THE ARCHIVE — THE 20,000 FORMULATIONS
The untold story of 19,700 failures and the 300 breakthroughs that created PrismPixel Glass™
ITHE TIMELINE
Red is what was scraped off. Gold is what stayed on the glass. Hover or touch a year to open its ledger; the ratio above each node is the cumulative success rate at that point in the journey.
2009
BATCHES FM-00001 → FM-00420
The first pane went opaque on the first pass. Everything after that was learning what light refuses.
BATCHES THIS YEAR
420 sprayed · 420 scraped · 0 kept
PARTICLE TYPES TESTED
Glass microspheres · Talc · Kaolin clay · Calcium carbonate
NANOPARTICLE CONCENTRATIONS
18–30% vol · 40–90 µm
ADHESIVE FORMULATIONS
PVA (craft grade) · Acrylic latex
SPRAY PRESSURE
1.4–3.6 bar
CUMULATIVE TO 2009
420 tested · 420 failed · 0 breakthroughs
100.00% failure · 0.00% success
IITHE FORMULATION EXPLORER
Each card is a pane that was sprayed, dried, walked around and judged. Filter by verdict, by era, or by what was in the suspension.
FM-00001
TESTED 2009
First pass. Craft-store glass beads in PVA; the pane went fully opaque. The medium announced its terms.
FM-00189
TESTED 2009
PVA clouded within a week in 60% RH.
FM-00224
TESTED 2009
Runs and sags at 30 cm; the binder never grabbed.
FM-00847
TESTED 2010
Excessive hotspot, delamination risk.
FM-01054
TESTED 2010
Coating lifted whole when the pane was tilted.
FM-01062
TESTED 2010
Runs and sags at 30 cm; the binder never grabbed.
FM-01353
TESTED 2011
Colour shift toward warm; TiO₂ scatter too strong.
FM-01962
TESTED 2011
The first image that survived a walk-around. Hotspot still there, but bounded.
MTF
0.41
EXTINCTION
71.0%
TRANSMIT.
29.0%
FM-02172
TESTED 2011
Uneven build — 22 µm at centre, 9 µm at edges.
FM-02672
TESTED 2012
Adhesion failure: crosshatch test graded 0B.
FM-02699
TESTED 2012
Excessive hotspot, delamination risk.
FM-03197
TESTED 2012
Nanoparticle agglomeration; visible speckle at 1 m.
FM-03455
TESTED 2013
Flaked on thermal cycling, −10 °C to 60 °C.
FM-03586
TESTED 2013
Flaked on thermal cycling, −10 °C to 60 °C.
FM-04414
TESTED 2013
Adhesion failure: crosshatch test graded 0B.
FM-05136
TESTED 2014
Viewing cone 150° with no visible speckle at 1 m.
MTF
0.52
EXTINCTION
76.7%
TRANSMIT.
33.7%
FM-05330
TESTED 2014
Adhesion failure: crosshatch test graded 0B.
FM-06108
TESTED 2014
Uneven build — 22 µm at centre, 9 µm at edges.
FM-06973
TESTED 2015
Adhesion failure: crosshatch test graded 0B.
FM-07084
TESTED 2015
Uneven build — 22 µm at centre, 9 µm at edges.
FM-07462
TESTED 2015
Binder yellowed within 72 h under UV.
FM-08079
TESTED 2016
Extinction 98.9%. Two batches later the binder lot changed.
MTF
0.53
EXTINCTION
82.0%
TRANSMIT.
35.5%
FM-08091
TESTED 2016
Uneven build — 22 µm at centre, 9 µm at edges.
FM-08776
TESTED 2016
Moiré against a 4K pixel grid.
FM-09718
TESTED 2017
Flaked on thermal cycling, −10 °C to 60 °C.
FM-09998
TESTED 2017
Solvent attacked the previous layer; wrinkle across the field.
FM-10588
TESTED 2017
Cure shrinkage cracked the film at 18 µm.
FM-11363
TESTED 2018
Trace TiO₂ corrected the warm shift to neutral D65.
MTF
0.63
EXTINCTION
85.2%
TRANSMIT.
38.6%
FM-11431
TESTED 2018
Bimodal ratio drifted; speckle returned at 2 m.
FM-11628
TESTED 2018
Edge halo under off-axis projection at 35°.
FM-12752
TESTED 2019
Transmittance and scatter balanced at a 15 µm build.
MTF
0.66
EXTINCTION
88.8%
TRANSMIT.
39.6%
FM-12926
TESTED 2019
Trace TiO₂ overshot; D65 neutrality lost by 180 K.
FM-13210
TESTED 2019
Bimodal ratio drifted; speckle returned at 2 m.
FM-14140
TESTED 2020
Crosshatch adhesion 5B; survived 200 thermal cycles.
MTF
0.65
EXTINCTION
90.0%
TRANSMIT.
43.0%
FM-14219
TESTED 2020
Sol-gel cured glass-hard but clouded on the third day.
FM-15427
TESTED 2020
Sol-gel cured glass-hard but clouded on the third day.
FM-15521
TESTED 2021
Crosshatch adhesion 5B; survived 200 thermal cycles.
MTF
0.72
EXTINCTION
92.7%
TRANSMIT.
42.6%
FM-15819
TESTED 2021
Edge halo under off-axis projection at 35°.
FM-16763
TESTED 2021
Haze 4% over spec — binder lot variance.
FM-16873
TESTED 2022
Trace TiO₂ corrected the warm shift to neutral D65.
MTF
0.71
EXTINCTION
93.5%
TRANSMIT.
45.8%
FM-16883
TESTED 2022
Hotspot extinction stalled at 96.8%; the last 2% would not come.
FM-18066
TESTED 2022
Sol-gel cured glass-hard but clouded on the third day.
FM-18478
TESTED 2023
Bimodal SiO₂ flattened the gain curve; viewing cone widened to 120°.
MTF
0.76
EXTINCTION
95.6%
TRANSMIT.
47.8%
FM-19369
TESTED 2023
First formulation to survive 1,000 h UV without yellowing.
MTF
0.75
EXTINCTION
96.5%
TRANSMIT.
47.1%
FM-19393
TESTED 2023
Bimodal ratio drifted; speckle returned at 2 m.
FM-19665
TESTED 2024
Extinction 98.9%. Two batches later the binder lot changed.
MTF
0.79
EXTINCTION
99.5%
TRANSMIT.
46.8%
FM-19745
TESTED 2024
Image held from 3 m; hotspot bounded to a 6 cm disc.
MTF
0.78
EXTINCTION
98.3%
TRANSMIT.
46.9%
FM-19926
TESTED 2024
Transmittance 51% but MTF fell below 0.7 — too little scatter.
FM-19983
TESTED 2024
Seventeen batches from the end. Binder lot variance; haze 4% over spec. Scraped.
FM-20000
TESTED 2024
The window stayed a window. 48.2% transmittance, 99.3% hotspot extinction at a 15 µm build.
MTF
0.82
EXTINCTION
99.3%
TRANSMIT.
48.2%
IIITHE ONE THAT STAYED
Nineteen thousand nine hundred and ninety-nine panes had been scraped back to clear glass. This one was not.
FM-20000
BREAKTHROUGH2024PARTICLE BREAKDOWN
| SiO₂ nanospheres | 180 nm | 4.1% vol |
| SiO₂ nanospheres | 320 nm | 2.7% vol |
| TiO₂ (rutile), trace | 240 nm | 0.4% vol |
| BaSO₄ | 1.1 µm | 2.1% vol |
| Acrylic-urethane hybrid binder, lot A | — | balance |
SPRAY METHODOLOGY
0.82
MTF @ 10 LP/MM
99.3%
HOTSPOT EXTINCTION
48.2%
TRANSMITTANCE
1.8%
HAZE
160°
VIEWING CONE
To my son, Charlie, who is now fifteen years old: I want you and the world to know that the fourteen years I spent upon this, the eighteen-hour days, the seven days a week that were poured into this work—they were all for you.
— THE RESEARCHER, ON THE NIGHT OF FM-20000
IVTHE CYCLE
There was no automation and no assistant. One gun, one rack, one projector, one razor. The cycle ran twenty thousand times.
Spray
HVLP gun, 1.3 mm tip. Three cross-hatched passes at arm's length.
Substrate
A pane of float glass, cleaned to a squeak and edge-lit.
Dry
Forty minutes on the rack. Flash-off, then a slow ambient cure.
Walk around
Projector on. A full circuit of the pane at one, two and three metres.
Evaluate
Hotspot, haze, speckle, colour shift, cone. A number for each; most were wrong.
Scrape
A steel razor and an hour. The pane returned to clear glass.
Repeat
Adjust one variable. Only one. Begin again.
THE SUSPENSION
Nanoparticles, binder and solvent, swirled by hand before every batch.
THE SPRAY
One pass too wet and the film sagged; one too dry and it dusted.
THE RACK
Forty minutes of not touching it.
THE WALK-AROUND
The projector tells the truth from every angle at once.
THE SCRAPE
Nineteen thousand seven hundred times, back to glass.
VTHE LEDGER
0
TOTAL FORMULATIONS TESTED
0
FAILURES — 98.5%
0
BREAKTHROUGHS — 1.5%
0
YEARS OF ITERATION
0
INDIVIDUAL RESEARCHER
Countless
PARTICLE COMBINATIONS