A little while back I tried to put words to why French polish keeps pulling me in - its honesty, the way it refuses force. That was mostly about the finish I chose. This is about the finish I didn't, and the strange road it took to end up on so many guitars.
Because if you spend any time around serious guitars, you hear the word nitro a lot, usually with reverence. Nitrocellulose lacquer is the finish of the golden era - the prewar Martins, the fifties Gibsons - and in the boutique world it's shorthand for "done right."
I get it. Some of my favorite instruments ever made are wearing it. But the story of how it got there is funnier than the reverence suggests.
Surplus guncotton
I should say this plainly: I am not a finish chemist or an industrial historian. I am a luthier with a bench, some old instruments to repair, and a habit of pulling on threads when a story sounds too polished. Nitro has so much romance around it now that I wanted to understand where that romance actually began. So I went looking, expecting the trail to run through guitar shops, old factory ledgers, maybe the finish rooms at Martin and Gibson.
It did, for a little while. But pretty quickly the trail left the guitar world. The names that kept showing up were DuPont, DUCO, General Motors, Oakland automobiles, spray guns, drying time, and factory throughput. The dates mattered too. The early 1920s were not just a charming prewar guitar moment. They were also a period when fast-drying industrial coatings were changing how manufacturers finished cars, furniture, appliances, and anything else that had to leave a production line quickly.
That changed the way I heard the word nitro. In guitar culture it often sounds like a tradition handed down from the mountaintop. In the research, it looked more like a practical industrial answer to a practical industrial problem. Not a bad answer. Not a dishonest one. Just a much less mystical one.
Nitro wasn't developed for instruments. It wasn't even developed for wood. It's cellulose treated with nitric acid - the same basic chemistry as guncotton - and after World War I, the chemical industry had enormous nitrocellulose capacity and suddenly no war to feed it. DuPont pointed that chemistry at coatings, and in the early 1920s its DUCO lacquers became part of the automobile story. General Motors was the big customer. Oakland was one of the first cars to wear it.
The pitch was not romance. It was speed. Varnishing a car took a long time; spraying lacquer moved the work along. That is the part that matters to me as a small builder. Guitar factories had the same pressure every factory has: finish the work cleanly, repeatably, and fast enough to ship. Nitro replaced varnish and French polish because it helped inventory move. Nobody had to sit down and decide it was what spruce deserved. It was what the paint salesman was selling that decade, and it happened to look pretty good.
The slow work
It is worth saying what the old method costs. French polish is a ton of work. There is no clever way around that. You make a pad, charge it with shellac and alcohol, and build the surface a little at a time. Not with one heroic coat, but with hundreds of small passes. The pressure has to stay light. The pad has to keep moving. Too much shellac and it drags. Too much alcohol and you cut back into what you just built. Too much ambition and the surface tells on you immediately.
A sprayed finish lets a factory separate skill into stations: prep here, spray there, cure over there, buff at the end. French polish keeps asking the same person to come back to the same surface, day after day, and pay attention. Body sessions, leveling, more bodying, spiriting off, waiting, looking in raking light, finding the low spots, doing it again. It is quiet work, but not easy work. The guitar has to be handled constantly without being bruised. Dust matters. Humidity matters. The mood of the surface matters.
That is exactly why it disappeared from production. It does not scale kindly. A factory manager would be right to hate it. But I am not managing a line of instruments. I am building one guitar at a time, and for that kind of work the slowness stops being a liability and starts becoming part of the care.
What it does with time
Ask anyone with an old lacquered guitar. The finish checks - that web of fine cracks - because the plasticizers that keep it flexible slowly migrate out, and the film shrinks and gets brittle. It yellows, not always evenly. It keeps off-gassing for years, which is why it melts against guitar stands and case linings. And it wears through wherever you touch it, chipping and flaking at the edges as it goes.
The vintage market has turned all of this into charm - checking as authenticity, wear as mojo. And honestly, on a '37 D-28, it is charming. But it's worth noticing that nitro's best day is the day it leaves the spray booth, and there's an entire relic industry devoted to faking the decline. That tells you something.
Shellac runs the clock backwards
I finish everything here in French polish - shellac, applied by hand with a pad, the same finish nitro pushed aside a century ago. Not for nostalgia. It's just the only finish I know of where time is on your side. A French polished guitar isn't really done when the last pad session ends. It's done months later, and then it keeps going.
The film keeps tightening. The last traces of alcohol take months to leave, and as they go the film shrinks and pulls down onto the wood and into the pores. A finish that looks a little soft at two weeks looks crisp at three months. The old polishers knew this - they'd finish an instrument, set it aside, and deliver later. The waiting was part of the job.
The resin keeps curing. Shellac slowly cross-links, getting noticeably harder over its first year. So while nitro spends decades losing plasticizers and going brittle, shellac spends its first year getting stronger. I find that trade pretty easy to like.
The look keeps deepening. Shellac's refractive index is close to wood's, so a thin film reads as depth rather than coating. As it settles closer to the wood, more comes through. The medullary rays in a quartersawn spruce top - that faint silk running across the grain - show more at month six than month one. Rosewood does the same thing, just darker.
Playing helps. Handling gives shellac a soft burnish nothing in a bottle reproduces. Old French polished instruments have a glow in the touched places - the finish keeping a record of hands. Nitro wears through where you touch it; shellac lights up there. Same years, opposite directions.
They age as a pair
Spruce starts out nearly white and a decade of light turns it honey and amber. Rosewood settles and deepens. Shellac warms right along with them, so the wood and finish drift the same direction, together. A French polished guitar at year five is simply better looking than the day it was strung up. The koa on my ukuleles might be the best demo of all - it darkens fast and the finish keeps pace - but it works the same on everything that leaves this bench.
And when a French polish eventually needs attention - they all do - the repair is invisible. New shellac dissolves into old with no witness line. I learned the other side of this the hard way, repairing the finish on my own 2000 Collings D2H. Wonderful guitar, and Collings' lacquer work is as good as it gets - which is exactly the problem. Every touch-up on nitro is a patch trying to impersonate the original, and the original never fully accepts it. Blending new lacquer into old is a fight at every stage, and under raking light the seam is always there if you know where to look. It was a nightmare, and it's part of why I finish the way I do. Shellac never puts you in that position: the new material becomes the old.
Where that leaves things
None of this makes nitro a bad finish. It's a great finish - for a factory. It sprays fast, cures fast, ships fast, exactly what it was invented to do, for cars. French polish takes weeks of handwork and can't be hurried, which is precisely why the factories dropped it.
But I build one guitar at a time, for one person, and at that pace nitro's advantages don't buy me anything. So I use the slow finish - the one that treats time as an ingredient instead of a threat. The commission list here runs long anyway. Turns out the finish doesn't mind waiting either.