*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

> With 2-photon, things are likely to be worse because even though the (squared) excitation volume is somewhat larger, the excitation is pulsed and at least 90% of the time you are not counting anything. And because of this low duty cycle, you are apt to be working at excitation levels a lot closer to singlet saturation than you should be (I assume that anyone interested in rapid changes must be interested in looking at living cells, not dying ones?)
You can use the low duty cycle to your advantage, though: Using a low number of excitation pulses per pixel by scanning very quickly (up to kHz frame rates) while using low excitation powers is the idea behind 2-photon LOTOS microscopy as used by Chen et al. from Arthur Konnerth's group (protocol paper: http://www.nature.com/nprot/journal/v7/n10/abs/nprot.2012.106.html). This would require AODs rather than resonance scanning to do it properly.

> This is why LaVision Biotech and many others trying to do 2-photon fast use multiple IR beams. But is does put a lot of power into (through?) the specimen.
Isn't the alternative of "High repetition rates with low power pulses" as implemented by Ji, Magee and Betzig (http://www.nature.com/nmeth/journal/v5/n2/full/nmeth.1175.html) potentially more promising (and easier to do)? This puts less power into the specimen while still giving the user many excitation pulses per pixel while scanning at high frame rates. I would assume that this approach combined with LOTOS might give users the best SNR at high speeds. In which case I would put my money into AODs… 

But it is possible that I am overlooking something, so please to correct me if I am!

Cheers, Chris