Category Archives: Tutorials

Processing workflow for lunar surface images using GIMP and G’MIC

This post is going to illustrate a post-processing workflow for lunar surface images using the open source tools GIMP and G’MIC. GIMP has been largely ignored by astrophotographers in the past since it only supported 8-bit colour channels. The long awaited GIMP 2.10 release in April 2018, introduced 16-bit and 32-bit colour channel support, along with many other important improvements that enable high quality post-processing.

Astrophotographers seeking to present high detail images of The Moon, have long recognised that capturing a single still image is not sufficient. Instead normal practice is to capture a high definition video at a high frame rate lasting for a minute or more, by attaching a webcam to a telescope in place of the eyepiece. A program such as AutoStakkert2 will then process the video, analysing the quality of each video frame, selecting the “best” frames, and then merging them to produce a single frame with less noise and more detail. The output of AutoStakkert2 though is not a finished product and requires further post-processing to correct various image artefacts and pull out the inherent detail. A common tool used for this is Registax which particularly found popularity because of its wavelet sharpening feature.

Use of AutoStakkert2 can be a blog post in its own right, so won’t be covered here. What follows will pick up immediately after stacking has produced a merged image, and show how GIMP and G’MIC can replace use of the closed source, Windows based Registax tool.

The source material for the blog post is a 40 second long video captured with a modified Microsoft Lifecam HD paired with a Celestron Nexstar 4GT telescope. Most astrophotographers will spend £100 or more on CCD cameras directly designed for use with telescopes, so this modded Lifecam is very much at the low end of what can be used. This presents some extra challenges, but as can be seen, still allows for great results to be obtained with minimal expense.

The first noticeable characteristic of the video is a strong pink/purple colour cast on the edges of the frame. This is caused by unwanted infrared light reaching the webcam sensor. A IR cut filter is attached to the webcam, but it is positioned too far away from the CCD chip to be fully effective. A look at a single video frame at 100% magnification shows high level of speckled chromatic noise across the frame. Finally the image slowly drifts due to inaccurate tracking of The Moon’s movement by the telescope mount and features are stretched and squashed due to atmospheric distortion.

100% magnification crop of a single still video frame before any processing

After the video frames are stacked using AutoStakkert2, the resulting merged frame shows significant improvements. The speckled noise has been completely eliminated by the stacking process which effectively averages out the noise across 100s (even 1000s) of frames. The image, however, appears very soft lacking any fine detail and there is chromatic aberration present on the red and blue channels

100% magnification crop after stacking top (50% by quality) video frames in AutoStakkert2

AutoStakkert2 will save the merged image as a 16-bit PNG file, and GIMP 2.10 will honour this bit-depth when loading the image. It is possible to then convert it to 32-bit per channel before processing, but for lunar images this is probably overkill. The first task is to get rid of the chromatic aberration since that has the effect of making the image even softer. With this particular webcam and telescope combination it is apparent that the blue channel is shifted 2 or 3 pixels up, relative to the green, while the red is shifted 2 or 3 pixels down. It is possible to fix this in GIMP alone by decomposing the image, creating a layer for each colour channel, then moving the x,y offset of the blue and red layers until they line up with green, and finally recomposing the layers to a produce a new colour image.

This is a rather long winded process that is best automated, which is where G’MIC comes into play. It is a general purpose image processing tool which has the ability to run as a GIMP plugin, providing more than 450 image filters. The relevant filter for our purpose is called “Degradations -> Chromatic Aberrations“. It allows you to simply enter the desired x,y offset for the red and blue channels and will re-align them in one go, avoiding the multi-step decompose process in GIMP.

G’MIC Chromatic Aberration filter. The secondary colour defaults to green, but it is simpler if it is changed to blue, since that is the second fringe colour we’re looking to eliminate. The preview should be zoomed in to about 400% to allow alignment to be clearly viewed when adjusting x,y offsets.

100% magnification crop after aligning the RGB colour components to correct chromatic aberration.

With the chromatic aberration removed the next step is to get rid of the colour cast. The Moon is not a purely monochrome object, with different areas of its surface have distinct colours which would ideally be preserved in any processed images. Due to the limitations of the camera being used, however, the IR wavelength pollution makes that largely impossible/impractical. The only real option is to desaturate the image to create an uniformly monochrome image. If a slightly off-grey colour tint is desired in the end result, that could be added by colourizing the final image.

100% magnification crop after desaturating to remove colour cast due to IR wavelengths

The image that we have at this stage is still very soft, lacking in any fine detail. One of the most popular features in Registax is its wavelet based sharpening facility. Fortunately there are a number of options available in GIMP now that can achieve comparable results. GIMP 2.10 comes with a “Filters -> Enhance -> Wavelet decompose” operation, while G’MIC has “Details -> Split Details (wavelets)” both of which can get results comparable to Registax wavelets operating in linear mode. The preferred Registax approach though is to use guassian wavelets, and this has an equivalent in G’MIC available as “Details -> Split Details (gaussian)“. The way the G’MIC filter is used, however, is rather different so needs some explaining.

Split details (gaussian) filter. The image will be split into 6 layers by default, 5 layers of detail and a final background residual layer. Together the layers are identical to the original image. The number layers together with the two scale settings determine the granularity of detail in each layer. The defaults are reasonable but there’s scope to experiment if desired.

Describing the real mathematical principals behind gaussian wavelets is beyond the scope of this posting, but those interested can learn more from the AviStack implementation. Sticking to the high level, when the plugin is run it will split the visible layer into a sequence of layers. There is a base layer “residual” and then multiple layers of increasingly fine details applied with “Grain Merge” mode. Taken together these new layers are precisely equivalent to the original image.

The task now is to work on the individual detail layers to emphasize the details that are desired in the image, and (much less frequently) to de-emphasize details that are not desired. To increase the emphasis of details at a particular level, all that is required is to duplicate the appropriate layer. The finest detail layer may be duplicated many, many, many times while coarse detail layers may be duplicated only once, or not at all. If even one duplication is too strong, the duplicated layer opacity can be reduce to control its impact.

GIMP layers. The default G’MIC split details filter settings created 6 layers. The layer labelled “Scale #5” holds the fine details and has been duplicated 5 times to enhance fine details. The “Scale #4” and “Scale #3” layers have both been duplicated once, and opacity reduced on the “Scale #3” duplicate.

It is recommended to work in order from coarsest (“Scale #1”) to finest (“Scale #5”) detail layers, and typically the first two or three levels of details would be left unchanged to avoid unnatural looking images with too much contrast. There is no perfect set of wavelet adjustments that will provide the right amount of sharpening. It will vary depending on the camera, the telescope, the subject, the seeing conditions, the quality of stacking and more. Some experimentation will be required to find the right balance, but fortunately this is easy with layers since changes can be easily rolled back. After working on an image, ensure it is saved in GIMP’s native XCF format, leaving all layers intact. It that then be revisited the following day with a fresh eye whereupon the sharpening may be further fine tuned with benefit of hindsight.

100% magnification crop after sharpening using G’MIC gaussian wavelets filter and GIMP layer blending

As the image below shows, even with a modified webcam costing < £20 on a popular auction site, it is possible to produce high detail images of The Moon’s surface, using open source tools for all post-processing after the initial video stacking process.

Complete final image after all post-processing

 

Building a Kodak Brownie digital camera

The Kodak Brownie was the first camera to really bring photography to the masses with a low purchase price. The simplicity of its design meant anyone could figure out how to use it with little difficulty – even by comparison with today’s cameras it is still easy to use, since it has essentially no controls to learn – just a shutter button, view finder and film winder. Millions of Kodak Brownies were made over the course of its 60 year lifespan, from 1900 onwards and the build quality & simplicity means many survive in good working order. The upshot is that a Kodak Brownie is a good option for custom modifications – easily available on ebay or in car boot sales, simple to hack and cheap enough that it doesn’t matter if things go wrong.

The original plan was to build a variant on my previous raspberry pi & webcam based pinhole digital camera, since I already had a second raspberry pi zero needing a purpose. A previous trip to the local carboot sale had yielded a Kodak Brownie Hawkeye for less than £5, which is the variant from the 1950’s with a case made out of bakelite instead of wood / cardboard. So the only key missing piece was a webcam. Since both the raspberry pi zero and kodak brownie had cost less than £5, that was set as the upper limit for obtaining a webcam. Trawling eBay listings found a number of sellers offering a variety of 50 megapixel cameras at this price point. These technical specs were clearly complete & utter lies – it was never going to be a 50 MP sensor for that price – but at the same time it was worth a punt to discover just what the cameras did offer. I first one I obtained turned out to provide 640×480, or a mere 0.3 MP with raw video only, no mjpeg, thus limiting the framerate too. IOW pretty awful, but only marginally more awful than expected. The plus side was that the case was easy to remove exposing a very compact circuit board which would be an ideal size for embedding.

The “50 megapixel” eBay webcam, which turned out to be 0.3 megapixels, prior to removing the case

Upon testing the webcam with an improvised pinhole plate, it was clear that the sensor had unacceptably poor low-light performance. While it could serve as a pinhole camera, it would only be usable outside in bright conditions. I wanted to build a camera that was more versatile, so the plan was changed to build a “normal” digital camera instead of pinhole digital camera.

With the key parts obtained, design and assembly could begin. An initial approach was to keep the Brownie’s original lens and position the bare webcam sensor behind it. To achieve sharp focus, the sensor would have to be placed at the same position that the film would be relative to the lens. Each film negative, however, was 60x60mm in size, while the webcam sensor was less than 5x5mm. Testing confirmed that the webcam would have an incredibly narrow field of view, making it near impossible to compose shots with the crude viewfinder mirror.

The alternative was to disassemble the Brownie and remove its own plastic lens. The webcam circuit board could then be positioned such that its own lens was would be right behind the shutter. The circuit board was just a few mm to large to fit into the required position, so a dremmel tool was used to carve a slot in the inside of the case, allowing the circuit board to slip into place.

The interior of the Kodak Brownie, showing the circuit board in place immediately behind the lens. The circuit protrudes through a slot cut in the wall and held in place with bluetak.

This allowed the webcam to have a field of view similar to that of the original Brownie. In fact the field of view was wide enough that it covered the entire shutter aperture so the resulting images showed a circular vignette.

Still image captured by the webcam when behind the Kodak Brownie lens. The sensor field of view expands beyond the maximum size of the shutter aperture

The second task was deciding how to position to Raspberry Pi Zero in the case. As luck would have it, the width of the Pi Zero is exactly the same as the length of a 620 film spool, so the film holders were able to grip the Pi Zero circuit board. In common with the previous pinhole webcam, two LEDs were to be used, one illuminating when the power is on and one illuminating when the webcam is capturing an image. Two small holes were drilled in the top of the Brownie case next ot the shutter button, through which the LEDs could poke. A spot of superglue held the LEDs in the correct position.

Inside of Brownie case showing the Raspberry Pi Zero board, USB power cable and LED status indicators

One of the key goals for the camera design was that the shutter button on the Kodak Brownie should be used to trigger image capture on the webcam. Ideally a single press of the shutter button would capture a single image. To achieve this, the mechanical shutter button needed to interface with the Raspberry Pi GPIO pins in some manner. After thinking about this tricky problem for a while, a solution involving a pair of bare wires and some conductive paint was conceived. One wire would connect to a programmable GPIO pin configured as an input in pull-up mode. The second wire would connect to a GPIO ground pin. A hole was drilled through the case coming out immediately below the shutter button, through which the wires could pass. Insulation was stripped off the wires and they were superglued in position below the shutter button. Finally a blob of conductive paint was applied to the underside of the shutter button. The result is that when the shutter button is pressed, the conductive paint shorts out the two wires, pulling the GPIO pin to ground. This change can be detected by the Pi Zero and used to trigger the shutter.

Shutter button on the Kodak Brownie, showing a blob of conductive paint, used to short circuit wires used to trigger webcam capture

The Brownie shutter mechanism is designed for film with a fixed shutter speed. It was not practical to synchronize image capture with the precise fraction of a second that the shutter was open. Fortunately the Brownie has a long exposure mode where the shutter remains open for as long as the shutter button is pressed. Normally this long exposure mode is activated by raising a second button on the Brownie, but this is somewhat tedious. A little bit of electrical tape applied to the shutter mechanism was able to hook it into permanent long exposure mode.

The Brownie shutter mechanism with spot of electrical tape used to fix it in long exposure mode permanently.

Testing of the shutter mechanism was revealed a small problem – the webcam takes  a second or two to automatically measure & adjust exposure to suit the lighting conditions. The result was that if an image was captured immediately after pressing the shutter button, it would often be totally underexposed. This prompted another slight change in design. Rather than capturing a single image immediately as the shutter is pressed, the software was written to wait a second after shutter press and then capture images continuously thereafter, one per second, until the shutter was released. IOW, it would be a timelapse capture device.

The only remaining task was power. The Pi Zero and webcam combination has very low power requirements, at most 200 milliamps, and it was already known that a USB lithium ion powerpack provides an excellent power source that lasts a really long time. The problem is that for most powerpacks on sale today, physical size is not a hugely important factor. To date it has been impossible to find one that is small enough to fit inside the Brownie case – it would need to have a longest dimension of 6cm to stand a chance of fitting once the USB cable is plugged in – 5cm would be even better. Having the battery inside the case also adds a requirement to put a physical power switch between the Pi Zero and the battery, unless you want to open the camera to turn it on/off every time. The simple solution was thus to just drill a hole in the case for the USB cable, leaving the battery on the outside.

Finished Kodak Brownie digital camera showing cable to external battery pack

With the hardware construction complete, attention turned to the software to control it. Rather than start from scratch, the previous code used for the pinhole webcam Arcturus, was used extended. First it was discovered that the cheap webcam didn’t provide MJPEG capture, which meant pulling in libjpeg to do encoding of the raw frames into JPEG still image files. Capturing images in raw format means the USB device has to transfer far larger quantities of data. While this wasn’t a problem on the laptop used for development, the raspberry pi was continually getting dropped / incomplete frames from the webcam. After countless hours debugging, the problem was discovered to be the driver for the USB controller in Linux mainline, used by Pignus (the Fedora fork for the Raspberry Pi arm6 borads). The Raspbian kernel by comparison has an out of tree driver for the USB controller which turned out to work fine. The remaining software changes involved wiring up support for using the shutter button to trigger capture via a GPIO pin. The end result was that upon pressing the shutter button it would capture still images continuously, 1 per second. After a sufficiently long series of stills had been captured, they could be turned into a timelapse video sequence.

Building a UV LED light box for cyanotype and lumen printing

The cyanotype and lumen printing processes are two of the more frequently practised alternative photographic processes, because of their overall simplicity and the fact there is little-to-no need for equipment. Most people will do their exposures outside in the Sun initially, but if you want to work with these processes on a regular basis it can be frustrating to have ideas ready to try, but be delayed by lack of a sunny day. While it is still possible to expose on cloudy days, the length of time required to obtain a suitable image is greatly increased. Working with the sun it is also difficult to figure out predictable exposure times as the intensity varies according to the daily weather, time of day and time of year.

The solution is to switch to an artificial UV light source which can be used any time of day or year, whatever the weather, always giving the same exposure time. Historically though, UV exposure units have been relatively expensive to obtain, £100 or more. The ongoing developments in LED lighting technology though have now opened up new possibilities for constructing a custom UV light source for minimal cost. In particular it is possible to obtain 5m long strips holding 300 UV LEDs from online marketplaces such as eBay, for around £15 / $20 (search eBay for keywords “5M UV 5050 SMD 300LED“).

 5M UV (395-405nm) waterproof 5050 SMD 300LED strip, powered by a 12 V @ 5 amp supply

5M UV (395-405nm) waterproof 5050 SMD 300LED strip, powered by a 12 V @ 5 amp supply

The 5M long 5050 LED strips are 1cm in diameter and can be cut every 3rd LED. If they are cut into groups of 15 LEDs, this will result in 20 LED strips, each 25 cm long. Arranged side by side, this allows for creating a light source that will evenly expose a 20 cm x 25 cm area which is practically perfect for both A4 and 8x10in paper sizes. If one didn’t mind lower intensity it would be possible to left a 1cm gap between strips producing a source suitable for 16x20in / A3 paper, at the cost of longer exposure times.

When cutting the LED strip up, it is important to cut exactly in the middle of the metal contacts between every 3rd LED, as it will shortly be necessary to solder wires onto the metal contacts. If using the waterproof coated LED strip, the rubbery coating will have to be removed from the contact pads after cutting, which is quite tedious and an argument in favour of the non-waterproof variants. With the strip cut into pieces, it is now time to connect them back together by soldering short (5-10 cm length) wires between the metal pads. While it is possible to wire them all together in series to form one long strip, this means the link wires will be carrying the full 5 amp current load and if any link goes bad it risks taking out the entire set of LEDs beyond it. A better bet is to wire them up in parallel, or perhaps grouped in a mesh giving multiple paths for the current, so the link wires only need handle a tiny current and there is redundancy. The important thing when soldering the link wires is to preserve the polarity between strips – ie connect positive to positive, and ground to ground.

All 20 strips with connecting wires soldered on. Mistakenly all 20 strips are in series. This was later resoldered to put them in parallel

All 20 strips with connecting wires soldered on. Mistakenly all 20 strips are in series. This was later resoldered to split them in 5 groups of 4 strips, each group in parallel, reducing current in the link wires to 1amp.

The vendors of the LED strips will typically also sell suitable power supplies. These mains powered units need to output 12 volts and be capable of supplying 5 amps to enable the LEDs to run at full brightness. Lower amperage PSUs will work, but the reduced LED intensity will obviously increase exposure times, so it is best to simply get the right specification of PSU from the start.

Mains PSU for the LED strip able to supply 12 volt at 5 amp

Mains PSU for the LED strip able to supply 12 volt at 5 amp

The power supply will likely have either a 2.5mm or 2.1mm plug, so a correspond matching socket needs to be purchased. While it is possible to just turn the device on/off at the wall, or by pulling the plug out, a better bet is to put a rocker switch inline with the positive power line between the plug socket and the LED strip. Again make sure the rocker switch is rated to carry 5 amps.

Power output is given by multiplying voltage by current, so 12 volt at 5 amps will produce 60 watts of power output. This is going to generate a reasonably large amount of heat and if something is not done about this, it will gradually degrade the LEDs shortening their lifetime. The obvious answer to this is to obtain a couple of 12 volt computer fans to fit in the case of the final light box. A fan that is approximately 4cm x 4cm in diameter will be ideal. They’re quite cheap so you might even consider using a pair of fans. The wires from the fan can be connected in parallel with the LED strips, since they’re conveniently driven from the same voltage. Do NOT connect them in series with the LEDs, as the 5 amp current draw of the LED circuit will kill the fans. Also be careful to get the positive/ground polarity right when connecting the fan, as reversing polarity will NOT make the fan run in reverse and likely kill the fan too.

12 volt computer fan, 4cm in diameter

12 volt computer fan, 4cm in diameter

The case for the light box will be made out of wood and comprise two pieces, a base which will hold the paper to be exposed and a slightly larger lid which will hold the LED panel. Both will have sides and be sized so that the base nests snugly inside the lid (or vica-verca). The top and bottom panels can both be cut from a sheet of 3mm plywood, the lid panel being 39×30.5cm and the base panel 36.5×25.5cm. These sizes are fairly arbitrary – the smaller simply needs to be about 2 inches larger than the size of paper to be exposed on each side. So for 8×10 paper, the smaller would want to be about 12×14 inches. For the larger lid, sides were cut from a length of 70x18mm timber, and nailed to the plywood panel. For the smaller base, sides were cut from a length of 36x10mm timber. In the timber sides of the lid, two 4x4cm holes were cut to hold the fans. Two holes were also drilled in the lid, one for the power supply plug socket and the other for the on/off rocker switch. When inserting the fans in the case, one should be oriented so that it sucks air into the case while the other should blow air out of the case, creating good airflow across the LED panel.

Light box lid showing the on/off rocker switch through the panel and power supply socket in the side

Light box lid showing the on/off rocker switch through the panel and power supply socket in the side

Close up of the lid, showing the computer fan inserted in the side to pull air across the LEDs for cooling.

The LED strip usually comes with a self-adhesive backing tape which is supposed to be able to stick the LEDs to most surfaces. This proved insufficiently sticky for me, so I applied super-glue instead. While the LED strips could be attached directly to the lid of the light box, it was thought preferable to attach them to a sheet of perspex or aluminium to allow the LED sheet to be separated from the case if needed. If using metal just be careful to avoid any short circuits with the link wires of the LED. Once the LEDs are attached, the sheet can be fixed to the inside of the lid with a couple of screws.

The larger lid, showing the metal plate with LED strips attached. At either end are cardboard shields to block UV light leakage through the fans.

The light box base

The smaller lightbox base, sized to be able to hold an 8×10 inch picture frame from a pound shop. Note a couple of screws sticking out of each side of the base, to prevent it sliding completely inside the lid when nested

When first turning it on, there was some UV light leakage through the cooling fans. Thus a couple of shields were cut from heavy duty cardboard and duct taped over the fan openings. With this in place there is no significant UV leakage from the light box, due to the closely nested lid and base. The UV LEDs are emitting at the end of the UVA spectrum, quite close to the start of the visible light spectrum, so the light is not a serious danger like UVB light would be, but it is none the less worth taking care to avoid accidental exposure.

In use the light box has proved to be intense enough to expose acceptable cyanotype images in as little as 5 minutes, and lumen images in anywhere from 10 minutes upwards depending on the visible effect desired. This is considerably faster than many commercially obtainable UV light sources that photographers have used in the past, which could take 15 to 30 minutes or even more. All together the cost of the complete box was probably around £45 – if you already have some parts in the shed such as plywood / timber pieces and a suitable power supply, then the price could be around £20-25. Either way, it will easily beat the cost of commercially produced light boxes and likely perform better too. The hardest part in construction is probably the soldering of the 50+ link wires between the LED strips. The case needs only minimal wood working skills – use of a saw and hammer. In summary creation of the light box is a very worthwhile use of time and money and will proof useful for years after.