Most reeds don't fail at the tip. They fail because the back was never balanced — and no amount of clever tip work rescues a reed sitting on an uneven foundation. After enough years at the desk I stopped thinking of scraping as removing cane and started thinking of it as four conversations with the blank, each one answering a different question.
This isn't a measurement chart — the cane decides the numbers. It's a way to keep your head straight while your hands work, so you always know which pass you're on and what you're listening for.
First scrape — find the tip
The first conversation is about response. With the reed soaked and tied, the goal is only to thin the very tip until the reed will crow — that rattling double-tone it makes when you blow across it alone. Don't chase pitch yet. You're looking for the reed to wake up and speak.
Work in light, symmetrical strokes from the tip back, a hair at a time, testing the crow constantly. The moment it crows a clear, buzzy double-C, stop. You've found the tip; everything after this is balance.
If only one side crows, you're already off-balance. Scrape the quiet side, never the loud one — you bring a reed into balance by lifting the weak corner, not by beating down the strong one.
Second scrape — open the heart
Now the question is resistance. The heart — the slightly thicker zone just behind the tip — is the engine of the reed. Too thick and the reed is stuffy and sharp; too thin and it spreads, loses focus, and goes flat. This pass is the most patient one. You're defining the heart by what you take from around it, not by carving the heart itself.
Blow the reed often, on and off the staple. You want it to feel like it's pulling air through, with a little spring back against the lip. If it feels dead, you've gone too far — set it aside and start another. A reed you've over-scraped is a lesson, not a tragedy.
"A reed is finished not when there's nothing left to take off, but when taking anything more would make it worse."Notebook, third winter of making
Third scrape — balance the sides
The third conversation is about symmetry. Hold the reed to a lamp and look through it: the two blades should show the same window of light, the same gradient from spine to rail. Most of us have a dominant hand and scrape one side heavier without noticing — the light tells the truth your hands won't.
Correct in tiny increments, testing the crow after every few strokes. By the end of this pass the crow should be even, centred, and stable when you change the air pressure. If the pitch wobbles as you blow harder, the sides still aren't talking to each other.
Fourth scrape — settle it
The last conversation is about stability over time. A reed straight off the knife is almost always too bright and a touch unstable; it needs to be played in and revisited over a day or two as the cane relaxes. The fourth scrape is really several short visits: a little off the tip if it's closing, a touch off the back if it's stuffy, nothing at all if it's right.
Resist finishing a reed the night you start it. Cane keeps moving for a day after you tie it, and the reed you call perfect at midnight is often flat by morning. Let it settle, then make the final, smallest cuts.
Keep a one-line log for every reed: cane batch, gouge, shape, and how it ended up. Patterns appear after twenty reeds that you'd never spot in two. This is the single habit that improved my reeds the fastest — more than any tool.
The whole arc
Find the tip, open the heart, balance the sides, settle it. Four questions, asked in order, each one only once you've answered the last. It won't make cane behave — nothing does — but it keeps you from the most common mistake, which is scraping everywhere at once and ending up with a reed you can't diagnose.
If you want the foundations under all of this, the cane article covers what you're actually cutting, and The reed knife covers the tool that does the cutting. The rest is reps.