So the start to this term did not quite go to plan. In the first week of term I was hampered in my progress by an OFSTED inspection and then the following week I was afflicted with a bout of appendicitis. This resulted in an admission to hospital and an appendectomy the following day (and I though it was just wind!). So now I am stuck at home unable to go to work and dosed up on painkillers.
However all is not bad, before my illness I was able to do a bit of testing with the camera to see how it would perform. I also managed to test some different solutes for making the crystals with my Home Educated students to see which would be best for our project.
Also the Kit has arrived which means I can start to properly set up the experiment I had planned.
The Kit on Test
The kit for this Road Test is the Raspberry Pi Model B+ Camera Kit in the UK this is branded U:Create and is available from CPC.
The kit comes with a Raspberry Model B+, A Raspberry Pi NoIR Camera board, an 8Gb SD card with NOOBS pre-loaded, a power supply with plug, a Wi-Pi 802.11n Wifi Adapter, a clear case and 36 page user guide.
I will go into more detail about the kit in the review but here I will say that the kit is a really good starter kit. With the pre-loaded OS and user guide it is really simple to get started even if you had not used the Raspberry Pi before. The user guide covers a range of topics from loading the OS through simple Unix commands to some project ideas including physical computing and Minecraft Pi. This is not enough to become an expert but a really good starting point for a beginner with the RPi.
For my planned project I am particularly pleased with the ability to mount the camera securely to the case giving a nice solid platform for photography (no more blue-tack!)
Testing - Crystals
With the Home Educated students (4 - 7yo) I planned to experiment with different crystals to see which would work best for this experiment. After looking around the kitchen and a discussion we decided on sugar and salt for the first attempts. These are thngs that are easily available in most homes so would be best for a Home Education based project.
After spending a large amount time thinking about what colours would be best (we added a small quantity of food colouring to make the experiment look more interesting. I finally solved the colour issue and was able to move on to making a couple of saturated solutions. One of sugar (red) and one with salt (blue). A small sample of each solution was added to a petri dish (a small bowl or saucer would also work). To make a saturated solution it is best if you can hest the liquid whilst you add the solute. In a school we would do this with a beaker on a tripod over a bunsen burner but in the world of Home Education in meant using a saucepan. All you need to do is heat the solution and then add the chosen solute, then stir and heat stir and heat until no more solute will be disolved into the mixture. The better the saturated solution the quicker the crystals will start to form so it is worth taking a bit of time to ensure the solution is properly saturated.
The aim was to test the kit and try different positions for the camera as well as seeing how long the crystals took to form. This would allow us to practice and see how long we would need to run the full timelapse sequence for and to see how best to mount the camera. As the tests were done before the arrival of the kit the camera i had was attached to the side of a monitor using blue tack and the raspberry Pi balanced as best as possible to support this position.
After around 48 hours the crystals had formed and the liquid mostly evaporated from the solution.
The salt crystals were the easiest to see on the Pi Camera and showed a more obvious formation. The sugar crystals did form but was much less visible to the Pi Camera. When we looked closely at the results the sugar had mixed with the colour to make a syrupy solution that did form crystals but they were all stuck together. When I cleaned the dishes afterwards they was still a thin layer of solution under the top layer of crystals which had made the crystals less easy to see.
Out of the two solutes we tried here salt was the best. It does have quite small crystals formed but they were more easily visible. That may have been a little due to the colour choice but overall the salt performance was better. We are looking to see if we can find copper sulphate at our local garden centre to try as well as that would be the solute of choice if I were doing this at school.
As for timings the process was fully complete after 24 hours with all the liquid evaporated from the salt and a solid shell formed over the sugar.
Testing - Timelapse sequence
The idea or doing the testing was also to experiment with the timelapse sequence interval and duration as well as experimenting with lighting and camera position.
After a discussion of how long it would take to form the crystals and how we should set up the camera the students decided on a overall period of 5 hours with a picture every 15 minutes. this produced the first video shown below (very rough composite knocked together using windows movie maker):
This was clearly not enough to show the results and the lighting (normal house light (which was turned off for a short while during the experiment by mistake) was clearly not sufficient. This showed the importance of a secure light source that would not disturb the household whilst we try to sleep.
A bright white light (an e-reader clip on LED based lamp) was added to give better illumination and the timelapse set off again.
This time (especially after the LED light was added the picture is much clearer. There was a little Blue tack release moment and a small hand cameo at one stage but the effect is much better.
However the camera appears to be too far away to get a really good view of what is happening (this is slightly later on in the crystal formation process as was still using the same solutions). The LED light gives a good illumination and also has the bonus of reflecting off the crystals as they form highlighting them in the view.
When the camera was placed closer it was not in focus so we will need to experiment with changing the focus of the camera (fairly simple just need to turn the bevel braking the glue then adjusting to suit) to get a picture with one petri dish filling the screen.
From the experiments we have discovered that the best lighting was achieved with the LED reading lamp and that some adjustments would need to be made to get a good focused picture that filled the screen. Of the solutes we tested the salt with a light colouration was the most likely to produce a good result.
The blue tack mounting method was a little too haphazard even fro testing, fortunately with the secure mounting of the camera in the case provided with the kit this should not be a problem in future tests.
Obviously the results are very rough at this stage and the production of the videos needs more finesse to produce a good end product but this was not a bad first attempt for the age of the students. it also would be good to have more pictures for the change time period so a smaller interval between pictures will be used. We also know that the experiment was mostly complete over a 48 hour period. This means we have a firm time period for the experiment and we can concentrate on experimenting with time intervals to get the best result vs file size compromise.
-Set up the focus of the camera and mount the case securely to get a frame filling in focus picture of the solution.
-Experiment with different time intervals
-Experiment with copper sulphate (if possible to obtain easily locally).
-Experiment with different ways of combining the pictures. FFMPEG, mencoder and AVconv are on the list to try directly on the RPi to have a fully RPi Project. I will also try Go Pro Cineform software mentioned in mikedavis blog Time Lapse of the Chicago Sky and have a look at other PC software options available.
-Try to get better to try the experiment with my School group . I have set up a short scheme of work to use with them and am keen to get the chance to try it out.