BCM304 - Valley Banshees, circling and rising

(7:46) Level 2 ('Halo') on Easy. In the last movie I showed examples of the valley Banshees circling downwards after spawning, but it's also possible for them to circle upwards, something I've known for years though never documented before. This movie covers a single example which lasted hours and led to a fascinating discovery. Not to be missed!

Released August 15th 2018, gameplay recorded August 12th 2018.


00:02 (Spawning and initial ascent) After spawning, the Banshee initially do some close circling, but then they move into a larger orbit, and they're slowly rising. You can see also that they're not diametrically opposite, like I suspect you always get with sinking. There's a clear 'lead Banshee'.

In the following commentary I'll use a particular notation to indicate 'evolution time' - the amount of time the Banshees have evolved since spawning. For example, +1:15 would mean 1 hour 15 minutes of evolution.

01:04 (Into the valley, and spotted) Around +0:13 I decided to go into the valley, getting an exit checkpoint on the way like in the previous movie. However, when I moved forward to be more directly underneath the Banshees, I got spotted. What you see here isn't the original time that happened, which took longer to occur and was a bit messy. I did some replaying from the checkpoint to get neater examples of being spotted (and in which I did a better job of destroying the Banshees), and I've used one of those.

01:56 (Ascending for hours) After reverting to the exit checkpoint, I now took care not to be spotted (i.e. I avoided getting too directly underneath them until they'd got a bit higher), and then I started getting different views of the ascent, partly with movie-making in mind. I ended up with a ton of footage covering chunks of the ascent (I wasn't recording the whole time - that wasn't practical) and I've edited together a variety of nice clips. As well as seeing the ascent, you can also enjoy the impressive sights and sounds of the valley!

03:30 Around +1:15, the Banshees were just about getting out of range for the readout in my zoomed sniper scope. A few seconds after this clip, in which I'm failing to get a readout, I moved a few metres and briefly got some readouts. The highest was 426.138, and that was the last I got.

04:15 This 'pushing 6 hours' clip was namely around +5:40. By now the Banshees were so distant it was hard to quickly spot them with normal vision; but things had been that way for a few hours already. The blue beams were helpful in guiding me towards their location.

04:35 (Breakup and diving return) Remarkably, the circling eventually broke up and the Banshees came back down! The first time this happened, I wasn't recording. I was doing stuff on my computer while the ascent continued. Then I suddenly noticed MC getting attacked on my TV screen, and I was like "Whoa, something happened!" Fortunately I'd checkpointed quite recently (just in case something interesting happened), so I was able to revert and let the evolution unfold again.

After two replays it looked like breakup would always occur after about 9 minutes; namely around +5:42. So on my third replay (which is actually where the 'pushing 6 hours' clip was from), I checkpointed after about 7 minutes (right after that clip) so I could then do replays in which the breakup would come after only about 2 minutes (more convenient). The footage you see here is of one such replay. The lead Banshee breaks off (that was always the case), the other follows, and after a brief bit of spiralling and interplay they get more separated and simply dive, which lasts about 80 seconds. I gave them a warm welcome of course!

06:03 (Stabilizing) My replays also revealed two alternative outcomes of breakup, both of which seem more common. The most common is illustrated here. The Banshees arrest the descent and stabilize into slow orbiting. Because it's slow, I expect they're rising, but I can't say for sure (I need to do more research).

06:30 (Circling return) The other outcome I saw is that the Banshees go into sinking circling, as seen here (which was the first time I saw it). This descent lasted just over 10 minutes. More specifically, it took them about 10 seconds to get into the circling, then there was almost exactly 10 minutes of that. Remarkable stuff! It was already impressive enough that they got to such an enormous height; but for them to also be able to come back down with yet more circling, that's pretty amusing!

During the descent I drove to a few other locations to get other views, before returning to the original area. But I haven't used any of those. Seemed better to just sequence some views from the same place, and keep it concise. The two short clips used are from about 6 and 8 minutes into the descent.

Closing remarks Newly created for the theme of this movie, I was using a set-up in which the Banshees seemed to always rise. Sometimes the rise was extremely slow however. In this case the rise looked to be relatively fast, which is why I let it continue. I hadn't expected things to go on so long however! Even though I wasn't recording constantly, I started running low om disc space and had to make room by temporarily transferring a load of stuff onto backup devices. I don't think any earlier has movie strained things that far - nor given me such a daunting editing job thereafter!

In regard to behaviour after breakup, the outcome appears to be dependent on how the Banshees interact when they end their initial short spiralling phase. They might happen to get back into some orbiting, which could either be fast and sinking or slow and probably rising, or they might get too separated in which case they end up diving.

Various questions remain. How fast can the Banshees rise? Do ascents always end with breakup (assuming you're not spotted), and are they due to contact with a ceiling or what? If there's actually not a ceiling up there, can Banshees go significantly higher than seen here? Probably the next thing I'll do on this topic is to see if I can get Banshees rising significantly faster than seen here. That would be helpful. Later I might try to estimate the height of breakups.