What is the standard format that composers are getting movies in these days?

[shnurgle]

Are they still delivering films to composers on videotape at 29.97 fps (ndf)? Have they switched to digital format for delivering to composers? If so what is the standard file format .avi? .mov? Is SMPTE still used in LA?

When 24fps film is converted to digital format, how does that effect the fps and what does that mean for the composer in terms of synchonozation? If I dump a quicktime movie into my sequencer that was xfered from 24 fps film, does that mean I can set my seq. to 24fps and be OK? In other words, when calculating tempos for a score based on a digital clip is it safe to assume that the tempos that work in the sequencer will work at the final dub for the final lock?
With SMPTE there is a universal timecode for the clip and what the composer gets will invariably work at the final dub because they are using the same timecode. With digital however, there doesn't seem to be a universal timecode, so how do you know for sure that your music will lock up exactly as you want it? If there are slight discrepancies in the fps and timecode what beginas as a perfect synch will be out of synch by the end of the clip.

many questions, I know. Thanks!

[pappagheno]

Are they still delivering films to composers on videotape at 29.97 fps (ndf)? Have they switched to digital format for delivering to composers? If so what is the standard file format .avi? .mov? Is SMPTE still used in LA?

When 24fps film is converted to digital format, how does that effect the fps and what does that mean for the composer in terms of synchonozation? If I dump a quicktime movie into my sequencer that was xfered from 24 fps film, does that mean I can set my seq. to 24fps and be OK? In other words, when calculating tempos for a score based on a digital clip is it safe to assume that the tempos that work in the sequencer will work at the final dub for the final lock?
With SMPTE there is a universal timecode for the clip and what the composer gets will invariably work at the final dub because they are using the same timecode. With digital however, there doesn't seem to be a universal timecode, so how do you know for sure that your music will lock up exactly as you want it? If there are slight discrepancies in the fps and timecode what beginas as a perfect synch will be out of synch by the end of the clip.

many questions, I know. Thanks!

what do you mean for digital ?

a quick time can be a converted PAL as well NTSC ...

[guy theaker]

Are they still delivering films to composers on videotape at 29.97 fps (ndf)? Have they switched to digital format for delivering to composers? If so what is the standard file format .avi? .mov? Is SMPTE still used in LA?

When 24fps film is converted to digital format, how does that effect the fps and what does that mean for the composer in terms of synchonozation? If I dump a quicktime movie into my sequencer that was xfered from 24 fps film, does that mean I can set my seq. to 24fps and be OK? In other words, when calculating tempos for a score based on a digital clip is it safe to assume that the tempos that work in the sequencer will work at the final dub for the final lock?
With SMPTE there is a universal timecode for the clip and what the composer gets will invariably work at the final dub because they are using the same timecode. With digital however, there doesn't seem to be a universal timecode, so how do you know for sure that your music will lock up exactly as you want it? If there are slight discrepancies in the fps and timecode what beginas as a perfect synch will be out of synch by the end of the clip.

many questions, I know. Thanks!

Can't answer all of these, but maybe some.

Quicktime file format (.mov) seems to be the standard for visuals these days. If you get yourself a copy of Quicktime Pro then that should cover most conversions between digital formats. As far as sync is concerned, once you have digital visuals it ceases to be an issue. However, it's best to set your sequencer to the frame rate of the .mov so that the diplayed frame rate matches that of the original picture - that way you can hit certain frames with your music, when needed.

But more, I cannot say...

G.

[Daryl]

Quicktime file format (.mov) seems to be the standard for visuals these days.....
G.
Not on PC it isn't...........!

D

[PaulR]

many questions, I know. Thanks!

Miles - just take alook at this to start with. Should be helpful and it's cheap.

http://www.macprovideo.com/tutorial/logicVideoEds

[guy theaker]

Not on PC it isn't...........!

D
Well, it is on MY PC! :)

Obviously, I'm not saying that every movie you get will be a .mov, but given the predominance of Macs within the professional sector (particularly the visuals sector) I think it's safe to say it's more than likely.

[pappagheno]

Well, it is on MY PC! :)

Obviously, I'm not saying that every movie you get will be a .mov, but given the predominance of Macs within the professional sector (particularly the visuals sector) I think it's safe to say it's more than likely.

I think depends also from what sequencer you are using..

on my pc quick tiem work well with Cubase..
working on a project that have footage coming from a dvd.. so had to do a lot of passage to get a proper quick time.. but if someone is doing it for u ..well even better

[Jeff Hayat]

Composers get their work in different formats. Quicktime files, VHS, SVHS, 3/4" Umatic...

As for the frame rate, you made the question more difficult than it is.

"When 24fps film..."

Film is transferred to video not at 24, but one tenth of a percent slower: 23.98. This will sync to video, wich doesn't really run at 30, but 29.97. Hence, the time code runs at a rate of 29.97. When the film is transferred to "digital format" it is not running at 24, but at 23.98; therefore, the digital file is not 24 but 23.98. Your sequencer should be running at 29.97.

24 = 30
23.98 = 29.97

Cheers.

[shnurgle]

When the film is transferred to "digital format" it is not running at 24, but at 23.98; therefore, the digital file is not 24 but 23.98. Your sequencer should be running at 29.97.

You mean the sequencer should be running at 23.98. You said the digital file is 23.98; sequencer and video should be identical, right?

Do all digital video forms .mov, .avi etc all run at 29.97? SI this df or non drop frame? How do you know what fps a file is?

[shnurgle]

when you use the term "video" what do you mean? Is that a blanket term for video tape, umatic tape, .mov files, .avi files? Do ALL of these "video" formats run at the same speed: 29.97 ndf?

[Jeff Hayat]

Ok, there is film and there is video.

B&W video runs at 30 fps, color video runs at 29.97 fps. Film is shot on loacation at 24 fps, which does not "translate" properly to 29.97 color video. Therefore, when the film is transferred to video, also known as a telecine transfer, the film machine - the telecine - is pulled down 1/10% from 24 fps to 23.98 (aka 23.976); the film is not running at the same exact speed at which it was recorded. If cutting (editing) digitally, the video tape (usually Beta SP) gets digitized into the Avid, and the Avid is set to pulldown mode; Avid users will say the film rate is set to .99 instead of 1. From the Avid comes your digital file, be it a .mov, .avi, .wmv, .whatever, and this file is running at 29.97/23.98. It is NOT running at the equivalent of 24 fps; if you were to play this file alongside the film in the cinema, you would start to see a drift in sync. How much? 1.44 frames per minute. This is of course assuming that the project was not shot Hi-Def, and that everything was done normally.

When you say 29.97 TC (TC = time code), you are referring to non-drop frame time code. Very rarely is non-drop frame time code referred to as such; you will rarely see '29.97 ndf', as it is assumed that 29.97 is in fact ndf. However, you WILL see 29.97 df if the project is drop-frame; this is needed to specify. Drop frame TC is the same exact thing as non-drop: it runs at the same speed and at the same frame rate, but drops two frames every minute except every tenth minute in order to be completely synchronous with the time of day clock for TV (it's not exact, but the miniscule discrepancy that spans a period of a few weeks the stations can deal with). This is why TV projects are df, and film is ndf. HBO is an exception; they will take either/or as they for some reason or another are not that concerned. Not sure about other cable stations, tho....

Your digital file should be 29.97/23.98. Your sequencer should be set to 29.97, and your music will now: 1) sync to the video in your sequencer (obviously), and 2) sync to the film as it appears in the Avid (depending on how you deliver the music).

Anything else? :)

[Crossingsound]

Ok,
hope this won't confuse you anymore, but having been a post engineer, and working on some films, some things that we took into consideration. If you get a video format, VHS, Beta or such, that is going to go BACK to film as the final layback, then there is the 24 fps difference that film runs at versus the 29.97 frame rate for color video. So, when working on audio that you know will be mixed back to film, We had to pull down our sample rate. So instead of working at 48k, we would actually be working at 47.9 someting something. Now this pull down of sample rate will slow your session down enough to compensate from the 24 and 29.97 frame difference. So when we bounced down our stems to d-88's, we then pulled up our sample rate, and the mix house would lock up the D-88's and pull the sample rate down again, and mix to picture.

Now working with a digital file, either quicktime, AVI, DV, etc, I think its best to find out where the final mixdown is going to go, back to film, or stay at video. then probally take note of above. If you have questions, maybe talk to the video house, and see what they say. keep your session the same frame rate as what they made the file in, 29,97 dp, or nd (probally nd), but keep in mind you may have to pull down your sample rate, otherwise you will get drift when working with the film.

this is how we handled the post projects on our end, not sure if things have changed as it has been a few years....

if working with a digital file, you can get a more advanced video card, and actually send that video signal out to a monitor, so you can watch it on a bigger screen then inside your DAW. conopus makes some decent firewire video cards...

working non-linearly with a DV file is the way to go, If I have a tape, I would want to convert it into a DV file anyway for the random access.

[Crossingsound]

Oh, a few udder things when working with DV files,

If there is no visual time code in the DV file, then hopefully they supplied a 2 pop at the top, so they can no where to sync the audio from your session. If there is visual timecode, It is best to make the video the same time code address as what is on the visual, in other words make 2 hours and 22 minutes, 30 seconds and 22 frames on the video, be the same location within your DAW, that way, if they take the session, your addresses are identical.

[Jeff Hayat]

.....we would actually be working at 47.9 someting something

52

48k pulled down is 47.952; 44.1 pulled down is 44.056.

[Crossingsound]

I am sure logic will be able to pull down your sample rate, you are only adjusting it by -.1 percent, which will not be noticable in pitch, but in the long run , will help ensure it won't drift when mixing back to the film medium. This should be a standard feature in all professional DAW's.

you are working to the video rate of 29.97, but pulling down your sample rate will basically be slowing your audio down just a touch, as to compensat for the frame differences between film and video. I would just make sure where your audio is going to go first, maybe yu don't need to pull down your audio, and someone else will do that, but be in communication ahead of time, and don't be afraid to ask these questions.

[shnurgle]

Thanks guys very much for your experienced input.
Your digital file should be 29.97/23.98. Your sequencer should be set to 29.97, and your music will now: 1) sync to the video in your sequencer (obviously), and 2) sync to the film as it appears in the Avid (depending on how you deliver the music).
So when the soundtrack is edited into the film in avid (still at 29.97) there is no concern that it will become out of synch when it is processed back to 35mm film for the theaters (to be projected at 48fps?
48k pulled down is 47.952; 44.1 pulled down is 44.056.
Riff, based on your statement ubove there should be no need to pull down sample rate though, right? Can you please indicate where this would be necessary?

I know I'm asking a lot of questions here guys, but believe me, I wouldn't if this info was available on the net. I've researched fps, video transfer, drop frame, and all the specifics of video's role in film. I can tell you how a 2:3 pull-down is achieved, why drop frame timecode was created for television, and exactly how many horizontal scan lines NTSC video has (525); but there is no information on fps and syncing music to video that is going back to film, so you wisdom is most greatly appreciated.

Thanks!
Miles

[shnurgle]

I am sure logic will be able to pull down your sample rate, you are only adjusting it by -.1 percent, which will not be noticable in pitch, but in the long run , will help ensure it won't drift when mixing back to the film medium. This should be a standard feature in all professional DAW's.

you are working to the video rate of 29.97, but pulling down your sample rate will basically be slowing your audio down just a touch, as to compensat for the frame differences between film and video. I would just make sure where your audio is going to go first, maybe yu don't need to pull down your audio, and someone else will do that, but be in communication ahead of time, and don't be afraid to ask these questions.
Thanks very much.

[shnurgle]

Miles - just take alook at this to start with. Should be helpful and it's cheap.

http://www.macprovideo.com/tutorial/logicVideoEds
Thanks Paul, that looks like a cool video.

[DaveWalton]

I can only speak to my participation in "minor league" projects, low budget independent films. I'd say that 95% of what I receive is in Quicktime format. The remaining 5% comes in WMV and/or AVI format. I did one feature film last year and my working copy was sent to me as a standard NTSC DVD. Because Sonar is fussy about the type of video files it works with, I convert everything I get to AVI format for working within Sonar. I've only had one film given to me with the timecode burned in.

I use a PC and I have Quicktime Pro as well as a handful of video conversion utility programs.

HTH,

Dave

[awpmusic]

100% of what I've had in the past year has been QuickTime - and I'm using Cubase SX3.1 on a PC - works perfectly. Simpler in the UK too - video is shot at 25 fps. As an added safety measure better to get your QuickTime with the TC burned into the picture.

[Daryl]

100% of what I've had in the past year has been QuickTime - and I'm using Cubase SX3.1 on a PC - works perfectly. Simpler in the UK too - video is shot at 25 fps. As an added safety measure better to get your QuickTime with the TC burned into the picture.
Most of the people I work with get VHS tapes......! I also find that QT is cr*p on my systems. I usually convert whatever I'm given to AVI, which is rock solid.

D

[PaulR]

Thanks Paul, that looks like a cool video.

No problem Miles. What I might suggest is - if you're pushing ahead big style with film and tv scoring now, it may be worth considering going to Logic 7.2.1. Look into that. If you go ahead - take out the hard drives from the Mac and put in two new ones and load the newest OS etc and start from scratch. Bit of a pain I know - but probably worth it for you.
You can keep the old ones in case of any problems of course.

Just a thought.

[Jeff Hayat]

"So when the soundtrack is edited into the film in avid (still at 29.97) there is no concern that it will become out of synch when it is processed back to 35mm film for the theaters (to be projected at 48fps?"

Ok, first off, I am assuming that 'projected at 48fps' was a typo; I am sure you know it is 24. Ok, now that that is out of the way - there is no concern at all. If the film is running at 1/10% slower, and the audio is conformed to this speed, the audio gets "bumped up" (technical term there) when the optical track negative is shot. This is the "processing" you referred to - it is one of the the final steps in making a film - the optical track negative gets married to the picture negative at the lab and that is what the prints are made from - the same prints that get distributed to the theaters.


"Riff, based on your statement ubove there should be no need to pull down sample rate though, right? Can you please indicate where this would be necessary?"

Where this would be necessary is when you have location audio, recorded at 24 fps/44.1 (or 48), and you need to sync up that audio to video. Well, the video is no longer running at film speed, it is running at video speed. So, you take the 44.1 location audio, pull it down to 44.056, and now the audio will sync to the video. Furthermore, any new non-location audio (say, score) will be recorded at 44.056, so that when it gets bumped up to film speed later on, the audio will now be running at 44.1. You could technically bump the audio from 44.1 to 44.xxx (I forget the number) but this is not how it is done generally.

Cheers.

[shnurgle]

Thanks Rif!
Where this would be necessary is when you have location audio, recorded at 24 fps/44.1 (or 48), and you need to sync up that audio to video. Well, the video is no longer running at film speed, it is running at video speed. So, you take the 44.1 location audio, pull it down to 44.056, and now the audio will sync to the video. Furthermore, any new non-location audio (say, score) will be recorded at 44.056, so that when it gets bumped up to film speed later on, the audio will now be running at 44.1. You could technically bump the audio from 44.1 to 44.xxx (I forget the number) but this is not how it is done generally.
So basically, every recording session in Hollywood for a theatrically-released feature film is recorded with a pulled down sample rate? Every time Zimmer goes into the studio with an orchestra, the engineer is recording with a pulled-down rate? Everytime he loads a video of a reel of the film into his sequencer and starts mocking things up with sample libraries, he sets his sequencer to a pulled down rate, right? (Because his score will eventually be synced back to 24 fps film for the theater)

In other words, it is UNUSUAL for a typical blockbuster score to be recorded with a normal sample rate? I wonder why this information was never once mentioned in any of the film scoring manuals.


so this is what I've learned:

If I am scoring a movie that will NOT be synced back to film (straight to vhs, DVD, or television, or multimedia) even if the original film was SHOT on 35mm at 24fps:

1.set the sequencer to 29.97 fps while recording the score (because regardless, whatever you're watching in your sequencer is SOME kind of video and therefore must be 29.97fps)
2.set the sample rate at a normal rate (44.1, 48, etc)
3.record and bounce down.

Your audio file will have frame acurate sync to video and will remain as such no matter what medium the movie ends up on (DVD, VHS, digital file etc) because all video is 29.97

If you are scoring a film that will ultimately be transferred back to 24fps film (for theater projection):
1. set the sequencer to 29.97 fps (again, because at this stage you are dealing with video)
2. pull down the sample rate by the correct factor while recording and composing (we do this so that score sample rate matches location audio sample rate when bumped back up to film)
3. Record and bounce down.

Then this "slower" audio file is then synced to video in Avid or Final Cut or whatever (still at 29.97fps) and when it is finally transferred back to 24fps film the picture as well as the sound moves slightly faster (negligable to the eye or ear) which brings the fps of the film on video from 23.98 to an even 24, and the sample rate of the audio from 44.056 to 44.1

The difference in pitch and tempo is impercaptable, so the music you wrote at the pulled down tempo doesn't suffer from this "bump up" in speed.

Where this would be necessary is when you have location audio, recorded at 24 fps/44.1 (or 48), and you need to sync up that audio to video. Well, the video is no longer running at film speed, it is running at video speed. So, you take the 44.1 location audio, pull it down to 44.056, and now the audio will sync to the video.
Right, but what if there is no location audio? And the video will be ultimately transferred back to film for theatrical projection? In this case, there would be no need for the sample rate pull down when recordin the score, because there will be no sample rate discrepancy. I mean, if you were scoring say, a silent film, or documentary with just score and narration. You would transfer the silent film to video, then you could record your score and post audio at a completly normal sample rate (fps set to 29.97 of course). The sound tracks would be synced to the video in Avid, and then the thing would be transferred to 24fps film. Because everything was synced in video and because both the audio and picture are being natually spead up by the same factor, the sync will reman perfect, right?

So in other words, sample rates must always be uniform. The sample rate "pull-down" facilitates this when there is audio that was recorded while the film was being shot at 24 fps. If there is no such audio, however, and all audio is done in post, then the pull-down is unecessary because sample rates are uniform already. If all audio is recorded at 44.1, when the video is transferred up to film, all audio will then be 44.XXX (whatever the number is--slightly faster), but it won't matter that it's not exactly 44.1 because all audio will now be 44.XXX and it will sync perfectly to film (as it synced perfectly to video at 44.1).

Concluding thoughts:
So ALL of this is necesary because there is currently no way to dump a video into your sequencer that is exactly 24 fps. That's the bottom line. If you could somehow transfer 24fps film to a medium that you could load in your sequencer and score to exactly 24fps (or if you could somehow hook your sequencer up to a 35mm, 24fps film projector!, or simply, if NTSC color video ran at an even 30fps, then film would play back on video at exactly 24 fps (not 23.98) so sync wouldn't be a problem.

So thanks to this invaluable thread, I think I now have a grasp synchronization. Thanks to all the contributors!

[shnurgle]

Just had an afterthought. I noticed that my sequencer seems to support many different frame rates. But if all US video is 29.97, under what circumstances could I ever be working at 24fps or 30 etc. in my sequencer?

[Daryl]

Just had an afterthought. I noticed that my sequencer seems to support many different frame rates. But if all US video is 29.97, under what circumstances could I ever be working at 24fps or 30 etc. in my sequencer?
Are you never going to consider working with films from other countries?

D

[awpmusic]

Are you never going to consider working with films from other countries?

I hope not! There's not enough work over here as it is!! :D :D :D

[PaulR]

Just had an afterthought. I noticed that my sequencer seems to support many different frame rates. But if all US video is 29.97, under what circumstances could I ever be working at 24fps or 30 etc. in my sequencer?

hehehe! Miles - you need to get that video.

[shnurgle]

Are you never going to consider working with films from other countries?
Never. I would sooner cross the Sahara sitting backward on a pregnant mule than even CONSIDER scoring a film made outside the US.

Seriously now,
24fps is used for film here in the States. Unless there is an international video format that runs at 24 fps, I don't see how you could ever utilize the "24" frame rate in your sequencer. It's not like you can hook Logic up to a reel of film and compose to 24fps...

[awpmusic]

It's not like you can hook Logic up to a reel of film and compose to 24fps...

eh? don't you - but how could I ever compose without the whir of the projector in my right ear??

[shnurgle]

Hey guys. Ok, I'm really not trying to be annoying here (waaaaaaaaaaay too late), but I keep reading conflicting information. I am under the impression after reading this thread and other research, that when film is digiized to video it is 29.97 non drop frame, and that "drop frame" is not used at all except for on TV (when it is necesarry that an hour equals an hour)

BUT

Reading Logic Pro 7's reference guide under syncronization and frame rates, it says that 29.97d (d=drop here) is the standard for color NTSC video, and that 29.97 non drop is "Extremely rare."
..But if all films go to video at 29.97 non drop (and if non drop is really the standard as many have said here) then WHAT THE SHNINGLEBERRIES IS GOIN ON??!! Someone get me a sedative and set me straight please. why does Logic think that drop frame video is the standard and that no one ever uses non drop?

[awpmusic]

Have a beer - it'll all become obvious!

I dunno mate - we only work at 25 fps here! :)

[Daryl]

I hope not! There's not enough work over here as it is!! :D :D :D
But we're not the only country with a proper frame rate for video; there are loads of others...!

D

[shnurgle]

I'm in agreement. Let's get the states using 25fps. Let's start a petition.

[awpmusic]

Let's get the states using 25fps. Let's start a petition.Might be a bit of an uphill struggle that one! :D

[shnurgle]

Seriously though, can anyone explain why logic thinks 29.97 drop frame is the standard for video in the states and it thinks 29.97 non drop is "extremely rare?"

Riff, anyone..?

[dinosound]

I'm in agreement. Let's get the states using 25fps. Let's start a petition.

Better start small and begin with changing over to the metric system!

[dinosound]

Seriously though, can anyone explain why logic thinks 29.97 drop frame is the standard for video in the states and it thinks 29.97 non drop is "extremely rare?"

Riff, anyone..?

It's simply wrong. I get stuff in nd all the time. I would say it's a 50/50 split between ndf and df that I receive a project on. I always make a point of asking when I receive a video. I also check the burn in just to make extra sure.

D.

[shnurgle]

It's simply wrong. I get stuff in nd all the time. I would say it's a 50/50 split between ndf and df that I receive a project on. I always make a point of asking when I receive a video. I also check the burn in just to make extra sure.

D.
man, if i had a dollar for every thing "wrong" about Logic I'd be so filthy stinkin rich, I'd just record all my scores live and skip the whole god-forsaken sequencer experience altogether.

I mean seriously, how can a manual for a program that convoluted afford to contain even a bit of wrong information? If it's all correct, it's still a learning curve and a half!

It's all about not f-ing releasing software (manuals included!) until it works right. And that goes for releasing new versions of software every year just to force everyone to shell out another $200/year to keep "current." If you're software doesn't work, you don't get to charge people for it! It's called criminal behavior. I'm so sick of the low standard with software. What if Honda released a car and every unit sold didn't work. Imagine if you bought a dvd that didn't work and you found out that all of them don't work, and that the company knew about it and sold them anyway!
But with software..."oh it's got bugs? That's ok, we'll release it before christmas and charge $1000 for it, and people can just download updates as they become available."

i-n-e-x-c-u-s-e-a-b-l-e

I'm done now.

[dinosound]

man, if i had a dollar for every thing "wrong" about Logic I'd be so filthy stinkin rich, I'd just record all my scores live and skip the whole god-forsaken sequencer experience altogether.

I mean seriously, how can a manual for a program that convoluted afford to contain even a bit of wrong information? If it's all correct, it's still a learning curve and a half!

It's all about not f-ing releasing software (manuals included!) until it works right. And that goes for releasing new versions of software every year just to force everyone to shell out another $200/year to keep "current." If you're software doesn't work, you don't get to charge people for it! It's called criminal behavior. I'm so sick of the low standard with software. What if Honda released a car and every unit sold didn't work. Imagine if you bought a dvd that didn't work and you found out that all of them don't work, and that the company knew about it and sold them anyway!
But with software..."oh it's got bugs? That's ok, we'll release it before christmas and charge $1000 for it, and people can just download updates as they become available."

i-n-e-x-c-u-s-e-a-b-l-e

I'm done now.


You're preachin' to the choir, brotha'!

[shnurgle]

sorry about the rant, just want to add that I've been very impressed w/ this compant'y releases (obviously)

[Jeff Hayat]

"So basically, every recording session in Hollywood for a theatrically-released feature film is recorded with a pulled down sample rate?”

Yes.

“Every time Zimmer goes into the studio with an orchestra, the engineer is recording with a pulled-down rate? Everytime he loads a video of a reel of the film into his sequencer and starts mocking things up with sample libraries, he sets his sequencer to a pulled down rate, right?

Probably – I’ll ask him next time we do lunch.

“(Because his score will eventually be synced back to 24 fps film for the theater)”

Yes.

”In other words, it is UNUSUAL for a typical blockbuster score to be recorded with a normal sample rate? I wonder why this information was never once mentioned in any of the film scoring manuals.”

Yes, and you got me there.

Ok, regarding how you set your sequencer: Ultimately, it shouldn’t matter – depending on other factors, and what your deliverable is.

Film is shot on 35mm film @ 24 fps
Film is sent to Post facility, and transferred to BetaSP – film runs @ 23.976 fps
BetaSP gets sent to picture department cutting room; picture editor assistant digitizes Beta into Avid
Editor and Director compile scenes, and create 1st cut
Picture editor assistant creates an Avid dump with pulldown on, and sends file, running at 29.97 video speed, to shnurgle
Shnurgle loads video file into sequencer, and starts to score

Now - if your sequencer is set to 29.97 you will not have a problem, and if your sequencer is set to 30 (or 24) you theoretically should not have a problem, unless your sequencer for some reason does not like to play back 29.97 video files when it is set to 30 or 24.* The reason I say this is because time is time, and if the score written on your sequencer @ 29.97 syncs to the 29.97 video file, the score written on your sequencer @ 30 will still sync to the 29.97 video file - when played back that way.

Write @ 29.97, play back at 29.97, and the score will sync to the 29.97 video
write @ 30, play back at 30, and the score will sync to the 29.97 video*
write @ 29.97, play back at 30, and the score will NOT sync to the 29.97 video

Hope that's clear. Now - how do you plan on deliver the score?

Oh yeah - one monkey wrench needs to get thrown into this whole thing - your sequencer, is it being referenced externally? And are you feeding it external time code?

“Just had an afterthought. I noticed that my sequencer seems to support many different frame rates. But if all US video is 29.97, under what circumstances could I ever be working at 24fps or 30 etc. in my sequencer?”

HD, or if someone gives you a Final Cut/Avid dump which was not pulled down to video speed.

“Reading Logic Pro 7's reference guide under syncronization and frame rates, it says that 29.97d (d=drop here) is the standard for color NTSC video, and that 29.97 non drop is "Extremely rare."”

You either read that wrong, the printer screwed up, or they have no idea what they are talking about. Reading that, I am glad I went with Cubase.

“I'm in agreement. Let's get the states using 25fps. Let's start a petition.”

Uhh, let’s NOT.

*This is assuming that the video file will play back at 29.97 if the sequencer is set to 30. If not, and the sequencer pulls up the video file, the video is now running at film speed (the equivalent of 24 fps) and you WILL have sync issues later on. And keep in mind, this only applies to INTERNAL video files; with external (say, vhs) video, it wouldn't matter how the sequencer is set.

Cheers.

[shnurgle]

Riff, just want to say thanks very much for all your input here! It is helpful and much appreciated.

"So basically, every recording session in Hollywood for a theatrically-released feature film is recorded with a pulled down sample rate?”

Yes.

“Every time Zimmer goes into the studio with an orchestra, the engineer is recording with a pulled-down rate? Everytime he loads a video of a reel of the film into his sequencer and starts mocking things up with sample libraries, he sets his sequencer to a pulled down rate, right?

Probably – I’ll ask him next time we do lunch.

“(Because his score will eventually be synced back to 24 fps film for the theater)”

Yes.

”In other words, it is UNUSUAL for a typical blockbuster score to be recorded with a normal sample rate? I wonder why this information was never once mentioned in any of the film scoring manuals.”

Yes
This is a revelation. Definitely the single most crucial bit of technical information I've ever learned about the process of film scoring and I've never even heard (or seen in print) it mentioned before.

Now, let's say the movie was shot in digital, but it will ultimately be transferred to 35mm film for theatrical presentation. In other words, a movie not originally shot on film, but going to film for the theaters. How does this effect the sample rate pull-down? In this case, I would think no pull-down is nec for recording the score. When the video is xferred to film it will in fact bump up a little faster, but because everything was in sync in video already, it will all bump up together and be in sync on film --it will just be a few seconds shorter than the video version.

And keep in mind, this only applies to INTERNAL video files; with external (say, vhs) video, it wouldn't matter how the sequencer is set.
Because external stuff has it's own time code that the sequencer slaves to, right?
This is assuming that the video file will play back at 29.97 if the sequencer is set to 30. If not, and the sequencer pulls up the video file, the video is now running at film speed (the equivalent of 24 fps) and you WILL have sync issues later on.
So in other words, keep the sequencer at what the video is to avoid any such problems. 29.97 video...keep it at 29.97.

Write @ 29.97, play back at 29.97, and the score will sync to the 29.97 video
write @ 30, play back at 30, and the score will sync to the 29.97 video*
write @ 29.97, play back at 30, and the score will NOT sync to the 29.97 video
I'm not quite sure I follow this. Per your footnote, you're saying that the sequencer will atually play 30 at 29.97, but why would it do this?

Riff, can you recommend as book that explains all this in detail (specifically what we're talking about with respect to film scoring, not video specifications only).

Thanks a million,
Miles

[Jeff Hayat]

“Because external stuff has it's own time code that the sequencer slaves to, right?”

Yes. Well, it should – this is how the sequencer will follow and stay in sync with the external source. I am not sure about other sequencers, but Cubase cannot sync to real TC; the TC needs to be converted to MTC and then fed into Cubase. I question the accuracy of this – I will believe it is accurate when I see it.

“So in other words, keep the sequencer at what the video is to avoid any such problems. 29.97 video...keep it at 29.97.”

Yes.

“I'm not quite sure I follow this. Per your footnote, you're saying that the sequencer will atually play 30 at 29.97, but why would it do this?”

Err, no – that’s not exactly what I said. Not sure about what it is you are getting at there….

“Now, let's say the movie was shot in digital, but it will ultimately be transferred to 35mm film for theatrical presentation. In other words, a movie not originally shot on film, but going to film for the theaters. How does this effect the sample rate pull-down?”

Not too sure about that one to be honest. How are they getting the video to film? Kinescope? How was the video shot? Was the video shot at 29.97? 30?

“it will just be a few seconds shorter than the video version.”

Yes, for a 2 hr. feature, the difference would be about 5 – 5.5 seconds.

A good book is the Time Code Handbook - I belive the author is Walter A Hickman, and that it is published by By Cipher Digital, tho I am not sure if you will be able to find a new copy.

Cheers.

[shnurgle]

“I'm not quite sure I follow this. Per your footnote, you're saying that the sequencer will atually play 30 at 29.97, but why would it do this?”

Err, no – that’s not exactly what I said. Not sure about what it is you are getting at there….

Fair enough let me try again. here's what you said:
write @ 30, play back at 30, and the score will sync to the 29.97 video*

*This is assuming that the video file will play back at 29.97 if the sequencer is set to 30.

I think you're saying is that the audio (so long as it's recorded at the same fps as it is bounced down or played back) will lock to the video (and ultimately to the film) the same way it did during composition. So 30 or 29.97 isn't truly relavant except that one might make your composing easier.
In other words, the reason we set our sequencers to a specific fps at all is to make our composing job easier for frame acurate sync during writing. But, theoretically if we had the intuitive skills, we could set our sequencers to any fps we want--17---2 million whatever--- and as long as what we wrote synced in real time to the picture (running at 29.97) and as long as we bounced down the audio file at the same fps as we recorded it, it would sync fine once married.
To put it another way: fps setting in a sequencer is there for the composer, but isn't really relavant otherwse. It sets our sequncers clock to match the frame count of the video exactly, but fps setting in the host sequencer has no bearing on real time of the video. If we change he fps, we change the way real time is counted on the screen, but obviously we are not effecting the actual time (length) of the video or music we write, just the way it is counted out.

Tell me I'm onto something here...

[Crossingsound]

shnurgle
here is a real life example of frame rate,

I was doing a post with sfx and vo to a beta, locked with house sync and timecode. I concluded that the tape was running at 29.97 DF. I started to place some sfx, and say I went and placed a sfx 2 minutes into the peice, went back to the top of the tape, watched from the beggining, and my carefully placed sfx was not right on cue, but drifted just a bit. So, the frame rate was not accurate on the tape (you have to get after video guys sometimes cause they don't mark tapes sometimes). when i stopped the tape, and played right before the sfx, it was in sync, but starting from the top, it drifted. So I changed my sesion rate to 29.97 ND and all was well with the world..

just a real life example of drift and the proper frame rate within a session.. not sure how tempo when composing and frame rate effect each other, proballoy not untill you bounce to audio..

[LEX]

Good stuff here. Long read. I skimmed a bit.

Just to clear a few things up.

Production sound is actually recorded at 30fps, not 24fps.

Avid editors work at 24fps, not 29.97. Think how confusing it would be for the negative cutter who got an EDL at 29.97.

Not every big film is done in pulldown. Back when running a bunch of machines together it was necesarry, but not anymore.
I have done a ton of movies this past year, mixing, all at 29.97 48k no pull.
All sync fine to my 35mm projector.

Part of the reason it is done is mostly habit. But it isn't necessary. All telecine and avid transfers can be done at 29.97 without the need for a pulldown. The Avid actually has a pulldown switch on it so you can do the output either way.
It use to be that in ProTools 5.1, when you pulled down the audio, the video pulled down as well.
Now, even simpler you can just pullup the video to film speed and leave the audio at 48k.
Depending on what the studio is going to run at, but that all needs to be determined before spotting, usually determined by the Post facility.
Always check with them on what their specs are. It will make all the difference.

29.97 and 30 are different. So if you scor in 30 and it is delivered and put into a 29.97 or 29 puldown, your score will drift.
If your video is running at 29.97 and you set your session to 30, the timecodes will be different.
So you place a hit on a specific cut of the film.
You change the frame rate to 29.97, now it comes early.
Working within you own system, everything is going to sync to whatever you set it for.
What you want to be aware of is that it is delivering it that is what is important.

AS far as formats, I request everything in DVCam format with an address track.
This way I can check the address track against the burnin to ensure they are the same.
Also, it is a digital transfer so there is no loss in the Analog convertion.
I do everything in NTSC DV, quicktime movies. The files are large, but the quality is better then many and it doesn't tax your CPU with having to decode another kind of codec like MJPEG.

The whole process can be a whirl wind of confusing. There is the whole 2-3 or 3-2 pulldown (for film to video transfers).
Quick and basic. 1 film frame goes to 2 video frames, the next film frame goes to 3 video frames and so on.
3-2 is different in the way it works 3 then 2 but on the b frame. It is confusing.
The norm is 2-3.

Glad you are finding this out snurg. After my rant a while back, which made us mortal enima's :D, I figured you just didn't care.

Trust me. Always get the specs of the post facility and run the sequencer the same way.

LEX

[Synesthesia]

Would it be fair to say that a possible very significant error can be introduced by *the way your sequencer/audio recorder plays back the video file*..

ie: if your protools session is set to a certain rate or nd/d and the video is different, then tools will alter

a: the way the video plays (either by just going frame by frame and ignoring absolute values)

or

b: the sample rate of the playback/record, so when you output a bounced bwav its suddenly playing back slightly faster/slower because it is now freed from the tyranny of the movie playback??

Thats actually a question!!

Thanks,

Paul:confused:

[Jeff Hayat]

“Production sound is actually recorded at 30fps, not 24fps.”

True – but 30 is the eqivalent of 24 in terms of speed.

”Avid editors work at 24fps, not 29.97.”

But bear in mind that any dumps that are given to the sound and music editors are not at 24fps - at least as far as video tapes are concerned.

“All telecine and avid transfers can be done at 29.97 without the need for a pulldown. “

Umm, if the telecine/Avid transfer is done at 29.97, that IS pulled-down.

“The Avid actually has a pulldown switch on it so you can do the output either way.”

True – see my above comment.

”It use to be that in ProTools 5.1, when you pulled down the audio, the video pulled down as well.
Now, even simpler you can just pullup the video to film speed and leave the audio at 48k.
Depending on what the studio is going to run at, but that all needs to be determined before spotting, usually determined by the Post facility.
Always check with them on what their specs are. It will make all the difference.”

True. In vers 5 that there was no option to not pull down the video; any imported video file automatically was pulled down. But when vers 6 came out, the option to not pull down, and/or pull up was added.

”29.97 and 30 are different. So if you scor in 30 and it is delivered and put into a 29.97 or 29 puldown, your score will drift.”

True – but there really is never an instance of pulling down 29.97 as it is already pulled down, therefore, there is really is no such thing as “29 pulldown”. The point to remember is always score with the end result in mind.

"The whole process can be a whirl wind of confusing. There is the whole 2-3 or 3-2 pulldown (for film to video transfers).
Quick and basic. 1 film frame goes to 2 video frames, the next film frame goes to 3 video frames and so on.
3-2 is different in the way it works 3 then 2 but on the b frame. It is confusing."

Actually, I don’t find it confusing. Well, maybe if someone is completely new to this….but you have it right – sort of. 1 film frame > 2 video fields, the next film frame > 3 video fields, etc. I am assuming you meant fields there, and not frames. And yes, that is where the 2:3 pull down comes in.

Film frame 1 > A1, A2
Film frame 2 > B1, B2, B3
Film frame 3 > C1, C2
Film frame 4 > D1, D2, D3
Film frame 5 > A1, A2

And so on….whereas A1, A2, B1, etc. represent video fields. It is done this way (starting with an A-field) because of the Avid…I am far from an Avid expert, but it has something to do with the fact that either the Avid doesn’t like to digitize/make a cut on anything but an A-frame, or the EDL is made easier when the sequence starts with an A-frame, or both...or something like that.

------------------------------------------------------

“Tell me I'm onto something here...”

Ok, you’re onto something here...

“So 30 or 29.97 isn't truly relavant except that one might make your composing easier.”

Well, actually, talent has a lot more to do with that than frame rate.

“In other words, the reason we set our sequencers to a specific fps at all is to make our composing job easier for frame acurate sync during writing. But, theoretically if we had the intuitive skills, we could set our sequencers to any fps we want--17---2 million whatever--- and as long as what we wrote synced in real time to the picture (running at 29.97) and as long as we bounced down the audio file at the same fps as we recorded it, it would sync fine once married.”

Umm, yeah – something like that. 2 million fps - DAYUM!!!!

“but fps setting in the host sequencer has no bearing on real time of the video.”

Well, that’s not necessarily true, depending on your sequencer. To use Lex’s example: if you load a 29.97 video into P-Tools, and are working at a hard 48 with no pulldown, P-Tools will pull up the video to film speed; the video is now no longer playing at video speed, it is playing at film speed…it is playing at 30 fps, the equivalent of 24 fps. This would be done to erase the need for a bump-up when the optical track negative is shot.

Hope this cleared it up a bit…

[Jeff Hayat]

"Would it be fair to say that a possible very significant error can be introduced by *the way your sequencer/audio recorder plays back the video file*.."

Yes - possible.

"if your protools session is set to a certain rate or nd/d and the video is different, then tools will alter

a: the way the video plays (either by just going frame by frame and ignoring absolute values)..."

Yes - the objective is to make sure the video is playing back at the proper frame rate, so that any music you write will sync to not only the video file within P-Tools, but other elements as well.

"or

b: the sample rate of the playback/record, so when you output a bounced bwav its suddenly playing back slightly faster/slower because it is now freed from the tyranny of the movie playback??"

Well, P-Tools won't automatically alter the sample rate of the playback/record - this is something you must set beforehand, so when you output an audio file, it will play back correctly - not too fast and not too slow.

Cheers.

[shnurgle]

Thanks everyone for contributing!
Hey Lex,
Thanks for the input, I was hoping you'd jump on board. ("enima" or not :) )

Avid editors work at 24fps, not 29.97. Think how confusing it would be for the negative cutter who got an EDL at 29.97.
OK guys, can we clear this up please?

I have done a ton of movies this past year, mixing, all at 29.97 48k no pull.
All sync fine to my 35mm projector.
This is encourageing to hear. But, if the location audio was shot at 48 to sync perfectly to 24fps film, then your score, if recorded to video at 48, would be out of sync with the location audio (the sample rates would be different) when bumped back up to film for theatrical presentation. (unles you're saying that the 29.97 video you are running is actually playing the film at 24fps and not 23.98 fps--which would be HUGE.

I would like to see Riff and Lex come to an agreement here. We're trying to figure out the empirical facts and you guys seem to believe (with equal degrees of passion and assuredness) the opposite things. I'm now getting the impresion that pulling down a sample rate recording a score is not so common...

What I know to be true when scoring a film:
You MUST speak directly with the post guys about each project to determine:
a) sample rate specs
b) if your sample rate should be pulled down while recording score to accomodate a transfer back to film (to match location audio)
c) the exact time code format of the video

What I'd like to establish/clarify for sure (please):
1. All video given to composers to score from is some form of 29.97 fps, right?
2. When I am given video to score from (of a film shot at 24ps) is the film on that video showing at 23.98 or 24? In other words, have they pulled down the fps of the film to 23.98 when coverting it to video in Avid, or have they (can they) keep it at 24fps? If they can keep the film showing at an even 24 fps, as Lex suggests they can, then clearly there would be no need for any sample rate pull-down when composing--WHICH WOULD BE F-ING AWESOOOOOOOOME MINE BREEEEEEEETHS! Keeping it at 24 all the way would be smoove, to say the least.. As I understand it, the ONLY reason film is pulled down from 24 to 23.98 when going to video is because 23.98 fits well on video (29.97) and 24fps does not. But, if they have found a way to make even 24fps play as such on 29,97 then we are indeed in business my friends.
3. Just to clarify, when we say video runs at 29.97 fps, we are differentiating that from the frame rate that the film is showing at (on the video). So we say that 24 fps film runs at 23.98 fps (or POSSIBLY 24FPS) on NTSC video which plays at 29.97 fps. And when we set our sequencers, we do it to 29.97 because it is technically the video, not the film, we are syncing to in the sequencer. SO even though the film may be going by at 24fps, we must recognize that the actual file we were given is a video and we must set the sequencer as such. Right?

[sinkd]

gentlemen all,

I found this which might be of some utility as a general reference:
104

Doesn't address DFTC, though.

DS

Switch to landscape before printing. (D'Oh!!)

[Crossingsound]

shnurgle ,

perhaps it is a good thing to understand why video is running at 29.97. If I remember corectly, video used to run at 30 fps, and when color Video came out, it slowed down just a titch because of the introduction of color to video, which is now running at 29.97 rather then straight 30.

[LEX]

Thanks everyone for contributing!
Hey Lex,
Thanks for the input, I was hoping you'd jump on board. ("enima" or not :) )


OK guys, can we clear this up please?

This is encourageing to hear. But, if the location audio was shot at 48 to sync perfectly to 24fps film, then your score, if recorded to video at 48, would be out of sync with the location audio (the sample rates would be different) when bumped back up to film for theatrical presentation. (unles you're saying that the 29.97 video you are running is actually playing the film at 24fps and not 23.98 fps--which would be HUGE.

I would like to see Riff and Lex come to an agreement here. We're trying to figure out the empirical facts and you guys seem to believe (with equal degrees of passion and assuredness) the opposite things. I'm now getting the impresion that pulling down a sample rate recording a score is not so common...

What I know to be true when scoring a film:
You MUST speak directly with the post guys about each project to determine:
a) sample rate specs
b) if your sample rate should be pulled down while recording score to accomodate a transfer back to film (to match location audio)
c) the exact time code format of the video

What I'd like to establish/clarify for sure (please):
1. All video given to composers to score from is some form of 29.97 fps, right?
2. When I am given video to score from (of a film shot at 24ps) is the film on that video showing at 23.98 or 24? In other words, have they pulled down the fps of the film to 23.98 when coverting it to video in Avid, or have they (can they) keep it at 24fps? If they can keep the film showing at an even 24 fps, as Lex suggests they can, then clearly there would be no need for any sample rate pull-down when composing--WHICH WOULD BE F-ING AWESOOOOOOOOME MINE BREEEEEEEETHS! Keeping it at 24 all the way would be smoove, to say the least.. As I understand it, the ONLY reason film is pulled down from 24 to 23.98 when going to video is because 23.98 fits well on video (29.97) and 24fps does not. But, if they have found a way to make even 24fps play as such on 29,97 then we are indeed in business my friends.
3. Just to clarify, when we say video runs at 29.97 fps, we are differentiating that from the frame rate that the film is showing at (on the video). So we say that 24 fps film runs at 23.98 fps (or POSSIBLY 24FPS) on NTSC video which plays at 29.97 fps. And when we set our sequencers, we do it to 29.97 because it is technically the video, not the film, we are syncing to in the sequencer. SO even though the film may be going by at 24fps, we must recognize that the actual file we were given is a video and we must set the sequencer as such. Right?

Editor with AVID cut at 24fps. All the EDL's I get from them are in 24fps, unless they remember to convert it to 29.97.

Production sound is recorded at 30fps, 48k. What matters is when the production is transfered for editing. Usually called a "Dialog Load".
This is where Post house specs come into play.
When the production sound is loaded, it can be loaded at either 30 or 29.97 and will sync up depending on the SR speed, 48 or 47952.

I think you are thinking too much on the film side and 24fps.
I suggest that you don't have to do film at 48k pulldown anymore. But that doesn't mean it still isn't being done.
But, if the specs don't call for pulldown, then don't do it.
My facility is set for 48k, 29.97. Yet I have done a few film that were in Pulldown.
Both sync'ed fine to the film print.

Again, it is the specs of the Post House (or editorial shop) that will set those numbers.

"You MUST speak directly with the post guys about each project to determine:
a) sample rate specs
b) if your sample rate should be pulled down while recording score to accomodate a transfer back to film (to match location audio)
c) the exact time code format of the video"
Sample rate specs is A and B. It will either be 48k, or 48k pulldown.
C) yes, the timecode format.

"In other words, have they pulled down the fps of the film to 23.98 when coverting it to video in Avid, or have they (can they) keep it at 24fps? "
What they are doing is just slowing it down so that the video runs at the same speed.
We run the 35mm at 240 pulses (24fps) with 29.97ndf, 48k no pull.

The project I was working on this week, which ironically had to be stopped because the composer who had 6 months to score still didn't have reels 3, 4 and 5 done, was mixed to the print and video.
35mm on the big screen, video on the TV, and internal video on a PT and they all changed at the same cut.

The worst thing that has ever happened is this.
The assistant editor didn't apply the pulldown to the Beta output. No one knew until we put the print up.
So there was drift.
So we had to change to pulldown.
We were able to change PT to 48kpulldown and adjust the Dialog, FX ect.

Worst case is that a little adjust ment, everything should sync up.

LEX

[shnurgle]

I suggest that you don't have to do film at 48k pulldown anymore.
But 48pulldown will sync to film differently than regular 48, so one must be right and the other wrong, correct? I don't see how you can "go either way" on this.

We run the 35mm at 240 pulses (24fps) with 29.97ndf, 48k no pull.
So, yes, it is possible to put film onto 29.97 video so that it plays at 24fps, not 23.98 fps. In which case a sample rate pull down would not be nec during recording becasue you are allready at 24fps and the film will be 24fps in the theaters, so there will be no "bumping up" during this process. I thought that there was no way to keep film at 24 on video. I thought you had to slow it down to 23.98 to accomodate video. But you're saying this is false and that you can in fact receive video where the film is running at 24 fps, not 23.98 fps.

35mm on the big screen, video on the TV, and internal video on a PT and they all changed at the same cut.
Because the film was digitzied at 24fps and not slowed to 23.98 for the video transfer right? Had the video been slowed it would be playing at 23.98 and would not have synced with the film plying at 24. So all this supports what I think you're telling me that film can be transeferred to video and keep its 24fps, doesn't HAVE to be slowed down to accomodate NTSC video.

[LEX]

But 48pulldown will sync to film differently than regular 48, so one must be right and the other wrong, correct? I don't see how you can "go either way" on this.
It depends on the video transfer. I don't exactly know what is done, but I have done them both ways. You can go either way.


So, yes, it is possible to put film onto 29.97 video so that it plays at 24fps, not 23.98 fps. In which case a sample rate pull down would not be nec during recording becasue you are allready at 24fps and the film will be 24fps in the theaters, so there will be no "bumping up" during this process. I thought that there was no way to keep film at 24 on video. I thought you had to slow it down to 23.98 to accomodate video. But you're saying this is false and that you can in fact receive video where the film is running at 24 fps, not 23.98 fps.
Yes this is true. Again, I don't completely know the process. Now this isn't going to make sense, but it is the best I can describe it.
IT is like you are at 30, 48k, but when you pull it down you are at 29.97,47952.
The difference over 1 hour is 3 seconds and 18 frames.
You should really worry less about what 24/23.97 in theaters. It will still be 24fps in theaters.
It is all done when Dolby does the optical track to the film. Yes, all the sound is on the film.
4 different formats. Dolby Digital, Dolby SR 2 track, SDDS and DTS. Audio is on the film print.
Because the film was digitzied at 24fps and not slowed to 23.98 for the video transfer right? Had the video been slowed it would be playing at 23.98 and would not have synced with the film plying at 24. So all this supports what I think you're telling me that film can be transeferred to video and keep its 24fps, doesn't HAVE to be slowed down to accomodate NTSC video.
Yes, but there is something they do on the video side, but I am not sure.
I think it is that everything is slowed but the video. I'd guess that in the transfer, the video is running at a pull up speed when recorded.
Then when played back at normal speed it will sync to the film.

Trust me it works both ways. The last step of a film mix is a printmaster. Dolby comes in to insure sync, 2 pops and 2 track levels.
They record it to their MO drive in the DMU and that is taken to be laid back on the film.
I have printmastered films at 48k, 29.97 and 48k pulldown, 29.97.

It works

LEX

[shnurgle]

Thanks again for the help.
I guess it all really just comes down to every single time you score a picture you need to have a chat with the sound editor, and go down the list of questions and get all the specs you need. I guess i was just hoping for some hard and fast rules about this sync mumbo jumbo, but if this thread's taught me anything it's that there are no rules for this stuff, every time is different, and people believe differently about it.

I just wish there was a standard way of doing this stuff so the composer could focus on writing music instead of spending 90% of his time trying to crack the astro-physics code, but hey what are you gonna do? I mean it just makes sense that if the technology EXISTS so that we don't have to pull down sample rates when we record scores, why isn' it used EVERY time? Riddle of the sphynx I guess.

I'm scoring a video segment at 29.97 drop frame. I dump the QT video into logic, line up my start points, set the fps in logic accordingly, and the burned in time code on the video dosn't sync to logic's counter...

[Crossingsound]

I'm scoring a video segment at 29.97 drop frame. I dump the QT video into logic, line up my start points, set the fps in logic accordingly, and the burned in time code on the video dosn't sync to logic's counter...[/QUOTE]

did you spot the video to the same time code location in logic? meaning did you place the video in the appropriate address within logic. So for instance if the video burn says 1 hours 10 minutes 30 seconds and 22 frames, did you sync that specific point in the window to that exact address within logic? I wish I could tell you how to do it in logic, but in pro tools, I zoom in, and create a sync point, write down the address while zooming in really close, then I spot (place) that sync piont to that exact time address location in pro tools, then the counters are in sync. does that make sence? or are you experiencing drift when you have the correct starating location..

[LEX]

It really isn't rocket science.

Just ask 3 things:

1) frame rate - 29.97ndf, df ect
2) sample rate - 48k straight or pulldown
3) delivery format - ProTools, OMF and 16bit or 24 bit.

You should also see how many tracks they can handle for your stems. 16, 24, 32 ect.

AS far as your 29DF session, doubt check the burnin. I get vids sometime that say NDF and are DF.

To check, every other minute 2 frames are dropped.
So go to every minute in the video and look for frame 29 at the end of the whole minute and nudge to the next frame.

IF the next frame is 01, then it is drop frame.

Also another quick way to check, but you still need to do the above it this:
NDF timecode is separated by colon's..
1:00:00:00
DF timecode is separated by semi colons..
1:00:00;00

Check that Logic has the ;, and the burnin has the ;.
Then check the frames at the end of the minutes.

LEX

[shnurgle]

did you spot the video to the same time code location in logic? meaning did you place the video in the appropriate address within logic. So for instance if the video burn says 1 hours 10 minutes 30 seconds and 22 frames, did you sync that specific point in the window to that exact address within logic? I wish I could tell you how to do it in logic, but in pro tools, I zoom in, and create a sync point, write down the address while zooming in really close, then I spot (place) that sync piont to that exact time address location in pro tools, then the counters are in sync. does that make sence? or are you experiencing drift when you have the correct starating location
Not sure if it's drift or not, the clip is only like 2 mins long so it's hard to tell, but the issue might be the start points. The movie start has 5 sec of preroll which is burned as negative timecode (-00:00:05:00). Logic doesn't let you set to negative time code, so I have to set the movie start at 00:59:55:00 to get the preroll. IS this problem? With drop frame would a five secodn countdown to 01:00:00:00 be dif than the countdown to 00:00:00:00? In other words maybe the video and Logic are dropping frames in dif spots because of this. I wish I could set logic to start at a negative timecode to match the burn on the video.

Lex, thanks for your tips man, I'll double check that the burn-in is dropping frames every two minutes and I'll check for th semi-colons, although I reset logic to 29.97 non drop and the counters were out of sync as well.
-S

[PaulR]

There's no such thing as 'negative' timecode Miles.

[shnurgle]

There's no such thing as 'negative' timecode Miles.
Hey Paul, nevertheless, the burned timecode on the video counts down the 5 sec preroll from -00:00:05:00 to 00:00:00:00, then counts te time forward in the typical positive fashion. I was a bit surprised to see this too, as I assumed the preroll would count from 00:59:55:00 to 01:00:00:00.

[PaulR]

Hey Paul, nevertheless, the burned timecode on the video counts down the 5 sec preroll from -00:00:05:00 to 00:00:00:00, then counts te time forward in the typical positive fashion. I was a bit surprised to see this too, as I assumed the preroll would count from 00:59:55:00 to 01:00:00:00.

I think a picture tells a thousand words Miles. You just may want to download that video mentioned before. It's a good starting point.

[Jeff Hayat]

Ok, I am back....

It really isn't rocket science.

Just ask 3 things:

1) frame rate - 29.97ndf, df ect
2) sample rate - 48k straight or pulldown
3) delivery format - ProTools, OMF and 16bit or 24 bit.


Good advice.

There's no such thing as 'negative' timecode Miles.

True - if you were to roll TC backwards from 00:00:00:00, the next frame would be 23:59:59:29 - NOT -00:00:00:01.

However - some visual counters may count this way to "make it easier for the user" - it is probably that they are assuming that the user may not know how TC really works, and they fear the user, upon seeing TC go fro 0hrs, > 23hrs, may think the display is not functioning correctly.

Ok - here's the deal: for film projects - films shot on film @ 24fps - you should be working at 29.97/44.056 (or 47.952). As stated earlier, this allows for an easy bump-up to go back to film speed. If you do work at a hard 44.1 (or 48), it can still be bumped up, but not as easily - depending on the post house/lab/equipment being used. Which brings me back to Lex's post - always do research and ask first. For video and tv projects - you should be working at 29.97/44.1 (or 48). The reason for this is, for video and TV, there is no bump-up concern, as it is not going back to film - regardless of it being shot on film at film speed, or shot on video at video speed.

I saw something briefly up there^ about Lex and myself needing to come to an agreement on something? Did we disagree on something? Hmmmm....

Cheers. :)

[shnurgle]

I saw something briefly up there^ about Lex and myself needing to come to an agreement on something? Did we disagree on something?
Yes. On the chief subject of the thread-whether or not it is industry standard to pull down sample rates when recording scores for feature films that are being transfered back to 24 fps film for theatrical presentation. You both expresed opposite views.

[Jeff Hayat]

Yes, it is standard to pull down sample rates when recording scores for feature films that are being transfered back to 24 fps film for theatrical presentation. Not always, but standard. Note that it is not "pulling down the sample rate" once the recording is done, but recording at a pulled down sample rate, and then pulling up later on. Also note, as stated earlier, this does not apply to tv/video work.

Cheers.

[LEX]

Ok - here's the deal: for film projects - films shot on film @ 24fps - you should be working at 29.97/44.056 (or 47.952). As stated earlier, this allows for an easy bump-up to go back to film speed. If you do work at a hard 44.1 (or 48), it can still be bumped up, but not as easily - depending on the post house/lab/equipment being used. Which brings me back to Lex's post - always do research and ask first. For video and tv projects - you should be working at 29.97/44.1 (or 48). The reason for this is, for video and TV, there is no bump-up concern, as it is not going back to film - regardless of it being shot on film at film speed, or shot on video at video speed.

I saw something briefly up there^ about Lex and myself needing to come to an agreement on something? Did we disagree on something? Hmmmm....

Cheers. :)
I think he means about pulldown/no pulldown.


AS all the post houses I deal with here in LA can do any convertion, 48PD, 48k hard to film.

It has become more common to keep everything at 48k hard then it was 5 years ago.
The die hard places like Todd AO are still doing 48k pulldown.

WE do both and both sync fine.

Be careful though. Even TV shows are being done at 48kpulldown. 24 is like that.
So the standard has become loose.

LEX

[shnurgle]

Thanks again guys for your insight on this. A very informative thread on a topic that seems to be little known by most.
Cheers :)