Tips on Working with HDV

[ This article was first published in the July, 2005, issue of
Larry’s Final Cut Pro Newsletter. Click here to subscribe.
Updated Jan. 2006 and June 2007. ]

Update – June 2007

I continue to learn new things about HDV, and Apple continues to improve their products. With the release of Final Cut Studio 2, and the ProRes 422 codec, much of what I talk about below, while still true, have improved.

Please read this article to learn more about the current state of my thinking about HDV

---

There’s no doubt that HDV has made an impact on our industry. However, HDV is not DV. In fact, there are significant differences to working with HDV, especially when compared to the video we’ve traditionally worked with.

The big issue is that HDV uses a compression system called Long-GOP MPEG-2. What this means in English is that a group of images, generally 15 for NTSC video, are all grouped together and treated as a single frame (a GOP, or Group-of-Pictures). So, instead of one picture per frame, which we are used to in DV and film, we have 15 pictures to one frame. This makes for really great compression, but it’s almost impossible to edit. The reason is that MPEG was designed to only allow edits at the start of one of these Groups of Pictures. This boundary is called an “I-frame.”

DVD’s use a similar compression structure, which is why it is often difficult to edit video which has been compressed into MPEG-2 for DVD use.

However, Apple has figured out a way to edit HDV natively, without converting it into a different file format. However, this native editing is not without trade-offs. So, here is a list of the ways HDV is not the same as DV:

1. HDV needs serious CPU power. You need a minimum 1 GHz processor and 1 GB RAM.
2. You can’t monitor HDV thru FireWire while editing, you can only watch HDV on your computer screen.
3. You can only output to tape using Print to Video. You can’t Edit to Tape.
4. You can’t up-rez HDV to HDCAM using FireWire, though you can do so using a capture card, such as an AJA Kona card.
5. The native resolution of HDV is 1440 x 1080. However, HDV only outputs at 1920 x 1080 interlaced using line-doubling, or 1280 x 720 progressive using down-sampling.
6. HDV takes significantly longer to render than DV, because the image size is four times larger.
7. FCP only renders HDV frames that changed during editing. A frame is changed if an edit occurs anywhere within the GOP, except on the I-frame boundary, or if you’ve applied an effect or transition.
8. There are timecode accuracy and frame-rate issues in capturing and outputting HDV.
9. When capturing HDV, you can’t control the deck or camera as precisely as you can with DV. So, in general, capture with larger handles than you would normally use.
10. HDV only supports 2 channels of audio in or out.
11. If you want to export an HDV sequence to include on a DVD, be sure to export it using File > Export > QuickTime Movie and make it SELF-contained. Reconform all I-frames to match the HDV compression structure. Remember, exporting puts it into an MPEG format.

To help with some of these problems, Apple has created an Intermediate codec which converts HDV from it’s long-GOP MPEG-2 structure into a frame-based video format. There are several benefits to using the Intermediate codec and one disadvantage:

1. You can monitor the results of your editing using the intermediate codec thru FireWire.
2. The intermediate codec is less resource intensive and better for editing on slower computers.
3. The intermediate codec creates files that are four-times larger than native HDV.
4. There are reported quality issues with the intermediate codec.
5. Converting between HDV and the intermediate codec is not loss-less, however, Apple says it’s testing shows that the conversion quality is better than competitive products. Look at the results and judge for yourself.

---

Bookmark the permalink.