Tutorial: How to Record a Modeling Session

Screen recording is a great way to share your knowledge with friends and demonstrate your skills to potential employers. It’s an easy and fast solution to present your ideas to the world and it’s fairly easy to learn too. All you need are good tools, some planning and few tips for a good start… (pdf version).

Planning

There are many screen recording workflows. The decision depends on a number of factors like: the outcome you wish to attain, your viewers attention span or even their internet connection speed. That’s why it’s essential to plan your work before you even decide what software to choose.

Timing

We are going to record a modeling session and this is often a lengthy and fairly repetitive process. That leads to a boring output and we don’t want to bore our viewers to death. A good way to avoid it is to speed up our video.

Time lapse is a process of recording video at a specific (often low) frame rate and applying a higher playback rate to the output. Personally I believe that 1:4 ratio (x4 time lapse) is the best choice for showing modeling sessions, so I will focus on it in this article.

Output

Now it’s the time to decide what output format we should choose. If your monitor resolution is higher or equal to 1280×1024 I would advice to use HDTV 720p format. This is a new technology and it’s supported by most of video sharing sites nowadays, besides you are always left with an option of compressing your video in smaller formats like EDTV or LDTV if needed. HDTV 720p stands for 1280×720 non-interlaced. I’s a wide-screen 16:9 aspect ratio, more or less compatible with most of modern LCD screen sizes. There’s one more advantage of using a standardized output – most of video compressors are optimized to produce the best quality versus size in these formats.

HDTV 720p is compatible with many different frame rate standards, but video sharing sites accept 24fps only. Since we speed up our video 4 times we need to record only 6 frames per second, but I would advice to use 12 instead. This way we will get more data to interpolate from in case of a dropped frames. As a result our output video should be smoother.

Recording area

Once we have our output video format chosen we have to decide what screen area to record. It’s up to you whether to use full screen recording or fixed region, but you should keep in mind that a perfect screen recording size should be equal or bigger than the output and it should have the same aspect ratio. This way we can avoid unnecessary cropping or stretching later on. My advice is to use fixed-region with the highest possible screen width and adjust heigh to meet the 16:9 aspect ratio as shown on illustration below.

recording areas

Keep in mind that big recording area may result in low performance. You can check the the frame rate in CamStudio window while recording. If it drops under 6 fps I advice to use smaller resolution.

Workflow

We will divide the process into two main steps: recording video and producing final output. That means we are going to grab a recording area in original resolution and save it using a lossless codec. After doing so, we will produce final video using more efficient compressor, smaller resolution and faster frame rate.

This approach is time consuming and it requires a lot of disk space to store lossless data, but will result in best video quality and will give us a flexibility of reusing recorded video with many different production settings.

Tools

There are many tools for both screen recording and video production you can choose from. We don’t need the ability to record many full screen frames per second nor professional video editing tool set, so we will try to accomplish our goals using freeware only.

• CamStudio – open source screen recording tool that should be able to record 6 full frames per second. It comes with a lossless video codec.
• VirtualDub – well known, free video processing utility. It’s biggest advantage is processing speed. I personally find it much faster than many commercial tools I had chance to play with.
• x264vfw – vfw version of a great x264 codec, licensed under GPL, supports multithreading

Preparations

Download and install software.

• CamStudio + CamStudio Lossless Codec
• VirtualDub 32bit
• x264vfw 32bit

CamStudio settings

1. Open CamStudio
2. Open Region > Fixed Region and enter your recording area dimensions
3. Click OK
4. Open Options > Video Options
5. From the Compressor drop-down list select CamStudio Lossless Codec
6. Use Auto Adjust checkbox to enable Slider below
7. Move the Slider until you get Playback Rate at 12 frames/second
   

   Note: We are forcing CamStudio to save 12fps video file, but most likely real recording frame rate won’t reach this level. I encourage to use 12fps anyway, as it leads to more fluent 24fps conversion.
   

8. Click OK
9. Open Options > Program Options > Temporary directory for recording > Use user specified directory … and choose a location for temporary files
   

   Note: Recording fullscreen requires a vast amount of disk space. 20Gb should be enough to record 1-2 hours session in 1080p format.
   

10. Open Options > Keyboard Shortcuts
11. I advice to remap Cancel Recording Key to reduce the risk of confusing it with Stop Key
12. Close CamStudio to save modifications

Long recording issues:

I faced a number of problems with recording long modeling sessions in CamStudio. I’m not sure if it’s a codec issue or a general rule in CamStudio but lossless videos longer than 45 minutes may get corrupted. Sometimes it is possible to recover those files using VirtualDub error mode, but there is no rule. The only workaround I’ve managed to find is to stop recording after 30 minutes, save file and start another one. There’s an option to stop recording after a certain amount of time in Camstudio. To activate it go to Options > Program Options > Preset Time and set 1800 seconds. Your recording will stop after 30 minutes and you’ll be able to start another one right away. You can easily merge two or more video files in VirtualDub using File > Append AVI segment…

VirtualDub settings

Output size:

1. Open VirtualDub
2. Open Video > Filters…
3. Click on Add… button
4. Double-click Resize on filters list
5. Filter: Resize window appears. Select Aspect ratio: Compute height from ratio and enter 16 : 9
6. Select Lanczos3 from the Filter mode drop-down list
   

   Note: You can use Show preview button to experiment with different filters.
   

7. Select New size: Absolute (pixels) and enter 1280 x 720
8. Click OK to close Filter: Resize window
9. Click OK to close Filters window

Frame rate settings:

1. Open Video > Frame Rate…
2. Select Change frame rate to (fps) and enter 48
   

   Note: This will speed up our video 4 times.
   

3. Select Convert to fps and enter 24
   

   Note: This will interpolate 48fps data and convert it to 24fps output
   

4. Click OK to close Video frame rate control window

Compression settings:

1. Open Video > Compression…
2. Select x264vfw – H.264/MPEG-4 AVC codec from the list
3. Select Force keyframes every and enter 24 frames
4. Click on Configure button
5. Select VirtualDub Hack
6. Open Rate control & Other tab
7. Set the Threads number if you want to use multithreading option
8. Click OK to close x264vfw configuration window
9. Click OK to close Select video compression window
   

   Note: As you can see there are many more compression settings not mentioned here. Personally I use more or less default configuration, but feel free to experiment.

Saving changes:

1. Select File > Save processing settings
2. Choose desired location and name the file “timelapse_x4_720p_24.vcf”
3. Click Save
   

   Note: This file will store compressor settings as well; You can create multiply files to quickly load different production settings using File > Load processing settings.

Quick Example

Recording video

1. Open your modeling application
2. Open CamStudio
3. Click on Record button
   

   Note: If you wish CamStudio to minimize on recording start, choose Options > Program Options > Minimize program on start recording. Alternatively you can minimize it manually and use keyboard shortcut to start recording.
   

4. Recording area frame appears
5. Choose the area you wish to record, left-click to activate recording, right-click to cancel
6. Start recording, you should be able to see flashing corners indicating that CamStudio is working
7. Customize your modeling application interface to meet the recording area nicely.
8. Stop recording using keyboard shortcut (F9 default), or by right-clicking on CamStudio icon in system tray and selecting Stop
9. Save AVI File window appears
10. Choose desired location and name the file “quick_example.avi”
11. Click Save

Producing final output

1. Open VitualDub
2. Select File > Open video file…
3. Open “quick_example.avi”
4. Select File > Load processing settings
5. Open “timelapse_x4_720p_24.vcf”
6. Open Video > Filters…
7. Double-click Resize on filters list
8. Filter: Resize window appears
9. Enter new size: 768 x 432

Note: From the compression point of view optimal width and heigh are multiples of 16. In case of 16:9 aspect ratio I would advice to use: 512×288, 768×432, 1024×576 or 1280×720.

1. Click OK to close Filter: Resize window
2. Click OK to close Filters window
3. Select File > Save processing settings
4. Choose desired location and name the file “timelapse_x4_432_24.vcf”
5. Click Save
6. Select File > Save as AVI…
7. Choose desired location and name the file “quick_example_x4.avi”
8. Click Save
9. VirtualDub Status window appears, it will close automatically when our example video is finished

Summary

As you can see, recording screen activity is a very simple process and can be done entirely using free software. I encourage you to experiment with all settings I haven’t mentioned here. Remember, you can save your own processing files in VirtualDub and create your own library of video production settings.

I hope this tutorial will inspire you to create time lapse videos and share your modeling knowledge with others. In case of any questions feel free to contact me.