|
|||||||||
|
Thread Tools | Search this Thread |
September 11th, 2015, 10:13 AM | #1 |
Regular Crew
Join Date: Jan 2008
Location: Zelienople, PA
Posts: 86
|
Relevance of drop-frame time code in modern video ...
Excluding legacy or scientific uses; do NTSC, PAL, drop-frame time codes and/or interlaced formats still matter in video today?
In the past, we had different regions, formats, time code, resolutions, voltages and audio sync considerations. Along with CRT displays and their whiz-bang vacuum tubes, electron beams and cathode rays. The stuff of the future! If you lived in the 1950s. Today, with digital acquisition, delivery and display; it seems like we've moved beyond all this. One pixel is one pixel, one frame is one frame, data is data and that's as complicated as it gets. 24 fps is the "film look" while 30 fps is the "video look". Anything beyond that is used for effect. Does it have to be any more complicated than that? |
September 11th, 2015, 12:13 PM | #2 | |
Inner Circle
Join Date: Sep 2003
Location: Portland, Oregon
Posts: 3,420
|
Re: Relevance of drop-frame time code in modern video ...
Ideally and theoretically, no, none of that stuff *should* matter.
As a practical matter, some of that *does* matter a great deal. I'm not saying it's right, but I and people I know are dealing with interlace issues, framerates, and various country's power supply differences' impact on video on a daily basis. Some of this is because you *can't* exclude legacy from the picture and expect to meet the needs of many markets. Other issues include the various slowmo/fastmo strategies, acquiring in a different format than you distribute in, what the hell is "filmic" anyways (it isn't framerate, IMO), and the general usefulness of a variety of originations in all sorts of programming. And don't get me started on 3D and 4k!!! 720p and 1080i/p would seem to narrow the options, but add in 24, 25, 29.97, true 30, 50, 60, 120 fps and beyond and even in modern HD formats there are more, um, "choices" than ever. And can we really exclude capture, editing, intermediate, mastering, and distribution codecs from such a discussion? Things were a lot more straightforward in many way back in the days of standard def, but simpler is *not* better. If you want to call me a crusty old curmudgeon and a stick in mud, I accept. Quote:
Timecode format - not so much for many people empowered by the great democratization of access to video making, but still extremely important in some workflows. Interlace - you gotta' be kidding, here. Interlace still hurts. Ow. "In theory, theory and practice are the same. In practice, not really." attributed to Yogi Berra
__________________
30 years of pro media production. Vegas user since 1.0. Webcaster since 1997. Freelancer since 2000. College instructor since 2001. |
|
September 11th, 2015, 01:24 PM | #3 |
Inner Circle
Join Date: Mar 2003
Location: Ottawa, Ontario, Canada
Posts: 4,220
|
Re: Relevance of drop-frame time code in modern video ...
Well the video look I expect is 29.97fps interlaced. In NTSC unfortunately has the same time code as 30P ( actually 29.97P ) but 29.97fps interlace is actually taking 59.94 exposures a second, hence the smooth motion . Half frames either odd or even fields with half the vertical resolution of a full frame. The frame rate is the start of an interlace sequence of 2 fields so the frame rate ( time code ) increments every 2 fields. This was so that CRT's and downstream systems would see this frame sync and know how to start the field sequence. The closest temporal motion in progressive is 60P ( 59.97P ) so for me now that we have progressive its 60P for the video "real" look, So 24fps for film and 60P for video look. 30P is in the middle and likely due in part to a technological limitation for some equipment !!! 30P ( 25P for PAL ) is better than 24 for TV and PC playback because it is easier to match the refresh rates of the displays. 60P is better for motion as well as display on TV's or PC screens.
|
September 11th, 2015, 03:04 PM | #4 | |
Major Player
Join Date: Mar 2010
Location: Raleigh, NC, USA
Posts: 710
|
Re: Relevance of drop-frame time code in modern video ...
Quote:
There's actually zero technological reason for any of this anymore. It's all digital, and none of it is tied to mains frequency, there are no color subcarriers any more, so drop-frame no longer "solves" the artifact it was implemented to solve. We could, if we wanted to, have a universal set of progressive frame rates (no, I'm not going to argue which set should be used), without drop-frame or interlacing, or a bunch of other spurious crap like overscan and title/action safe areas we have because of overscan. Or the aspect ratios that came to us because of the width of 35mm film, and the spacing of the perforations on that same 35mm film. We could actually be rational. We could create standards for digital that make sense for digital. But we don't. And we won't. Because... politics. Fights over control. And "we've always done it that way". And of course "not invented here". I mean really -- we can't even decide what resolution "4k" actually is -- is it, you know, 4k? As in 4096? Or is it really QHD, which is really 3840? While committees fight over the standards, the manufacturers are rolling out product. We will of course end up with dual standards (if it's just two we should count ourselves lucky). Business as usual. Sigh... EDIT: Sorry to be a downer, it's just the OP hit upon a particular pet peeve of mine. |
|
September 11th, 2015, 03:28 PM | #5 |
Inner Circle
Join Date: May 2006
Location: Camas, WA, USA
Posts: 5,513
|
Re: Relevance of drop-frame time code in modern video ...
Drop frame time code exists due to the fractional frame rates. And to be pedantic, the correct rates in 60 Hz regions is 24/1.001, 30/1.001, and 60/1.001, rather than 23.976, 29.97, and 59.94. This doesn't really matter until you try writing firmware that deals with continuous audio and video...
If we were just shooting for clips on the web, we could transition to 24, 30, and 60 Integer rates and be done with it. When you try to route all of your signals through a broadcast or cable head-end and need a master sync generator with a single reference, it's another story. You need to choose fractional (1/1.001) or integer rates. And then Sunday rolls around and all of the networks need to exchange sports highlights. This is kind of like driving on the left or right side of the road. You can do either one or the other. And you can switch from one standard to the other. But you can't mix them, and when you make the transition, everybody had better do it on the same day and at the same time. Many people would like to switch to integer rates. It's more elegant. The math is easier. When encoding a broadcast signal, the number of audio samples per video frame is constant. But with the whole industry working with fractional rates and millions of hours of fractional content, switching is easy to say and hard to do. So yeah, we're stuck with it in the foreseeable future.
__________________
Jon Fairhurst |
September 11th, 2015, 06:31 PM | #6 |
Regular Crew
Join Date: Jan 2008
Location: Zelienople, PA
Posts: 86
|
Re: Relevance of drop-frame time code in modern video ...
Thanks for confirming my suspicions. Seems as if the old standards aren't technically required anymore, unless your client requires them.
In my case, the question came up because I'm looking to buy a new camera (retiring my ole Sony Z1U). I was surprised to find cameras still offering drop-frame time-code and 60i. Which had me scratching my head and wondering how relevant that stuff was anymore. In my case, all of my work is internal and all content will be delivered via YouTube. I suppose my standard is simply 1080p and a straight 30 fps. |
September 12th, 2015, 06:04 PM | #7 | |||
Inner Circle
Join Date: Jan 2006
Posts: 2,699
|
Re: Relevance of drop-frame time code in modern video ...
Quote:
But the sound subcarrier frequency relative to vision had to be fixed. Come colour, and the subcarrier had to be a fixed multiple of line frequency (and hence frame frequency), to avoid disturbing dot patterning. So the frame frequency then had to be accurately defined as well - gone were the days of mains locking. Fixing at exactly 60Hz would have been the obvious thing - but unfortunately that led to a subcarrier frequency which interfered badly with sound subcarrier. Would produce severe interference on existing monochrome receivers. The sound frequency couldn't be changed without rendering all existing sets obsolete, so the colour subcarrier was altered slightly - which made the frame frequency not quite 60Hz. None of this mattered at all - until someone thought of timecode. And without drop frame, the problem was that the timecode count didn't accurately match the real running time of the programme, and by an amount which is significant for network timing purposes over the length of an average programme. Quote:
Such are the joys of legacy....! (In the UK life is far easier with framerates based on exactly 50Hz, so easy maths and no drop frame timecode. But I well remember my mother wishing that the UK had switched to driving on the right just after the Second World War. At the time it would have been quite easy (all signposts taken down, road junctions quite simple etc) but at the time it didn't seem worth the effort. Nowadays, with the numbers of cars taken to Europe, and vice versa, it is seen very differently - huge practical benefits if we drove on the same side of the road as the rest of Europe. But I digress.... :-) ) Quote:
|
|||
| ||||||
|
|