Suggestion for grainy-anime footage

Well, the title says it all. I have this kind of footage which is animation with film-grain. I would like to keep the grain because it's a precise director choice. The average bitrate (estimated) will be at about 10Mbps (I wish some more at the end). I've tried with CineVision and its Film Grain Optimization but without great success (sometimes the grain simply vanish with an ugly effect). To optimize it a little I'm thinking to stay with L4.0 using 1 slice (the bitrate is not so high and with VBV there could be some spikes greater than 24 Mbps for short time if necessary for the compression). I have some doubt regarding AQ and psy-rd settings. In particular, psy-rd trellis (for what I've red) seems not good for animation but it's enabled in the grain preset so I'm asking if someone have some advice/experience on film-grain animation with such bitrate. Regarding AQ, is the AQ2 (new) suitable for such kind of footage? Thanks.

Well, I think if your source is very grainy and you want to preserve that grain, then most of the "complexity" of your source is in the grain. In other words: What's "under" the grain doesn't matter that much anymore. So your source actually isn't a typical "anime" source. Instead the the "grain" tuning of x264 should apply. Also from what I know, AQ doesn't help for gain retention at all. The purpose of AQ is to move bits from "complex" area to "flat" area. But if you have a lot of grain (and preserve it), then, as said before, most of the complexity is just grain - and the grain is everywhere in the frame. So all blocks have (more or less) the same complexity in that case. Consequently we don't want AQ to move bits away from certain blocks, as that would kill grain in those blocks, while it is retained in others...

Thank you very much!

Just out of curiosity, what is the title name of that grainy anime? First thoughts were about Kino no tabi or Texhnolyze, but your target bitrates are about BD source, so i'm a bit confused. Sorry for offtopic in advance.