stuart noble wrote in message ...

PVA seals wood, and wiping off the excess doesn't prevent that,

It seems it does if done carefully, see other message,

so any stain
will penetrate less in that area and give you a different colour.

Exactly!

A stain based on a strong solvent might cut through the pva but these
are hard to come by nowadays. Naptha, toluene, xylene etc.

A tinted varnish, which is designed to sit closer to the surface,
should give you less colour differential.

I wasn't being accurate when I said "stain". I'm actually using a
coloured wood treatment product which will be followed by Diamond Coat
or similar.

I would usually use clear wood treatment followed by tinted varnish,
probably diluted with clear varnish since the colours are often too
strong, but good quality varnish is hard to find in France and very
expensive. Also what I am trying to do is match the new oak to older oak
already in the building - don't know why I'm bothering really, since
timber in the main house has all ended up the same colour, but I didn't
really want to wait 300 years or so :-) The varnish is much better over
PVA, but I have had problems with that too on timber from Douglas fir,
over some french wood filler which also seems to have a component which
soaks in.

BTW what is the name of timber which comes from Douglas fir? We used it
because we had alot of it going to waste in our woods, where it came
down in the storm of 1999, and it is really quite nice, a reddish
colour, looks like something between red deal and pitch pine.

You could pva the whole section where the problem occurs. Sounds
silly, but it would give you a uniformly less porous surface to work
with. IME allowing dyes to freely penetrate isn't a good idea anyway,
especially the darker shades.

Not a bad idea, that, but it's too late now since ballusters, handrails
and outside and tops of stringers are already done!

Thanks for all the replies
Holly