Keep in mind that it may take me a while to accomplish this, so I am going to sort of use this post as a wiki and edit it as I go. The thread will be open to questions and discussion, because while I am using Mac-only software for a significant portion of this process, the concepts are the same and I know a good bit about encoding even if I may not use the same software that you are using. So feel free to ask questions and discuss things in this thread so that it is not only The MordEth Way™.
The Fuji FinePix F10 that I have (inexpensive 6.3 megapixel digital still camera) also has the ability to record video, and so John and I decided some time ago to record his phonographs for YouTube. With a 1 GB xD picture card, I can fit approximately 15 minutes of video on my camera at 640 x 480 pixels (standard definition).
Our current procedure works like this:
- We set up with a tripod and film as many videos will fit on my storage card.
- I then dump the video off onto either my laptop, John’s computer, my iPod (it makes a great portable external hard drive), or another means of storage.
- We repeat the first two steps until we either run out of light (e.g. the ‘Victrolas in the Park’ series) or otherwise have done whatever we want to get done.
- I take all of the recorded videos from the camera (which are recorded 30 FPS in an AVI container) and put them on my laptop.
Note: Your camera will quite possibly record in a different format, so you may need to handle it differently. - I prefer to use QuickTime to trim the beginning and end of the raw video from the camera. You can use QuickTime on either Mac or Windows, but you can use other software for this.
- Optional: If the video that I am using is intended to be one of more than one cut used in a video (a number of our newer videos switch camera angles), I have a video file from which I export just the audio as AIFF. This will be added back to the video later.
- If I did not yank my audio to mix in later, I take a still of the first and last frame in the video to use for fading. The version of iMovie that I am still using has certain bits of weirdness when you go to cross fade content with audio and content without audio, and so I fade a single frame still instead of the video. Look for it in our videos.
- I then export the video from QuickTime using the h264 codec. This is the best quality codec that you can use. Depending on the aspect ratio (keep in mind that I am doing standard definition), I use 1600 kbps for my bit rate, with a 2-pass encode, 30 FPS, at the best quality. If the video still has audio, I encode at 192 kbps, again at the best quality.
You should play with these settings until you figure out what looks best for you, and if you do not want to add any more bells and whistles to your video, you are done at this point.¹ - I import my video and audio pieces (the latter if applicable) into iMovie. Since iMovie uses QuickTime using the h264 natively, this only takes a few seconds, rather than the minutes the encode in the step above takes. Quality takes time.
- I then order my pieces. We have pre-made splash graphics for John’s YouTube videos, and these are saved in an iMovie template so they can be dragged into place. The cranking sound (and other sound effects) used in the beginning of a video is also saved in the template. We also scan the label (and not photograph it), and import PNG images (for optimal quality) of the record label(s) into this template. The usual order is: (fade in) opening splash, (cross fade) artist/composer image (if applicable, and if so, fade afterward), record label image(s)—and if multiple, fade between them, (cross fade) the still used for fading (if applicable)—with no between it and the video, the video, the end still (again, if applicable) and a cross fade followed by the credits images. We add an audio file that I made of John winding one of his machines during the intro stills.
Note: I will be adding screen captures to illustrate this because I think seeing is easier than reading about it. - Once all of the pieces are in place, I watch the transitions to make sure I did not mess up anything (e.g. flip the stills used for fading). Always check your work.
- I then encode for YouTube in iMovie. Since Apple uses QuickTime’s core for all of its video handling, it is exactly the same as encoding in QuickTime, but this time around we want to encode using the DivX codec (or Xvid if you don’t want to pay for DivX—for all practical purposes they can be used interchangeably). DivX produces better results for YouTube. For standard definition (640 x 480), I use at least a 1100 kbps bit rate, 2-pass encode, 30 FPS, at best quality (if you are using the DivX codec and not Xvid, this would be ‘insane quality’). For audio, DivX leans towards MP3, so I recommend 192 kbps MP3 audio. Here, it’s actually better if you ripped out your audio originally, because doing one lossy encode is better than doing two, but on YouTube no-one is likely to notice the difference.
- You are now done, watch your video to make sure it encoded well.
- If the video looks good, upload it to YouTube, tag it thoroughly, and enjoy sharing your phonographs with others.
As I mentioned before—stay tuned for screen captures and further illustration of my process—and note that it can (and no doubt should) be fine-tuned for what you are using. I am using ‘home’ software for the Mac and not ‘professional’ software, even though I have it—for what we are doing, this works quite well enough.
Also—if you are interested in having me help you with splash graphics for your channel or videos, please let me know. I would be glad to, although I’d ask that you please credit me for them.
Your friendly internet daemon,
MordEth