Cinematography Mailing List - CML
    advanced

60i _vs_59.94

 

28th June 2004

The Sony F900 has options to shoot 60I and 59.94. Any idea what the difference is?

Dan Coplan

Cinematographer/Editor/DVD Authoring
www.dancoplan.com


Dan,

Let me be the first to flame you here then...

59.94 looks like twice the NTSC drop-frame rate. So I figure if your project is to be down-converted to NTSC then that's the setting to use.

60 would be for film out or HD that wasn't destined for SD television.

Yeh, pity we didn't get rid of the silly drop-frame legacy when HD
standards were written.

Cheers,

Clive Woodward,
Its all PAL here, choice is 50I or 50I, or sometimes 50I..
Perth, Western Australia.


Dan Coplan wrote:

>The Sony F900 has options to shoot 60I and 59.94. Any idea what the >difference is?

Well, the first difference is that one runs at 59.94 FIELDS per second while the other runs at 60.00 FIELDS per second. Or think of it as 29.97 interlaced frames per second versus 30.00.

It might be obvious that, if you are shooting for normal television applications in NTSC markets you should use 59.94 and not 60.000.

The other difference is that, in theory there's a minuscule difference in sensitivity between the two. Probably insignificant.

I can't think of any other worthwhile difference.

Martin Euredjian
eCinema Systems, Inc.
www.ecinemasys.com


>>59.94 looks like twice the NTSC drop-frame rate. So I figure if your >project is to be down-converted to NTSC then that's the setting to use. >60 would be for film out or HD that wasn't destined for SD television.

No, 59.94 is not a frame rate, it's a field rate and it's the same as NTSC. You could shoot either 59.94i or 60i and downconvert to NTSC (59.94i) just like you could shoot either 24P or 23.98P and downconvert to NTSC with a 3:2 pulldown. But since NTSC is 59.94i (29.97 fps), 24P HD material ends up being 23.976 fps (just as when 24 fps film is transferred to NTSC) and 60i HD material ends up 59.94i anyway. So you might as well shoot 23.98P or 59.94i.

It seems to me that the reason there's these options (23.98P and 59.94i) has more to do with audio sync issues when editing sound in an NTSC environment.

23.98P and 24P look the same when transferred to film - it's a 1:1 transfer, one frame of HD transferred to one frame of film. It just gets projected at 24 fps even it had been shot at 23.98P. The same goes for 59.94i versus 60i transferred to film for 24 fps projection.

At least, this is my understanding of the situation.

David Mullen ASC
Cinematographer / L.A.


So why would you ever shoot 60I? I understand the 59.94/29.97 concept and I know people always talk about 60 fields/sec and 30 frames/sec as a shortcut in speech, but I'm still baffled why 60I is an option along with 59.94.

Dan Coplan
Cinematographer/Editor/DVD Authoring


Clive Woodward wrote:

>>59.94 looks like twice the NTSC drop-frame rate.

I think I just realized where the question/confusion might come from. It's the fact that nomenclature in HD is inconsistent. Progressive standards are listed by using their frame rates, i.e.: 23.976, 24.000, 25.000, 29.970, 30.000, 50.000, 59.940 and 60.000, either "P" or "PsF".

However, the interlaced standards, for some reason, list the FIELD rate! So, you see 59.94 and could think that this is some weird new flavour of video to contend with when, in reality, it's just 29.970 frames per second interlaced.

Martin Euredjian
eCinema Systems, Inc.


>>So why would you ever shoot 60I? I understand the 59.94/29.97 >concept and I know people always talk about 60 fields/sec and 30 >frames/sec as a shortcut in speech, but I'm still baffled why 60I is an >option along with 59.94.

Isn't 60i HDTV broadcast true 60i, not 59.94i? I guess if 60i HDTV was your only presentation format, you wouldn't need to shoot in 59.94i.

David Mullen ASC
Cinematographer/ L.A.


Dan Coplan wrote :

>>So why would you ever shoot 60I? I understand the 59.94/29.97 >concept and I know people always talk about 60 fields/sec and 30 >frames/sec as a shortcut in speech, but I'm still baffled why 60I is an >option along with 59.94.

Ah, that's what the question was about!

I think the original intent was to solve the drop/non-drop problem. Outside of that I'm not sure I can tell you why or when you might want to use 60.000i outside of a very special project.

Martin Euredjian
eCinema Systems, Inc.


David Mullen ASC wrote :

>>"Isn't 60i HDTV broadcast true 60i, not 59.94i? "

No all Broadcast HD in the 1080 line interlace format (NBC, CBS, & PBS) is 59.94. This matches the Standard Definition image running on the analog channel. As long as analog transmission continues we will have 59.94 fields of interlace television. If this was not the case the analog transmission of a show would end after the end of the Digital transmission. It is a difference of 108 frames per hour.

Also all 720 line Progressive (As ABC and FOX) is 59.94 FRAMES per second.

Bill Hogan


Dan,

Maybe if the answer to "what frame/field rate do we shoot at?" is 30 Psf, then IF you wanted to shoot at a higher rate (for slomo) or you wanted to shoot interlaced (for whatever reason) in the same project, then 60i would be the logical choice, not 59.97i.

Randy Miller, DP in LA


Dan Coplan

>>So why would you ever shoot 60I? I understand the 59.94/29.97 >concept and I know people always talk about 60 fields/sec and 30 >frames/sec as a shortcut in speech, but I'm still baffled why 60I is an >option along with 59.94.

Techno Geek Plain English Reply :

Guys! Guys! Jeez Louise. Its troubling to me that such a basic issue has all these pundits theorizing about how many video angels can dance on the head of a pin. Holy Moly! This is a textbook issue literally of black and white. There are no shades of gray, and no opinions.

The 29.97/59.94 scan rate has its roots in making color TV compatible with black and white TV back in the 1950's. There was a slight offset in frequency (29.97 vs 30 frames) to make the color sub-carrier compatible with black & white scan rate of 30 frames per second. (PAL is 25 fps, so there is no issue in European countries.)

I won't bore you with the techno babble mumbo-jumbo (which I frankly can't remember the fine points of myself anymore), but the bottom line is: We here in the USA are stuck with shooting 29.97/59.94 to make broadcast color video compatible with the old black & white standard. Unless you are shooting something that will never enter the broadcast environment, to be on the safe side, shooting 29.97/59.94 is probably a good move.

Anyone who wants to chime in about the fine points of this is free to do so, and if I am in error feel free to correct me. Let the flames begin.

Lew Comenetz
HD Video Engineer


Martin Euredjian wrote:

>>I think the original intent was to solve the drop/non-drop problem.

Martin, I'm surprised at you!

59.94 was introduced as a result of the move to color. In order to piggyback a sub-carrier on the black and white video signal, a frequency was arrived at that allowed such an implementation but retained backward compatibility for black and white receivers. Bill Hogan or any of the video engineers here could explain this more accurately and in more detail than myself, but that's the basic gist of it.

Drop frame and non-drop frame code have absolutely nothing to do with frame rates. You can have 30 DF and 29.97 NDF. The drop frame technique is a counting scheme in which 2 frame numbers are skipped in the counting sequence every minute except the minutes that are multiples of 10 in order to stay accurate to real time. It has nothing whatsoever to do with the frame rate itself.

Mike Most
VFX Supervisor
IATSE Local 600
Los Angeles


>> We here in the USA are stuck with shooting 29.97/59.94 to make >broadcast color video compatible with the old black & white >standard...shooting 29.97/59.94 is probably a good move.

Ok, so 59.94 it is - enough said on that. But I'm still curious about the 60I option.

Is it good for anything? Is it just there so Sony can say they have *another* frame rate option on their camera? Would you be screwed if you shot 60I for a standard 59.94 show?

Dan "Curious for Curiosity's Sake" Coplan
Cinematographer/Editor/DVD Authoring


The reason is that HD in Japan is 60.00 hz All Japanese produced HD equipment must be compatible with the NHK 1080I 60 Hz standard

Dave Satin
Video Engineer


Michael Most wrote :

>>I think the original intent was to solve the drop/non-drop problem. >Martin, I'm surprised at you! 59.94 was introduced as a result of the >move to color.

Of course, I know that.

>>Drop frame and non-drop frame code have absolutely nothing to do with >frame rates.

Sure it does. That's the only reason drop-frame even exists!

With 29.970 frames per second drop-frame timecode was invented in order to keep some semblance of real time.

At 30.000 frames per second non-drop has no reason to exists.

Using drop-frame timecode with 30.000 fps material (if even possible) is laying out the groundwork for trouble.

My original statement was intended to say that the 60.000i (30.000s frame per second) standard is a way to, finally, do away with the ridiculous drop-frame timecode mess.

Martin Euredjian
eCinema Systems, Inc.


Martin Euredjian wrote :

>>Using drop-frame timecode with 30.000 fps material (if even possible) >is laying out the groundwork for trouble.

It's used all the time for sound recording when the camera is running at 24fps. This keeps them in sync when telecine runs the film at 23.98 and the sound is resolved to video rate (29.97).

>>My original statement was intended to say that the 60.000i (30.000s >frame per second) standard is a way to, finally, do away with the >ridiculous drop-frame timecode mess.

Understood. However, the "mess" is really not much of a mess anymore, particularly in terms of post production. Since nearly all sound houses use video masters, the use of 29.97 sync has become commonplace. In fact, it would probably create even more of a mess if it were abandoned at this point.

Mike Most
VFX Supervisor
IATSE Local 600
Los Angeles


Michael Most wrote :

>>Using drop-frame timecode with 30.000 fps material (if even possible)

>>It's used all the time for sound recording when the camera is running at >24fps.

Right, of course, I wasn't thinking audio here but in the context of a program. As you said, both the film and audio get slowed-down to NTSC rates and it all works. It's an artifice, really, to make it all jive in the NTSC world. Without the slowdown using NDF in a 30.000fps environment would be quite a mess.

>However, the "mess" is really not much of a mess anymore, >particularly in terms of post production. Since nearly all sound >houses use video masters, the use of 29.97 sync has become >commonplace.

The fact that we figured out how to use it doesn't mean that it isn't a mess. Just like the interlaced-vs-progressive issue. I feel pretty strongly that both interlaced, fractional frame rates and NDF timecode need to die and find a nice warm place in a museum somewhere.

But, that's another topic.

Martin Euredjian
eCinema Systems, Inc.


Michael Most wrote :

>>I feel pretty strongly that both interlaced, fractional frame rates and NDF >timecode need to die and find a nice warm place in a museum >somewhere.

Interesting idea. Quite feasible too. Of course it means throwing out the quazillion TV sets in the US and elsewhere, changing all the transmitters, repeaters, satellites, ground stations, most cameras etc. Did I leave anything out?

Of course, our ancestors should've standardized the whole world
properly, starting with alternating current frequency, voltage, making PAL standard, motion picture fps and have foreseen the difficulties with DV. Did I leave something out?

Robert -It's MY opinion : who minds don't matter: who matter don't
mind - Rouveroy csc
The Hague, Holland


Robert Rouveroy wrote:

>>>Michael Most wrote :

>>>I feel pretty strongly that both interlaced, fractional frame rates and >NDF timecode need to die and find a nice warm place in a museum >somewhere.

Robert, Michael didn't say that, I did.

>>Interesting idea. Quite feasible too. Of course it means throwing out the >quazillion TV sets in the US and elsewhere, changing all the >transmitters, repeaters, satellites, ground stations, most cameras etc.

You are exaggerating a bit. You know that industrious individuals always come up with ways to deal with change. Besides, "all" is a bit extreme. Most satellites, for example, use packaged digital transmission operating in the 10 to 20 GHz bands.

This is one of the issues that was at the core of the ATSC/Grand Alliance/FCC process to define HDTV. Debated for about ten years. It started as an analog system and then became digital with terrestrial delivery. The infrastructure will be blown-up and replaced no matter what.

That's what the whole channel allocation issue was about. For those who do not want to buy a new TV, converters will be available.

Progress doesn't happen without a little chaos.

Martin Euredjian
eCinema Systems, Inc.




Copyright © CML. All rights reserved.