A (possibly) final question

Now, I am literally at the end of the game making process and I have an important question. I need to be able to export the game video into a Windows on-screen recorder of some sort to make a professional trailer for it.

I have tried a couple of video capture programs but boy do they… draw vacuum. You could boil ice water in a bell jar with them. Anyway.

Does anyone have a recommendation of a recording program capable of capturing the contents of a screen window? It would be great if the program was open-source or at least not expensive. I don’t need a great deal of video in terms of time, but it has to preserve the overall quality of the game when I assemble the the trailer.

Thanks for all the help that you guys have provided so far, I truly appreciate that.



If you want to record a video while preserving quality, the best bet would be to use base.movie(), which saves it as a bunch of png files. It will slow down the Panda clock as is necessary so that the resulting movie will reflect the actual game speed even though the game might be lagged significantly as a result of using this function.

Fantastic, thank you so much.

Okay, a quick look at the manual shows no information about this module. Where do I get some information about it? It sounds ideal.


I imagine that this is the method to which rdb refers.

(Note that “base” generally refers to a specific, automatically-created instance of the “ShowBase” class; even without that, it’s possible to search for a method in the API reference, and while the links provided tend to not work for me, it does at least usually give the relevant owner class, I believe.)

So, here it is. I am remaking my trailer and using the movie API call to record the video. I got the call to work with minimal problems, put it in a specific function call to make it record the action at that point (pretty simple) and then watched the game slow down when the function kicked in. Marvelous.

I am watching the animation unfold slowly as expected. It shows exactly what I expect. I spend an hour generating 10 seconds of video at 12 seconds per frame, thirty frames per second. Seems pretty darn slow to me, considering I am using a Qosmio. Whatever.

At the end of getting the action where it should be and snapping a frame every so often, I halt the process and load the frames into Virtual Dub to create the movie clip. Nothing.

What I mean is that each and every frame is exactly the same. No action. Yet I watched it unfolding on the screen.

I took a short test at another point and got the same results. It rendered faster but the output frames are all identical. The video() function is spitting out the same image over and over. Now what???

Never mind, I looked at the source an figured out what the *&?! was happening. I figured it out.

So, you can see a trailer of the game at www.evolveabot.com by clicking the large logo.

It’s not meant to be amazing graphics or anything but it is a game that does something unique. It uses evolution to create robot control software. You might enjoy a brief look at it there.


Congratulations on getting it done, and good luck with it! :slight_smile:

May I offer a little critique on the trailer? (Before I do, let me note that I write this on the assumption that the trailer is intended to be a full trailer, not a teaser; if I’m mistaken, then the below may well not apply.)

For me, at least, it spent far too much of its time on the introduction featuring the industrial robots; when we finally saw some gameplay, it was only a brief look at a single screen, with lots of buttons at the bottom and little real indication of what was going on. As it stands I have little idea of what the player actually does in your game, aside from comparisons with other games that are visually similar.

I really do recommend including more gameplay: show viewers how it plays, what they get to do in the game, and what’s exciting about it: that is–I think–what’s most likely to convince a potential customer to actually buy it. Conversely, I’m inclined to recommend paring down the intro sequence, making it quicker (without reducing the amount of backstory conveyed).

(It might also be a good idea to embed your trailer into your page, rather than one click away: the more that potential customers have to do to become interested in your game, the more opportunities that have to decide that it’s not worth the bother.)

Thanks, I will be expanding things immediately.