I am getting here and there something that appears to be spruce endgrain contamination from the fish glue i use. It happens with the rosette and the purfling. It is not a problem if I do not need to sand the piece too much, especially with the frnech polish on the coloring is very hardly visible even if knowing what to look for, but if I need to take off some thickness, the color problem turns into a mechanical problem towards the center of the piece: can even run into sort of void pockets, even worse than the pictures below...(click to zoom in)
It sort of looks like routing tearout filled with glue/dust, but no, the channel was perfectly and smoothly cut with a gramil.
Strange that I did not notice the problem on my first 3 guitars. The 4th has had some of this issue, and now thinning the top of my 5th and there it is again at the rosette.
![[xx(]](./images/smilies/smiley11.gif "Dead")
I believe it happens because the fish glue has a very long open time, so it has plenty of time to soak in, perhaps moving some color from the hardwood too.
It is obvious that in the future I need to seal the rosette and purfling channel walls with either hide glue or shellac. Wonder which would be better?
So this thread is also sort of a warning for those that plan to try fish without sealing first.
) easily as well. On your channel, I respectfully suggest that your gramil is crushing some of the end grain. Spruce is very fragile in that area because the long tenacious just does not want to separate from itself and will tear rather than cut fairly easily. So there is probably a combination of things going on, the long open time of the glue likely does allow for some migration into the exposed ends of the endgrain cells (fix this by sealing the channel with shellac) and I suggest that there is a bit of fibre damage occurring during gramilling (is that a word?). This can be fixed by ensuring your gramil is VERY sharp and reducing pressure, allowing the fibres to be entirely sliced.