Quote from Victor :

...
I guess it requires something like depth maps to indicate the amount of blur on each position on the screen .. just guessing here ..

That's what I said in my posts, that would be barely possible.

oh, right. yeah those maps seem to make sense. But yeah LFS would have to export that for proper maps that cover all of the lfs 3d world. But if lfs can do that, then it can also apply the depth of view effect.
So no, this is not possible at this time.

Well that´s a pity but there´s nothing I can do about it...I will have to do it manually as I have already tried...
two layers of video track...upper layer is masked according to shape of car and sharpened...lower layer is blurred but constantly along whole image which doesnt look the best but at least partially fakes the effect...if this is done properly then it looks a bit better than without DoF but if lfs could do it by itself that would be the best option...
you can start working on it while scawen works on tyre physics and race which one of you will be the first one to finish their work (just kidding)

One way would be to render the Z-buffer to a separate surface and export the screenshots as RGBA PNGs where the Alpha channel is used for the Z-buffer. On the other hand, since Record can read SMX, it could simply determine the depth itself and the dimensions of the cars can be taken from their respective models.

Quote from morpha :

On the other hand, since Record can read SMX, it could simply determine the depth itself and the dimensions of the cars can be taken from their respective models.

well, yes, but not very precise in all places. Most scenery is missing from the smx's, so those wouldn't be included in the depth map. Maybe it would still give a nice result though, but it wouldn't be an exact match with LFS's detail.

Either way, I think for me this goes a bit over my head atm. You're welcome to give it a shot though, if you think you can do it LFS Record's source is available at http://lfsrecord.codeplex.com/SourceControl/list/changesets (vs2010 c# .net4 solution)

Is there away for LFSRecord to animate the cameras without taking the focus from LFS? It would be useful to be able to record the car sounds from the position of the camera, even if only to record them using a third program.

I took a look at 2.07 today, and saw the nifty looking Audio Layer. Are you able to elaborate on what the intent is for this layer?

Quote from JayEyeBee :

Is there away for LFSRecord to animate the cameras without taking the focus from LFS?
It would be useful to be able to record the car sounds from the position of the camera, even if only to record them using a third program.

I took a look at 2.07 today, and saw the nifty looking Audio Layer. Are you able to elaborate on what the intent is for this layer?

Firstly the audio-output i'm working on is mostly useful so people can put a music track under the animation so they can sync it to the music.
Eventually I hope to be able to record sound from LFS itself, and then you can put that sound on the timeline too.

A new patch is available : version 2.0.7
Biggest new thing is audio output and audio tracks. You can import audio files (wav/aif) via the File->Import menu and then drag the imported audio file onto an audio track (audio layer) to use it.

Download (Edited to link to 2.0.8 as 2.0.7 had a bug)
=======
Full : http://www.lfs.net/LFSRecord/LFSRecord2.0.8.zip
Patch : http://www.lfs.net/LFSRecord/LFSRecord2.0.8_Patch.zip
(you can only use this patch if you have previously installed 2.0.5 or up)

Changes for 2.0.7
=============
New : Copy (ctrl-c) and Paste (ctrl-v) keyframes
New : Layers are color coded in timeline view
New : Can import audio. Supported formats are WAV and AIF
New : Audio playback supported. Settings in Project Settings window.
New : Audio Track layers. Drag and drop audio media on top of the animation area of the audio track to add audio regions.
Change : Importing media files is now via one generic menu command (autodetects media file type)
Fix : Moving LookCamera's position and lookat in viewport (while pressing ctrl), where either or both had an animation was messed up.
Fix : Undo/redo of moved keyframe would be wrong if it had moved beyond another keyframe

Quote from Litro :

Eh, looks like yes, LFSRecord doesn't help to me record videos without FPS lags. VirtualDub is the same.
LAGs even more than with fraps

That's cause bitrate of uncompressed video is bigger than your HDD can read at, just compress your video using either H.264 with crf=18, or, if your GPU doesn't supports hardware H.264 decoding, then use Xvid with quantizer=2.

Quote from E.Reiljans :

That's cause bitrate of uncompressed video is bigger than your HDD can read at, just compress your video using either H.264 with crf=18, or, if your GPU doesn't supports hardware H.264 decoding, then use Xvid with quantizer=2.

Sorry, but I don't understand any of these words. I know i'm noob in software things.

I'm having some difficulty with the audio output. I import a .wav and I drag it down to an audio layer. I let the animation run, and I while I'm connected to LFS I can not hear the file. When I disconnect insim, I can hear the file, but it is sped up.

Quote from JayEyeBee :

I'm having some diffculty with the audio output. I import a .wav and I drag it down to an audio layer. I let the animation run, and I while I'm connected to LFS I can not hear the file. When I disconnect insim, I can hear the file, but it is sped up.

awrh right. Fixed that no sound issue - I'll release a new patch in the morning.

As for the file being sped up; can you check in Project Settings -> Sound which sample rate is being used? And compare that to the sample rate of the file as shown on the audio media object?
If they don't match then atm they'll be either played back slower of quicker. Ive' not done on the fly sample rate conversion, though I'm still thinking about it...
In any case, if you match the sample rate of the file in the project settings -> sound options, then the file should be played back ok.

Quote from Litro :

Sorry, but I don't understand any of these words. I know i'm noob in software things.

In VirtualDub,

1. open .bmp's
2. Video -> Frame rate... -> Change frame rate to -> input 25 -> ok -> ok
3. Video -> Compression -> Xvid MPEG-4 Codec -> Configure -> click on "Target bitrate" (might be not needed (that button might not be present)) -> set quantizer to 2.00 -> OK -> OK
4. File -> Save as AVI
5. profit

Quote from Victor :

In any case, if you match the sample rate of the file in the project settings -> sound options, then the file should be played back ok.

D'oh. Looked at that the second after I hit 'Submit', sorry.

New patch available : v2.0.8

Fixes the no sound bug in v2.0.7 when playing back a replay while being connected to LFS.

Download
=======
Full : http://www.lfs.net/LFSRecord/LFSRecord2.0.8.zip
Patch : http://www.lfs.net/LFSRecord/LFSRecord2.0.8_Patch.zip
(you can only use this patch if you have previously installed 2.0.5 or up)

Quote from E.Reiljans :

In VirtualDub,

1. open .bmp's
2. Video -> Frame rate... -> Change frame rate to -> input 25 -> ok -> ok
3. Video -> Compression -> Xvid MPEG-4 Codec -> Configure -> click on "Target bitrate" (might be not needed (that button might not be present)) -> set quantizer to 2.00 -> OK -> OK
4. File -> Save as AVI
5. profit

I can't find this.

Hmm.. what other encoders you have there? Mind posting a screenshot?

Quote from E.Reiljans :

Hmm.. what other encoders you have there? Mind posting a screenshot?

Here

you could install an additional codec for xvid / H264 DirectShow support. Then you'll have xvid / H264 to select from in that list.
Something like this : http://www.free-codecs.com/download/ffdshow.htm

(actually that would show up as "ffdshow Video Codec" in your codec list)

Quote from morpha :

One way would be to render the Z-buffer to a separate surface and export the screenshots as RGBA PNGs where the Alpha channel is used for the Z-buffer. On the other hand, since Record can read SMX, it could simply determine the depth itself and the dimensions of the cars can be taken from their respective models.

I imagine you are refering to this effect: http://www.videocopilot.net/tutorial/2d_depth_of_field/
I tried faking it not with a ramp but with a solid rectangle and couldn't but it's the same idea. Problem is I'm not sure if you can read the z-buffer from the outside.

Quote from Victor :

you could install an additional codec for xvid / H264 DirectShow support. Then you'll have xvid / H264 to select from in that list.
Something like this : http://www.free-codecs.com/download/ffdshow.htm

(actually that would show up as "ffdshow Video Codec" in your codec list)

..or actual Xvid codec (some ffdshow versions are unable to encode to MPEG-4 (Xvid) :s ).

You can get it here: http://www.koepi.info/Xvid-1.2.2-07062009.exe

Hi guys, i have a problem. I have a 30 min replay and i need to record only last 3 mins, but i can't choose a comp length more than 300sec. Help me plz

On ReplayControl layer, set replay offset to 'replay_lenght - 3 minutes'

Whiskey TY man that is not what i want but you helped me very much) It seems my replay is broken O_o because another one works fantastic.

Could somebody tell me about AE Export? When i export my project i don't see the vid. Is it all right? If not, what's a for this feature

AFAIK the AE export option only includes information about camera position and such things, so you can do 3D effects that moves exactly like the camera. You still have to record though LFS Record and import it into any video editing program