Needs less detail!

[Grazza]

To avoid VPOs during the playback of certain demos in Doom2.exe, that is.

I've noticed that a couple of demos from the last year that VPOed when played back with Doom2.exe will actually work OK if you choose low detail. This seems to reduce the number of visplanes sufficiently and there is no overflow.

The two cases I have tested are:
Anima's ks15-719.lmp for Kama Sutra
My 2s05-102.lmp for 2sectors
Both times were improved upon in the demo projects that they were for, but it's possibly useful information for the next time this occurs.

BTW, the IDRATE cheat in prboom may be a useful way to test maps for VPO issues, and shows how the resolution can affect the number of visplanes. I'm not sure exactly how the numbers it displays correlate to the likelihood of a VPO in Doom2.exe though.

BTW2: During the testing of this stuff, I noticed that Gusta's s202-210.lmp used -longtics, and despite a spechits overflow plays back OK with Doom2.exe v1.91 (as long as you "watch" it in the automap to avoid VPOs, heh).

[myk]

Interesting! It also makes sense with a post Graf Zahl made about Doom95 possibly suffering some unexpected VPOs at higher resolutions. Another thing to try is reducing the screen size, which would also diminish resolution (doubly so, if both methods are used.)