PDA

View Full Version : Proper framerate nomenclature



Mike Harvey
09-28-2008, 04:27 PM
So this came up in another thread, but figured it was worthy of it's own... what is the proper framerate nomenclature? I was always taught i n school and at work that it was (resolution)(progressive or Interlaced)/framerate)... i.e. 1080i/30 or 720p/24.

I see a lot of folks calling 1080i 1080i/60... including manufacturers. That denotes to me that it is 1080 interlaced at 60fps. Which means that there are 120 fields. I've never heard of a camera doing that in the HMC150's price range. Actually, I've never heard of anything shooting in 1080i (or 480i, for that matter) other than at 30fps (at least in this country). I know some can record different framerates and lay it down to a 1080i/30fps stream... but interlaced in this country was pretty much tied to 30fps.

Am I wrong in thinking this?

frubsen
09-28-2008, 04:33 PM
So this came up in another thread, but figured it was worthy of it's own... what is the proper framerate nomenclature? I was always taught i n school and at work that it was (resolution)(progressive or Interlaced)/framerate)... i.e. 1080i/30 or 720p/24.

I see a lot of folks calling 1080i 1080i/60... including manufacturers. That denotes to me that it is 1080 interlaced at 60fps. Which means that there are 120 fields. I've never heard of a camera doing that in the HMC150's price range. Actually, I've never heard of anything shooting in 1080i (or 480i, for that matter) other than at 30fps (at least in this country). I know some can record different framerates and lay it down to a 1080i/30fps stream... but interlaced in this country was pretty much tied to 30fps.

Am I wrong in thinking this?

1080i/60 = 60(59.94) fields per second which plays back equivilent to 30(29.97) interlaced frames per second, although don't get that confused with 30(29.97) progressive frames per second.
1080p/60 = 60(59.94) progressive frames per second

there techically is no HDTV standard of 1080i/30 because that would infer 30 fields per second.

there is however 1080p/30 and 720p/30 which is 30 progressive frames per second however this does have the same look as 1080i/60.

Mike Harvey
09-28-2008, 04:52 PM
1080i/60 = 60(59.94) fields per second which plays back equivilent to 30(29.97) interlaced frames per second, although don't get that confused with 30(29.97) progressive frames per second.
1080p/60 = 60(59.94) progressive frames per second

there techically is no HDTV standard of 1080i/30 because that would infer 30 fields per second.

there is however 1080p/30 and 720p/30 which is 30 progressive frames per second however this does have the same look as 1080i/60.

So basically, when interlaced it's listing the fields, progressive it's frames. Thanks, though I think it makes more sense to not deal with fields and leave it at frames, so as not to confuse people. Seams a bit unnecessary to change everything when it's interlaced.

But that's just my opinion (and we all know what body part opinions are like :D )

Barry_Green
09-29-2008, 07:02 AM
It's listing the refresh/scan rate. Whether frames or fields, there are 60 motion samples per second.

Technically, when you get right down to it, there's no such thing as a "frame" in interlaced video anyway, so the idea of using "frame" nomenclature is a mistaken process that's tied to NLE systems, not broadcast and not the signal itself.

Jim Simon
10-01-2008, 10:49 PM
there techically is no HDTV standard of 1080i/30 because that would infer 30 fields per second.

No, that designation properly means 30 interlaced frames per second. Just as 1080p/24 means 24 progressive frames per second. The number after the slash denotes frame rate.

I have to second Mike's issue with the cross labeling. It would be easier if specs only listed frames, not fields. We know that if it's interlaced, there are two fields per "frame". Spelling it out only causes confusion, I think.

For example, when I first heard of 720p/60, I thought somebody goofed, because who ever heard of a 60 fps format that wasn't slow motion?

Luis Caffesse
10-01-2008, 11:34 PM
For example, when I first heard of 720p/60, I thought somebody goofed, because who ever heard of a 60 fps format that wasn't slow motion?

Except that no one goofed, and 720P really is a 60fps format (in DVCproHD anyhow).

So the naming convention still stands - as Barry pointed out -
You can think of it as:

Spacial Resolution / Capture Method / Temporal Resolution (or refresh rate).

720P/60 makes perfect sense.
As does 1080i/60

I suppose it can seem confusing at first -but once you know the letter denoting the capture method then you know if the number is referring to frames or fields.

PKraft
10-03-2008, 03:45 AM
So the naming convention still stands - as Barry pointed out -
You can think of it as:

Spacial Resolution / Capture Method / Temporal Resolution (or refresh rate).


Luis, with all due respect, IMHO both you and Barry are not right for several reasons.

First and above all, we speak about "framerate" and not "scanrate" or "refreshrate". In
consequence it denominates a certain number of frames in a given timeframe.
In that case frames per second or fps.

Second, following Barry's assumption of scanrate, we would not talk about, say,
1080i24 but rather 1080i48 for interlaced material and 180p24 for progressive
material. However, I have never ever seen anything like 1080i48.

Third, there is an direct relation to the denomination of timecode, which only refers to
fps but not to a scanrate or whatever, for the sake of clarity.

Four, there are several technical papers, recommendations, etc, both from the EBU
as well as from SMPTE, ISO or other comparable bodies that only speak of framerates
but never of scanrates in that context.

For example: http://tech.ebu.ch/docs/tech/tech3301.pdf

With reference to:
• SMPTE Standard 330M-2004 for Television — Unique Material Identifier (UMID)
• SMPTE RP 205-2000 — Recommended Practice on Application of Unique Material Identifiers
in Production and Broadcast Environments
• EBU Technical Statement D92-2001 - System Requirements for the unique identification of
material in broadcast production
• EBU Technical Recommendation R108-2001 - High-level rules for systems implementing the
SMPTE UMID.




Hope this helps. pe.

Luis Caffesse
10-03-2008, 08:19 AM
First and above all, we speak about "framerate" and not "scanrate" or "refreshrate". In
consequence it denominates a certain number of frames in a given timeframe.
In that case frames per second or fps.
True - and keep in mind that for years people refered to NTSC as 30 fps when it was really 29.97. And even today we talk about 24p when in fact it is 23.98 frames per second.

My point is that term 'framerate' may not in fact be used in the most accurate way by people. While the word literally does mean a specific number of frames in a given timeframe - it seems many people have taken it to mean a certain number of samples taken in a specific timeframe.


Second, following Barry's assumption of scanrate, we would not talk about, say,
1080i24 but rather 1080i48 for interlaced material and 180p24 for progressive
material. However, I have never ever seen anything like 1080i48.

I dont' think I've ever seen 24 frame interlaced material - it does not exist as far as I know, which is why we've never seen 1080i48 used. 1080i is only seen at 30 frames per second, or 25 frames per second in PAL countries. Each of which are expressed as 1080i/60 and 1080i/50 respectively. I suppose if someone had a camera that could shoot 24 interlaced frames at 1080 we would in fact refer to it as 1080i48.


Third, there is an direct relation to the denomination of timecode, which only refers to fps but not to a scanrate or whatever, for the sake of clarity.

I'm not certain I followed you on this.
Could you clarify?
EDITED TO ADD:
Ah, I see what you're saying now - that timecode only tracks frames, and not fields?
Okay. I guess I don't see that as having a direct effect on how a given framerate is expressed in writing or amongst people.


Four, there are several technical papers, recommendations, etc, both from the EBU as well as from SMPTE, ISO or other comparable bodies that only speak of framerates but never of scanrates in that context.

The pdf you linked to didn't seem to specifically address the nomenclature issue - and in fact in the Metadata section on page 13 actually made reference to both Frame Rate and Field rate.

Either way - I don't really want to split hairs on this - I was only trying to clarify the rationale behind how everyone I know refers to framerates and recording formats. For right or for wrong this is how they are denoted amongst pretty much every cameraman and editor I work with.

Cranky
10-04-2008, 09:47 PM
Scanning systems are identified with three major parameters:

* Frame size in pixels is defined as number of horizontal pixels x number of vertical pixels, for example 1280 x 720 or 1920 x 1080. Often number of horizontal pixels is implied from context and is omitted.
* Scanning system is identified with the letter p for progressive scanning or i for interlaced scanning.
* Frame rate is identified as number of video frames per second. For interlaced systems an alternative form of specifying number of fields per second is often used.

If all three parameters are used, they are specified in form frame size scanning system frame rate. Often, one parameter can be dropped if its value is implied from context. In this case the remaining numeric parameter is specified first, followed by the scanning system.

For example, 1920x1080p25 identifies progressive scanning format with 25 frames per second, each frame being 1920 pixels wide and 1080 pixels high. The 1080i25 or 1080i50 notation identifies interlaced scanning format with 50 fields(25 frames) per second, each frame being 1920 pixels wide and 1080 pixels high. The 1080i30 or 1080i60 notation identifies interlaced scanning format with 60 fields (30 frames) per second, each frame being 1920 pixels wide and 1080 pixels high. The 720p60 notation identifies progressive scanning format with 60 frames per second, each frame being 720 pixels high, 1280 pixels horizontally are implied.

50Hz systems have only four scanning rates: 25i/50i, 25PsF, 25p and 50p. 60Hz systems operate with much wider set of frame rates: 23.98p, 23.98PsF, 24p, 24PsF, 29.97i/59.94i, 29.97p, 29.97PsF, 30i/60i, 30p, 30PsF, 59.94p and 60p. In the days of standard definition television, the "fractional" rates were often rounded up to whole numbers, like 23.98p was often called 24p, or 59.94i was often called 60i. High definition television allows using both fractional and whole rates, therefore strict usage of notation is desirable. Nevertheless, 29.97i/59.94i is almost universally called 60i, likewise 23.98p is called 24p.


I dont' think I've ever seen 24 frame interlaced material - it does not exist as far as I know Not native interlaced, but PsF. AFAIK, Canon use it internally fot its "F" modes (actually, 23.98PsF). Real 24PsF is used for film-to-digital transfer.

Luis Caffesse
10-04-2008, 10:41 PM
Not native interlaced, but PsF. AFAIK, Canon use it internally fot its "F" modes (actually, 23.98PsF). Real 24PsF is used for film-to-digital transfer.

True, but that's a whole different animal, hence having it's own identifier of PsF.
That's not the same thing that would be denoted by calling something 24i (if such a thing existed).