The Rathaus GigaPan

Posted on June 9, 2017 by Brian Lawler

I have a GigaPan device. It’s a computerized camera mount that moves a camera to take a photo, then moves the camera, and takes a photo, and continues doing that until a series of images are taken that complete a very large panoramic image.

I have done a lot of these over the years, and I am very fond of the device (I am not fond of carrying it and its tripod).

This is a screen shot of the GigaPan Stitch software with my 992 photo panoramic image ready to stitch. Completing the image took hours. Completing the project (retouching, uploading, etc.) took 15 hours, some of which I spent sleeping while the machine carried on without me.

The largest photo I have ever made was at Cal Poly, very near my home in San Luis Obispo. I climbed to the top of Cal Poly’s local mountain (unnamed, above the big “P” on campus) and I made a 360-degree panoramic image with the GigaPan rig that eventually stitched into an image that was too large for Photoshop to open (Photoshop’s limit is 300,000 pixels in either dimension). I wrote a blog about that. Over the years I have made numerous other GigaPan images that are high in resolution, but not so high as to break the software.

On Wednesday this week my students and I descended upon the Marienplatz in the center of Munich to make a number of photos. I brought the GigaPan and I positioned it in a fourth floor window opposite the city’s famous Rathaus, a building that was built in 1867 (and onward) in the city center.

Here is my GigaPan on its Really Right Stuff carbon-fiber tripod in the window of the business on the Marienplatz in Munich. My camera bag, with four other lenses in it is hanging on the hook in the center of the tripod to prevent it from moving.

The building is a landmark partly because of its famous glockenspiel, which plays a musical and animated program at 11:00, 12:00 and 5:00 each day. Hundreds of people stand in the plaza each day to witness this charming presentation.

My objective was to make the highest resolution photo ever taken of the building, something I believe I have now done (I welcome anyone to challenge this claim).

Several weeks ago I contacted a company that has its offices in the building opposite the Rathaus and asked for permission to use a window on their floor of the building for a few hours while I took my photos. They were open to this idea, and welcomed me into their lovely office.

The crowd in Marienplatz waits for the glockenspiel to make its musical presentation. Up on the 4th floor, circled, is my GigaPan and camera. It took more than 90 minutes for each of my two GigaPan photo series on this day. I captured a total of more than 2,000 frames – each 63 MB in size when open.

On Wednesday I arrived, three tripods, three cameras and two backpacks in hand. I set up in the window of their offices and put my GigaPan rig in the window. Two of the tripod legs were outside the window frame on the narrow ledge and one was inside, which was a bit scary. I really didn’t want to see my equipment topple into the square, so I weighed the tripod down with my camera case (there is a hook in the center of my Really Right Stuff tripod that works well for this), and this made me more confident that it was safe.

Meanwhile, my students took the other tripods and cameras to make street-level panoramic images with those. We were experimenting with a new 3-stop neutral density filter on one camera. It looks almost opaque, and cuts the light coming through by half of half of half (which is 1/8). The reason this is a valuable accessory is that it allowed us to make 30-second exposures in broad daylight. By doing this, most of the people on the plaza disappeared from the photos, making almost eery images of the buildings without crowds of people nearby (More on that in future blog).

This is a look out of the window where I was working. My students are down there in the center-right with another two cameras and tripods, taking street-level panoramas and time-lapse videos.

Up in the fourth floor window I set up the camera and the GigaPan. My lens choice was the 100-400mm Canon telephoto. I set the zoom to 300mm, which was appropriate for this shot. At that focal length, the photos for the image would be close enough to zoom in and see tremendous detail, and it would allow me to finish the project in a reasonable amount of time.

The GigaPan does its work by calculating the lens’s field of view in degrees, then building-in overlaps between adjacent images so that the stitching program can put them together successfully. It asks you to position the image in your viewfinder at two spots, and by motor movement, it calculates the angles. I performed this setting, and it calculated the field of view at 4.1 degrees vertical.

After you establish the field of view, the GigaPan instructs you to identify the upper-left corner of your image, then the lower-right corner. I did both, and the GigaPan determined that I would need 992 images to complete my shot. The physical limitations were the frame of the window in which I was working, otherwise I had an unhindered view of the plaza.

This is a reduced-resolution view of the final GigaPan image. There is mayhem in the plaza in the image because people were not photographed in their entirety (the camera was making multiple passes horizontally). As a result, there are many partial people in the image, but no one was injured in the making of this image! I promise.

The GigaPan then steps you through a series of questions to stop you from making dumb mistakes: Is the camera on? Is the zoom locked? Is focus locked? Is exposure locked? Is the strobe off? Are you sure? Yes, I indicated, and it began its work. The LCD display counts-off the photos as they are taken, and it indicates how much time remains. I needed one hour and 21 minutes to make this series of shots. The sky was cloudy with blue showing-through in patches.

About twenty minutes into the series, the sky turned dark gray and it rained. People ran from the plaza and took cover under the umbrellas of stores along the Marienplatz. About ten minutes later the sun came out again and the sky turned to a lighter gray. The photos continued all the while.

At the end of the first series of photos the sky turned a beautiful blue with scattered white clouds. I decided to do it over, so I started it again; that is the image I ended up using. In the end I captured over 2,000 photos of the Rathaus in two series.

After I completed the photos, my students and I took some group photos in the plaza then headed back to Hochschule München to begin work on the images.

The taking of GigaPan images is the easy part; the processing of GigaPan images is most tedious. Back in California I have a Mac Pro cylinder computer with six processors and 128 GB of RAM and a GPU, all of which combine to make processing of projects like this go very fast. Here in Germany I am using a MacBook Air with an Intel 1.8 GHz i7 processor and 4GB of RAM. It’s a nice lap-top computer, but it is certainly not as fast as my machine at home. This was going to take some time.

Some of my photography students in the Marienplatz at the end of our photography day there. We were quite successful with what we set out to accomplish.

In the classroom we worked on the street-level panoramas, which will turn out beautifully. I took the GigaPan photos back to my apartment for a long night of processing.

Just downloading 2,000 camera Raw photos to the machine took several hours. I use Adobe Bridge and its Get Images from Camera tool to import, rename, and simultaneously convert from Canon CR2 to Adobe DNG. This takes time, but I prefer this work flow because it is obsolescence-proof. I watched a documentary about things that are underneath London (Netflix) while this was going on. Then I watched a second documentary about zzzzzz (I fell asleep) until the job was done.

GigaPan does not know what a camera Raw file is, so in order to do the stitching I must convert all the images from DNG to TIFF in order to stitch them (cynics are sniping right now saying, “Why don’t you just use JPEG? It’s just as good and it wouldn’t require all that work!”). They are wrong; JPEG images seriously compromise the quality of the gradients in blue skies. The damage is clearly visible and it shows in these ultra-high resolution images. I prefer quality over efficiency (and I don’t think of JPEG as efficient) so I spend the time necessary to do the job right.

Converting from DNG to TIFF took about one hour using the Image Processor in Photoshop (thank you, Russell Brown!). The folder containing the images consumes 23.4 GB of disk space (just the DNG raw files).

So, late, late last night I had my 992 photos in TIFF ready to stitch. I set up the software and told it to start work. Five hours later I awoke and it was still working. One hour after that it was finished.

The image stitched beautifully. The final steps take a long time also: first one must export the GigaPan from the stitching program to the local drive. GigaPan creates a Raw file that Photoshop can read; it’s an RGB interlaced image. In this case the image is 124,076 x 84,884 pixels. Exporting that image took about three hours.

I also save my GigaPan images and their work files (there are thousands of them), this keeps a record all of the stitching parameters in case I ever need to do it over. It is possible to discard the TIFF images once the stitching is done, as they are redundant.

Once the Raw file is written to the disk, I open it in Photoshop to do minor (sometimes major) retouching and color correction. This image benefitted from some Vibrance and Saturation adjustment. Just opening the photo (at about 27 GB) took 17 minutes. Cropping the scraps off the edges took another 15 minutes. There are lots of photographic mishaps in the crowds at the bottom of this image. I will spend some time cleaning those up before I print the image.

I haven’t totaled it up yet, but this image, like most of my ultra-high-resolution GigaPan images, has taken about 15 hours.

Uploading the image to the GigaPan took about three hours. It is available for public viewing, zooming and exploring now.

You can see the finished, navigable GigaPan here.

I hope you enjoy seeing the result of our work on this image.

 

Posted in Adventures, Education, Panoramic Photography, People, Photography | Tagged , , , , , , , , , | Leave a comment

Panoramic cameras and images, Part I

Posted on June 3, 2017 by Brian Lawler

The first true panoramic camera was patented by William J. Johnston in 1904. Century Camera Company brought that camera to market in 1905 as the Century Cirkut Camera. The company later became a part of the Eastman Kodak Company, which made the cameras until the mid-century, when they spun it off to the Graflex Company to meet the terms of a federal law prohibiting monopolies.

A Century Cirkut panoramic camera, probably a No. 8. date unknown. Note the gear wheel as the camera base. The motor, located in the housing, drove a small gear on a driveshaft that turned the camera on the tripod base. Meanwhile, the same motor pulled a roll of film through the camera.

The Cirkut camera is a clever combination of mechanics and optics, where the camera rotates on a tripod, driven by a spring motor. As the camera turns, a roll of film inside the camera passes across a narrow vertical slit to make a continuous strip of film with an image on it. Only a tiny amount of the image is recorded by the film at any time. The image on the roll is a record of what happened in front of the camera for a very small amount of time. The camera does not record anything else.

These cameras have been used over the last century to record breathtaking landscapes and group photos with hundreds of people in them, each one recorded with precision.

This is more modern version of the Cirkut No. 8, this one from the 1920s (judging by the lens). The drive gear can be seen in this image. They are beautiful instruments, and they take amazing photos.

The key to the success of a panoramic image taken with one of these complex cameras is that the rotation, the point-of-rotation, and the speed of the film must all be perfectly coordinated for the image to come out without distortion. If the film runs too fast, the photo will be disproportionately narrow. If the film moves too slowly, the image will be distorted. If the camera turns too fast or too slow, the same errors occur.

The mesh of the gears must also be perfect to avoid gear-chatter, which is an inconsistent movement of the camera (or the film) as it runs. The spring motors on most of these cameras, most of them being nearly 100 years old, do not run smoothly enough, and modern users of Cirkut cameras often replace the spring motors with smooth DC motors that run on battery power.

This illustrates the mechanism of a Cirkut camera. I have taken some liberties with the size and shape of the parts in the interest of clarity.

A Cirkut camera owner will have a box of gears with various numbers of teeth to set the speed of the camera’s rotation. Choosing the correct gear is done by consulting a chart where focal length and distance to the subject are the variables. The result is the size of the gear used the drive the camera’s rotation speed. Outdoor photos, where there is a lot of available light, take just a few seconds. Indoor photos require sometimes extremely long exposures – 30 minutes or more – to collect enough light to record the image on film.

In one historic panoramic photo I had the opportunity to study, the entire military base at Camp San Luis Obispo was photographed using a rotating Cirkut camera. It was obviously taken during daylight hours. While retouching the scan of that photo I noticed that in its over 50 inches of imagery there is not a single human being in the image! This would indicate that the camera was moving so slowly that anything moving in the scene moved too quickly to be recorded on the film. There are also no moving trucks or cars or trains. Nothing. My guess is that the photo (which also has some visible blurring) probably had an effective shutter speed (there is no shutter in a panoramic camera) of 30 to 45 seconds – enough time for a person to walk through the scene, or a truck to drive by without being recorded by the camera.

This is a scanning set-up I configured for scanning original Cirkut negatives. I made wings for my Epson scanner with register bars along the back edge. Then, moving the film in increments through the scanner, I made high-resolution scans of the film in sections, later putting them back together in Adobe Photoshop using its Photomerge function – which works well in this application. This particular film was 8-inch material. The length of each image varied from about 30 inches to over 72 inches.

Getting film for a panoramic camera is not impossible, but it will eventually be so. Batches of large roll film suitable for panoramic images can be purchased today, but must be ordered in adequate quantity and with enough lead time to allow the film manufacturer to coat, roll and package the film for these specialized cameras. In the not-too-distant future it will become impossible to purchase film in this format.

For those who have such a camera, and who can get film, the process is straight-forward. It requires attention to detail and careful planning, and it takes time. The camera is set up, focused (the panoramic film holder and motor mechanism are removed from the camera for focusing) and prepared. A light meter is consulted for proper exposure. As there is no actual shutter on a panoramic camera, exposure time is determined by how long the vertical slit is exposed to the film surface. This is a function of film speed, which is a function of rotational speed. The Century Cirkut camera has a chart showing the effective shutter speed at different motor speeds.

After the exposure is made on a Cirkut camera, that long roll of film must be processed. Numerous sizes of Cirkut cameras exist, measured in inches on the vertical slit. There are the Number 6, 8 and 10, the three most common, that record on rolls of six-inch, eight-inch and ten-inch film, respectively. There was even a Number 16, which recorded images onto 16-inch rolls of film!

For processing, some photographers use a technique called dip-and-dunk, where the roll of exposed film is submerged in diluted developer in a large, deep tank. The film is rolled back and forth during the long development time, giving every point on the film emulsion an opportunity to be in contact with the solution long enough to be processed evenly and correctly. Others had the opportunity to roll the exposed film into a canister that held the film in a spiral frame, and allowed the development chemicals equal access to the film’s surface during processing. This is a far better system, as it is easier to control and the results are much better.

After development, a quick trip through a “stop-bath” of slightly acidic solution, will stop the development. Finally, the film is immersed in a fixative, typically a bleaching chemical like sodium thiosulfate, which dissolves the undeveloped grains of silver in the emulsion, and clarifies the non-image areas of the photo. The fixative is a critical step to stop the light-sensitivity of the film, making it permanent and usable for photographic prints.

After the three-step development process, the film is washed thoroughly to remove these processing chemicals, then allowed to air dry. The result: a strip of film 6, 8 or 10 inches in height and as long as necessary to record the image. Many Cirkut panorama films are 50 or 60 inches in length.

There is no way to enlarge a panoramic film image, so the method of delivery has always been to make a face-to-face contact print of the Cirkut negative onto photographic paper in a large pressure contact-frame. This is also done in a darkroom, as photographic paper is sensitive to white light. The unexposed paper is placed in a frame, face-up, and the negative is placed in contact with the paper, emulsion-down. A glass cover is placed on top, typically with clamps surrounding the glass to apply an even pressure to the sandwich of film and paper.

Then, under red or ochre “safe” light, a contact exposure is made using a pinpoint light source overhead, which allows light to pass through the panoramic negative to the photographic paper underneath. The resulting print is typically processed by hand in a large tray of developing chemicals. Gentle agitation of the chemicals on the photo paper allows the image to be developed evenly. After several minutes of processing, the print is be stopped with an acidic bath, the fixed by immersion in a tray of sodium thiosulfate to fix the image and dissolve the unexposed parts of the emulsion.

Washing the prints, followed by air drying results in beautiful, long panoramic photos with amazing amounts of detailed information. It is a process involving hours of work, and the results are usually excellent.

Many of these cameras still exist, though few of them are in working condition. If you find one that is not in a museum, snatch it up (they sell used for as much as $25,000 in perfect condition).

In another blog I will write about more modern cameras and panoramic methods.

 

Posted in Art, History, Panoramic Photography, Photography, Photoshop techniques, Scanning, Software, Technology | Tagged , , , , , , , | Leave a comment

Alois was here

Posted on May 23, 2017 by Brian Lawler

Several of my students joined me and my colleague Prof. Martin Delp today for a tour of the Landsampt für Digitalisierung, Breitband und Vermessung, the Bavarian office for digitization and surveying. Here, the government creates maps, collects and distributes digital geographical data, and prints maps, charts and various related things for data analysis, mapping software (including cell phone mapping) and aviation “sectional” charts. for the state of Bavaria.

Alois Senefelder, discoverer of the lithographic printing process. Born in the Czech Republic, he brought his invention of planographic printing to Munich and the Western world.

This facility is also houses the Bavarian archive of lithographic stones created between the early 1800s and the late 1900s for mapping the state. There are 26,000 slabs of Solnhofen Limestone stored underground here in a vault that is a designated national treasure of the people of Germany.

Bavaria is mapped on litho stones at a scale of 1:5000. There are 18,000 stones (each one a “plate” for lithographic printing) at this scale, and many more at double scale of 1:2,500. Every square meter of Bavaria is written on stone, and curiously, not a single stone is missing or broken, this despite the building being bombed at the end of World War II by the Allies. They are all still intact.

Professor Delp and Sebastian Söckler explore the operation of a 150-year-old lithographic press in the Landsampt for Digitalisierung office.

On the street level of the same building are a few hand presses, each capable of printing images from letterpress, intaglio (copper engraved plates), or lithographic plates (either limestone or aluminum). We were given an opportunity to see these presses, and to touch and explore several lithographic stones on display there.

Our host was Sebastian Söckler, a graduate of the Hochschule München, who is now a printing engineer at the facility, and who is very knowledgeable about lithographic printing, platemaking and imagery. He was an extraordinary host who explained how images are imparted to stone, how the stones are prepared for printing, and how the presses work.

The collection of lithographic stones is stored below ground in a special warehouse that protects this priceless archive of original images for lithographic printing.

The Landsampt für Digitalisierung has printed and scanned all 26,000 lithographic stones, and the archive is a public resource. It’s possible to order a print of the scan of your neighborhood anywhere in Bavaria.

Invented, or more correctly, discovered by Alois Senefelder in the late 18th century, the lithographic process does not use a raised or recessed image, but instead uses a polished limestone surface with an image made by either chemical or mechanical etching on the surface. The image is made to accept ink (the oleophilic component) while the non-image area of the stone (the hydrophilic area) is left to hold water in its microscopic pores (about 250 grains per cm. or 1,200 per inch “resolution”). It is the balance of ink and water that makes lithography work.

This is a typical image on a stone – this one is of Moosburg – that shows the detail of these beautiful maps. I reversed the imge in Photoshop to make it legible;
lithographic stones are imaged backwards.

To print from a litho stone, one sponges water onto the surface, then follows that with a roller covered with oil-based printing ink. The water and ink are kept in balance by adding more water or more ink to the surface. Once the image is inked, paper is positioned on top of the plate (directly in contact with the stone) and pressure is applied with a bar that forces down on the paper as the carriage holding the stone is moved underneath.

After impression, the paper is removed with its image (and a considerable amount of water) on it. Allowed to dry, the water evaporates and the ink dries in place, making a printed work of tremendous quality.

The office runs their antique presses for occasional public events. We were able to dampen the plate today, but not to print, as the press was not set up for printing.

Martin Hämmerle, one of my students from Hochschule München, waves the hand-fan over the litho stone to dry it. The press is a 100-year-old machine
in the collection of the state agency.

We were then taken to the underground storage of the slabs of limestone, each one weighing over 400 lbs. On the wall is a chart showing the state of Bavaria and its mathematical coordinates that correspond to a specific limestone slab. Those slabs are stored in rolling storage shelves that have the stones placed vertically in rows. Over 200 square meters of space house the collection which is kept cool and dry to preserve the images on the stones.

It was interesting to think that the inventor of lithography, Alois Senefelder, once worked on the stones in the archive. That Senefelder himself made some of those images gave me an awesome feeling. He was here! Well, perhaps he was not not here exactly (this is a modern building), but his hands made the images on these stones. these are not replicas; these are the real thing.

It was another day of discovery for me and for my students. We stood at the place where lithographic printing was first developed. It will leave a lasting impression on all of us.

Posted in Adventures, Art, History, Printing and Printing Processes, Scanning, Technology | Tagged , , , , , , , , | Leave a comment

Take the Ten Fonts Survey

Posted on May 18, 2017 by Brian Lawler

Greetings, Blognosticator readers,

I am launching a survey about popular type fonts.

Here are the rules:

You are stranded on a deserted island. You have with you a computer and the complete Adobe Creative Suite (not the Creative Cloud, as that requires Internet access). You have numerous island residents nearby who have decided that it’s time to publish a newspaper and some tourist information in case you are ever rescued from the island.

You have power (solar, of course), a printer of some kind, and supplies to print your publications.

You only had time to upload ten type fonts when you left on this journey. With care and deliberation you chose those ten, and they must last you for the duration of your stay on the island.

Please tell me which ten you would take with you, and I will eventually tally them and write a blog about the results. There are nearly a quarter million of you readers out there, so I know that this will be fruitful.

Thank you for participating. This will not result in any response from me until the survey is complete. I will not use your contact information for anything other than to tell you the results of the survey.

Please enter font families. Helvetica is a font family that includes numerous weights and widths. Garamond in a family with numerous weights. I’m most interested in the font families than the specific faces. An exception might be something like Times Italic, which (in my onion) is a really lovely font that is significantly nicer than Times Roman. I’ll sort out the details as the submissions come in. Thank you!

Ten Type Fonts Survey
You can enter font families here. Helvetica, for example, would include numerous weights and widths. Neue Helvetica is a different family.
Font 1
Font 2
Font 3
Font 4
Font 5
Font 6
Font 7
Font 8
Font 9
Font 10
Your name (optional):
Your e-mail (optional):
Posted in Typography | Tagged , , , , , , , , , | Leave a comment

Ink is a terrible thing to waste

Posted on May 17, 2017 by Brian Lawler

I’m teaching this semester in Munich, Germany, at the Hochschule München, or more correctly, Hochschule für Angewandte Wissenschaften München. It’s just a few blocks from the city center on Tram route 20. We have about 20,000 students on several campuses.

It’s a technical university where they teach in the same proud tradition as we do at Cal Poly – Learn by Doing. It’s all hands-on, all the time, and that shows in the many interesting projects that one sees on campus and off.

My classes are one in photography, one in advanced typography, and one called Projekt. It’s a senior-level course in project planning and execution. My students are dedicated, respectful and eager. I’m having a great time here.

This is the gamut comparison between the Epson on Tuesday afternoon and the FOGRA 39 standard. The Epson is slightly larger, but I expected more.

In the photo class I needed to print panoramic photos, and there is an Epson 9900 Stylus Pro printer in the lab (along with three smaller Epson 4800 printers). The 9900 features more base colors of ink than the others, and it is wider, so I wanted to use it for my course. Unfortunately, the printer had not been used for quite a while.

And, as everyone knows, an ink-jet printer sitting idle is the Devil’s paradise. I started it up and immediately learned that the ink was low on the light-light-black channel, the cyan channel, the orange channel and the green channel. Fortunately an inventory of inks was on-hand and soon I was running. But then the waste cartridge on the right side was filled and the machine stopped. A day later another was found and I was back in the game.

This, again, is Epson (Monday) vs. FOGRA. It was acceptable, but not at all what I would have expected from an ink-jet printer with CMYK plus orange and green base colors. The hardest thing for me to accept was that there would be more cyan in FOGRA than on the Epson; that was illogical.

The quality was marginal. I was getting streaks and breaks in horizontal lines. I own an Epson of this same size at home, so I am experienced in the troubles that ink-jet printers can have. I have used that printer continuously for many years and have had almost no trouble with image quality of head clogging. The more you use an Epson, the better it works. This one, however, was sitting and some of the ink-jet nozzles were clogged.

I ran a cleaning cycle on the Hochschule München Epson. It ran better. Then I ran an ICC profile patch set and profiled the printer. The resulting gamut of colors was OK, but I suspected that something was amiss. My printer at home is a 9600, which has “only” seven inks, and an amazing gamut of colors, one which encircles all CMYK printing devices and makes really beautiful photo prints.

This is Epson Wednesday afternoon – after several cleaning cycles – compared to the same printer the day before. It was definitely getting better.

This printer in Munich has an even larger gamut I suspected because of the presence of Orange and Green inks in the 12-ink stable (only 11 are used, as there are two exclusive dark black inks).

On Tuesday I ran out of Cyan altogether, so I headed to the very impressive Calumet Photo store in downtown Munich and bought a new cartridge of Cyan (€143.55 with VAT) and took it back to the university. Soon I was running on all 11 cylinders again. I ran another ICC profile, and this time the gamut looked better.

But I was still troubled by the fact that the overall gamut of colors was only slightly larger than the gamut of FOGRA 39, a CMYK gamut for printing presses. So, I ran diagnostics and discovered that I was getting significant cut-outs on various color nozzles: green, orange, cyan. The streaks persisted. I ran another cleaning cycle, which uses a tremendous amount of ink (which is about €1250 per liter). Things got better, but I was still getting errors of clogged nozzles, so I ran another cleaning cycle. This was starting to hurt.

…and this is the Epson Wednesday afternoon compared to FOGRA 39. This is much better than when I started. I suspect that there may still be more color possible on this machine, but that may take another day of cleaning and printing.

I ran the nozzle alignment test, which seemed to work perfectly, so I moved back to the cleaning cycles and test prints.

By Wednesday afternoon I had the machine performing nearly perfectly. The streaks were almost gone, and prints were starting to look quite good. I decided that my next cleaning cycles would be to run some of my student work. Rather than just pump ink through the nozzles and into the waste cartridge, I wanted to put that ink onto photo paper.

I ran three panoramic images, and they looked good. Another round of profiling revealed what my eyes were telling me earlier: the gamut of colors was just not good enough for that machine. Analysis of the new gamut showed that I was finally using all the colors on the printer – strong cyans were showing, much more green and considerably more red and orange were finally showing up in the charts.

Today I will run some more work through the machine in hopes that that few misfiring nozzles in cyan, orange and green will clear up and give me the results for which the Epson 9900 printer is so famous.

It has taken three days of solid effort, probably €500 worth of ink, and a good part of a roll of GMG proofing paper to get the machine running well again. It has been worth the effort in seeing the results.

June 22 addendum: Yesterday I turned the printer on to run some student work, and I sent a job to print and NOTHING came out! All of the ink-jet nozzles were clogged. NOTHING! So I ran a Power Clean, which is a utility run on the machine itself. I ran an errand while this was running, and I returned to find that about 98 percent of the nozzles are now working – the best yet. I printed several jobs and the output is spectacular. I don’t know what caused the failure, but the results are great now. I plan to run one more profiling test today to see if the gamut is better.

June 23 addendum: After yesterday’s success in getting the ink jets clear, I decided to run a calibration/profiling test of the printer now that it’s printing with all of its many colors of ink. The results are impressive, and they justify the effort I have taken to restore this machine to full working condition.

This is the gamut of the fully functional Epson 9900 printer. That’s more like it! My experience with ink-jet printers over the years has shown me that this is more normal than exceptional.

When compared to the first day’s gamut, the colors look like this:

The wire frame shows that there is a tremendous amount of green and cyan (left edge) printing now that was not in the gamut in May. This is exciting!

The big problem, though, is how do I keep the machine from drying-up again?

 

Posted in Color Management, Photography, Printing and Printing Processes, Technology | Tagged , , , , , , | 4 Comments

The Bishop Peak Portrait Project – complete!

Posted on May 15, 2017 by Brian Lawler

Thank you all for your patience while I built the Bishop Peak Portrait Project, a series of thousands of photos of San Luis Obispo’s most prominent mountain. I started the project in November, 2015 with the construction of a weatherproof box for the camera and electronics, and then started shooting the photos in March, 2016.

The photography continued for a year to the end of February, 2017 when I turned it off and removed the equipment from the roof of the Kennedy Library at Cal Poly.

In the end, 365 of those photos were prepared and built into a large display for the Baker building at Cal Poly.

During the months between the beginning and the end, I machined the panels from an aluminum laminate material called AluPanel, developed a system for machining the individual photos for the display, and cut those to fit. Then we put it together.

The story starts here if you want to go back to read all the details.

This week Cal Poly Facilities workers put the panels up in the Baker Science Center at Cal Poly, and my colleague Rob Brewster (one of the technicians on the project) sent me photos.

I hereby present the completed Bishop Peak Portrait Project:

More photos will follow of course, but I’ve never seen it, and will not be able to see it until late August when I return from my teaching post in Germany.

I am very proud to see the results, and once again thank those who helped make the project happen:

Dr. Phil Bailey, Emeritus Dean, College of Science and Mathematics
Dr. Derek Gragson, Associate Dean, College of Science and Mathematics
Rachel Henry, Communications, College of Science and Mathematics
Rob Brewster, Technician, College of Science and Mathematics
Doug Brewster, Technician, College of Science and Mathematics
Emma Wilson, Graphic Communication student and project assistant
Patrick Kammermeyer, Kennedy Library communications specialist
Catherine Trujillo, Kennedy Library, and campus art coordinator
Eric Johnson, Graphic Communication Dept. IT consultant
Tim Hastings, Facilities Department
Sarah Sayeed, Kennedy Library Information Technologies lead
Dale Kohler, Kennedy Library and Cal Poly ITS lead
Jim Eckford, Rancho Burro technical support
Ashala Lawler, technical support
Bryn Forbes, CNC support and assistance

 

Posted in Art, Photography, Photoshop techniques, Printing and Printing Processes | Tagged , , , , , , , , , | Leave a comment

Street Art provides an opportunity
for repositioned panoramas

Posted on May 9, 2017 by Brian Lawler

My wife and I traveled today to see a neighborhood in Munich that is famous for its street art. It’s called the Tumblingerstraße area, and it features some very sophisticated works of graffiti – much better than graffiti – that shows some amazing talent and creativity.

I am not a fan of graffiti; I prefer that artists not paint on buildings and train cars, etc. However, this neighborhood has become the canvas for some talented artists who use the urban landscape as their palette. Though there is a municipal sign at the edge of the neighborhood prohibiting graffiti and spray-painting (itself defaced by spray paint), the street artists who do this work must use scissor lifts, or at least ladders, to reach their canvases, and one does not do that kind of art in the middle of the night unseen by the Polizei. This work must be at least condoned by the authorities.

This example is over 20 yards wide, about 12 feet tall. I took seven horizontal images to make the complete image, and stitched them together using Photomerge as described here.

My objective today was to capture some photos of this work, so I took my trusty Canon 5D Mark III camera and my “street lens” (16-35 f2.8) with me to make the images. Tumblingerstraße is a moderately-trafficked street bordering an industrial neighborhood and the Isar river. A huge brewery is located just off the street, and there is a lot of train traffic that passes overhead.

There are cars parked everywhere along the street, making it difficult to step far enough away from the art to photograph each piece in one frame; the sidewalks are wide enough for people and bicycles, and barely wide enough for me to make multiple-exposure “panoramic” images of the art. The term “panoramic” infers that the camera is in one position, and that it turns on its own axis to make a long horizontal photo. In this case, the camera does not stand still and turn. Instead, the camera moves horizontally in increments, making a series of square photos that are later stitched into a large, seamless image that represents the original art in its entirety. I will call these repositioned panoramas.

They are quite difficult to capture and to stitch, and there is only one way I know to get them to come out correctly – straight and undistorted – and that is to use Adobe Photoshop’s Photomerge feature to assemble them into whole images.

Photomerge is very good at this. I used it extensively four years ago to assemble section-scanned film negatives of Kodak Cirkut panoramic camera film. These negatives measured about ten inches tall by over four feet in length. I scanned them on an Epson scanner using its transparency function, and then assembled the sections in Photoshop using this technique. It worked flawlessly. I had an advantage on that project though – the film had no perspective, and the scanning bed stayed absolutely flat and square for each scanned section. Assembling these into coherent images was relatively easy for Photoshop, and the resulting photos were extraordinary.

This is my shooting pattern. I stand square to the wall, find a horizontal reference  in the shot (edge of the wall in this case), then take a series of images with careful horizontal indexing – I take three steps between each frame.

Today’s adventure was unaided by a tripod. I have one, but I did not take it with me to Tumblingerstraße. Instead I hand-held the camera and made very careful series of photos of each image I wanted to capture. I checked focus and exposure, then attempted to find something straight to use as a reference. Typically, I used the edge of the wall at the sidewalk for a horizontal reference in each frame. Then I took a photo, stepped three measured steps to the right, took another, and repeated this until I had a series of images that covered the art with significant overlaps.

Four sequential horizontal images taken today to capture one work of art. Notice the overlap in the images – this is critical to the success of merging the photos into one.

Back in my apartment I assembled the photos in Photoshop using one of the techniques that Photomerge will practice to get sequential images to match and merge into one.

To get this to work, one must first be extremely careful to take effective photos. Unless you can go back and take the photos again, it’s best to get it right the first time. Check square, check focus, use manual exposure (and don’t change it!), don’t change the distance between the art and the camera (this can be challenging). The process does allow for some error, but it must be minimal.

I try for about 20 percent overlap between the photos. Photoshop needs a lot to find common areas and to blend the images successfully. I open all of the contributing images for a single work of art, then choose File>Automate>Photomerge to reach the Photomerge pane. Then, in that pane, I choose the technique that works best for this type of work: Reposition. The other six options will generally not work to assemble rectangular images correctly. A couple of them won’t work at all for this kind of photo.

Photomerge in Adobe Photoshop, showing the selection of the Reposition mode, and the selection of the Open Files in the center. This is the most effective way to accomplish the task of making these images.

Choose Add Open Files then select them all: click, shift-click on the list to include them. Then click OK. It takes several minutes at the resolution at which I work (63 MB per image), but the result can be spectacular.

This is the composite image from the four source images shown above. Once completed, it needed only to be cropped to finish the image.

Note that the most visible difference between a true panorama and a sequential, repositioned panoramic image, is that these photos are straight, and do not display the arcs of a normal single-point-perspective panoramic photo. Photomerge has two modes that will do that, but then your straight lines become arcs, and that is unattractive for this kind of work. Notice in these images that the rows of bricks are straight from left to right.

In the end I recorded a faithful photo of each of the major works on display in the neighborhood. I am very pleased with the results. As with all of my panoramic images, the greatest amount of time is spent in front of the computer, as opposed to the relatively small amount of time spent on the scene.

I found myself enchanted by the street art today, and I applaud the creative people who make this work.

Please don’t do it on my apartment building!!

Addendum December, 2021:

I am back in Munich, and have spent a lot of time at Tumblingerstraße recently, photographing the artwork there. This time, my Master’s degree students and I have been documenting the artwork, and my students have been interviewing some of the artists. It has been very successful.

I have written a new blog about a more precise method I used this year to photograph the street art (note that I am not calling it “graffiti” anymore).

You can read the new article here.

Posted in Adventures, Art, Panoramic Photography, People, Photography, Photoshop techniques, Software | Tagged , , , , , , , , , | Leave a comment

Van Gogh’s Père Tanguy in perspective

Posted on May 5, 2017 by Brian Lawler

Museums often have no-photography rules, depending on their policy about allowing visitors to take photos, or the legal arrangements they have with art owners, copyright holders and other things that go beyond my pay grade.

In the Galleria dell’Accademia in Florence, Italy, a few years ago I noticed that the museum had finally given up on its prohibition of photography of the statue of David. On my prior visit they played a continuous recording in numerous languages saying that photography was prohibited. My guess is that the museum staff threatened to quit if management didn’t stop playing the recording.

Here I am, having a close photographic experience with Vincent Van Gogh at the National Portrait Gallery in Washington, D.C. in 2008. This gallery allows photos, but does not allow tripods. This painting does not have glass in its frame, making the relief of the brushstrokes visible.

Since the tourists ignored the prohibition anyway, the museum acquiesced to the ubiquity of cell phone cameras. They just gave up. It probably doesn’t matter anyway because the value of a photo of the statue is exactly nothing. These photos are just mementos, photos that people show to others saying, “Look – I took a photo of the David!”

The fact that millions of people have taken (previously illicit) photos of that statue does not diminish the value of the original in any way. Michelangelo would be proud.

(I know that most museums prohibit flash photography, and I know that the powerful ultraviolet light emitted by some strobe flash tubes can potentially damage the colors in works of art. That is a topic for another day.)

Taking photos of fine art in ambient light does no harm to the originals, nor does it hurt the marketing opportunities that museums have to sell postcards. I buy them too, and just like you, I never send them to anyone. They end up in boxes in my storage locker with postcards from other trips to other lands.

This is my off-axis photo of Van Gogh’s Père Tanguy portrait in the Rodin Museum in Paris. I was able to eliminate all but three pinpoint reflections in the glass that protects the image.

I suppose that an unscrupulous photographer could take an illicit photo of a painting, then print postcards to sell on the sidewalk outside the museum to take business away from the museum. That scenario is unlikely, but in case you are considering this as a career, following is my technique for making a darn good reproduction photo of a masterpiece in a museum.

This is the same image, squared-up in Adobe Photoshop. I use the Distort tool to align the image with a grid of guidelines I put into the image frame. The blue was set as the background color so you can see how much distortion was needed to get the image square.

There are two Van Gogh paintings on display at the Rodin Museum in Paris. On my first visit to that museum, nine years ago, I noticed that the paintings were not secured to the wall, only hung on a cable like the paintings that hang in my home. I also noticed that the two paintings were not covered with glass, making the three dimensional brush strokes easily visible. To me, that is what makes Van Gogh’s work so fun to see in person.

On my visit to the Rodin this week I was pleased to see that they have now secured the frames to the wall with very sturdy connectors. I was disappointed to see that they now have glass in front of the paintings (this is good, really, because it protects them from people and the environment). The problem with the glass is that it hides the paint strokes from the viewer, and it increases the reflections on the painting.

This is the reproduction of the Père Tanguy portrait that is available for free download from the Musée Rodin. The museum had the advantage of taking this image in a carefully controlled photo studio.

This is my Père Tanguy portrait reproduction after correcting the distortion in the top image. It started as a 62 MB image, and ended up being cropped to about 18 MB. That is enough resolution to reproduce the image on postcard-size works. Note that my image missed a little bit of the right-hand vertical.

I wanted to photograph the image of Père Tanguy, the art supply dealer Van Gogh painted in lieu of payment for pigments and canvas. Mssr. Tanguy was, according to the information at the Rodin Museum, a nice fellow who supported starving artists, and willingly posed when they wanted to trade a painting for the raw materials of their craft. The indoor lighting at the museum shone down from the ceiling in pinpoints of light that reflected off the new glass. It was impossible to take a good photo with the reflections in it.

Instead, I took the photo from an angle that provided the fewest reflections. And, as cinematographers say, I left the rest for “post” (this means that I’ll leave it to the next guy to fix it.)

It turns out that I am that next guy, so it was up to me to fix the photo. I shoot in camera Raw, and then I convert as I import into my computer into DNG format (this is my insurance against obsolescence). I opened the DNG, then made some small tonal adjustments and opened the image in Photoshop. There, I dragged guidelines to the edges of the image as I would like it to be.

Then, using the Marquee tool, I selected a rectangular area of the photo large enough to expand it to fit the space and square the photo to these guidelines. Choosing the Distort tool under Image>Transform>Distort. I usually compress rather than expand the corners when I do this, meaning that one corner (the upper-right in this example) should be brought inward rather than taking the upper-left corner and dragging it upward. This results in a reduction of pixels in the image, and that might be better than an interpolation of new pixels caused by stretching the image. As I pull or push the corners of the image, I add the Shift key to keep the control point from wandering vertically while I am pushing it horizontally (or vice versa).

In the end, I have a nice, square image of Pere Tanguy, free of reflections (mostly) and ready for contrast adjustment and finishing. The Rodin Museum provides the dimensions of the original on their web site, figures which make it possible to return the image to its proper proportions. I set my canvas to these proportions, then scale the image non-proportionally until it fits (with Snap turned on).

I downloaded the Museum’s low-res version of the original image (they no doubt have a photo studio for this work), opened it in Photoshop next to mine, and then I adjusted mine to match better in contrast, color saturation, and black values. The final product is over 18MB, and at that size I could print postcards with good enough quality to set up outside the museum and sell my wares. But I’m too busy for that, and they own the original. It is an interesting exercise, one that is useful for photographing highly reflective art (I used this technique once for photographing badges).

I appreciate the fact that Musée Rodin does not mind when techies like me come to visit, carrying professional cameras and intent on making nice images of their masterpieces. They are generous with their originals, and I am grateful that they seem to have no cares about my taking photos in their beautiful museum.

Posted in Art, History, Photography, Photoshop techniques, Printing and Printing Processes | Tagged , , , , , | Leave a comment

Der Druck ist nicht tot

Posted on April 21, 2017 by Brian Lawler

At least in Germany it’s not.

As I stepped off the train in Heidelberg last Monday afternoon an American asked me for directions. At this I was not very helpful because he and I had been in Heidelberg exactly the same length of time – about ten seconds.

We chatted as we walked into the station. He mentioned that he was visiting the city where his parents met 60 years ago. I told him I was there to visit the printing press manufacturer named after the city we were visiting.

He lamented that publishing is dead. So sorry, he said. He, himself once helped out on a press. Did I know that the press operators used to know just how much ink to add to a unit to get the color right? Yes, I admitted that I knew that, and I had done that very thing myself.

We said our pleasantries and parted ways, he to find the place of his parents’ meeting, and I to the Schnellpressenfabrik to see multimillion-dollar printing machines assembled and prepared for their new owners around the world.

This is a small selection of the many, many magazines for sale at the bookstore located in the Heidelberg Hauptbahnhof. There were another five racks this size in the store. This is eye-candy for printers!

A few minutes later my wife and I visited a nice magazine store in the Heidelberg Hauptbahnhof; we were both looking for something to read – an English language magazine perhaps. As we wandered the aisles in the store I realized that I was standing in the middle of a well-stocked example of the fact that publishing is not dead. I snapped a few photos with my iPhone, while admiring the retail end of the publishing industry.

I am going to take a guess that this single store had over 1,000 magazines for sale. In the photo section where I was standing, there were 49 magazines in German on the subject of photography. There were several in other languages also, but they were in the foreign languages section, which featured another 150 titles.

I walked around and picked a beautiful tourist magazine about the city of Heidelberg. I picked that up and decided to buy it. I also grabbed a copy of the week’s Economist, printed in England (I think) for the European market. I get this magazine back home, and I think that the American edition is printed in the U.S.

I reflected on the man’s comments just minutes earlier on the train platform. So sorry. And I realized that he is absolutely wrong. Ill-informed is probably the better term.

I wondered to myself where he had read that publishing is dead.

Probably in a magazine.

The next morning my wife and I were treated to a tour of the huge Heidelberg printing press factory in Wiesloch-Walldorf just outside Heidelberg. There we saw presses large and small – even the small presses are large these days. We were taken to a building 620 meters in length and 120 meters in width where the presses are assembled. It’s a combination assembly line and individual assembly facility where 4,000 printing units are made each year and shipped around the world.

Ralph Schönfeld and Clarence Penge, our wonderful hosts at the Heidelberg plant in Wiesloch, outside of Heidelberg, Germany. They are standing on a press with eight printing units and four coaters. The machine is 93 feet long. It will be used in a packaging plant.

We visited the gear-making building and saw the production area where cast steel components are machined to a precise size, then measured for quality-control on their way to final assembly in a huge multi-color Heidelberg press.

We walked through and over a printing press so large that the feeder is way over my head. This one is destined for a packaging printer in North America. It is over 90 feet in length. The press sheet it prints is 63 inches in width. I suspect that some of the magazines I saw in the bookstore are printed on machines like this (actually it’s more likely they are printed on the smaller 40-inch machines).

My wife Ashala stands at the delivery of one of the huge Heidelberg presses in the demonstration room in Wiesloch. This machine prints on sheets of paper over five feet in width. It is also the fastest press made by Heidelberg – it prints up to 18,000 sheets per hour, which translates to five sheets per second. The amazing thing about this press is that it’s designed to be operated by only two people.

I came away from this experience knowing that printing (and publishing) are doing pretty well. After some lean years fighting our way through the recession, the industry is on the go – especially packaging, and the printers of the world are buying new printing machines to fill the stores in all the Hauptbahnhofs in all the world with beautiful colorful magazines.

And I hope the man from the train found the place where his parents met. I also hope he bought a magazine from the store in the station as he was leaving town.

 

Posted in Adventures, New technology, People, Printing and Printing Processes, Technology | Tagged , , , , , , , , | 2 Comments

Learning letterpress in a 14th century town

Posted on April 19, 2017 by Brian Lawler

One of the three classes I am teaching in Munich is Advanced Typography.

The students are in their third year in the Print Media program at Munich University of Applied Sciences. Five of my students are exchange students from my home university, Cal Poly.

At Cal Poly I have access to the Shakespeare Press Museum, a delightful collection of 19th and early 20th century letterpress equipment and type. When I teach this course at Cal Poly we just walk down the stairs to the museum so that I can show the students how to set type and print with the letterpress equipment in the museum. At Munich there is no museum, and only one letterpress.

My colleague Martin Delp, who spent last spring quarter teaching at Cal Poly, suggested that when I wanted to teach the students about letterpress printing, we should visit his friend Oskar Bernhard, a letterpress printer in the town of Nördlingen, Germany. That town, which was founded in the early 14th century, is about two hours from Munich by train.

Oskar Bernhard explains the marks on a letterpress proof for my Advanced Typography students. From left to right: Emily Hoehenrieder, Shauna Rinaldi-El Abd and Christine Wiechers.

Last Thursday, my students joined Prof. Delp and me at the Munich central train station and we boarded a train headed for Nördlingen. When we arrived in that city, we walked over to Herr Bernhard’s atelier and began a most impressive day in a working letterpress shop.

This is a look from the balcony of Oskar’s workshop. On the left is his offset proofer. At center is the Heidelberg K-series cylinder press. At the lower-right is his Heidelberg “Tiegel” press. The students are gathered at the back of the shop setting type for their projects.

Bernhard, a veteran of the typographic industry, has an impressive collection of machines – several letterpress proofers, one Heidelberg “Tiegel” (in America these are called the “Windmill”), a large Heidelberg cylinder press, and a very large offset proofing press. He uses these machines to produce limited edition letterpress books, specialty cards and booklets, and a variety of commercial jobs.

Shauna Rinaldi El-Abd and Ashley Boehmer, both Cal Poly students in the exchange program at Munich, operate the letterpress proofer in Mr. Bernhard’s shop. They are positioning a sheet of paper to be printed on the press.

He has a Monotype machine – a typecasting machine from about the 1950s. The keyboard is still in working condition, but his caster is in need of repair. The rest of his type is foundry type and some wood type in sizes up to about three inches. It’s a beautiful shop, very neat and orderly. And, unlike many letterpress shops, there are not very many orphaned projects sitting on the counters. He is meticulous about putting type away.

The students learned how to use a composing stick, how to add (real) leading to increase the space between lines, and how to quad-out their lines to properly center them. They chose their type from the many drawers of hand-set type in Mr. Bernhard’s collection. He was most gracious in allowing the students to invade his workshop and use his precious equipment.

With Oskar Bernhard’s historic building in the background, my students gather for a group photo in Nördlingen. In back are me, Franz Eder, Martin Hämmerle, Christine Wiechers and Oskar Bernhard. Second row: Antonia Schaefer, Jessica Dimulias, Skyler Ulep, Front row: Sophie Wegenknecht, Ashley Boehmer, Shauna Rinaldi El-Abd, Emily Hoehenrieder and Prof. Martin Delp. Photo by Ashala Lawler

It was fun and highly educational. Each student had a chance to set type, to position it on the press and to print several sheets of finished work on Mr. Bernhard’s proofing press. They had an experience that one cannot get in a textbook, nor from a PowerPoint presentation.

After our day in Oskar Bernhard’s workshop, we had a pizza, then took a guided tour of the town, the only city in Germany that still features a wall around the city to keep marauders out. At the end of the day, we sent the students back to Munich on the train, and we stayed behind for a nice dinner and a second day in the historic city.

At Cal Poly we have the motto “Learn By Doing.” One of the reasons that Munich’s program is such a good fit for our exchange program is that its system is also based on hands-on learning. With our visit to Nördlingen and to Oskar Bernhard’s workshop, the students had a worthwhile learn-by-doing experience that they will long remember.

Note added May 10, 2017:
My students and I created a thank you poster to send to Herr Bernhard. We included hand lettering by Sophie Wagenknecht and a translated version of This is a Printing Office by Beatrice Ward. Each student signed the poster and we mailed it to Oskar Bernhard.

Posted in Education, History, Typography | Tagged , , , , , , , , | 1 Comment