but since it's mounted on the bike, that's not going to be too far off.

Now if I only could get flickr to consistenly import the location data. It's working only sometimes and I don't know why.

As Christmas approaches quickly, my wife and I are trying to figure out what presents to give to friends and family, and to ourselves. We figured it'd be time to bite the bullet and get ourselves a new digital camera. We own a rather old Casio 3 MP compact DSC that has served us well for a couple of years but now the case is seriously starting to fall apart. The latch of the battery housing lost its lock and now it doesn't want to keep shut and can only be held in place with some duct tape. Anyway it's a pretty old camera, slow and with very limited features, its only advantage being, well, it's compact.

Now, while we could just get a similar DSC to replace it, unfortunately we both are a bit gadget crazy. Also, my wife owns an old Minolta SLR that makes awesome pictures but could be replaced by something more fancy. It's quite battered, the case is cracked a bit so she put some black insulating tape over the crack to keep light from coming in. And now we've got to decide what we want to get.

The options are close to infinite. Should we go for a real DSLR or is a "bridge" camera with a 20x Megazoom the better choice? And since we also love travelling with our motorbikes and taking pictures on the road, wouldn't a point-and-shoot compact camera be a much better idea? This weekend I've looked at a lot of reviews and narrowed down the options to only a few, but it's still tough.

For a compact DSC, the Canon PowerShot G10 could be a good choice. But it's got a too small sensor with way to many pixels (14.7MP) and quite consequently it suffers from noise a lot. Looking at the sample pictures the reviewer took, already at ISO 200 you can see the effect of noise filtering done by the camera, taking away sharpness and details. However, there's the predecessor G9 with a slightly "smaller" sensor at 12.1MP which has better noise figures. The G9 or G10 would be the ideal travel companion on the motorbike. Still, it's not an adequate replacement for the Minolta.

We know from experience that a small SLR camera can be handled easy enough even while sitting on the motorbike, balancing the bike with one foot on the ground and the other foot on the brake pedal, even with the helmet on. You just use a wide angle lens and point the camera to the approximate direction and press the trigger. A life preview like on compact DSCs helps but is not strictly mandatory. Changing lenses to go from wide angle to zoom is out of question, however.

This is where we enter the domain of Megazoom bridge cameras. I looked at two models, the Nikon Coolpix P80 and the Canon PowerShot SX10IS. The fixed lense ranges from wide angle to extreme zoom and their sizes are similar to compact SLR cameras. But again, a small sensor and high pixel count drive up the noise figures. The SX10IS is much better than the P80 there, but its image quality is just about par to the compact PowerShot G10, this leaves the long zoom as the main advantage. But again, on a motorbike its very difficult to use a long zoom even in good light.

The Panasonic Lumix G1 is an interesting candidate. It's a so called "Micro FourThirds" camera with a big 4-by-3 sensor at 12.1MP. It's not a real DSLR, because there is no mirror box and hence no optical view finder. But leaving out the mirror box and using a electronic view finder allows to make a very compact case. Not that compact however. It's not much smaller than compact DSLRs, but it's got a twist-able live preview display which is an advantage over regular DSLRs. The image quality is stunning! Due to the big sensor the noise figures are much better than of any of the MegaZoom or 10+ MP compact cameras and the quality of the lense is much better as well. The lenses are however also the weak spot: They are exchangeable and currently there is just two options, a variable 14-45 mm and a 45-200 mm. I already stated that changing lenses on a motorbike is a no-go. But I also said that a long zoom is difficult to handle anyway and looking at our trusty Minolta, we find that it also has a vario lense with 28-70 mm focal length, which is about equivalent to the 14-45 of the G1. And we know from experience that a compact SLR works well with the bikes.

Enters another contender, the Nikon D40. This entry level DSLR has a 6.1MP Nikon DX sensor(crop factor 1.5) and is amazingly cheap to have. The noise figure of this camera is really, really good. It doesn't even feature any ISO below 200, it's just not necessary! You can expect it taking very good pictures especially under low light conditions, think sightseeing and not needing to carry a flash for photographing in churches, crypts or in fading daylight. It's also quite well established and has a wide range of lenses to chose from. It's currently sold as a kit of body and 18-55 mm plus 55-200 mm vario lenses. Cons: it's larger than the Lumix G1 and doesn't have a life view display. On the positive side there's the image quality and the wide variety of lenses, making it the perfect replacement for the old Minolta.

It's still a tough choice. On one side compact cameras can be carried wherever you go and taking out the camera to shoot is quick and easy. But then, my wife also owns a Nokia N95 with a built-in 5MP autofocus camera, we have used it on our last summer holiday trip with good results. So is there really a need for another compact DSC? MegaZoom cameras are very versatile due to the wide zoom range. However the image quality appears to suffer from the technical constraints and the long zoom is no advantage while on the motorbike. Still, in the coming days we'll also look at new model, the Canon PowerShot SX1, but it's hard to tell if the use of a CMOS sensor can do anything about the image quality. The Micro FourThird Panasonic Lumix G1 is quite a cool device. The electronic view finder is stunning, image noise figures are not so bad and the quality of the lenses is very good. But will it survive as a concept so that more lense options will be available eventually? Experts seem to be undecided. And since it's not really small, why not go all the way and buy a real DSLR like the Nikon D40?