View Full Version : HDV To HD SDI
Matthew Pugerude December 15th, 2005, 03:33 PM Is convertion the same as up rezing or are these two different animals. I am looking at the Miranda as a part of my work flow because it converts to HD SDI. I do Doc's and Corprate stuff. I am wanting to get into the HD realm but most of my budgets don't allow for full HD Acquistion so the HDV is really appealing to me then I saw this product for Miranada so I thought that i could shoot HDV and up rez it to HD.
http://www.miranda.com/product.php?i=323
With
http://www.blackmagic-design.com/products/hd/
Work flow idea (please give me your opionins good and bad)
1. Shoot HDV (Sony or JVC I have not made a choice yet)
2. hook up to Miranda with firewire
3. hook up Miranda to Deck link pro HD with deck contorl 422
4. edit converted HDV to HD and
5. export to HD deck (rental of course) from Deck link pro HD
Still waiting to here if the 422 will work with the Miranda and the Deck link
A. J. deLange December 17th, 2005, 01:18 PM No, it doesn't do uprezzing. HDV is already in full HD resolution but it is heavily compressed in the temporal domain i.e. there are 1 or 2 full pictures per second in the HDV stream and the rest of the information is motion data which allows a decoder to reconstruct the other 28 or 29 frames. Judging from their data sheet the HD-Bridge will take in a DV stream over fire wire and convert it to an SDI stream and to analog component outputs. In other words it contains the necessary circuitry to reconstruct the missing frames and put an uncompressed HD signal out to SDI (digital format) or Y, Cb, Cr analogue. Thus this box could be used, for example, to connect an HD monitor to an editing station or to drive an HD recorder which wants uncompressed input. Note that it is not magically replacing the information that was thrown away when the HDV stream was encoded. It is decoding the stream, replacing the lost frames with reconstructed frames and putting the result out in the same format as an uncompressed stream.
Ken Hodson December 17th, 2005, 03:45 PM Actually the Sony's use a 15GOP, so 2 new frames per/sec, while JVC's use a 6GOP for 5 new frames per/sec.
Keith Wakeham December 17th, 2005, 03:49 PM Your workflow will work, but you will gain nothing in quality over capturing the HDV. HDV is compressed to 25 Mbps, while HD-SDI is close to 1.5 Gbps, so it bloats the data about 60 times and then you either need a fast and large raid array or use an intermidiate codec.
All HD-SDI single links are 4:2:2, so since HDV is 4:2:0 the miranda device rebuilds it to 4:2:2, but that doesn't add anything, it lots identical.
If you really need to finish to a tape format such as HDCAM, D5, DVCPRO HD, or any others you should probably shoot HDV, edit in native HDV or use an intermidiate codec, and use something like decklink or Kona/Xena to output to your rented deck. So you would bypass the whole miranda thing. It is unnecessary
Miranda is more for an enviroment where a dedicated HD-SDI editing suite and don't want to use firewire. (I interned at a place with a symphony suite that would capture composite @ 4:4:4 uncompressed because they didn't use miniDV in the suite).
Douglas Spotted Eagle December 17th, 2005, 03:51 PM We have an HD Bridge, and it's a wonderful tool, very easy to cross convert with, and just simply EASY to use.
FWIW, of course it's true the JVC has 5 I frames per sequence and Sony has only 2, however Sony and Canon have 16 error correction units vs JVC's 10. In a conversion, this is a much more important factor. (Error correction)
Partly related to that, Sony and Canon have 2 search streams embedded in the error tracks, meaning you can scan forward/backward and view frames at 8x/24x. This is pretty helpful when scrubbing, IMO.
The HD Bridge supports both formats.
Douglas Spotted Eagle December 17th, 2005, 03:58 PM Your workflow will work, but you will gain nothing in quality over capturing the HDV. HDV is compressed to 25 Mbps, while HD-SDI is close to 1.5 Gbps, so it bloats the data about 60 times and then you either need a fast and large raid array or use an intermidiate codec.
All HD-SDI single links are 4:2:2, so since HDV is 4:2:0 the miranda device rebuilds it to 4:2:2, but that doesn't add anything, it lots identical.
If you really need to finish to a tape format such as HDCAM, D5, DVCPRO HD, or any others you should probably shoot HDV, edit in native HDV or use an intermidiate codec, and use something like decklink or Kona/Xena to output to your rented deck. So you would bypass the whole miranda thing. It is unnecessary
Miranda is more for an enviroment where a dedicated HD-SDI editing suite and don't want to use firewire. (I interned at a place with a symphony suite that would capture composite @ 4:4:4 uncompressed because they didn't use miniDV in the suite).
I'd have to disagree.
Converting to uncompressed or lightly compressed prior to editing is a much more robust method of working vs using an intermediary and particularly editing native HDV. Editing native HDV is not a good practice in any event.
*IF* and that's the big one...IF you're delivering on HDCAM, D5, or DVCPRO HD then convert straight from the cam to 4:2:2 and edit from there. This is our in-house practice using either the HD bridge or the Decklink Multibridge (component out) If you're not delivering on HDCAM,D5, DVCPRO HD, then I'd completely agree with Keith's comments.
Keith Wakeham December 17th, 2005, 05:06 PM I'm sorry, I guess I didn't really mean so much as an intermidiate codec (because that would imply going back to HDV). I meant that if you capture uncompressed with the miranda device that it will be similar (depending on if the miranda does just decompression or it it actually processes the footage) to capturing HDV and then exporting it as uncompressed.
Some of these devices may filter the chroma channels linearly between samples to make the colour smoother, but can be done in software also.
So If your setup with an editing suit to do uncompressed or lossless compressed HD then the miranda is probably a good choice, but from a technical perspective it does gain quality (and with me being an engieering student, if it doesn't gain quality it can't justify it).
Kevin Shaw December 17th, 2005, 05:35 PM Where you might gain quality would be from capturing the signal out of the camera before it's compressed to the HDV format, but then you have to decide how you're going to capture and edit the HD-SDI data. Depending on your specific needs that may not be worth the extra trouble; standard methods for capturing and editing HDV material are pretty good already. (Especially if your final output is regular DVDs.)
Ken Hodson December 18th, 2005, 01:22 AM I'd have to disagree.
Converting to uncompressed or lightly compressed prior to editing is a much more robust method of working vs using an intermediary and particularly editing native HDV. Editing native HDV is not a good practice in any event.
*IF* and that's the big one...IF you're delivering on HDCAM, D5, or DVCPRO HD then convert straight from the cam to 4:2:2 and edit from there. This is our in-house practice using either the HD bridge or the Decklink Multibridge (component out) If you're not delivering on HDCAM,D5, DVCPRO HD, then I'd completely agree with Keith's comments.
I think I stand in the middle of the two opinions.
There is going to be a HDV to (Pro format) conversion, regardless. When is the best time to do this in my opinion would be after any effects work, rendered to uncompressed first, then output to (Pro format). To first convert the HDV to another compressed format, especially with lower resolution, would be counter productive. If one was to be doing basic editing, I don't think it would matter quality wise which way you went, but direct to (Pro format) and editing in it, would be very efficient.
So when you convert all depends on what type of production your doing.
Matthew Pugerude December 18th, 2005, 03:18 PM I think I stand in the middle of the two opinions.
There is going to be a HDV to (Pro format) conversion, regardless. When is the best time to do this in my opinion would be after any effects work, rendered to uncompressed first, then output to (Pro format). To first convert the HDV to another compressed format, especially with lower resolution, would be counter productive. If one was to be doing basic editing, I don't think it would matter quality wise which way you went, but direct to (Pro format) and editing in it, would be very efficient.
So when you convert all depends on what type of production your doing.
Thanks every one and Ken that is what I was thinking I Do Corporate and Doc's. Corporate pays the bills the Doc's make my production life livable but I would like the option to go "full" HD for those really special Doc's that could get some air time at possabily my local PBS station that has a HD channel. I bet most of the time I would use Cineform to edit the HDV stream for the Corporate stuff. I guess the next question is Does Blackmagic do HDV I can not find anything on their site that says yes we do HDV. And if they do HDV Do I need Cineform to work with Premier Pro. Now which way do I go. What about dumping the finished program for a say a HDV Deck or Camera (the big man upstairs forbid I have to do that) to a HDcam deck any favarable results in doing that or not and good idea at all?
|
|