More important than ever

Hey all - I know everyone is busy, as am I. Our Sci, our new company, is getting off the ground and it’s taken all my effort.

But I wanted to share as to why I think this single pixel camera idea is even more relevant -->

Until the last few years, if you wanted a ‘single pixel’, you could find inexpensive ones in the visible range (350 - 950nm like the mini spec I shipped with the kits), however, higher or lower than that was often quite hard for two reasons - 1) if you need a light to illuminate the image, LEDs in the 200 - 300nm (UV) range and 1050nm and above (Mid-IR) were very expensive, and 2) if you needed a sensor to detect those ranges it was even more expensive.

Furthermore, there are all kinds of things that correlate and are visible in the mid-IR and UV-B and UV-C range - in many ways much more so than in the visible range. So making these ranges cheap opens a massive number of applications.

But things are changing:

  1. There are now UV LEDs in the 260 - 290nm range for < $50/led, and in a few years they’ll be <$10/led (here’s an example datasheet -
  2. There are now mid IR detectors from 1350 - 1850nm in the $100 - $200 range, which have 30nm resolution - so they are like low-resolution spectrometers in the mid IR range!!! And, the word on the street is, soon this will be 1350 - 2150nm !
  3. There are now mid IR LEDs in the 1300 - 2100 range as well.
  4. Even in the normal visible and NIR range, components are getting more integrated and cheaper - sparkfun has 2 breakout boards which cover the range from 400 - 850nm - if you bought the modules directly you could get it for < 15 dollars. These modules include the detectors, LED controller, and signal conditioning (that’s a lot of circuitry you’d otherwise have to build yourself).

So some of these parts aren’t really cheap yet (<$10 dollars), but they are getting close, so imagine this:

A single pixel camera which imaged from 260nm to 2100nm for under $2000 dollars. That is completely possible in the next 2 years.

There are so many pieces of crazy expensive lab equipment this device could replace, from plate scanners to reflectance spectrometers and beyond. Besides the fact that it could be used in field and outdoor environments for monitoring way way way beyond the current capacity on the market.

Anyway, I’m sorry I haven’t been able to do more active development on this, but I want to at least note to folks that creating the backbone for future sensors in a general purpose single pixel camera setup is more relevant and useful as ever, and wherever I go I’m pushing folks to try to fund the development so I can spend some time on it, and I’d suggest you do the same!


Thanks Greg! I’ve passed this on to a couple of people I think might be interested.
From the perspective of funding funding, are there one or two applications you think would be most interesting/useful/impactful? I’ll keep an eye out for opportunities.


Let me think about it. Maybe I’ll post a review here of the applications, as just saying ‘there are lots’ isn’t super convincing :slight_smile: I guess it’s lit review time.

I agree it is perspective. I had opportunity work with the device on basis X-ray fluorescence. Impressive.
Yes, it is X-ray, but no less interesting range are UV-A,-B, -C and IR (Raman spectrometers)

The commercial space is expanding it’s capacity into the longer-wave NIR -

Still closed and very expensive, but I can say from my work with food and soil the 1000 - 2000nm range is critical for identification of all kids of useful stuff.

1 Like

Interesting realization on InGaAs detector: “Acquire spectrum transmittance data across a 900 to 1700 nm wavelength range in less than 1 second with >6,000:1 signal-to-noise ratio (SNR)”

1 Like

For inspires: … “Handheld spectroscopy is still in its infancy”

Yeah, this is mostly about consumer physics ScIO which is immensely hyped but hasn’t panned out that much. I know some of the folks involved, and can tell you there is nothing magical there.

Handheld raman, on the other hand, is super exciting. The ‘infancy’ part definitely applies but the potential is huge. There is a person who’s working on a low-cost raman, I think that’s a great avenue for work in GOSH.