There are several misconceptions about how ATR works.
Dr. Lecter, my post (and some from others) explained why cylinder volume is irrelevant to the ATR and does not need to be programmed in. The calculation is based on change in pressure over time. A small cylinder will drop in pressure faster than a large one for the same volume of gas consumption. The computer registers the rate of drop and calculates your ATR accordingly. If you have a large cylinder, the rate of drop will be slower and (no surprise) the computer will show a longer ATR. Vice versa for a small cylinder. The computer does not need to "know" the cylinder volume to do this; it only needs to know the starting pressure.
Dale C, Boulderjohn and Dr. Lecter, the computer in fact does take into account sudden changes in breathing rate. I think the sampling rate on mine is every 5 (maybe 10) seconds. The adjustment for changes breathing rate is very fast, maybe it will use 30 to 60 seconds of sampling to re-calculate just so one or two deviant samples do not control, but it is fast, and accurate in my experience. It is far more reliable and a much bigger safety margin than trying to do mental calculation while suddenly under stress or a heavy workload.
Kharon, I use SPF 50 on my computer
---------- Post added July 31st, 2015 at 12:15 PM ----------
That 500 psi is the contingency gas. For example should you need to share air with an OOG buddy it is the gas available to get them to the surface. Surely that depends on depth, buddy sac etc. is this calculated on the fly by the computer or does it need to be planned?
No so much. Remember, I am talking about rec diving, with no hard ceiling or mandatory stop. The 500psi reserve is "your buddy's" gas because if you follow the ATR you will not be breathing it. The ATR will send you up with more than 500 psi from depth. As an example, it might send you up with 800 psi if you are at 90 feet. That means you will use 300 doing a standard ascent and safety stop, which leaves 300 for your buddy, and still 200 at the surface. It is only if the ATR would send you up at over 1000 that the concern would arise (ie if it sent you up at 1100, you would need 600 and only 500 would be there for your buddy). This has never happened to me, or even come close.
Plus, my planning for an ascent with an OOG buddy gives a lot more cushion than what the computer is assuming. I have only had to do this once (from 140' long ago) but that experience taught me that getting up while still alive is the top top priority. So, on a rec dive, where we are not in deco, I am going to push the ascent rate to 60 fpm until I hit 30' or so, and then 30 fpm to the surface, with no safety stop. This does two things. First, it rapidly increases the available gas as you get shallow quickly, and it decreases gas consumption by getting to the surface directly and without a stop. The minor increase in DCS risk is an easy trade-off for getting out of the far more lift-threatening situation.
I did an exercise in "rock bottom" calculation a while back, and found it very imprecise and based on a host of assumptions (square profile, assumed elevated SAC rates unchanging during the ascent, assumed standard ascent and assumed safety stop). The bottom line was that on AL80s at depth, such calculations might significantly cut dive time short of NDL on the actual dive itself. I decided not to plan every single dive of my life as involving an inevitable ascent for two divers at high consumption, at regular rates and through safety stops. I would have lost many precious hours underwater had I done this, and I am not going to start now. the AI computer has actually improved my knowledge regarding my gas use in real dives and I am comfortable with the planning and even the 500 psi surface reserve setting. If, for some reason, I had to dive to 130 on an AL80, I might get more conservative by setting the surface reserve at 750.