Monster Generator timings in original DM/CSB

[Gambit37]

<pedant>
I've done some timings of monster generator recharge times using FTL PC DM. In DMute, if you view a generator at hex level, you'll see that the 8th byte controls the delay between each monster being generated. DM (and CSB) won't allow a generator to create a new monster until this delay has elapsed, regardless of how many times you might activate the generator.

I built a test dungeon and changed the timing of the generator to different values to see if I could find a pattern. I didn't have a stop watch so could only time these using a clock, which means the timings aren't that accurate.

8th byte= 0 : No delay
8th byte= 10 : ~5 seconds
8th byte= 20 : ~7 seconds
8th byte= 30 : ~9 seconds
8th byte= 40 : ~11 seconds
8th byte= 50 : ~13 seconds
8th byte= 60 : ~15 seconds
8th byte= 70 : ~17 seconds
8th byte= 80 : ~20 seconds
8th byte= 88 : ~2 mins 5 seconds (DMute default)
8th byte= 90 : ~3 mins 50 seconds
8th byte= A0 : ~7 mins 40 seconds

I stopped timing at this point as it was getting very boring... However, you'll notice
there are two patterns here. Using values 10-80 seems to increment the timer by
approximately 2 seconds each . At value 88 things change, and a default time is created (2 mins 5 secs approx). I timed 90 and A0 and discovered a new pattern; that the A0 timer doubled the previous 90 timer. It's therefore logical that if byte 8 is in the range B0-E0, these would be the recharge times:

8th byte= B0 : ~15 mins 20 seconds
8th byte= C0 : ~30 mins 40 seconds
8th byte= D0 : ~61 mins 20 seconds
8th byte= E0 : ~2hr 2 mins 40 seconds
8th byte= F0 : ~4hr 5 mins 20 seconds

This seems ridiculous, but I did try timing F0 and got bored after 17 minutes, so it's
possible the delay really is that high! Has anyone else done any work on this? I would be interested to see what interim times might be. I got quite bored, so perhaps someone else wants to try values such as 38 or A8. Do they fall between the previous and next timer?

The reason for doing all this is to make RTC more accurate. Having looked through the .TXT files it seems the highest recharge time is 250. This equates to about 42 seconds, much lower than the default time for DM/CSB. George, is that the highest the recharge timers can go?

If so, it explains why much of RTC CSB is very difficult because monsters are generating too quickly. I have posted another response in the CSB bugs/oddities thread giving specifics for the DDD. When I get more time, I'll check all the other generators to give you exact timings! </pedant>

[George Gilbert]

The delay value is stored as an int so can go up to about (2^32)/2...which equates to a bit over 11 years of game play. Should be enough for the toughest dungeons...just

[Gambit37]

11 years???!!! Blimey, I was worried about turning 30, I certainly don't relish the prospect of still playing when I'm 40 odd, only to be wiped out by a Golem I hadn't expected to be there...

Highly illogical I know, but you get the picture!

[beowuuf]

Quick check in DMute, i used a self targeting system with zero time, using an 80/88 activation effect to constantly target both a toggling lever behind the generator and the generator itself so i got an internal game 'tick' to measure these things with

I found (counting is bad, plus graphical updates and other timing effects seem to knock the count somewhat) that as you say, 0 - 80 seem to be seperate from higher numbers

Basically each 10 gives you 16 'ticks' (approximately) as you would expect, so 80 takes 128 activations to recharge

For 90 i counted 508, which was probbaly 512, which is 8 times the previous 64!Then at 88 i counted 219 which was strange, as i was expecting a quadrupalling effect of 256 at least!

At this point i got bored, but that is my findings so far!

[Gambit37]

Beowuuf, could you send me your test dungeon, as I tried to do the same thing but it didn't work (I'm not a DMute expert like you).

My method was different: My dungeon contained a party-activated pad that activated a generator. The monster that is generated activates it's own pad that launches a fireball from a shooter, killing the monster. I then keep walking back and forth over the pad until another monster is generated. I time between the point the first monster appears and the point the second monster appears. It's nowhere near as accurate as yours, but probably accurate enough to get better timings for RTC.

[Gambit37]

Thanks for the dungeon Beowuuf! I changed it a bit, to combine your approach with mine, so I'm now in the process of timing everything much more accurately. This will take some time (!) but when it's done, I will post the revised times. I'm now using a stopwatch on my iPaq, and it looks like my original timings are a fair bit out.

[beowuuf]

np, hope it helps refine the times