|
|||||||||
|
Thread Tools | Search this Thread |
May 3rd, 2013, 07:37 PM | #1 |
Regular Crew
Join Date: Apr 2013
Location: Luckey, OH
Posts: 196
|
29.97p vs 30p
Is there a difference between 29.97p and 30p, or are people saying 30p just rounding for convenience?
|
May 8th, 2013, 01:29 PM | #2 | |
Major Player
Join Date: Mar 2010
Location: Raleigh, NC, USA
Posts: 710
|
Re: 29.97p vs 30p
Quote:
|
|
May 8th, 2013, 02:02 PM | #3 | |
Inner Circle
Join Date: Sep 2002
Location: Vancouver, British Columbia (formerly Winnipeg, Manitoba) Canada
Posts: 4,088
|
Re: 29.97p vs 30p
Quote:
Not even just for broadcast but for when you need an EXACT run time to fit to time allotted.
__________________
Shaun C. Roemich Road Dog Media - Vancouver, BC - Videographer - Webcaster www.roaddogmedia.ca Blog: http://roaddogmedia.wordpress.com/ |
|
May 8th, 2013, 04:56 PM | #4 |
Inner Circle
Join Date: Jan 2006
Posts: 2,699
|
Re: 29.97p vs 30p
It's worth reminding exactly why 29.97fps came about - TV in the US started off (in monochrome days) as 30fps, so why the change?
It is all to do with the introduction of colour, and an unfortunate choice (with hindsight) in the early days for the sound subcarrier frequency for transmission. Come colour, the subcarrier frequency has to be an exact number of multiples of line frequency, and that has to be exactly 525x the frame frequency. Unfortunately, the first choice of colour subcarrier frequency was found to interfere badly with the sound subcarrier frequency in some of the early monochrome receivers. Rather than making many TVs unusable overnight, it was decided to shift the frequency used for colour subcarrier - which meant a consequent slight reduction in line frequency, and hence frame frequency - hence 29.97. That caused no problems at first - but issues arose with the introduction of timecode. Easiest implementation is to update the seconds after every 30 frames - but with a 29.97fps system, that means actual run time and displayed timecode will no longer match, and to a significant degree after an hour or so. Hence drop frame timecode was introduced as a "bodge" to make timecode once again correspond to actual running time. As Bruce says - 29.97 is an artefact of analogue days, of the early days of NTSC colour. So why is it still needed? In a word, legacy. It would be quite feasible to engineer a true 30fps non-drop frame timecode system in the digital HD world. The reason it's not done is compatability with SD systems, as when such a programme was broadcast SD, it would HAVE to run at a frame rate of 29.97 fps - or the colour subcarrier would be too high a frequency. (Subcarrier has to be that exact multiple of frame rate.) And if HD/SD simulcasts are happening, that means both have to be 29.97fps - that or do very messy 30-29.97 frame rate conversion. Hence the decision to make the HD standards 29.97 etc - pure compatability with analogue NTSC, for reasons which date back to the early 1950's. |
May 8th, 2013, 07:06 PM | #5 |
Regular Crew
Join Date: Apr 2013
Location: Luckey, OH
Posts: 196
|
Re: 29.97p vs 30p
So there is a difference...but if I have a Sony camera advertising 30p and a Canon spec saying 29.97, should I be concerned if using them together?
|
May 9th, 2013, 08:15 AM | #6 | |
Inner Circle
Join Date: Jan 2006
Posts: 2,699
|
Re: 29.97p vs 30p
Quote:
I seem to remember the exception is 24fps, when you can get a "true" 24p version intended for film out. |
|
May 10th, 2013, 02:57 AM | #7 |
Inner Circle
Join Date: Feb 2006
Location: Belfast, UK
Posts: 6,152
|
Re: 29.97p vs 30p
Except in NTSC land where they shoot at 23.98fps, for time code reasons. I gather 24fps can cause real problems in post, not so in PAL land where you can do true 24fps. It's something to be discussed with your post production people.
|
May 10th, 2013, 08:11 AM | #8 |
Major Player
Join Date: Sep 2011
Location: Merritt Island, Florida
Posts: 865
|
Re: 29.97p vs 30p
Mr. Heath...that was the best explanation of why we in 'NTSC' land use 29.97. Thank you.
Best regards, J. |
May 10th, 2013, 11:11 AM | #9 |
Major Player
Join Date: Jun 2008
Location: Milwaukee WI
Posts: 691
|
Re: 29.97p vs 30p
Byron,
Just rounding numbers for marketing purposes/simplicity. No worries, shoot away! Thanks
__________________
Jeff Pulera Safe Harbor Computers |
May 10th, 2013, 11:20 AM | #10 | |
Inner Circle
Join Date: Jan 2006
Posts: 2,699
|
Re: 29.97p vs 30p
Quote:
For feature film work, where the end product is for laying back on to film, the NTSC TV compatability is longer important, but it's highly desirable to avoid the complications of drop frame time code. In other words, the timecode seconds get updated every 24 frames - no exceptions. To keep the situation of an hour of material being represented by an hour of timecode, that means a 24fps framerate - not 23.98. If there are post problems, I think the stories I've heard are when the video gets recorded as true 24p, but separate audio is at 23.98, and yes, it's the timecode that causes the problems. If EVERYTHING is at 24p, and non-drop frame timecode, it's fine. Incidentally, I earlier said "..... TV in the US started off (in monochrome days) as 30fps,.....". In practice, it's more likely to have been locked to the local mains frequency - so a very nominal 30fps!!! The reason is because power supplies in early TVs were far less sophisticated than nowadays and mains hum would show as a brighter band on the picture. Any frequency difference between framerate and mains would mean the band would roll through the picture. Lock the two together and the stationary band is much less objectionable. Come colour, and a varying framerate would have meant a varying subcarrier frequency which would not be acceptable. NTSC and PAL colour systems meant TV framerates could no longer be locked to mains - so power supplies had to have far better specifications to avoid hum! |
|
May 10th, 2013, 11:51 AM | #11 |
Inner Circle
Join Date: Feb 2006
Location: Belfast, UK
Posts: 6,152
|
Re: 29.97p vs 30p
At the HD workshop I attended in US, 23.98fps was the default setting, the one wouldn't cause any problems.In fact, you were warned against using 24fps - this was with Sony F900 and Varicams. I guess it's because so much of the gear is set up for broadcast work.
Fortunately, it's not an issue in the UK, although it might be easier to shoot at 25fps, rather than 24fps. Again it's something to be checked in advance. We ran into problems on a film shoot with 24 fps, because the AVID needed a separate license to use 24fps. |
May 10th, 2013, 01:48 PM | #12 | ||
Inner Circle
Join Date: Jan 2006
Posts: 2,699
|
Re: 29.97p vs 30p
Quote:
Out of interest, what form of timecode did they recommend? It's obviously possible to shoot non-drop frame timecode and 23.98fps - the only implication will be that timecode time will no longer be real time. But maybe the advantages of non-drop frame for editing could outweigh the disadvantages? Quote:
|
||
May 10th, 2013, 02:48 PM | #13 |
Inner Circle
Join Date: Feb 2006
Location: Belfast, UK
Posts: 6,152
|
Re: 29.97p vs 30p
They were using standard NTSC type time code, this workshop was centred around US industry practise, although the students were international.
23.98fps was regarded as the best "universal" frame rate for distribution to both the US market and international sales in thr PAL market. Just play back at 25 fps. |
| ||||||
|
|