Hello, my name is Erin St Blaine and I love making things that light up.

When I'm not writing tutorials for the Adafruit Learning System, I spend my time creating beautiful things, and exploring the meeting of art and technology whenever I get the chance. Over the last several years I've been focusing on creating larger scale LED artwork, home decor and chandeliers. I started out using Arduino with FastLED and learned to code the hard way, but nowadays there are so many software packages out there that I've been learning and exploring them. This article is about my experience of LED mapping my newest chandelier commission using WLED, PixelBlaze and LEDLabs.

What is Pixel Mapping?

Pixel mapping is a technique used to control and program individual LEDs or pixels within a lighting display, allowing each light to be addressed and manipulated independently. By mapping out the exact location of each pixel in a 2D or 3D space, artists and designers can create intricate patterns, animations, and effects that synchronize with the physical layout of the lights. This method is essential for creating dynamic, visually stunning displays, as it enables precise control over color, intensity, and timing across complex setups like LED walls, chandeliers, and sculptures.

A pixel map is needed when the LED layout is not precisely rectangular. Basically, a pixel map forces a non-rectangular shape into a rectangle that can be divided into rows and columns so the animations lay out correctly in the physical space.

My first foray into pixel mapping was making an LED Festival Coat using WLED. This was a fairly straightforward map -- the layout was generally rectangular, with just a few "holes" in the rectangle to account for -- I needed the map to fill in the arm hole areas so that my animations would look even across the whole coat. I used WLED for this, and found it to be a little mind-bending and tricky even with a simple map. 

After the success of that project I decided to give mapping my chandelier a try. The chandelier is shaped like a hot-air balloon with eight spokes and no rectangles at all. This looked to be a challenge but I knew the end result could be absolutely stunning if I succeeded.

Starting with QT Py Pico and WLED

My first attempt involved using the QT Py Pico microcontroller running WLED, a popular open-source software for controlling LED strips. WLED is fantastic for creating beautiful, dynamic lighting effects and offers a lot of customization options. However, pixel mapping with it can be a bit tricky, especially if you're working with a complex setup like a chandelier.

Since WLED doesn't have an automated pixel mapping feature, I had to resort to manual mapping. This meant spending hours painstakingly charting each LED on graph paper, making sure every pixel was accurately mapped. While this process gave me a deep understanding of my setup and allowed for precise control, it was incredibly time-consuming. 

The mapping process involved a large piece of graph paper and a pencil with a really good eraser. It took me three tries to get my head around making this map, with hours of painstaking counting, noting, erasing, and guessing.

Once I had the physical map, the next step was to enter the data into a spreadsheet, add placeholders for all the blank spots, and export it as a comma delimited file for import into WLED. I had about 350 LEDs in my map, consisting of seed pixels wrapped around the rope I'm suspending the main body of the chandelier from. By the time the map was finished there were over 2,000 data points - each square on the graph paper has to be added to the list and "skipped" by WLED.

At this point I ran into a significant issue: memory allocation. WLED couldn't handle the data load of over 2,000 pixels on one data pin. Even though I only had 350 actual lights, WLED ran out of memory at about halfway through the light strand, at which point it stopped sending data to the lights. And this was before adding the lights inside the chandelier to the map.

For a smaller project or one that's largely rectangular with just a few odd pixels out, WLED mapping will work. It's free, easy to install and set up, and very easy to use. There are a handful of 2d animations available, enough to have a lot of fun making 2d animations happen. Also check out the LED Top Hat tutorial and the MonsterMatrix Tutorial to learn more about what's possible when you map pixels with WLED.

Here's a video of standard 1d (non-mapped) animations running on WLED.

Trying Out PixelBlaze

I also experimented with PixelBlaze, a microcontroller known for its ease of use and real-time, web-based interface. PixelBlaze has a built-in pixel mapping feature, which seemed like a perfect fit for my chandelier project. However, I soon discovered that I needed an expansion board to handle the number of LEDs I was working with, since my lights are soldered on two different pins. Unfortunately, this meant I couldn't fully explore PixelBlaze's potential without additional hardware. I did get it loaded up and running on the seed pixels on the outside of my chandelier, but without an additional $40 expansion board I couldn't map the whole thing.

That said, PixelBlaze's software is highly intuitive and offers a lot of creative freedom. The expansion board requirement is a bit of a drawback if you're working on a large project, but for smaller installations with only one ouptut it could be an excellent option. I'm looking forward to exploring this software more, but for this project I moved on pretty quickly without diving in too deep. 

LEDLabs

Frustrated but undeterred, I decided to explore other options. I'd used LEDLabs by Christopher Schardt, the creator of the Paraluna and Mariposa large-scale art pieces, prominently featured at Burning Man and a lot of other festivals. I've done crew for Christopher at a few festivals, and he turned me on to his software program, which is designed specifically for mapping non-rectangular LED creations. 

LEDLabs is a very robust software program that runs on iPads and allows for camera-based pixel mapping. Although LEDLabs isn't freeware, it was worth the investment for the time it saved me. With LEDLabs, I was able to map the entire chandelier in about five minutes using the iPad's camera. The software significantly reduced the hassle of manual mapping. Moreover, LEDLabs can layer onto the QT Py through WLED, which can export data in Artnet format. My chandelier runs WLED most of the time, and when I open my iPad and fire up LEDLabs, it seamlessly takes over. 

However, this software program is definitely not for beginners. WLED is so easy to use and get decent results that it feels like a box of crayons. LEDLabs software feels like an easel with oil paints: you can get absolutely stunning results that are worlds away from what WLED offers, but you have to invest a whole lot of time into learning the software and it's easy to get it wrong.

LEDLabs is designed with live performance in mind. Schardt's art pieces are an immersive experience, with full-length LED shows that run for hours on end, with surround sound and extreme persistence-of-vision visuals. This software was designed to run big shows. My iPad needs to be awake and manned (or at least running a playlist) the whole time. For a chandelier, which needs to be ambient most of the time, it may be just a little bit overkill. But it "fails" gracefully back to my simple WLED animations as soon as I quit the app, with no change in hardware, which I really love.

And the mapped animations are simply stunning. Sound reactivity and touch reactivity are built right into the software. You can even use the iPad camera to project realtime video onto the mapped LEDs. It basically turns your project into a low-res video screen, with loads of control over the video clips and the ability to sync everything to music. 

Here's a video showing mapped animations running LEDLabs.

Comparing the Tools

• WLED (QT Py Pico): Offers very easy usage and a solid open-source platform, but manual pixel mapping can be laborious, and memory limitations can be a significant hurdle. I am still using it as my base software and running non-mapped LED animations on the chandelier with impressive results.

• PixelBlaze: User-friendly with real-time mapping and creative controls but may require additional hardware for larger installations. I want to play with this more, but the hardware was significantly more expensive and harder to get my hands on.

• LEDLab: Efficient and fast with camera-based mapping, making it perfect for large and complex setups. However, it's not free and requires an iPad, which could be a limiting factor. It's also much harder to learn and use, but the end results are absolutely stunning.

Conclusion

Each of these tools has its strengths and weaknesses, and the right choice depends on the specific needs of your project. While WLED offers a lot of flexibility, the manual mapping process and memory constraints can be frustrating. LEDLab's efficiency and ease of use make it a standout choice for more complex projects, but it comes at a cost. PixelBlaze may strike a good balance but may require additional investment in hardware.

In the end, it's about finding the right tool for the job and balancing your time, budget, and project requirements. The journey of pixel mapping my LED chandelier was a lesson in patience, experimentation, and the joys of seeing a project come to life with the right technology.

See more artwork at www.erinstblaine.com