Blog 002 | Lume-M

This project explores a steganographic approach to data transmission using ambient light. Lume-M is different from Li-Fi and high-frequency optical methods because it doesn't need special high-speed sensors to "record and decode." I made in that that way so the data can't be lost by modern video platforms (WhatsApp, YouTube, etc.) by changing the light at a slow, human-perceptible rate.

Rather than aiming for reliability or completeness, the experiment was designed to observe failure modes, unexpected behavior, and trade-offs that emerge when theoretical examples are applied to real, resource-limited devices. Instead of aiming for high-speed data transfer, the experiment was designed to use another approach with some development to hide the data while maintaining the stability and survivability of the hidden data. When tested in a real-world situation with little to no environmental control, to know how we can transmit data.

Why I Built This

I built this experiment to move beyond simplified examples commonly found online and test what actually breaks when constraints are real. Many demonstrations assume ideal conditions, fully transparent, and never tried to hide data in visible source.

I wanted to understand how we can transform those demonstrations — specifically where Morse code is visible to all users into a manner so that it looks like an innocent ambient light but actually transfers data.

Constraints & Assumptions

1. No physical or wireless sync exists between the transmitter (LED) and the receiver (Camera).

2. Designed for standard smartphone cameras and variable frame rates (VFR).

3. The transmission must survive "Aggressive Transcoding" (WhatsApp compression or YouTube downscaling).

4. The environment is assumed to have non-static lighting (reflections, screen glare) which the decoder must filter out.

5. The output must mimic usual Smart Home or ambient patterns to avoid human suspicion.

High-Level Approach

This experiment was done to examine how we may use our daily life's light item into a morse transmitor utilizing basic and affordable microcontroller like ESP. Now, with ESP and any object with controlled light brightness/source, you can transport data from ambient light to room lighting or even table lamps without raising any human suspicion.

The approach was changed multiple times to reduce errors and simplify logic, allowing the data encoding and decoding more accurate finally selecting the Closed-Loop-style State Machine for best result so far. Still place for development.

What Broke / Didn’t Work

To complete this experiment, I tested various approaches, in both encoding in different streams and in different methods like Manchester encoding. What started with curiosity and dot dash later transformed into color-based morse, frequency-based morse, temperature-based morse, and finally ended up with time-based approach with grayscale.

This exposed a gap between examples that appear simple on paper and are very inaccurate in real life conditions and if we have even more specific requirements, so it all returns to the foundational approach.

Adjustments & Decisions

After repeated inaccurate results, I shifted focus away from new and creative ideas to more simpler and effective way, using time-based morse encoding. This mimics standard mood lighting, making the data transmission look like a natural decorative effect.

To prevent "bit-slip" (where one missed pulse ruins the whole message), the decoder analyzes the frame-gap between pulses to identify letter boundaries and word spaces:

1. Gap > 25 frames: End of Letter.

2. Gap > 50 frames: End of Word.

Key Takeaways

This project helped me to understand that some time instead of finding creative ways to solve a problem we can use and improve what we already have. Traditional approaches are robust and traditional for these reasons.

Instead to trying to find a new method just use the existing with some major upgrades then it can replace the new !founded approach.

// Personal Note
                 →  It's not bad to sometimes use what is left behind. — 
                Because maybe it can make the work out more efficient and easily.