You spray a clean waterborne finish over white oak. The bare wood looked fine. Within the hour, brown streaks rise up through the coat where there was nothing before, following the open grain like water finding a low spot. You didn't drag dirt across it. You didn't miss a step. The wood is bleeding into the finish on its own, and watching it happen feels like the panel is ruining itself in front of you.
Most makers who've seen this know the cure. Shellac first. They seal the oak with a wash coat, topcoat over that, and the bleed never shows. What they usually can't tell you is why it works, and that gap matters, because a rule you can't reason about is a rule you'll misapply at the edges. You won't know how heavy a seal coat needs to be, or which other woods do it, or why the same finish bled badly one day and not the next.
So here's the why, in two facts that sit right next to each other. Tannins, the brown stuff in oak and cedar and redwood, dissolve in water, and they dissolve best in water that runs slightly alkaline. A conventional waterborne finish carries its resin in water that runs slightly alkaline, around pH eight or nine. Put those two sentences together and you've got the whole problem. The finish isn't just sitting on the oak. Its carrier is close to the ideal solvent for pulling tannins out of the wood, so the moment the wet coat lands, it starts leaching brown up into itself. What you're seeing isn't the finish failing. It's the finish dissolving the wood's own chemistry and floating it into the film.
The part worth keeping is that this is a wet-film event with a clock on it. The tannins only migrate while the coat is still open and liquid. As a waterborne film cures, the top skins over first, and once that skin forms, the brown can't get through it anymore. The bleed is decided in the first stretch of open time, before the surface even looks dry. That's why a thin, slow coat laid down in a humid shop bleeds worse than a clean, fast one. You're holding the window open longer. You can't always control the weather, but knowing the clock is running changes how you work the coat.
Now shellac makes sense, and it makes sense on two counts, not one. Its carrier is alcohol, and tannins barely dissolve in alcohol, so over a shellac seal there's no solvent left to give them a ride into the topcoat. They stay put in the wood. On top of that, the cured shellac is a continuous skin that blocks the water pathways the tannins would travel on any later wetting. A solvent-borne oil or alkyd primer works for the first reason just as well, since mineral spirits won't dissolve tannins either. The fix was never magic. You're swapping out the one solvent the tannins love for one they ignore.
One more thing, because it rides along with the same chemistry. Keep steel wool off tannin-rich woods. A stray steel fiber plus a little moisture plus tannin makes a black iron stain, and it can form inside the film, where sanding it out just drags the contamination into the next coat. If you've been knocking back oak between coats with steel wool out of habit, switch to a synthetic pad and you've closed a door you didn't know was open.
None of this makes waterborne finishes worse than anything else. They're not inferior. They're just chemically mismatched to a handful of woods, and shellac fixes the match. So when you've got oak or cedar or redwood on the bench and a water-based topcoat in the can, don't wait to see if it bleeds. Seal it first, lay the coat clean and quick, and keep the steel wool in the drawer. Match the carrier and you beat the bleed before it starts.