AVS Forum banner
Status
Not open for further replies.
1 - 13 of 13 Posts

· Registered
Joined
·
1,015 Posts
Discussion Starter · #1 ·
Just curious . . . What is the process used to transfer a film to HD? Also, is the program then broadcast over satellite or air from a special "tape?"
 

· Registered
Joined
·
19,587 Posts
Well, since no one has responded, I'll give you the very rough version, then someone who knows better than me will be disgusted by my amatuerish attempt and then be motivated to give you the real deal...


I think that they use a machine called a telecine or datacine (pronounced tele-sini or data-sini), which runs the film through one frame at a time and uses a tight beam of light to scan across the film frame onto a CCD (one per color I assume?) So each frame is then turned into a very large bitmap, which can then be lightly compressed and stored onto disk.


When all the frames are on disk, they can then go back and use MPEG2 encoding software to do a couple passes over those stored frames and come up with the optimum bit allocation strategy (since MPEG is a variable bit rate compression scheme), and then store it onto tape in MPEG2 format, in 1080i format usually if its bound for HD presentation, or usually 1080p 24f/s for archival purposes.


Ok, now someone who really knows can correct me and fill in the details.
 

· Registered
Joined
·
180 Posts
Sounds pretty close to what i thought/think it is?

Don't they just archive the raw bitmaps somewhere for remastering/storage?
 

· Registered
Joined
·
1,144 Posts
Quote:
Originally posted by Dean Roddey
I think that they use a machine called a telecine or datacine (pronounced tele-sini or data-sini), which runs the film through one frame at a time and uses a tight beam of light to scan across the film frame onto a CCD (one per color I assume?) So each frame is then turned into a very large bitmap, which can then be lightly compressed and stored onto disk.


When all the frames are on disk, they can then go back and use MPEG2 encoding software to do a couple passes over those stored frames and come up with the optimum bit allocation strategy (since MPEG is a variable bit rate compression scheme), and then store it onto tape in MPEG2 format, in 1080i format usually if its bound for HD presentation, or usually 1080p 24f/s for archival purposes.


Ok, now someone who really knows can correct me and fill in the details.
You were correct about the name of the machine (telecine). But you were incorrect about everything else.

Modern HD telecines run the film in real time, as a projector does. The film is scanned as it moves by either a moving beam of light (in a flying spot scanner) or a line array CCD (a CCD that is approx. 1920 pixels across by one line high). It is recorded in real time on an HD videotape recorder, of which there are several formats, but the most common for feature mastering is D5 HD, usually set up as 1080/24p. Television programs use either D5 or HDCam, but D5 is usually preferred because it utilizes considerably less data compression. There is no "mpeg2 encoding" step, HD videotape machines operate just like any other videotape machine - you record it, then you play it back.
 

· Registered
Joined
·
619 Posts
mmost,


I thought that the raw pre-transmission bit rate of HD was somewhere over 100 Mbs. Doesn't that require some encoding to get it to the 18 Mbs of the HD signal?


Rick
 

· Registered
Joined
·
1,144 Posts
Quote:
Originally posted by cpto
mmost,


I thought that the raw pre-transmission bit rate of HD was somewhere over 100 Mbs. Doesn't that require some encoding to get it to the 18 Mbs of the HD signal?

Mastering and broadcasting are two very, very, very different things. The poster asked about the mastering process, how film is transferred to HD video. What you're talking about is what's done to compress and broadcast an HD data stream over the air. That is a process that is done in real time at the time of broadcast by the broadcaster, and has nothing to do with how the original master is created.
 

· Banned
Joined
·
3,772 Posts
From what I understand a typical HD "datastream"

starts out uncompressed (~1.5Gbit/sec) and is

often encoded down to some more managable

rate (approx 45-200Mbit/sec) for editing/storage.

Once it is edited, it is often encoded down even

more (
 

· Registered
Joined
·
1,144 Posts
Quote:
Originally posted by PVR
From what I understand a typical HD "datastream"

starts out uncompressed (~1.5Gbit/sec) and is

often encoded down to some more managable

rate (approx 45-200Mbit/sec) for editing/storage.

Once it is edited, it is often encoded down even

more (
 

· Premium Member
Joined
·
13,008 Posts
MMOST-


To clear up some conflicting info I have from Cintel and Glimmie-

Is a DVD Mastering tape made from the D5 in a film that is being telecine to tape or do they make a second run and make an initial master twice, once to D1 and another pass for D5? I realize that different houses, procedure may vary but what is considered the SOP by your experience?
 

· Registered
Joined
·
1,144 Posts
Quote:
Originally posted by Don Landis
MMOST-


To clear up some conflicting info I have from Cintel and Glimmie-

Is a DVD Mastering tape made from the D5 in a film that is being telecine to tape or do they make a second run and make an initial master twice, once to D1 and another pass for D5? I realize that different houses, procedure may vary but what is considered the SOP by your experience?
I'm not heavily involved with feature mastering these days, so I can't speak with a great deal of authority on the subject. However, it is much more economical and technically sensible to downconvert from the HD master for the standard definition master, in part because a telecine solution requires another pass of color trims, whereas the downconversion does not. From an economical standpoint, telecine time is far more expensive than downconversion time, and offers little in the way of technical superiority for this particular purpose.
 

· Registered
Joined
·
9,224 Posts
There are basically two way to master features. The old way and still the most commen way is multiple passes on D1 tape. Remember there are many different versions. OAR, letterboxed, pan&scan, etc. Then once again for PAL. Even though D1 tape supports both 525 and 625, it still requires the recording to be made in each line rate. The machines cannot convert between the two rates on playback. These each require a telecine pass to tape. The color corrections for the most part will copy across these different versions and line rates but some tweaking is still needed in tube based telecine machines such as the Cintel.


The other way and more sensible way is to make an OAR master at 24 frame video and extract everything you need from that. This process makes the most sense with HDTV and is typically how HDTV feature mastering is done today. But this process outdates HDTV. The idea of 24p video telecine was around since 1979 when IVC and Rank Cintel actually teamed up to make a combination system. Rank did the telecine and IVC the 2" helical machine. IVC went bankrupt soon after becayuse it's machine was too good. The R&D plus manufacturing costs killed them. Laser Pacidfic did early composite analog conversions with modified Ampex 1 inch VTR's in the mid 1980s. The idea never caught on for composite primarily from the inability to do pan and scan after the fact. Early quasi analog DVE's such as the Ampex ADO had rather poor quality when expanding or compressing the aspect ratio. Warner Bros internal telecine department perfected this process with D1 and DCT format VTR's that could run at 48i. This trick never went outside of Warner still due to the diffiuculty of making clean conversions. Even with the newer expensive processing equipment running in a pure digital chain the results still had artifacts and many European users would not accept masters from that process. HDTV with the vast oversampling for SDTV overcomes these limitations to this "one master tape" process.
 

· Registered
Joined
·
1,144 Posts
Quote:
Originally posted by Glimmie
The idea of 24p video telecine was around since 1979 when IVC and Rank Cintel actually teamed up to make a combination system. Rank did the telecine and IVC the 2" helical machine. IVC went bankrupt soon after becayuse it's machine was too good. The R&D plus manufacturing costs killed them
That would be the IVC 9000, also marketed as the Rank Cintel 9000. I was working at a company that no longer exists (Studio Television Services) in Hollywood back in the late 70's where we had 2 of those beasts, and then later at Bluth Video Systems in Burbank (another company now gone), where we had about 4 of them. I wouldn't call them "too good," I would call them difficult, finicky, non-standard, and definitely not 24 frame based. They were dual standard, as I recall (NTSC and PAL), and had a very complex analog time base corrector that encompassed about 9 boards and only about 3 people on the planet could properly align, as well as vacuum column loops that allowed for very high speed shuttling but also very high speed destruction of tape. They allowed high bandwidth video recording, but had really lousy audio performance. We actually recorded audio double system (on multitrack audio tapes, usually either 1/2 inch 4 track or 1 inch 8 track) using a Q Lock synchronizer. Editing on these machines was done with a cue tone based editor, much like the later generation quad machines. Ahhh, these kids today with their digital this and digital that don't know how good they have it............


Sorry for this rather technical (and nostalgic) post that probably about .5% of you find either interesting or amusing, but geez, Glimmie, you triggered some memories...
 
1 - 13 of 13 Posts
Status
Not open for further replies.
Top