Reverse Telecine Vs Deinterlacing at DVinfo.net

Anil Dasari · June 26th, 2007, 03:32 PM

This might seem like a dumb question, but I'm a bit confused about it.

Is reverse telecine (or pulldown removal, or 2:3 pulldown .... whatever) the same as deinterlacing? I recently edited a 24P Cinemode capture using Adobe Premiere Elements. I checked the 'always deinterlace' option before exporting, and the result showed no interlacing artifacts. If there's more to pulldown removal than a simple deinterlacing, how come there is no 'combing' effect in my output?

Gerry Goto · June 26th, 2007, 04:30 PM

Well if you "remove pulldown" you will remove all the added "filler" frames from the 60i wrapper that the 24p is encased in, leaving you with 24 complete progressive frames per second.

If you de-interlace, all that does is blend the interlaced frames together giving you 30 progressive frames, and in doing so its also keeping all the "junk/duplicate" frames that were added to the 24p to make it 60i.

In the end regardless of what you do you still end up with the 24p film-like effect. It really depends on what you are planning to do with your footage. If you are going to post to the web, then fully progressive is what you want,or if you are like me and will just be burning to dvd then leaving it as is in the 60i wrapper will work just fine, though you could go progressive with that as well if you wanted.

It should be noted that if you de-interlace footage you shot in regular HDV(60i), then it would become 30P. For 24P footage, because it is being recorded to tape, it has to encased in the 60i wrapper, so it is adding an extra frame here and and extra frame there, so you still get the same look whether you remove pulldown or not. So basically if you want your footage progressive, 24P needs to have the pulldown removed, and 60i needs to be de-interlaced.

If you have the software and its easy enough for you to remove pulldown, then by all means do it, itll make everything look a little better and be easier to edit as well as youll be dealing with less frames overall which will be less taxing on your editor. Hope this helps....

If you spend some time reading up on here you'll find alot of threads on dealing with pulldown removal and the software you'll need to do it.....it's our obsession!

Anil Dasari · June 26th, 2007, 05:57 PM

So a deinterlace on 60i gives me 30P, and a pulldown removal on the same gives me 24P? How would the software (that does the pd removal) know which frames are 'fillers' and which are not?
I've seen another thread in this forum that talks about some pd remval flags, and how HV20 doesn't embed those flags in the captured data, and how it's not possible to remove pulldown without the presence of those flags. Folks are flooding Canon with email requests to do something about it. At the same time, I hear that applications like Vegas, After Effects, Cineform, etc have the capability to remove pulldown. How are these applications doing the job without those flags?

Mike Dulay · June 26th, 2007, 06:23 PM

Its done by detecting cadence from the fields/frames between pictures. Flags are explicit while this method is less so. Cineform's and the "free" methods work without flags.

http://forum.doom9.org/showthread.php?t=45456&pp=20
"Just to clear things up CCE doesn't look at the RFF flags when attempting to inverse pulldown. Just like Decomb and other IVTC filters it examines the video and tries to detect a pulldown pattern. It probably does this by looking for repeat fields, but to account for analog sources it has to allow some differences in the repeat fields. So, low motion video may be mistaken for telecined material."

Gerry Goto · June 26th, 2007, 07:04 PM

You are on the right track here.....

"so a deinterlace on 60i gives me 30P, and a pulldown removal on the same gives me 24P?"

It depends what you "shoot" in.... both framerates (HDV and 24HDV) will be captured on the tape in 60i, however with 24P (24HDV) the pulldown is added by the camera to it so that it fills the 60i "wrapper" on the tape. Thats why you need to remove the pulldown to get at the 24P.

If you shoot in HDV(60i) then its already 30 frames progressive but it is split between 2 alternating fields to creat the image, hence the 60 "frames" interlaced, so by deinterlacing that you get 30 complete frames progressive.It has to do with the way standard def tvs work, as opposed to HDTV's or film which are progressive with one complete frame after another.

HDV needs to be in a 60i wrapper no matter what when its being recorded to tape, its just the way the format standard is. Thats why it adds the pulldown when you are shooting in 24 frames, it needs there to be 30 frames to work.

While the HV20 does not add flags it is still possible to remove the pulldown and thats where it becomes a whole lot more work once you understand how it all works, but it can certainly be done with a little hard work. Flags would make it very easy work, and that is why we all bug Canon to do something about that.

June 26th, 2007, 03:32 PM	#1
Anil Dasari Regular Crew Join Date: Apr 2007 Location: San Jose, CA Posts: 47	Reverse Telecine Vs Deinterlacing This might seem like a dumb question, but I'm a bit confused about it. Is reverse telecine (or pulldown removal, or 2:3 pulldown .... whatever) the same as deinterlacing? I recently edited a 24P Cinemode capture using Adobe Premiere Elements. I checked the 'always deinterlace' option before exporting, and the result showed no interlacing artifacts. If there's more to pulldown removal than a simple deinterlacing, how come there is no 'combing' effect in my output?

June 26th, 2007, 06:23 PM	#4
Mike Dulay Major Player Join Date: Mar 2007 Location: Somerville, NJ Posts: 304	Its done by detecting cadence from the fields/frames between pictures. Flags are explicit while this method is less so. Cineform's and the "free" methods work without flags. http://forum.doom9.org/showthread.php?t=45456&pp=20 "Just to clear things up CCE doesn't look at the RFF flags when attempting to inverse pulldown. Just like Decomb and other IVTC filters it examines the video and tries to detect a pulldown pattern. It probably does this by looking for repeat fields, but to account for analog sources it has to allow some differences in the repeat fields. So, low motion video may be mistaken for telecined material." __________________ DIY, 35mm, HV20: http://www.primitivebuteffective.net

June 26th, 2007, 04:30 PM	#2
Gerry Goto New Boot Join Date: Apr 2007 Location: Calgary Alberta Canada Posts: 8	Well if you "remove pulldown" you will remove all the added "filler" frames from the 60i wrapper that the 24p is encased in, leaving you with 24 complete progressive frames per second. If you de-interlace, all that does is blend the interlaced frames together giving you 30 progressive frames, and in doing so its also keeping all the "junk/duplicate" frames that were added to the 24p to make it 60i. In the end regardless of what you do you still end up with the 24p film-like effect. It really depends on what you are planning to do with your footage. If you are going to post to the web, then fully progressive is what you want,or if you are like me and will just be burning to dvd then leaving it as is in the 60i wrapper will work just fine, though you could go progressive with that as well if you wanted. It should be noted that if you de-interlace footage you shot in regular HDV(60i), then it would become 30P. For 24P footage, because it is being recorded to tape, it has to encased in the 60i wrapper, so it is adding an extra frame here and and extra frame there, so you still get the same look whether you remove pulldown or not. So basically if you want your footage progressive, 24P needs to have the pulldown removed, and 60i needs to be de-interlaced. If you have the software and its easy enough for you to remove pulldown, then by all means do it, itll make everything look a little better and be easier to edit as well as youll be dealing with less frames overall which will be less taxing on your editor. Hope this helps.... If you spend some time reading up on here you'll find alot of threads on dealing with pulldown removal and the software you'll need to do it.....it's our obsession!

June 26th, 2007, 05:57 PM	#3
Anil Dasari Regular Crew Join Date: Apr 2007 Location: San Jose, CA Posts: 47	So a deinterlace on 60i gives me 30P, and a pulldown removal on the same gives me 24P? How would the software (that does the pd removal) know which frames are 'fillers' and which are not? I've seen another thread in this forum that talks about some pd remval flags, and how HV20 doesn't embed those flags in the captured data, and how it's not possible to remove pulldown without the presence of those flags. Folks are flooding Canon with email requests to do something about it. At the same time, I hear that applications like Vegas, After Effects, Cineform, etc have the capability to remove pulldown. How are these applications doing the job without those flags?

June 26th, 2007, 07:04 PM	#5
Gerry Goto New Boot Join Date: Apr 2007 Location: Calgary Alberta Canada Posts: 8	You are on the right track here..... "so a deinterlace on 60i gives me 30P, and a pulldown removal on the same gives me 24P?" It depends what you "shoot" in.... both framerates (HDV and 24HDV) will be captured on the tape in 60i, however with 24P (24HDV) the pulldown is added by the camera to it so that it fills the 60i "wrapper" on the tape. Thats why you need to remove the pulldown to get at the 24P. If you shoot in HDV(60i) then its already 30 frames progressive but it is split between 2 alternating fields to creat the image, hence the 60 "frames" interlaced, so by deinterlacing that you get 30 complete frames progressive.It has to do with the way standard def tvs work, as opposed to HDTV's or film which are progressive with one complete frame after another. HDV needs to be in a 60i wrapper no matter what when its being recorded to tape, its just the way the format standard is. Thats why it adds the pulldown when you are shooting in 24 frames, it needs there to be 30 frames to work. While the HV20 does not add flags it is still possible to remove the pulldown and thats where it becomes a whole lot more work once you understand how it all works, but it can certainly be done with a little hard work. Flags would make it very easy work, and that is why we all bug Canon to do something about that.