Experimental Video Rendering - UPDATED 8.1.2011

Downloads of unfinished development versions

Moderator: noisy

Re: Experimental Video Rendering - released

Postby noisy » 01 Jan 2011, 23:55

Flusirainer wrote:
noisy wrote:
I just replaced the FS2004 version with an updated version 2.151, please install it and see if it works correctly then.


I still have HyperCam2

http://www.youtube.com/watch?v=YqKQXwvlfgw :roll:

So you didn't try the updated version?

I wonder if you really understood how the rendering is supposed to work (from your test videos I can see that so far it didn't work correctly, would be interesting to know if the update solves that), did you read my description in the announcement?
When the video rendering works correctly, you no longer have to worry about low FPS, as it slows down the simulation as much as necessary to capture every single frame. So you can turn up both FS graphics quality and video quality as much as you like. You can't compare that to programs like Fraps or HyperCam as it works completely different.
Matthias Neusinger
Author of FS Recorder
noisy
Site Admin
 
Posts: 952
Joined: 16 Jun 2006, 17:33
Location: EDDN

Re: Experimental Video Rendering - released

Postby noisy » 02 Jan 2011, 00:22

Variometer wrote:
noisy wrote:
Variometer wrote:I just found a little tool, which helped me to make the window title disappear like it does in full screen mode so that I can get the full resolution during capturing, and so there's no reason for me to use full screen mode for recording any longer.

Can you tell what tool that is, might be interesting for others as well?


The tool is called FS-SIMtimizer, and originally it's used to avoid some problems with the fs ATC-Window, but it also has an inbuild function called 'Optimized fullscreen', which can be used to eliminate the window title of a maximized FS window. The author of this tool is Rüdiger Specht and you can download the tool at http://www.friendlyflusi.at/index.php?p ... &dataID=23

Cheers, Andreas

Thanks. Maybe I can later add such a 'windowed fullscreen' mode for rendering.
Matthias Neusinger
Author of FS Recorder
noisy
Site Admin
 
Posts: 952
Joined: 16 Jun 2006, 17:33
Location: EDDN

Re: Experimental Video Rendering - released

Postby lancia1163 » 02 Jan 2011, 03:34

Hi, this is my first post here.

First, thank you very much for this new feature and for the whole FSRecorder-program, which without FSX would not be the same.

At the moment I'm developing some new camera's for FSX in a program, which i want to provide for free as soon the application is finished.

I'm working on a new spotcamera ( http://www.youtube.com/watch?v=biT_MtQL0oU ) and a new flybycamera, both fully controllable, but without FSRecorder it would not be possible to make this.

I of course also tried your new version of FSRecorder.
Nevertheless camera-control seems to be more difficult with your new video-rendering.
I made a video with your new feature and with my flybycamera as you can see here: http://www.youtube.com/watch?v=bsBb8QpHlvE
(notice that the stutter in this video is at least rendered by the camera-application)
With some modifications in my application i can slow down the camera-movements so they can follow more or less the slower playback when doing videorender.

Nevertheless it would be very useful if you could enter a constant framerate when video is rendering, where the playback never goes above.
Now , due to the variable framerate during rendering, camera-control is harder to do.
I don't know if this would be possible.

Anyway, thank you again an a Happy New Year to all of you!


ps. sorry for my bad English.
see also http://www.youtube.com/user/lancia1163?feature=mhum
lancia1163
 
Posts: 4
Joined: 13 Nov 2010, 13:20

Re: Experimental Video Rendering - released

Postby Variometer » 02 Jan 2011, 09:20

lancia1163 wrote:Hi, this is my first post here.
At the moment I'm developing some new camera's for FSX in a program, which i want to provide for free as soon the application is finished.

I'm working on a new spotcamera ( http://www.youtube.com/watch?v=biT_MtQL0oU ) and a new flybycamera, both fully controllable, but without FSRecorder it would not be possible to make this.


I don't want to discourage you in your development, because it's always good if someone is trying to improve things (for free), but I think your planned spot camera as shown in your clip is already one of fsrecorder's native function in it's previous version. I already used this floating spot camera in one of my clips http://www.youtube.com/watch?v=8KetbsCqmNs - Or am I missing something? This camera also is working fine in the newest version of fsrecorder in combination with the new rendering function, which can be seen at http://www.youtube.com/watch?v=-0TA5ghjksw

Cheers, Andreas
Variometer
 
Posts: 8
Joined: 30 Dec 2010, 19:09

Re: Experimental Video Rendering - released

Postby noisy » 02 Jan 2011, 11:04

Hi lancia1163,

I know what you mean, the varying simulation speed during rendering can be problematic for manual camera control. E.g. if you try to do manual panning with the mouse or a joystick axis during rendering, the panning will not be smooth in the video.
I had some thoughts on that already, one thing I want to add in version 2.2 is a smooth panning option for FS Recorder cameras, which could reduce the problem. But I also recently wondered if an option to stabilize simulation speed during rendering (as you suggest) could make sense. However, I first have to try it to see if it can be done, and I guess that it will slow down rendering quite a lot.

Some questions about your camera program for my understanding: How does manual camera control work in your program? And how does your program compute frame-by-frame camera movement, do you measure elapsed time yourself?
Matthias Neusinger
Author of FS Recorder
noisy
Site Admin
 
Posts: 952
Joined: 16 Jun 2006, 17:33
Location: EDDN

Re: Experimental Video Rendering - released

Postby riatjunkie » 02 Jan 2011, 16:36

Hi Matthias

Firstly, to answer your suggestion re wmv screen resolutions being multiples of 4, I don't think that that there is any restriction in that respect.

However The results of some brief tests may be of interest.

To recap, I am running FSX under Windows 7 with a triple monitor setup (giving total resolution of 5040 x 1050 x 32) via a Matrox THTGO. cpu is a Q6600 quad @ 3GHz, My 2 Video cards are in SLI mode and are GEforce GTX460.

All the settings for FSrecorder rendering were done while replaying a .frc file in full screen (5040 x 1050).

If I check the FS Window size box, a video is produced, but will not play in WMP 12.

If I Uncheck both checkboxes, I can request a resolution of 5040 x 1050, and a video is produced, but the horizontal resolution is capped at 4096, the vertical resolution stays at 1050 (the file properties confirm this, and the file will play in WMP 12). Also the aspect ratio changes because the FS window is not resized, and the result is a vertically elongated video image which is also not smooth but slightly jerky.
Regardless of which lower resolutions I choose with both boxes unchecked, I get similar results, with distorted image (unless I choose the correct aspect ratio I guess), and jerky video.

If I check the Resize FS Window to Video Size, I can produce perfectly smooth, undistorted video (which will play perfectly in WMP 12) in all resolutions up to 4096 x 1024, including 1920 x 1024, which is great for HD TV.
Beyond the value for the vertical resolution I get a message from FSRecorder that rendering has stopped because the FS Window was resized (not by me!).
If I request a horizontal resolution greater than 4096, the value is capped at 4096.

So there seems to be a cap on the horizontal and vertical resolutions that your rendering module will accept (on my system at least - this is not a complaint, just an observation! Your system produces video resolution way in excess of anything I would need!)

Hope this has been of help

Donation sent yesterday

Regards

Dave Nash
riatjunkie
 
Posts: 10
Joined: 12 Jan 2010, 13:36
Location: Birmingham UK

Re: Experimental Video Rendering - released

Postby noisy » 02 Jan 2011, 18:31

riatjunkie wrote:Firstly, to answer your suggestion re wmv screen resolutions being multiples of 4, I don't think that that there is any restriction in that respect.

At least they have to be even, odd resolutions don't work (if you enter an odd width/height, FS Recorder automatically uses an even value).

If I check the FS Window size box, a video is produced, but will not play in WMP 12.

Maybe a width of 5040 is just too large, don't know what the maximum supported size of WMV is.

If I Uncheck both checkboxes, I can request a resolution of 5040 x 1050, and a video is produced, but the horizontal resolution is capped at 4096, the vertical resolution stays at 1050 (the file properties confirm this, and the file will play in WMP 12). Also the aspect ratio changes because the FS window is not resized, and the result is a vertically elongated video image which is also not smooth but slightly jerky.
Regardless of which lower resolutions I choose with both boxes unchecked, I get similar results, with distorted image (unless I choose the correct aspect ratio I guess), and jerky video.

Yes, I reduce width/height to 4096 if greater values are entered. If you uncheck the Resize FS Window option, frames are resized by the encoder, and the scaling is not of high quality. And yes, you have to use the correct aspect ratio to avoid distortion.

If I check the Resize FS Window to Video Size, I can produce perfectly smooth, undistorted video (which will play perfectly in WMP 12) in all resolutions up to 4096 x 1024, including 1920 x 1024, which is great for HD TV.
Beyond the value for the vertical resolution I get a message from FSRecorder that rendering has stopped because the FS Window was resized (not by me!).

The problem here is that beyond that vertical resolution the total window size (including the title bar) exceeds your screen resolution, but Windows doesn't allow that and reduces window size again. I'll work on that later, maybe I can hide the title bar to fix this.

Hope this has been of help

Thanks for the feedback.

Donation sent yesterday

Got it, thanks a lot :D
Matthias Neusinger
Author of FS Recorder
noisy
Site Admin
 
Posts: 952
Joined: 16 Jun 2006, 17:33
Location: EDDN

Re: Experimental Video Rendering - released

Postby noisy » 02 Jan 2011, 18:44

Here's a video demonstrating that with a little extra work you can add the FS sound to a rendered video:

http://www.youtube.com/watch?v=OWkxKe0y1HE

(the original video is about 10 minutes long and rendered at 1920x1080, 60fps, for the upload I reencoded only a short part of it at 1280x720, 30fps).

I'll post details on how I did it later.
Matthias Neusinger
Author of FS Recorder
noisy
Site Admin
 
Posts: 952
Joined: 16 Jun 2006, 17:33
Location: EDDN

Re: Experimental Video Rendering - released

Postby lancia1163 » 02 Jan 2011, 20:29

Variometer wrote:
I don't want to discourage you in your development, because it's always good if someone is trying to improve things (for free), but I think your planned spot camera as shown in your clip is already one of fsrecorder's native function in it's previous version. I already used this floating spot camera in one of my clips http://www.youtube.com/watch?v=8KetbsCqmNs - Or am I missing something? This camera also is working fine in the newest version of fsrecorder in combination with the new rendering function, which can be seen at http://www.youtube.com/watch?v=-0TA5ghjksw

Cheers, Andreas


thanks for your reply.
I must admit that I never got the fsrecorder-spotcamera working well. Now that i tried again i got it to work.
I've used spotcamera mode "Follow (fixed heading)" with smooth positioning and smooth zoom.
a videoclip made with FSRecorder can been seen here : http://www.youtube.com/watch?v=Ss3_zqC8Eps
As you see the camaramovements are not as smooth as they should be.
In your last mentioned videoclip the heli goes out of the center of the screen. How do you fix this with the fsrecorder-spotcamera or is it done by video-editing.

b.t.w. My congratulations for those most beautiful videos of FSX that can been seen on the net!
Last edited by lancia1163 on 02 Jan 2011, 20:33, edited 1 time in total.
lancia1163
 
Posts: 4
Joined: 13 Nov 2010, 13:20

Re: Experimental Video Rendering - released

Postby lancia1163 » 02 Jan 2011, 20:31

@ Matthias

thanks for answering my question and and to consider my suggestion.

All controls in my app. are made by the mouse (panning en moving camera left right up down + circle around the plane) in combination with the mousebuttons.
Calculations are done at real time (yes, this can cause stutter) depending on aircarft-movements and -accelerations and send to fsx through simconnect.
I know accessing FSX's memory directly works much smoother, but I do not dare you to ask how you fix that :)

as I have written above, I did not know the capabilities of your spotcamera.
I suppose you use the regular spotview from FSX and spins the camera randomly around the airplane , depending on the aircraft-movements and accelerations.
lancia1163
 
Posts: 4
Joined: 13 Nov 2010, 13:20

PreviousNext

Return to Alpha Releases

Who is online

Users browsing this forum: No registered users and 1 guest

Copyright © 2004-2010 Matthias Neusinger
cron