Despite all the useful words given so far, Niels Berkers' player isn't
actually an Aeolian!  The pictures show this quite clearly.  The Bremar
piano tells you instantly that it's not Aeolian, as does the lead
tubing.  The 5-part wind motor is a giveaway as well, obviously not
Aeolian's characteristic design.

However, the pictures are rather fuzzy and don't home in on any of the
characteristic bits such as Tempo plate and keyslip controls, or the
stuff under the keybed, so it's hard to identify just what it might be.
A few more clasps might help.  It's got Themodist-like accenting
ports in the tracker bar, which narrows the field.

Anyhow, the make's not relevant to trying to figure out what's wrong
-- this is a perfectly normal-looking 1920s player.  A few thoughts on
individual dead or hard-to-play notes, following the air path from
tracker bar.  I know I'm repeating some ideas, but to be thorough...

First, bleeds and pouches, easy to test: blow down the tracker bar
hole!  It ought to be easy to spot ones that let more air through.
If no air goes through the pipe's blocked.  It's easy to dent a lead
tube or trap a rubber one.

Then, valve travel.  Assuming no grit in the works, do the top valve
faces make firm contact with top seat when they operate?  If their
pusher buttons are set too high the pouches won't move the valve body
enough, so they'll leak horribly when played.  Easiest to spot when the
cover cloth's off; the pouch should have plenty of spare travel when
the valve stops moving.

Then pneumatics.  Let's assume there are no holes in the cloth, and
that the pneumatics close easily by hand and (just as importantly)
open fully afterwards.  A less obvious problem is when they are not
glued properly onto the decks so air leaks across the top when the
note plays.

For all of the above, if it's a leak associated with an individual
note, a quick way of spotting this is to put a suction gauge somewhere
onto the stack (connecting to the top or bottom tracker bar tube is
the easiest if the stack's in the piano).  A leaky note will cause
stack tension to drop a lot more than a good note -- there's always
some drop from the air flowing though bleeds.  It's a useful test
to do on any player to help make them play evenly, and critical on
reproducing pianos (particularly Duo-Arts with their greater
sensitivity to air flow through the expression system).

As for rolls stalling at the end of play on high suction.  If this is
common, then think motor or governor fault.  Assuming that motor valve
faces and seats are smooth and well-graphited, the valves are set, and
the motor has plenty of power (it ought to be hard to stop by hand),
look at the governor.

For stalling under high suction, I'd check the regulator valve: lack
of lubrication can make it stick.  Also look at regulator travel.
All governors have a finite range which they can control, after which
the regulator valve will simply close and stop the motor completely --
many governors have a stop in them to prevent this, and it must be set
properly.  Having the regulator spring too weak or its initial set
point wrong can increase regulator travel and introduce this type of
problem, as can recovering the regulator with the wrong amount of cloth
so it's all out of range.  The exact cause can take a bit of puzzling
out at times!

On the other hand, if only occasional rolls stall, you'll probably find
these rolls have rough paper.  Maybe a result of damp, or just duff
paper?  I've had a couple of English Aeolian rolls like this.

Julian Dyer