PerroneFord
Contributor
It's late and I can't sleep, so I thought I'd offer a post detailing some posting work I am doing with a member here on SB.
In this modern world of HDV (which should be dead in a year or two with AVCHD taking it's place) editing has gotten more difficult and costly. Those who need to edit their HDV files have essentially 3 choices and then must create a workflow around it.
Choice 1. Stay in HDV. Capture your native files, drop them onto the timeline, cut, color, and output into your desired format. The drawbacks here are that HDV is notoriously hard to work with, it's slow, it doesn't hold up well as it gets edited and rendered and changed, etc.
Choice 2. Capture into 2 formats, the original HDV and a smaller "proxy" file. The thinking here is that you leave your HDV files untouched on the drive, and use the smaller proxy files for editing. When the time comes to render, you replace the proxy files with the originals, which preserves all the edits, color changes, titling, etc. And you get a nice clean render. The drawback here is that you are going to need more disk space, and things get complicated with multiple versions of files running around.
Choice 3. Capture (or immediately convert) into an intermediate format. This can be an uncompressed AVI or QT file, or better yet a file type designed for the work like Cineform. Preferably a 10 bit codec so that all the color space is preserved. File types like Cineform offer faster editing speeds than native HDV, hold up to many generations of edits and renders, and preserve all your colors. The drawback is that they are larger than HDV files and not all editors support them.
So this week, I have been doing posting work for a cave diver / videographer here. He has a Mac, I don't. He needed to send me either uncompressed or mildly compressed video so that I could get clean video to post with. Our solution was for him to buy a large external HD, and for me to purchase some software called MacDrive which lets me mount a Mac formatted drive on my PC. Works like a charm. Be mindful if you go this route that uncompressed or mildly compressed 1080i files are about 150GB per hour.
After playing with different workflows this is what I settled on.
1. Copy Mac Native QT files (photojpeg compressed) to my hard drive.
2. Convert 1080i QT files to 720i AVI Huff compressed files Since no edits need to done, just coloring and exposure cleanup, this is fine.
3. I made the decision to go with Proxy files, so I immediately cut a 1/2 size AVI proxy file at 640x360.
4. The new 720i file is deinterlaced, denoised, and cleaned up. A new 720p working file is created and the old QT and AVI files are deleted to save space.
(All processes to this point are done in batch overnight as pre-production steps)
5. Open the proxy file in the editor, which in this case is Sony Vegas. Apply an envelope to the file on the timeline that has my color correction tools on it.
6. Split the file scene by scene which will let me correct exposure and color on a per scene basis.
7. Color grade the scenes (this takes about 3-4 hours per 10 minutes of video)
8. Cut a working version of the file to a streaming format and upload to a streaming server for the client to see. Burn timecode onto the file so that the client can comment on the work and can ask for changes based on specific time in the video.
9. Apply any final corrections to the proxy version.
10. Do a media replace in Vegas which will replace the low res file with the 720p working file. Render a 720p or 540p streaming version for web use.
11. Render a 720p master file in a mastering format for archival purposes.
12. Render a 720p version for use on HD like Blu-Ray. (mpeg4)
13. Write all formats to specific folders on client disk for future use.
While I realize this workflow won't work for everyone, it is a viable workflow for doing professional type work when you can't afford mistakes or problems. The original shipped files are never touched other than to make a working copy. At no time during the process is there ever just one copy of the file. At each step, the best version of the file for the job is used. At each step in post, the client can see the work either as a proof, or as a full sized version.
For most here, this would be done with one video at a time. For those who take on work for others, you may have a number of files, so a clean method of working and naming files and folders is crucial.
Something else to note. If you are rendering for the web, or for non-finished format, you might want to overlay your name, logo, or some other identifier on the file. This allows you to identify your files if they should find their way to someone else's hands without your permission.
Examples:
This is a link to a small proof file with timecode burn. I'll leave this up for a few weeks or months if it's ok with the client. File is encoded at 512kbps, 24fps (drop frame), and 640x360.
mms://199.242.67.132/wd
A copy of that same file can be found on my Vimeo site without the timecode burn. The 720p version has not been finalized yet so it is not available.
Wizards Den on Vimeo
I hope this post helps someone along the way who needs to understand HDV workflow choices.
In this modern world of HDV (which should be dead in a year or two with AVCHD taking it's place) editing has gotten more difficult and costly. Those who need to edit their HDV files have essentially 3 choices and then must create a workflow around it.
Choice 1. Stay in HDV. Capture your native files, drop them onto the timeline, cut, color, and output into your desired format. The drawbacks here are that HDV is notoriously hard to work with, it's slow, it doesn't hold up well as it gets edited and rendered and changed, etc.
Choice 2. Capture into 2 formats, the original HDV and a smaller "proxy" file. The thinking here is that you leave your HDV files untouched on the drive, and use the smaller proxy files for editing. When the time comes to render, you replace the proxy files with the originals, which preserves all the edits, color changes, titling, etc. And you get a nice clean render. The drawback here is that you are going to need more disk space, and things get complicated with multiple versions of files running around.
Choice 3. Capture (or immediately convert) into an intermediate format. This can be an uncompressed AVI or QT file, or better yet a file type designed for the work like Cineform. Preferably a 10 bit codec so that all the color space is preserved. File types like Cineform offer faster editing speeds than native HDV, hold up to many generations of edits and renders, and preserve all your colors. The drawback is that they are larger than HDV files and not all editors support them.
So this week, I have been doing posting work for a cave diver / videographer here. He has a Mac, I don't. He needed to send me either uncompressed or mildly compressed video so that I could get clean video to post with. Our solution was for him to buy a large external HD, and for me to purchase some software called MacDrive which lets me mount a Mac formatted drive on my PC. Works like a charm. Be mindful if you go this route that uncompressed or mildly compressed 1080i files are about 150GB per hour.
After playing with different workflows this is what I settled on.
1. Copy Mac Native QT files (photojpeg compressed) to my hard drive.
2. Convert 1080i QT files to 720i AVI Huff compressed files Since no edits need to done, just coloring and exposure cleanup, this is fine.
3. I made the decision to go with Proxy files, so I immediately cut a 1/2 size AVI proxy file at 640x360.
4. The new 720i file is deinterlaced, denoised, and cleaned up. A new 720p working file is created and the old QT and AVI files are deleted to save space.
(All processes to this point are done in batch overnight as pre-production steps)
5. Open the proxy file in the editor, which in this case is Sony Vegas. Apply an envelope to the file on the timeline that has my color correction tools on it.
6. Split the file scene by scene which will let me correct exposure and color on a per scene basis.
7. Color grade the scenes (this takes about 3-4 hours per 10 minutes of video)
8. Cut a working version of the file to a streaming format and upload to a streaming server for the client to see. Burn timecode onto the file so that the client can comment on the work and can ask for changes based on specific time in the video.
9. Apply any final corrections to the proxy version.
10. Do a media replace in Vegas which will replace the low res file with the 720p working file. Render a 720p or 540p streaming version for web use.
11. Render a 720p master file in a mastering format for archival purposes.
12. Render a 720p version for use on HD like Blu-Ray. (mpeg4)
13. Write all formats to specific folders on client disk for future use.
While I realize this workflow won't work for everyone, it is a viable workflow for doing professional type work when you can't afford mistakes or problems. The original shipped files are never touched other than to make a working copy. At no time during the process is there ever just one copy of the file. At each step, the best version of the file for the job is used. At each step in post, the client can see the work either as a proof, or as a full sized version.
For most here, this would be done with one video at a time. For those who take on work for others, you may have a number of files, so a clean method of working and naming files and folders is crucial.
Something else to note. If you are rendering for the web, or for non-finished format, you might want to overlay your name, logo, or some other identifier on the file. This allows you to identify your files if they should find their way to someone else's hands without your permission.
Examples:
This is a link to a small proof file with timecode burn. I'll leave this up for a few weeks or months if it's ok with the client. File is encoded at 512kbps, 24fps (drop frame), and 640x360.
mms://199.242.67.132/wd
A copy of that same file can be found on my Vimeo site without the timecode burn. The 720p version has not been finalized yet so it is not available.
Wizards Den on Vimeo
I hope this post helps someone along the way who needs to understand HDV workflow choices.
