Replays that are viewable in WMP etc without the use of Game/Screen Recorders
Hi guys,
This set me thinking when I was viewing replays which sometimes may contain some interesting happenings in game or for investigations. Until now, replays are saved under the MPR format which can only be viewed in LFS itself(if I'm not wrong).
I'd like to suggest a video/format converter that converts the MPR replay to other media types like AVI/WMV/MP4/MOV etc. File formats which are compatible with media players and is a great ease to edit and upload.
If this were to be added to the LFS, game movies will no longer need to be recorded first by screen recorders but can be readily converted in-game straightaway into viable formats.
MPR is NOT video data, you guys have to understand that. MPR is telemetry data, no visual data at all is stored in MPR files. What LFS does when playing an MPR is simulate the recorded events again, it's not so much a playback as actual re-simulation. There is no way to convert MPR files to video files.
Think about if it wasn't a resimulation, you couldn't go into shift-u and get that perfect picture, or even look at a different driver or look at somthing in a different view if it was a plain old video. Its not practical to have a video replay instead of a recreation, because then your very limited on how many different things you can see in a plain old video.
I think the suggestion in the first post is flawed. What I'm saying is export what is shown on the screen from inside LFS, this means we don't have to use external software for making videos.
So you want Scavier to waste time developing code that records the screen to a video? Even though it's been done before, and that software is available for free? You can easily record videos from LFS (I've done it loads myself), without wasting months on Scawen coding it.
Wouldn't work. LFS can't know how what you want to view.
The most that could be done is to add a built-in feature to record to a movie-file when viewing a replay, but as Tristan said, there are already many great programs that can do that, so it would be waste of Scawen's time.
I agree that it's not in focus currently.. But it's still in my mind part of a finished piece of software.
Stop putting down ideas because it takes away the dev's time! This is an improvement suggestion forum, and this certainly is an improvement.
The exporter should be REALLY simple to code (and I mean REALLY). There's libraries to do this. You can use the library, or use some opensource code that won't have license consequences.
It knows what to export, because it would take WHAT IS SHOWN ON THE SCREEN and export it.
Well just buy GTR Evolution and be done with LFS then. Again, stop putting suggestions down because of other software. Thanks.
By the way, FRAPS is not an improvement to LFS. FRAPS also costs money. I have FRAPS, but it is not an improvement to LFS.
Why should he? There's no point in doing something that's been done before, especially if there are free solutions available. Besides, there're enough crappy videos out there already, no need to make it even easier for those you can't be bothered to spend just a few minutes to configure their screencapturing software.
But it's a pointless improvement because it's already been done well enough.
Why not ask Scawen to make a word processor inside LFS to save us having to load MS Word? Why not ask Scawen to make a first-person shooter in LFS to save us having to load whatever FPS is currently king?
There is no point duplicating stuff that's already been done purely for the sake of saving you money or time.
If he DOES do this (which he won't), I hope he charges an extra few bob for it. Then you'll wish you'd just used Fraps.
Fine that's your opinion. I still feel that it's a pretty nice feature to have, recording races within LFS. It's not first priority, it's very low on the priority list clearly. But for me, at some point it's there.
By the way, the things you compared with this were completely ridiculous, just so you know
Why not? It's not like you can totally forget about the time that it takes to create a certain functionality. You have to always take in to consideration if the time spent doing something is worth it. Besides this section for the forum is not just for people to throw all kinds of ideas in to the air and leave them there. It's about also discussing about the ideas and to see what other people think. These threads really don't make any ultimate decision about should these things really be done or not anyway, but if a dev comes to a thread like this it's probably important that they can see opinions from both sides. The devs can probably think for themselves and discard opinions that are not true in their own mind.
It's foolish to think that you can throw an idea out there and then tell people not to have negative opinions about it. It just doesn't work like that.
You didn't read the rest of what I wrote.. But I don't really care so much about this topic that I want to discuss it. I'm just saying, it's pretty nice to have sometime.
I only quoted a small relevant part because I wanted to address a common misconception that many people seem to have. My post wasn't only targeted at you or this thread, but in general.
If you can think of a way of reading the contents of the entire screen to memory, converting it into the correct video format, writing it to disk about thirty times every second AND allow LFS to function without any slowdown on low-end computers... you'll be worth millions by the time you leave school!
That's why LFS lets you slow down replays. If you want it to create this video live of course that wouldnt work. There is an extra problem with the original idea however - do you want a video just from inside your car with none of the flexibility of chosing more appropriate camera angles?
The suggetion to me is fundamentally flawed. Any video file created automatically is going to have limitations in camera angle, onscreen captions etc.
What LFS does very well is provide an interface for us to get the angle we want - whether we're doing it by hand with the shift-u camera, or doing it by insim - it allows us to record LFS at reduced framerates so we can screen capture on low end computers (This is how I did the early STCC videos), and the replay format itself is so versatile we can recreate any camera angle or even change our mind.
With this in mind, given the vastly superior replaying functionality of the MPR/SPR formats, to recreate any part of the simulation for playback. How could a video file be considered an improvement?
Well the answer lies in what you want to use the video for: Are you broadcasting a race? Well in that case you'll likely already be using a 3rd party camera control tool designed for that purpose, or are you making an artistic video for the showcase? Well you'll probably be setting up each view in particular detail using shift-u.
LFS cannot be expected recreate either of these scenarios automatically, there are other uses for video aswell of course - would LFS need to support these too?
Regarding compression, yes there are free libraries available to make video files but ultimately we would end up in the situation where some people want .mov, some people want .mpg, others want particular compression settings or ratios etc. You can get away with a certain amount by running it through windows and having the user chose, but we come back to what to output...
Would LFS need setting up as a video editor? It's never going to have the full power and potential as a video editor as a dedicated program would.
In short, what you would end up with, is a meens of editing a replay and specifying camera angles to create a final result that is both inferior in quality to standard LFS graphics (as videos are), with less features, and a much larger filesize.
The price for this new feature would be extensive development of internal editing tools, new camera functionalities, and end users still having to download and install video encoders for creating whatever filetype they want (hense, still using 3rd party software anyway), and the final result is never going to be as good as using proper tools for the job...
And all of this would need doing when LFS already has exceptional replay and both automatic and manual camera control interfaces.
I agree completely with Becky Rose. I just can't see how a built-in screen capture that you could use while WATCHING a replay (and thus being able to control cameras) is such a big deal to make.. I used to make hobby projects like that in C++ in like half a day when I was into coding
But I do agree that the half day would be better spent improving other aspects of the game.
