Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

30p vs 60i...why ever shoot 60i?

S

sbcooler

Well-known member
Joined
Apr 5, 2008
Messages
153
Reaction score
0
Points
0
Location
New York
Why would anyone ever film or video in 60i when for the identical same data rate you could do 30p?

If you ever needed to convert 30p to 60i, you could easily inerlace each frame into 2 fields and get to 60i.

If you have HIGH motion in the frame, 30p will allow all objects to stay together and you won't notice the inerlacing (there is none)...so high motion, like sports from close up, should look better in a progressive output.

If you are shooting low motion like an interview, again 30p and 60i will not look much different. Motion will be smooth either way because there isn't much motion.

So why does any camera get made today that shoots 60i...I mean whatever the codec, say HDV...why 1080i60...why not everything 1080p30? Again it's the same data rate...same memory to store the same amount of signal.

Can someone explain why one would ever want to shoot 60i over 30p?

Thanks-

PS One last thing...compositing, moving into after effects or doing anything is always simpler and easier using a progressive frame. So many advantages..I see no disadvantages.
 
Even though it may not seem like it at first, interlacing was actually invented as a type of 'compression' if you like. The idea was to create extra temporal resolution for the same bandwidth, and in doing so reduce flicker artifacts. Basically the eye is more sensitive to temporal resolution (fps) than it is to spatial resolution (pixels), so by doubling the amount of images you see in a second - by separating a 30p source into 60i 'fields', you get a smoother look and more 'bang for your buck'. The way of thinking has obviously changed over the years as we have become more accustomed to the look of film. The funny thing is that we have fallen in love with many of the 'defects' if you will, of film ..

Don't get me wrong, I am a progressive scan lover .. I will never shoot interlaced again in my life if I have a choice, it's an interesting story though.. and good to understand.
 
Also, you can't 'interlace' a progressive signal as in an interlaced signal each field is taken 1/60th (or 1/50th) of a second after the last, hence the gained temporal resolution.

I just wanted to add one thing though.. I do agree with you wholeheartedly - INTERLACING SUCKS ! PROGRESSIVE IS KING ! :ph34r:
 
Yeah I have read about why TV went with 60i...and the gain in temporal compression from a higher frame or field rate actually. I have this older digital cam that takes video at 15fps and it actually sucks...320 x 240...but the frame rate is not horrible..kind of like strobe light effect...although when there is not a whole lot of motion near the cam it is ok...now 24fps which is most motion pictures looks fine for the most part...maybe like you said we are used to it.

When I watch the Godfather, I have never thought about the frame rate or motion..it looks natural and fine to me. So at 30p, which is 6fps more, I think motion is fine for everything pretty much. Of course sports and close up camera movement is better at 60p...and maybe that is ideal...but for now and the next few years...I can deal with 30p always and like you all NEVER want to do interlaced.

One more thing. I HATE rectangular pixels...mostly I don't have issues with it...but every once in a while it creeps up somewhere where I don't first notice it and later in the job it rears its ugly head after I've spent days compressing a final version.

Just like interlacing, rectangular pixels (also for compression I'm sure) is just wrong and a pain. If I mix footage of DVCPHO HD (1,280 x 1,080 rectangular pixels so 16:9), DV (720 x 480 rectangular pixels) and then photos with square pixels and other video with square pixels...I have conversions all over the place.

And granted Final Cut Pro is great and usually takes care of things..but if I am compositing images in other programs, taking stuff back and forth...just another possibibilty for errors to occur.

So my hope is everything I touch next year is square pixels with no interlacing! Oh and i-frame compression only!
 
It's pretty simple - you shoot 60i when you want 60 discrete fields per second. You shoot 30p when you want 30 discrete frames per second.

There's no correct way to go from 30p to 60i. It would be better to shoot 60p and interlace in post.

So why shoot 60i when you can shoot 60p? Obviously, 60i takes up less space... So when you've got limited data rates, 60i may be preferable if you need the temporal resolution.

That said, I wish we always had the bandwidth, and that interlace was abandoned as a temporal compression solution.

I second your dislike for rectangular pixels also. I get ratio rage all the time.
 
On related subject:
I still do not understand why in HD broadcast we have 29.97 versus 30?
I understand that 29.97 for NTSC and 49.95 instead 30 or 50 was created in order not to interfere with chroma carrier.
Could anybody tell me if I can record 30p to be played most efficiently on the current HDTVs from Blu Ray or via HDMI interface.
Or I have to record 1000/1001 so it is compatible with MPEG2 decoders speed wise?
What is easier for Blu Ray player to decode 29.97p or 30p?


If HDTV is running at 30p can it synchronize to 29.97 or it has to slip convert somehow 29.97 to 30 and vice versa?
Any help will be highly appreciated.

Andrew
 
interlaced scanning and output are likely going to be phased out since non-CRT displays are progressive by nature and the advantages to 60i (i.e. slow-motion sports playback) become easier to re-create from 60p footage.
 
To make a 30p signal into a 60i (true the temporal compression is not the same) all you do is take the 1 frame and divide it into 2 interlaced fields, odd and even...now frame 1 is frame 1 odd say, and frame 2 even...etc...30p fills 60i...the image stays together perfectly because each field is derived from a progressive frame so fast moving objects stay together.

And that is the same data rate...30p and 60i...of course 60p is better motion but it's twice the data rate of 30p and 60i...so if you limit is such that you can do 60i, then go for 30p and never look back. The flicker issue isn't noticeable with todays equipment so there is no benefit to 60i over 30p as I can see. Zero!
 
sbcooler said:
To make a 30p signal into a 60i (true the temporal compression is not the same) all you do is take the 1 frame and divide it into 2 interlaced fields, odd and even...now frame 1 is frame 1 odd say, and frame 2 even...etc...30p fills 60i...the image stays together perfectly because each field is derived from a progressive frame so fast moving objects stay together.

And that is the same data rate...30p and 60i...of course 60p is better motion but it's twice the data rate of 30p and 60i...so if you limit is such that you can do 60i, then go for 30p and never look back. The flicker issue isn't noticeable with todays equipment so there is no benefit to 60i over 30p as I can see. Zero!
Click to expand...

No, you can only have a true 60i frame if it was captured that way. You are mixing up aquisition frame rate vs record frame rate. 60 seperate pictures are taken, and the 1/2 data is used from each frame. 30 is only 30 seperate pictures.

30p can have its own problems........and most appear with fast motion as there is only 30 frames per second, as opposed to 60. Most shows shoot either 60i (to save money on the camera rental) or 24p so they can get a Pal and NTSC master tape. (Universal Aquisition Master)
 
As a stock footage company, interlaced footage is a huge headache. It is almost useless for compositors because it cannot be easily scaled, time-stretched, rotated, repositioned, etc. And, no matter how carefully we create certain demos to play on DVD, there is always the wayword interlaced clip that is out of phase, and sometimes the problem is intermittent, which does not make sense but it happens. BAD!

It causes 90% of our tech support calls.

The only benefits I have seen of interlaced footage is to match the existing look other footage you are editing with, and for taking HDV 1080i and extracting 60fps Standard Def slow motion.

Phil
www.artbeats.com
 
Andrew said:
I understand that 29.97 for NTSC and 49.95 instead 30 or 50 was created in order not to interfere with chroma carrier.
Click to expand...

Actually, 29.97 was done so that chroma carrier did not mingle with the sound career. TV signal has "holes" in its spectrum and color information is filled/hidden in these holes, which was a marvelous design NTSC committee came up with so that color and existing monochrome TVs could co-exist. However, they had to delay the horizontal frequency of 30 HZ monochrome tv from 15750 slightly to 15736 (numbers from the top of my head, but in the ball park) for the 29.97Hz color signal.
 
sbcooler said:
One more thing. I HATE rectangular pixels...

Just like interlacing, rectangular pixels (also for compression I'm sure) is just wrong and a pain.
Click to expand...

Rectangular pixels don't creep up in NTSC and PAL because of compression. Luma information in NTSC signal is restricted to about 4.2 MHz and PAL to about 5.5 MHZ (top of my head number). Therefore, sampling theorem would say to sample NTSC signal at 4.2 * 2 = 8.4 Mega samples / sec and PAL at 5.5 * 2 = 11 Mega samples / sec. However, in a desire to even up some numbers (both NTSC and PAL have 720 widths in sampled data) a sampling frequency of 13.5 Mhz was chosen. Hence, both NTSC and PAL signals are actually over-sampled.

NTSC:
====

Number of pixels on a scan line (including horz. blanking) = 13500000 / (30 * 1000 / 1001) / 525 = 858.

Number of "active" pixels = 858 * 52.5 / 62.5 = 720.

This is how you get the 720 for 720x480/486 dimensions of NTSC/DV footage.

Similar calculation can be done for PAL.

And, interestingly, by design, the active pixel width for PAL is also 720 (720 x 576).
 
Spiff said:
It's pretty simple - you shoot 60i when you want 60 discrete fields per second. You shoot 30p when you want 30 discrete frames per second.
Click to expand...
But a field in 60i is only half a frame... and what good does half a frame do anyone in 2008?

If you need 60 discrete frames per second, shoot 60p. If you only need 30, shoot 30p. There's no reason to ever shoot interlaced in this day and age, and it's a crying shame that the interlaced mechanism has been carried over to the HDTV format.
 
Yes it is a crying shame interlaced has been carrier out in these cameras. NO need for it all. 30p is a nice compromise between 24p and 60p with manageable data rates...as most are used to this data rate from 60i...again 30p takes up the same room as 60i.

You know if you really get technical you can take a 60i picture say in HD 1080 vertical lines and take out half..to 540...now you have smooth motion, no artifacts from interlacing and no scan lines to see or worry about. Of course your resolution just got cut in half...so on HDV instead of shooting 1440 by 1080 you really are shooting 1440 x 540. That's a tight picture, but nowhere near as good as a true 1920 x 1080 square pixels!

Red corrects all that...and finally the more professional cams are starting to (Pannasonics HPX 3000 and Sony's EX-1). Finally true 1,920 x 1,080 square non interlaced footage. Of course Sony has the interframe compression so that is just plain wrong again for fast changing video. Pannasonics got the right kind of codec I think with that AVC INTRA.

And RED has the best codec so far. Of course this take none of the most important or equally important parts of the footage which is the lense, sensor, etc..

Go RED!
 
I believe the other reason for interlaced formats has to do with the speed of cathode ray tube etc from the days before LCD/Plasma etc The old TVs were limited on how many horizontal lines they could draw because lines at the top of the screen would fade before lines at the bottom could be drawn, causing a very bad flicker, so fooage was split into odd/even lines so the refresh took half as long, but twice as often, which meant watchable TV. Why interlaced footage is still created I don`t know, but a lot of people still have old TVs.
 
Not to be pointing anyone away from here, but I just wrote about this stuff at my blog. With pictures! Has to do with DV and uncompressed and the problems Final Cut has with exporting in an interlace format.

http://www.bayareafilmmakers.com
 
On related subject:
I still do not understand why in HD broadcast we have 29.97 versus 30?
Click to expand...
I believe most HD broadcast is this way so it is possible to simulcast SD easily. If you have 30.000fps HD and 29.97 SD simulcast, then eventually the two will run out of sync and you have to do something to fix that.

Or you can just stick with running everything at 29.97, which is not a big deal. It can be a bit of a pain in the butt, but it can be dealt with and the end result will be equally good. It doesn't affect the picture quality that the consumer sees.

2- Whereas with interlacing, the end result is not equally good compared with going with a broadcast system without interlacing.

“I am amazed that anybody would consider launching new services based on interlace. I have spent all of my life working on conversion from interlace to progressive. Now that I have sold my successful company, I can tell you the truth: interlace to progressive does not work!”.

This remark was followed by a spontaneous round of applause from the audience. When the person was asked to give his name, he replied “Yves Faroudja” (who is widely acknowledged as the world’s foremost expert on format converters).
Click to expand...
 
You must log in or register to reply here.
Back
Top