YOU CAN MEASURE THE TEMPERATURE OF A FLAME WITH A CAMERA, FOR 1,208 DOLLARS
A 1,208-dollar optical splitter that puts four simultaneous views on one camera sensor, so you can measure flame temperature pixel-by-pixel without synchronising separate cameras.
by Abinash Sahoo, Ryan D. DeBoskey, Venkateswaran Narayanaswamy
ScienceOpen-hardware
Built with3D printing
- difficulty
- ●●●●●
- time
- several weeks
- license
- CC BY 4.0
- repo
- repo FINISHED0 stars
●●●●● · several weeks · CC BY 4.0 · 0 stars · repo FINISHED
WHAT YOU’LL NEED
Jump to section
COMPAREE VERDICT
This is not a weekend project. It is a validated research instrument published in a peer-reviewed hardware journal, and it solves a real problem: commercial image doublers from LaVision cost around 8,000 to 11,000 dollars. The quadscope uses 3D-printed holders with off-the-shelf silver mirrors to route up to four views of the same scene onto a single camera sensor simultaneously — the key to two-colour pyrometry, where you measure flame temperature by comparing brightness at two wavelengths pixel-by-pixel. The 1,208-dollar figure is for the splitter alone; with the narrowband spectral filters needed to replicate the validation it comes to 2,678 dollars. You will also need a high-speed camera (not included), an optical bench, and genuine alignment skill. The validation was a non-premixed methane/air flame imaged at 500 Hz with a Photron SA-X2. If you are setting up combustion diagnostics or any experiment where perfect simultaneity matters more than the cost of multiple cameras, this is exactly what you need and the documentation is complete. If you have never aligned a multi-mirror optical path before, expect this to take longer than you think.
IN THE REPO
GOOD TO KNOW
- —Design files (SolidWorks and STL) are on Mendeley Data under DOI 10.17632/3pmyf5vghd.1; the assembly and alignment procedure is in the paper.
- —Bill of materials with suppliers and prices is in Table 2 of the paper.
- —No firmware — this is an entirely passive optical assembly.
- —Detailed assembly procedure, alignment protocol and validation methodology in the HardwareX paper (DOI 10.1016/j.ohx.2025.e00723).
- —Licence is CC BY 4.0, no commercial restrictions.
- —The 1,208-dollar cost excludes spectral filters (add 1,470 dollars for those), a compatible high-speed camera, and an optical bench.
Parts to buy
6 itemsFrom our check of the build. Exact quantities and part numbers are in the creator’s BOM.
Can I build this?
Build at your own risk. Projects involve tools, electronics and sometimes mains voltage — follow the creator’s safety notes.
More builds like this
All projectsGallery
Start here
Navigation into the creator’s own docs — we don’t rewrite the guide, we route you to the source.
- 1.Read the full HardwareX paper (Section 4 has the bill of materials, section 5 the build instructions, section 6 the alignment and image-mapping procedure.)
- 2.Download the design files from Mendeley Data (SolidWorks and STL files for the view splitter, mirror adapter and filter holder. Image mapping is done in your own software (the authors used LaVision DaVis; MATLAB or the open-source CalVi are alternatives).)
- 3.Source the bill of materials(Table 2 in the paper lists every part with supplier and price. Thorlabs is the primary vendor; the paper names Edmund Optics and Newport alternatives.)
- 4.Prepare your optical bench and camera mount(You need a stable platform and a compatible camera — the validation used a monochrome Photron FASTCAM SA-X2 at 500 Hz with a Nikon 50 mm f/1.2 lens.)
KNOWN ISSUES
- The 1,208-dollar cost is for the splitter only — spectral filters add 1,470 dollars, and you still need a high-speed camera and optical bench, so budget accordingly before ordering parts.
- Alignment is not forgiving: the four guiding mirrors sit at about 45 degrees and are tilted by hand until all four views of the dot grid overlap, and over-tightening the set screws can crack the thin mirrors - the paper advises half, then quarter turns.
- The four views share one sensor, so each gets a quarter of the resolution. That is fine for pyrometry where you are computing ratios, but do not expect full-frame sharpness in all four quadrants.
- Narrowband interference filters are wavelength- and angle-sensitive. If you change the beam path or use a different lens, you may need to recalibrate or respecify the filters.
- The paper's validation was on a lab burner in a controlled environment. If you are measuring something that moves, vibrates, or emits soot that coats optics, expect extra maintenance.
- No firmware and no electronics mean there is nothing to debug in software, but it also means you cannot tweak alignment digitally — every correction is mechanical.
Can I use this with a regular DSLR instead of a high-speed camera?
Yes, for static or slow processes. The splitter itself does not care about frame rate — it is passive optics. You lose the simultaneous high-speed advantage the paper demonstrates, but you still get four views on one sensor.
Do I need all four channels, or can I build a two-way version?
Two-colour pyrometry only needs two wavelengths, and the paper's optional filter holder carries two spectral filters. A two-view splitter is not described, so dropping two beam paths would be your own modification of the CAD. The 1,208-dollar figure is for the full four-view unit; with the paper's spectral filters it is 2,678.
What if I cannot source the exact Thorlabs parts?
The paper names alternatives: Edmund Optics 25 x 35 mm silver mirrors (3 mm thick instead of Thorlabs' 1 mm, so you adjust the SolidWorks holder), and Edmund or Newport for posts, post holders, clamps, base plates and mirror holders. Bandpass filters can come from Thorlabs or Edmund.
How do I know if my alignment is correct?
Section 6.1 of the paper walks through it: position the splitter in front of the lens (a sheet of white paper behind it helps), set the four guiding mirrors at about 45 degrees, then tilt each kinematic mirror until all four views of a dot-grid target line up on the sensor. Image mapping (distortion correction) is done afterwards in software.
Can I measure things other than flames?
Absolutely. Any process where you need simultaneous multi-spectral or multi-polarisation views on one sensor is a valid use case — plasma diagnostics, explosions, LEDs under test, anything where separate cameras cannot be synced tightly enough.
Community builds
No community builds yet — be the first, we feature the best ones.
Discussion1
FROM THE COMPAREE TEAM
The paper maps the temperature of a methane flame — peaking around 2,400 kelvin — with all four views captured on one high-speed camera at 500 frames per second. What would you point this at?
Abinash Sahoo, Ryan D. DeBoskey, Venkateswaran Narayanaswamy
Built at the Department of Mechanical and Aerospace Engineering, NC State University, to make simultaneous multi-spectral imaging accessible without commercial beam splitter costs. Published in HardwareX as a validated, reproducible research instrument.
DISCLAIMER
- Comparee is not the author of the projects featured here. All rights to each project belong to its creator — every page links to the original source, and we never host creators’ files.
- Information is provided without warranty and may become outdated as projects evolve. Prices are indicative bands only — always check the creator’s parts list for current costs.
- Building and operating any project is at your own responsibility. Protective equipment, safe workshop practice and compliance with local regulations are the builder’s responsibility.
CompareeTEAM23d agoedited
Practical notes from our verification: the HardwareX paper is unusually complete — SolidWorks and STL files for the view splitter, mirror adapter and spectral filter holder (CC BY 4.0, in a Mendeley Data repository), a bill of materials table, and a step-by-step assembly and alignment procedure. The 1,208 dollar figure covers the splitter and mirror hardware; with the optional spectral filters needed for two-colour pyrometry, the paper's total rises to 2,678 dollars. The camera is not included — the validation used a Photron FASTCAM SA-X2 high-speed camera at 500 Hz — and image mapping was done in LaVision DaVis, with MATLAB or the open-source Python tool CalVi named as alternatives. The fiddly part is alignment: the final step is tilting four kinematic mirror holders by hand until all four views of a dot-grid target line up. For context, the paper lists LaVision's commercial image doublers at around 8,000 and 11,000 dollars. Correction (4 October 2026): we re-checked this page line by line against the project's own repository, documentation and videos, and fixed errors in earlier versions.