I'm up to about batch #50 of hefes (and quite a few weizenbocks). I've tried all the tricks and the abuse of yeast-stress, under-pitching, hot fermentation etc..
I think everyone has a point. Some of my best results were pitched on the low side without a starter, but that said, I used very fresh top cropped yeast.
(about 50ml of top cropped yeast for a 19L batch that I kept in the fridge for less than a week before using)
Apart from pitching on the low side with very health yeast, all my other experiments at yeast stress were failures with off-flavours/aromas present.
If you're bottle conditioning the most important thing is to not use sugar to prime. Instead use 10-15% of the original wort. (keep it in the freezer).
To improve the esters I keep as little trub getting into the fermenter as possible. I strain the cooled wort into a tall fermenter and leave it for around an hour before racking off to a wide fermenter. Stopped getting the occasional soapy/fatty off-flavour from doing that too.
The other thing that notably increases the banana ester is to produce a wort high in glucose.
To do this you need to produce the maltose at 62c/143.6f using half your grain and then mix it with the other half to rest at 38c / 100f for an hour.
(The enzymes that produce the glucose get denatured at hot (62c) temperatures so thats why you only use half the grain).
Another way to do that is via a decoction.
My perfect mash programme for hefe is now:
Mash in at around 4.1L/KG (very wet mash, soft water with no additives or acid malt, high PH) at 30C/86F for half an hour to wet the grains and get all the starch and enzymes into the liquid. (stirring occasionally).
Pull 1/3 thick mash from the bottom and quickly heat to 62C/143F for 40mins to produce the maltose.
Mix it back (to 35c) and raise to 38C/100F for an hour to produce the glucose.
Pull thick 1/3 decoction again and raise to 70C/158F for 10 mins for some dextrins.
Boil thick for 20mins (for the melanoidins and some colour) while raising the thin mash to 66C/150F.
Carefully mix back to keep below mash-out temp, rest at 70C/158F for 10 mins. (add the hulls)
(boiling the decoction will release more starch that needs to be converted).
Then mash-out at 78C.
Takes all day. Totally worth it.
For more clovey hefe's I find using more (70%) rather than 50% wheat helps and also raising the glucose (38C) rest closer to the ferulic acid rest (43C),
but the main factor for accentuated clove is a lower fermentation temp.
I've also used a variant of this mash for pale weizenbocks with high attenuation (Weiphenstephaner Vitus clone).
Managed to get 80% attenuation out of Wy3068. (Lowered the 70c rests to 65c and "mashed out" at 74C not 78C)
For dark weizenbocks I skip the decoction boil for a shorter mash and then add the specialty malt at recirculation.
I also ferment at 22C for weizenbocks (max 3068 temp) using top cropped yeast and a starter with decent aeration.
As long as you don't stress the yeast you don't get the boozy higher alcohols or the horrid almond/marzipan esters.
I also use 10% of the original wort for priming weizenbocks. Sugar is for the belgians.
Despite all this mash abuse and fermenting in a semi-open fermenter I've had no issues with oxidation.
I still can't make a bottle conditioned IPA as I'm very sensitive to oxidised aroma and flavour.
The trick with getting more banana from removing as much trub as possible actually came from some super-taster here
who kept complaining about banana in their beers even when they were using US-05.
Their solution was adding more trub to the fermenter and so I tried the reverse as I always assumed trub was beneficial
and I noticed an immediate increase in esters.
