Question about PAL tape-to-film and audio speed

[RicardoEscallon]

I am about to start a film project originated in DV-PAL. They are going to tape-to-film it, so frame rate is going to be 4% slower in the film version.

I prefer to have first a telecine of the tape-to-film negative, but the producers of this movie tell me that the people in charge of the tape-to-film will take care of slowing down the audio.

How can be this done in a laboratory?

Can the Dolby MO be slowed down? Or they take the PrintMaster before the codification, correct audio speed and then do the Dolby encoding?

Thank you,

Ricardo

[cpi]

A good solution is to do a tape-to-film, making sure the film is going at 24 not 25 fps. Then telecine back to video to get a 24/25 (film speed, ebu tc) vid reference for all your audio work. When you master the Dolby MOD, the biphase feeding the ds10 (mod encoder unit) will be set at an interval of 24 (24x2) - sound will be recorded without pitch or other change, will match your film, and everyone will hear and see your film in the cinema as you made it.

If your film will be 25 fps, all your audio work should be done at 25 against the vid ref. The ds10 will be fed with something like a 25x2. When played in chase it will sound fine, when free you will hear the pitch change that will be on your film. Most film productions accept this as it saves in costs. One alternative is pitching up your programme pre ds10, understanding that it will pitched down when transferred to optical.

Folks in the US can't get away with this sort of stuff. NTSC to film speed difference is too great. But at least there is all this pull-up pull-down fun.

Hope this helps.

[Branko]

It's essential to understand that all video-to-film systems (film recorders) transfer frame to frame, which means burning 25 film frames for every second of video material. If such a print is screened at 25 fps, your sound will be in sync with it.
What we do in our studio is: mix at 25 fps, using video master or 35 mm print running at 25 fps, bounce (re-record) the LtRt to disc, then pitch up the LtRt for 4.1% and record the print master at 25 fps, too. DS 10 will follow the projector's speed, and the pitch will be right when screened at 24.
It is possible that a lab can do this for you, but I'd always prefer to change the pitch before SR encoding and to listen to the processed audio at correct speed, once the processing is done, just to hear the results. Dolby Labs also prefer that the encoding is done in the studio, on calibrated monitors and that there's no change to Print Master after it reaches the transferring facility.
Of course, this applies to analog SR prints, if you are to finish in 6 track digital, you'll need 6 ch of high quality, phase coherent pitch shifting. And that's expensive!
Branko

p.s.

A good solution is to do a tape-to-film, making sure the film is going at 24 not 25 fps

cpi, how do you do this?

[RicardoEscallon]

This film is going to be finished in Dolby Digital 5.1.
Do you think that Pitch'n'time will do right the job of pitch shift /time correction on the 6 channels without phase problems, or are there any other tools I can use for this purpose?

Isn't it easier just to "pull down" the telecine video from 25 to 24 fps using a Film Composer or Cinema Tools, time expand the production sound as a first step and keep working at film speed ?

CPI: I don't understand what you mean, can you be more specific

A good solution is to do a tape-to-film, making sure the film is going at 24 not 25 fps.

Thanks,
Ricardo

[Branko]

Ricardo,
I would also prefer to work at the correct final speed, which means that time stretching is the way to go. You could load your dailies in Avid's film project and it will automatically pull down the picture to 24 fps. Audio should be timestretched, preferably re-recorded to new DAT tapes and then logged and digitized. The advantage is you can see the real result immediately.
As for Serato, it is a great tool for music, phase coherence is remarkable, yet it fails to deliver correct results for human voices. To me, they sound phasey and less present, compared to original. Funny thing is, I get much better results with Digi's Time Comp/Exp. plugin, when dealing with production dialogue track.
For high-quality pitch shifting, I use Harmo, made by Genesis. It's a 2-channell box, and more units can be linked for true multi-ch operation, but the cost is really high. Dolby is coming out with a new 8 ch box, so you'll be able to process both 6track and 2track at the same time. I wonder what the price will be....
Branko

[Frank Kruse]

Originally posted by RicardoEscallon:
I am about to start a film project originated in DV-PAL. They are going to tape-to-film it, so frame rate is going to be 4% slower in the film version.

I prefer to have first a telecine of the tape-to-film negative, but the producers of this movie tell me that the people in charge of the tape-to-film will take care of slowing down the audio.

How can be this done in a laboratory?

Can the Dolby MO be slowed down? Or they take the PrintMaster before the codification, correct audio speed and then do the Dolby encoding?

Thank you,

Ricardo

I would not time stretch because the sync error is very high. I would pitch the sound +4.166667% then transfer to dolby-mod. later when the film is played back in the cinemas it will be 4% slower but your sound will be in the right pitch. by this you avoid sync drift through stretching.

frank.

[RicardoEscallon]

Branko, have not you seen the sync problem that Frank tells us about, after time expansion?

I would not time stretch because the sync error is very high.

Frank, How bad can that be?

We can not forget that this is going to be mastered in Dolby Digital, so I may be wrong, but pitching up 4.166667% will not work with the digital version, right?

Ricardo

[Frank Kruse]

of course the trick has to be done *before* you master. for example. you have your final mix in stems (music, fx, dialog) or 5.1 on a MMR-8 or other dubber. pull it into PT and pitchshift (without changing length) all track with serrato (don´t´forget to select "multi-mode).

after that do your transfer to the dolby mod. of course you will hear the wrong pitch during mastering but later in the theater it will be 100% again because the projector will run 24fps, so 4% slower than your film was shot.

the super expensive boxes like the mentioned french system or the upcoming dolby box do exactly that but in real time except they cost about 5 to 10k euro per channel!!!

frank.

[Frank Kruse]

BTW, I just did a test comparing a TC M6000 and the described method with PnT 2.0 and the M6000 was unusable PnT was pretty descent!

frank.

[Frank Kruse]

the pitch shift option for the M600 is called VP8. here´s a quote from the tc-site:

"As a result of intelligent splicing, the algorithm may be used for demanding applications like real-time pitch change
of entire 5.1/7.1 stems when converting frame rates between video and film (25 to 24 fps.and vice versa).
Channels may be locked in various ways to preserve phase relationship while applying pitch hange.This feature is in high demand -and currently not available in 5.1 -for telecine,film and post processing.The up and down pitch range is +/-200 cents."


frank.