John K.
Veteran
/ Moderator
Dec 3, 2005, 12:24 AM
Post #1 of 2
(2324 views)
Shortcut
|
|
Email from CEO about HDV Editing/Specifics
|
Can't Post
|
|
Below is an email response I received from Jörg Sprave (CEO of MacroSystem) a couple of days back that answered my questions and discusses exactly how our editors will handle HDV editing and the rational for his approach and the route MacroSystems is taking. He gave me permission to share this email with others who may be interested.
John
**** Email Response ****
Hello John,
We have invested a lot of thinking in our HDV editing concept and believe it is one of the best on the market. HDV editing takes more time than DV editing, but let’s look at why we are going the Native HDV route.
First, HDV has a much higher resolution (1.5 Megapixel for HDV vs. about 0.45 Megapixel for NTSC DV). That is about 4.3 times more. Of course it takes longer to apply any kind of effects to such a larger image. It is pretty linear - so you are right in your assumption that if it comes down to applying effects to HDV material, some units will take longer to work with it. Just let me say that no matter if you work in native HDV or in any other format - as long as you want to keep the HDV resolution you have to live with this drawback.
Second, there is the MPEG-2 issue. Yes, there is only one I-Frame in every 15 frames of video. But the other frames can be reconstructed based on the I-Frame. That's what the camera does when it plays HDV material back. It can be done in editing the same way. This reconstruction takes time, true. It takes less time if you use some kind of I-Frame-only-standard. But even such a standard has to be converted from the original HDV format.
OK, now that we have the two reasons for the slower HDV rendering pointed out, let's analyze them and see what kind of strategies an editing approach can theoretically employ.
As long as you want to keep the main advantage of HDV (the higher resolution), then nothing can be done about the effects. It simply takes longer to render them. But as long as you keep working in native HDV, all the parts of your video that have no effects (usually the vast majority of frames in your project) can be handled without any rendering time. They will just be copied from the tape to the drive and then back. If you convert the HDV to a different format, then every frame has to be converted. Native HDV editing is much faster than non-Native-editing, because of the lack of lengthy format changes. You loose some time when applying effects because for those frames, you have to decode and re-encode the MPEG-2 data. But you win back far more time because there is no general format conversion.
There is also one more disadvantage when you choose to convert HDV into some other format: The new material takes much more space on your drive. Count 2 to 3 times the space requirement.
OK, now we know that it is better to go the native way - it is faster, takes less space and is unbeatable in image quality.
For serious editing in HDV, a hardware codec is very helpful. There is no such HDV hardware in your Prestige. But you may already own a codec: It's built into your camcorder. The HDV cameras can play back HDV material in the DV format, in real time. There is no software codec, not even for the fastest machines around, which can do that.
So our concept employs the camera codec for editing. We read in the material in DV first, then you switch the camcorder to HDV and read the material in a second time. Both versions share the same time-code, so they can easily be sync'd by the system.
Then, you go ahead and edit the DV material, with the same speed as always. But the system saves your edit decisions, and applies them to the HDV material in background. Finally, you have a DV, a DVD, and (possibly after some extra waiting time) the HDV version. The release of SmartEdit 5 early next year will allow for our HDV editing abilities.
Finally, you brought up a question that you said another online videographer was discussing with you, that being: "How can you color correct unless it's on I frames?" With a direct answer, you can't. You can't apply effects to any kind of compressed video. You always (even for DV editing) decode the compressed material into uncompressed pixels, apply the effect to them, then you re-encode the new frames back. But, we do not use another format to edit the video and we only decode and re-encode when there's an effect to be rendered.
Hope you find this information useful. If you wish, you may share my answer with anyone else (or other Cassie users) who may be curious.
Regards,
Jörg Sprave
CEO MacroSystem
(This post was edited by John K. on Dec 3, 2005, 12:43 AM)
|
Thanks for this information,I hope santa do a fast delivery