"composite master" explanation requested

[loster]

This isn't really a newbie question but I happened across some anime related thread where people were discussing component versus composite masters. From my understanding older animation was transfered to film and for TV and video usage DVD etc was telecined from that film. The video would then have been transferred to D1, Betacam or similar correct? Why would the production company waste time with a conversion to composite video? Didn't the professional formats even in the 1980s support separate color components not high-definition obviously?
http://www.mania.com/aodvb/showthread.php?t=94316

[edDV]

This isn't really a newbie question but I happened across some anime related thread where people were discussing component versus composite masters. From my understanding older animation was transfered to film and for TV and video usage DVD etc was telecined from that film. The video would then have been transferred to D1, Betacam or similar correct? Why would the production company waste time with a conversion to composite video? Didn't the professional formats even in the 1980s support separate color components not high-definition obviously?
http://www.mania.com/aodvb/showthread.php?t=94316

D1 was slow to be adopted in the 80's since each D1 VTR cost about a quarter million dollars. TV animations were very low budget with much production moved to Korea, etc. for labor saving. Most animation cells were transferred to 1" Type C analog tape or later to cheaper D2 composite digital which could be short segment recorded from frame stores (with telecine added for NTSC). This was perfect for composite NTSC Laserdisc release.

Since NTSC was the prime target market, PAL was mostly standards converted from NTSC masters. Component digital mastering didn't become economic until devices like the Abekas A64 DDR allowed automation of the transfer process. Still a D1 machine was required to record the master. This still required about $500K capital investment.

In the 90's component DDR's got cheaper and Digital Betacam became available mid 90's. But like most other TV production, conversion to component digital was driven more by anticipation of DVD, 16:9 and high definition. The reasoning was that program inventories would have longer life if mastered to ITU-601 standards.

As for Anime, Japan was on a more advanced track driven by DVD and their analog HDTV system. There was more incentive during the early 90's to master in component digital.

[loster]

Thanks for the through explanation and taking the time to answer my question I'm understanding this much better.
I have a hunch that anime fans made up terms. Is the master being composite the main factor in quality, or would other factors such as film quality and equipment and/or setup make a larger difference?
I did come up with another related question, what about the usage of analog component isn't betacam component?
The quality of the master could depend on the contract correct (ie to save duplication costs use a cheaper medium for foreign markets)?

[lordsmurf]

I have a hunch that anime fans made up terms.

The various amateur "scenes" often do. Very likely.

[edDV]

Thanks for the through explanation and taking the time to answer my question I'm understanding this much better.
I have a hunch that anime fans made up terms. Is the master being composite the main factor in quality, or would other factors such as film quality and equipment and/or setup make a larger difference?

I'm not that familiar with Anime production in those years but more so the Saturday cartoons and TV animation specials back then. 80's - 90's Anime production seems targeted to analog or digital composite masters for Laserdisc production. These masters would also allow direct copies for VHS or broadcast TV. There was no 24 fps cinema distribution market for these as best I can tell. So the "film" cells were only a production step, not a finished product.

Saturday morning cartoons had a few drawn frame elements but these were layered with only some layers receiving animation. As the 80's progressed much of the layer compositing was done in digital frame stores, less on film. Therefore there is no film version of the finished work, only a tape and production elements that would need to be reprocessed.

I did come up with another related question, what about the usage of analog component isn't betacam component?

Analog Betacam and MII were recorded Y (R-Y) (B-Y). It was possible to record components in and components out. This use of Betacam or MII had a small following mostly for those interested in multi format (NTSC/PAL) production since component didn't have the SC/H frame sequence* requirements of composite and could be adapted to 24p. Since it was analog recording, any multi-generation editing had losses.

* SC/H explained http://www.leitch.com/resources/applicationNotes/sch.pdf
In short, composite editing was subject to horizontal picture jumps if the subcarrier to horizontal sync pulse relationship was not perfectly timed. Component editing didn't have these restrictions. Component could be edited frame by frame without H shift.

[loster]

Thanks a bunch any recommendations on a book about this stuff?
Yes your probably right about the composting/layering being done on a computer. I have an old Central Park Media DVD which was originally produced about this time that looks like the layering was done on a computer setup and looks very poor as if they didn't bother too much because the target was VHS at least in the United States R1. Also I think that your pdf link is dead.

[edDV]

The link worked for me. The whole point of the SC/H issue is composite NTSC can only be edited to 4 field (two frame) increments without a horizontal picture jump. Composite PAL requires 8 fields (four frames) due to the phase reversal each field. Editing composite in matched layers is a major pain unless all sources are SC/H matched. Analog or digital component can be edited at the frame level. There is no subcarrier.

Present day users may wonder why component digital wasn't widely used in the 80's except at the very high end*. You need to realize PC computer memory cost ~$1000 per GB. A frame of RGB memory was 3x that, or 4x if alpha was used. Frame buffers were used for TV animation only when their use produced cost savings vs. film technique.

Composite video was sampled 4x subcarrier (i.e. 14.3 MHz for NTSC, 17.7 MHz PAL) so frame buffers were larger then. CCIR-601 rationalized NTSC and PAL to 13.5 MHz sample rate.


* 4:2:2 digital was first used for effects editing, highest end film transfers and animation clip storage.

[loster]

Thanks again, my internet connection must be flaky. $1000/GB I was expecting even higher costs, no wounder there was limited usage.

[edDV]

Thanks again, my internet connection must be flaky. $1000/GB I was expecting even higher costs, no wounder there was limited usage.

Here is a copy (1MB)

BTW this isn't important anymore because nobody layers composite video today. You would first decode to component.

[loster]

BTW this isn't important anymore because nobody layers composite video today. You would first decode to component.

Thanks I kind of expected that, although you always get people claiming old tech is now "artistic".

[edDV]

It was certainly more complicated and pricy.

I guess the point is when component processing is prohibitively expensive, the engineers find another economically realistic way to get the result. They certainly perfected the state of art for mixed composite and component animation or effects layering.

[loster]

Ok thanks so much. Hopefully the last question is that, are composite artifacts visible with the professional formats or are the expectations simply higher because of what is possible with DVD?