Unfortunately we did not hit our fundraising target, and it's now a dead project. I'd love to see this new Kickstart project succeed, especially the online spectral database. Commercial databases of spectral signatures are expensive (thousands of dollars), and they shouldn't be.
Part of our problem, I think, was that we were attempting the project while taking classes full time, and working.
We were planning to release two different designs at different price points. One would have used inexpensive ruled transmission grating, while a more expensive unit would have used a concave holographic aberration-corrected reflective grating. The latter would have had better system efficiency, less "smear" across pixels, and better linearity. Each would have made use of a linear CCD array with good sensitivity (the same chip used in the Ocean Optics units). Concave gratings add a couple hundred bucks to the price, but are well worth it if you need to do UV.
A challenge we didn't get to solve was how to make something DIY that could be used to detect UV. It's difficult to sputter a uniform phosphor coating or remove a sensor window without specialized equipment.
"Part of our problem, I think, was that we were attempting the project while taking classes full time, and working."
I think your project was fantastic.
I see the main problem is the price. Most of the normal people can't justify expending $800 on an spectrometer. They don't even know what they could use it for.
People that could justify it because they need it will buy tested equipment anyway, they have money but not time.
So you need to fill the gap. The new project is making it super cheap so they could get traction like Reprap machines. At first they are terrible, but with enough people and huge market improvement is so fast like with the PC that at some point in the future it could get better than commercial closed systems.
Start small, if you can't detect UV it does not matter, remember it is 100x better than Newton ever had.
The $800 price tag was just a kickstarter only price, since we were still largely in the development phase we weren't sure of final pricing. Now it looks like a UV capable scientific instrument with a chinese grating would cost around $400, if you want the same optics but U.S. made it would be around $900.
I'm no optical scientist, but seeing as the diffraction grating being used is recycled from a DVD and the sensor is a consumer webcam, I wonder how accurate/reproducible the results would be.
Gratings like this (or a CD) will work, but they introduce plenty of artifacts. Scratches and material nonuniformities can scatter light in a way you don't want, and reduce the (already limited) amount of light hitting the right spot on the sensor.
I applaud the approach, but it seems like a challenge to achieve useful accuracy across many devices. I am a little skeptical about the 3nm claims.
First, the (bandpass) Bayer color filter arrays on cameras limit their color resolution pretty sharply. It might be a better (if less elegant) approach to use the ambient light sensor and scan it across the spectrum generated by the diffraction grating.
Also, the resolution of a spectrometer is in part determined by entry slit width. Old timers and DIY'ers use abutted razorblades to make a clean, straight slit. Problem here is that there is a trade-off. You need a wide slit to get light in for these insensitive webcam sensors. But then you lose spectral resolution because your slit is wide. Alternately, you narrow the slit to get more spectral resolution, but then your image is dominated by noise. I'm sure these guys have found the sweet spot.
Second, the spectral response of the color filter arrays varies both from camera to camera and even within same-model sensors (I personally own over 40 of one model of Canon Powershot, and the differences in color response are fairly extreme). This could in principle be at least partly calibrated out, if you had each device take a picture of a calibration standard.
Third, the image signal processors are closer and closer to the sensor these days. It may be impossible to get data which hasn't at least been somewhat mucked with, meaning some people will have a useful spectrometer and some won't.
This goes doubly for phones with bad imaging pipelines, where the image processing is out of your control. Lots of spatial and spectral processing going on. It will be interesting to see, for example, how the auto white balance and sharpening algos deal with having zero natural image content in the captured data.
Aside from the within-sensor variation/image processing problems it would definitely be usable for relative data - I see that they used a relative approach on their UV detection of a bluing dye in detergent. Nice.
All that said, I run a somewhat-similar project using cheap cameras to do work previously only reserved for hard-core SLRs and scanning line sensors, and I've found that over and over again, something is vastly better than nothing and often "good enough" is better than expensive, complicated, and pain in the ass. I've also dealt with a lot of people laying out a bunch of complicated technical "Here's why it won't work" talk and have learned to counter with my own "But look, it does work" answer.
Blah blah: I think it's important for the capture software captured as much metadata about the image sensor as possible. Their software looks great and a database of spectra would also be great. It would be AWESOME if the Shazam-style auto identification worked at all. Good luck, I'll probably be supporting the project.
>First, the (bandpass) Bayer color filter arrays on cameras limit their color resolution pretty sharply. It might be a better (if less elegant) approach to use the ambient light sensor and scan it across the spectrum generated by the diffraction grating.
The diffraction grating spreads the different frequencies physically across the sensor. So the camera sensor doesn't need any color information, you only need position and brightness information.
Right, but the different physical locations on the sensor have different spectral sensitivities due to the bayer CFA. Ideally, you'd have panchromatic pixels with no bayer pattern.
Oh I hadn't thought of that. I assume some calibration would be required for each device anyway, I wonder how hard it would be to detect and correct for this effect?
using a 2D sensor, you could actually just block one half of the slit at a time with your sample, and let the other half of the slit pass through unchanged... that way you could get a calibration shot for every sample, assuming the dynamic range of the sensor was high enough that the calibration lines didn't bleed (or conversely that your sample didn't get buried in noise because exposure wasn't high enough)
People have done it for a variety of purposes (natural image statistics, astronomy, etc). I can post references later. Problem is that it almost universally requires you to photograph a standard light source or a standard reflectance target under a standard illuminator. Or a monochromator. So you can calibrate your cheapy device but you need an expensive/scarce standard.
It's an interesting problem and I look forward to seeing how they solve it.
Back in the day, standard illuminants were candles made from spermaceti--the wax substance from the cavity of a sperm whale (it was also used as space-grade lubricant until very recently). Nowadays, CIE Illuminant D represents a sort of model of natural daylight. I bet a blue sky sunny day could be used for calibration, if the instrument is pointed straight up and a function corrects for lon/lat and time of the year.
For reflectance measurements, the highly-lambertian Spectralon material is basically just well-controlled Teflon sheet. Some PTFE sheet from McMaster might do the job for DIY purposes.
You want a "peaky" light source to build the spectral response model, xenon or deuterium flashbulbs are cheap on amazon... I would lean towards using deuterium because I think the relative power of the peak to the blackbody floor (the rest of the light freqs) is less than the xenon, so you'd have less chance of the peaks dominating your sample with a less sensitive webcam type sensor
B&W security cameras are inexpensive and lack Bayer masks. I've used one to make a spectrometer before. Pop the lens off the front and you're good to go.
Edit: some of them have ethernet or other fun connectivity options. Better for getting data out of a DIY instrument than screen capture.
Of course, any understanding of spectroscopy should include some background of the physics behind what you're measuring. Any modern physics book would be fine: Krane, etc.:
I think daniel_reetz put it right with "good enough". This isn't a lab or scientific instrument. But its cheap enough to use as a teaching tool, to get people interested and in-the-know about light and its connection to common life.
What kind of spatial resolution would be needed to get a good scan using a diffraction slit when moving the sensor across the spectrum?
Likewise, how wide is a typical slit when using razorblades?
If you had access to the unprocessed RGB data from the camera, wouldn't this provide the same information (for the visible spectrum) without needing to split the light?
The spatial resolution would depend on proximity to grating. The father away, the larger the pixels or scan integration areas (or, step size) could be. Most compact spectrometers fold the optical path using mirrors to increase the path length. Some also include a concave mirror to match the Rowland circle and project a more linear distribution of frequencies.
When I do this with a Project STAR spectrometer and my cameras and point it to a candle I get two distinct dips in the spectral distribution. It should be even like a blackbody. I have made a parser for the RGB-picture to turn it to a B&W power distribution and I thought I'd might do a simple software filter to make it completely even, like a calibration of sorts. Like so: http://www.flickr.com/photos/defdac/7874867702/in/photostrea...
I use the power distributions coupled with lumen and wattage-specification as input for my home made spectrally based global illumination renderer to calculate PAR in aquariums.
Do you think this will work?
Best of luck with the project. I think the really good idea is the open-source spectral library.
You need to educate your readers a bit more, though. They need to be aware of the kind of spectroscopy the machine does, i.e UV-Vis, and what information it provides.
I might try DIYing it, but what would be really interesting is an IR spectrometer, under 250$. Tough.
What are some of the cool experiments one can do with a spectrometer, other than the ones mentioned here (I especially liked the wine analysis application)?
If you can transmit light through a sample curvette to the spectrometer, you could use it to measure optical density (i.e. the amount of stuff in the way that blocks or scatters light). You could use that to measure the concentration of particulate in engine oil, or the strength of coffee.
If you can detect at 405nm, you could measure blood coagulation time/extent (separate blood cells from plasma, measure attenuation at 405nm though plasma).
You could cludge together a pulse oximeter if you can detect at 660nm and NIR (905, 910, 940nm) and have light sources to emit at these wavelengths. Absorbtion at these wavelengths changes as hemoglobin picks up and loses oxygen. The ratio at 660nm and one of the other wavelengths could be used to show blood oxygenation.
If you were into photography or cinema, you could measure the emission spectrum of a light source and fit a blackbody curve to it to find its color temperature--useful for white balancing.
If you wanted to match paint colors, you could use a spectrometer to do (along with some additional calculation and knowledge about your pigments).
If you perform a flame test, you could use a spectrometer to learn about elemental composition.
http://www.openspectrometer.com/
http://www.kickstarter.com/projects/makerhaus/open-spectrome...
Unfortunately we did not hit our fundraising target, and it's now a dead project. I'd love to see this new Kickstart project succeed, especially the online spectral database. Commercial databases of spectral signatures are expensive (thousands of dollars), and they shouldn't be.
Part of our problem, I think, was that we were attempting the project while taking classes full time, and working.
We were planning to release two different designs at different price points. One would have used inexpensive ruled transmission grating, while a more expensive unit would have used a concave holographic aberration-corrected reflective grating. The latter would have had better system efficiency, less "smear" across pixels, and better linearity. Each would have made use of a linear CCD array with good sensitivity (the same chip used in the Ocean Optics units). Concave gratings add a couple hundred bucks to the price, but are well worth it if you need to do UV.
A challenge we didn't get to solve was how to make something DIY that could be used to detect UV. It's difficult to sputter a uniform phosphor coating or remove a sensor window without specialized equipment.