29.97i vs 29.97p

This isn't necessarily fcpx inquiry.. but I can't seem to find any solution or answer about this.

I've shoot a short interview with Sony NX3 (1080 29.97i / 48Khz) and iPhone 6S+ 1080 30p / 44.1Khz).
Both cameras were shoot together as a full shot and a CU shot.

When I imported both file I've noticed that the Sample rate and the frame rate are different
so I've conveted iPhone clip to (1080 29.97p / 48KHz).

Now my question is...

What's the difference between 29.97p and 29.97i?
is the difference only the fields?
if so, do I have to worry about sync?

if not, what should i do to fix that?

duusfcp wrote: …What's the difference between 29.97p and 29.97i?
…
if so, do I have to worry about sync?
…

tons of wiki on that topic
progressive means, each of the 29.97 frames per second is stored a single solid frame
interlaced means, each of the 29.97 frames per second is split into two fields, each containing the odd/even lines of a frame, so therfore also named as 60i/59.94i ...

and, no, no worry, a second is always a second in FCPX, independently it contains 24, 30, 120 or 7.543 frames - usually manual converting does do the harm ...

And: Welcome to fcp.co!

Karsten Schlüter wrote:
tons of wiki on that topic
progressive means, each of the 29.97 frames per second is stored a single solid frame
interlaced means, each of the 29.97 frames per second is split into two fields, each containing the odd/even lines of a frame, so therfore also named as 60i/59.94i ...

and, no, no worry, a second is always a second in FCPX, independently it contains 24, 30, 120 or 7.543 frames - usually manual converting does do the harm ...

And: Welcome to fcp.co!

While Karsten's answer is totally correct regarding the video pictures, the audio part of an AV file may behave different.
While a rose is a rose is a ... and a second is a second is a ... in real world (and real time) native NTSC people don't live in real world and time :dry:
A displayed second in NTSC world and a real second differ by 1001/1000 - but the audio is always real world.
If you force an audio to live on the dark side of the world (NTSC) the sample rate won't match.

"1080 29.97i / 48Khz) and iPhone 6S+ 1080 30p / 44.1Khz)"
So the first 29.97i / 48Khz is NTSC and the real world sample rate must 48Khz/(1001/1000) = 48.048 KHz to match.
The second 30p / 44.1Khz is real world and the real world sample rate is 44.1Khz.

As Karsten said a second is a second and therefore the NTSC audio is missing 48 Hz = 48 samples in real world per second.
As I mentioned above this might be no problem, especially with short clips but you might have issues with longer takes.

- Andreas

thanks Andreas for chiming in ... excellent info! (as usual...)

I neglected the 'details' on purpose, 'cause it instantly makes me dizzy,
e.g. what happens to that 0.03 frame each second lacks?!? :blink:
and why that odd number 44.1 (do! read 3d ago an article why that... dizzyness again...)
...

that "a second is a second" is meant for simple folks like me, k.i.s.s. ....
in most cases, dump-it-into-FCPX works perfectly ... ok, 999/1000th perfectly

Fcp.co is not really compatible with sun shine. But for people in the UK with no sun it's okay I guess.
I reply later.

- andreas

Karsten Schlüter wrote: thanks Andreas for chiming in ... excellent info! (as usual...)

I neglected the 'details' on purpose, 'cause it instantly makes me dizzy,
e.g. what happens to that 0.03 frame each second lacks?!? :blink:
and why that odd number 44.1 (do! read 3d ago an article why that... dizzyness again...)
...

that "a second is a second" is meant for simple folks like me, k.i.s.s. ....
in most cases, dump-it-into-FCPX works perfectly ... ok, 999/1000th perfectly

"cause it instantly makes me dizzy" :lol:
Me too and 10,000s of other people as well. Even people like my friend Maurizio Argentieri (Sound Designer Jesus Christ Super Star) hates NTSC sampling trouble. Some people from Skywalker Studios told me once 'Andreas, look on the bright side of life' in so far we both get paid jobs handling this stuff.
Again you're not alone.

Regarding your questions.

"what happens to that 0.03 frame each second lacks"
You answered your question yourself: "a second is a second". An NTSC second isn't a real world second. It takes 1*1001/1000 = 1.001 seconds. And in case of 30 frames per NTSC second it is 29.97 frames per real world second. in case of 60 it's 0.06 frames missing. :woohoo:

"why that odd number 44.1"
44.1 KHz is CD quality sampling rate.

Anyway you and the OP should take it easy – in many cases X makes a good job.

- Andreas

Andreas Kiel wrote: …"why that odd number 44.1"
44.1 KHz is CD quality sampling rate.
…

I know, but the story behind, why 44.1 and not 44.2, 43.2, 44.0, or whatever - THAT made me dizzy :lol:
Read it twice in hope to understand.... not really. But not in need doing so...
Same with cars: pushing the pedal = faster. Why? …more gas, kinda? Anyhow.-

Andreas Kiel wrote: …

Anyway you and the OP should take it easy – in many cases X makes a good job.

I'm tiefenentspannt concerning that....
my trust in the engineering skills at Cupertino is ... huge.
Have to admit I'm a butcher, the last .5% I'm never interested in... ok, 10% ...

"30 frames per NTSC second it is 29.97 frames per"

No, it is 29.97 FIELDS, not frames. Someone mention it broke the original frame into 2 fields, no it doesn't. NTSC starts with alternative fields, there is never a frame to work with at all. A frame only exists physically once it is converted to a progressive format.

Rule of thumb, you can go easily from progressive to interlaced. Going from interlaced to progressive can be problematic.

Real world, I've never had issues with audio going from i to p. Only with the visual artifacts.

At the end of the day, I am baffled why anyone created a progressive codec that does interlaced numbers. I'd love to find the person responsible and punch him full force in the mouth.

BenB wrote: …No, it is 29.97 FIELDS, not frames. …

<ough> with all respect, and well knowing this gets soon academic/wording-only but :..

29.97i translates into 29.97 frames per second, each made of two fields.
so, technically 59.94 fields per second....

... imho ...

Yes, NTSC in FRAMS is 59.94, not 29.97 frames, correct.

It gets more fun than that. It can still have 60 independent moments in time per second. 60 completely different images per second... and it usually does. This really takes a hit on the overall resolution though most won't be able to perceive it.

Once there was a time, when sending signals via radio waves, had issues. But that was long ago. Today; I wish interlaced video would just be banished to hell for eternity...

BenB wrote: No, it is 29.97 FIELDS, not frames.

What? Nonsense. Interlaced NTSC is of course 29.97 or 30 FRAMES per second. Consisting of two FIELDS per frame. Essentially 60fps, only with half the information tossed out per frame.

BenB wrote: Rule of thumb, you can go easily from progressive to interlaced.

Oh, that's something I'd love to see. Especially the part about from where you're getting the missing half (alternate field) of the information. Do tell. Because otherwise there is zero difference between the before and after. None. Therefore: completely pointless.

BenB wrote: Going from interlaced to progressive can be problematic.

Huh? Unless you're using a specialty filter, all you're doing is throwing half of the image out. Not exactly rocket science or much of a "problem".

BenB wrote: Real world, I've never had issues with audio going from i to p. Only with the visual artifacts.

Artifacts? So simply being left with half the original image resolution is an "artifact" for you?

BenB wrote: At the end of the day, I am baffled why anyone created a progressive codec that does interlaced numbers.

What does that even mean, "doing interlaced numbers"? :huh:

BenB wrote: I'd love to find the person responsible and punch him full force in the mouth.

I know how you feel.

BenB wrote: I wish interlaced video would just be banished to hell for eternity...

Yeah, why use a system that gives you (perceptually) TWICE the framerate for HALF the bandwidth when you can just jack everything up by twos, huh? Yeah, to HELL with that, huh?

You took the words out of my mouth.

5 thumbs up for this reply.

One thing to add:
You don't need a field recorder to record interlaced video!!!

- Andreas

AndieMoepse wrote:

BenB wrote: Rule of thumb, you can go easily from progressive to interlaced.

Oh, that's something I'd love to see. Especially the part about from where you're getting the missing half (alternate field) of the information. Do tell. Because otherwise there is zero difference between the before and after. None. Therefore: completely pointless.

It's not completely pointless. If you have something you made in 480p 29.97 and you want to show it on something that only allows an interlaced signal (like NTSC broadcast back in the day) then going to interlaced is not pointless. Also if I have a 1080p 60fps program that I want to put on Blu-ray, then I will need to go to 1080i since Blu-ray is not powerful enough to handle 1080p at that speed. The loss of vertical resolution will be regrettable, but far from pointless.

AndieMoepse wrote:

BenB wrote: Going from interlaced to progressive can be problematic.

Huh? Unless you're using a specialty filter, all you're doing is throwing half of the image out. Not exactly rocket science or much of a "problem".

Throwing half the information out is a HUGE problem. Professionals don't like doing that.

AndieMoepse wrote:

BenB wrote: Real world, I've never had issues with audio going from i to p. Only with the visual artifacts.

Artifacts? So simply being left with half the original image resolution is an "artifact" for you?

Fits the definition.

AndieMoepse wrote:

BenB wrote: I wish interlaced video would just be banished to hell for eternity...

Yeah, why use a system that gives you (perceptually) TWICE the framerate for HALF the bandwidth when you can just jack everything up by twos, huh? Yeah, to HELL with that, huh?

Definitely. Interlacing is stupid. It had its uses for the time but I still wish it was never invented. Upper and lower fields? No thank you. I can see interlacing in action and it has always bothered me, even back in the VHS days. I noticed it as a child. Interlacing is evil and it belongs in hell.

I worked with interlace on tape for many years, lived through the tape to digital transition, I know something about what I'mt talking about. If you disagree, fine. But, my personal, real world experience is just that. Sorry for trying to help. I'll love you all to your discussion.

A frame is not two fields combined. It was not broadcasted as frames.

As i've probably got more experience editing in frames and fields than most (8 field colour sequence anybody?) I can see right and wrongs in the answers in this thread. I'm not going to go through correcting though as I haven't got the time and i'm sure you don't want to end up reading paragraphs of my boring analogue video engineering thoughts.

However, what I don't like is the tone of this conversation. People often comment on how the quality of answers on here is always high without the normal bickering you find on other forums, one of the hoofed kind for example.

Please keep it nice and friendly here - I'm close to locking the thread

Peter

Now where did I leave my pierce box?

A frame is not two fields combined. It was not broadcasted as frames.

It is if it's PsF, progressive segmented frame, a progressive frame encapsulated in an interlace stream.

Interlace is still used for broadcasting because it is half the bandwidth of 50p and because modern TVs do such an amazing job of deinterlacing there's not a human alive that can tell the difference between 50i and 50p broadcasts. It has its merits.

Most people referred to 50i footage as having a frame rate of 25 and no one really bothered thinking too deeply about it until I suppose HD came along and there was a provision for 50i and 50p so a distinction was drawn in the nomenclature i.e

resolution I or P framerate e.g 1080i25 or 1080p25. Some people were writing 1080i50 to denote typical PAL frequency interlaced 1080 which was wrong. The convention is to write frame rate not field rate.

peteramwiggins wrote:
Please keep it nice and friendly here - I'm close to locking the thread

I totally agree.
And I excuse myself for any stupid joke I made during the time I was posting here.

Many of the discussion here is just a discussion about "nomenclatura".
Unfenswinger makes it pretty clear in his thread. PSF is a progressive, ISF is an interlaced frame, These words are used very seldom but describe it very good.

If everybody is talking about his business experience – my is 43 years, if anybody is interested.

And by the way: 25 years nd 4 weeks ago (NAB) QT was presented first time to selected developers and customers. one year after I started working with it

Best
Andreas

me no engineer, far from that!, but wording is .. was my domain.
basically it should be called 540i ... not 1080i
and PAL (NTSCT) isn't just a framerate ..... :lol:


@Andreas
Would be a nice read, if you would share some memories from 25y ago ....
and not that cryptical please as at finalcutprofi.de ...
reminded me way too much of Kai Krauses weird byteburg blogs .... :silly:

... what happened to that guy, anyhow? hmmmm.....

(massive threadjack in progress!!)