PAL to NTSC - best way to speed pitch up audio by 4%

[electrohed]

A friend needs to convert a 25 fps PAL movie to a 24fps NTSC. What's the best way to deal with the audio? I used to use TC spark to varispeed up by 4% (pitch and time) but now I'm not sure how to do this with Logic.

The audio is 50 mins long. A sample rate conversion in audacity resulted in a time drift over the course of the movie.

anyone out there know a reliable method?

ta

[Hugh Robjohns]

A friend needs to convert a 25 fps PAL movie to a 24fps NTSC.

Film runs at 24fps. NTSC runs at 30 or 29.97fps.
There is clearly some level of confusion here.

What's the best way to deal with the audio?

Depends very much on the equipment available.

The audio is 50 mins long. A sample rate conversion in audacity resulted in a time drift over the course of the movie.

It would -- sample rate conversion doesn't change the run time, whereas a simple change of frame rate implicitly does.

But without timecode or video referencing, sync drift is going to be a problem using any simple 'varispeed' approach.

hugh

[illegal colors]

Film runs at 24fps. NTSC runs at 30 or 29.97fps.
There is clearly some level of confusion here.

Yes, there is some confusion here, buddy.
Broadcast NTSC is something like 30fps.
MPEG2 encoded NTSC signal on DVDs can be approximately 30fps or 24fps.
Any Region-1 DVD player will convert 24 to 30 when necessary.
Most film-based releases these days are 24fps.

[electrohed]

Thanks for the thoughts so far.

OK... forgetting the whole PAL/NTSC thing for a moment...

what I understand is that the movie has been converted from 24 fps to 25 fps. The audio now sounds 4% lower in pitch. I'm looking for a solution to pitch it up by 4% while remaining in sync with the visuals.

I have Logic Pro, Ableton and Audacity at my disposal...

My thoughts are to go back to the original audio and varispeed it up by 4% (pitch and time) and then dub that on to the new video.

- otherwise I think my options are to use Ableton to pitch up the audio or get hold of Serato Pitch'n'Time.

ta

[Chevytraveller]

What format is the movie in?..

assuming it has been encoded to some form of MPEG then the audio should be embedded with the video.. therefore changing over to NTSC should be a simple matter of transcoding the file.
This removes the issue of doing the audio separately..

[Hugh Robjohns]

what I understand is that the movie has been converted from 24 fps to 25 fps. The audio now sounds 4% lower in pitch. I'm looking for a solution to pitch it up by 4% while remaining in sync with the visuals.

Speeding it up by 4% should be trivially simple using virtually any varispeed/pitchshift algorithm.

Keeping it in sync is impossible without some kind of timing reference from the original video suchh as timecode and/or sync video referencing.

My thoughts are to go back to the original audio and varispeed it up by 4% (pitch and time) and then dub that on to the new video.

That will work, but you'll probably find the need to perform lots of mini-edits to tighten up (or slack off) the sync every so often to compensate for the inevitable sync drift.

Hugh

[electrohed]

Thanks guys. Here's the problem defined more clearly

"Film can be shot at any speed, but most commonly it’s 24fps in the US. In Europe, film and video is generally shot at 25fps.

NTSC television is 29.97fps, but for films shot at 24fps, there is an established process for converting to 29.97fps so that they can be broadcast or displayed on standard NTSC televisions. (This is part of the MPEG encoding process for DVD)

I need to get a 24fps master from my 25fps so I can author a 24fps DVD. Rather than just dropping one frame every second, the recommended way to do this is to basically slow down the 25fps by 4% to 24fps. That way, you get a frame for frame conversion, without stuttering. The unfortunate thing is that everything plays back 4% slower. Most of the time, this isn’t a problem. In fact, a lot of commercial films just live with it. Consequently, running times for the same film in PAL vs NTSC regions are generally 4% shorter...fact. Anyway, I’ve already done the conform to 24fps and have my master.

However, for this project, I need the audio to sound as close to the original as possible, so I’m looking to pitch it up. Of course, the duration of the film will be 4% longer that the PAL original, but that’s fine. The audio just needs to have the original pitch.

I’ve read that a few people have done this through ProTools with good results. As I say, I tried pitch shifting in Audacity, but it somehow left me with a new piece of audio that drifted by a couple of seconds towards the end of a 12 min segment."

So, I actually need to slow the audio by 4% but keep the original pitch.

[Hugh Robjohns]

"Film can be shot at any speed, but most commonly it’s 24fps in the US. In Europe, film and video is generally shot at 25fps.

Film can be shot at any speed... as can video, in fact. It is the replay standards that define how it looks. Film for cinema release is always replayed at 24fps.

Film shot as the basis of a TV programme can be shot at 25fps to avoid having to convert to 25fps for PAL transmissions, but is usually shot at 24fps as the basis for NTSC programmes, followed by a transfer process often referred to as '3:2 pull down' to convert to 29.97fps for NTSC.

(This is part of the MPEG encoding process for DVD)

MPEG encoding is entirely separate to the issues of frame rate conversion, although all of these processes (and more) may well take place during DVD authoring.

I need to get a 24fps master from my 25fps so I can author a 24fps DVD.

Ah... so actually you are trying to slow the material down by 4%, NOT speed it up as you originally stated.

Rather than just dropping one frame every second, the recommended way to do this is to basically slow down the 25fps by 4% to 24fps.

Yes... obviously! Otherwise you'd have regular glitches in the soundtrack.

The audio just needs to have the original pitch.

I'm surprised this is necessary, but if so, use a (good) pitch shift algorithm and raise the pitch by 4%.

I tried pitch shifting in Audacity, but it somehow left me with a new piece of audio that drifted by a couple of seconds towards the end of a 12 min segment."

This is very likely. It's all very well setting the dials for 4%, but without common and synchronised clock rates, some degree of drift between pictures and sound is inevitable. A couple of seconds in 12 minutes would be much as I would expect.

As I've said before, you need to maintain synchronisation using common clock sources and/or timecode throughout the processing (which is the proper way to do it)... or accept the fact that you're going to have to perform numerous tuck edits to pull the sync back in before the drift becomes noticeable.

Hugh

[electrohed]

thanks again.

Do you consider the Logic time and pitch machine to be a 'good pitch shift algorithm' or should I look at buying Serato.

I'm currently using Ableton to slow down the original audio by 4% to see how it fits with the conformed video. If it drifts slightly I guess it can by tightened up using markers at appropriate cue points.

Has my friend missed out some usual part of the conversion process relating to audio during the conversion from 25 to 24fps?

[PaulD]

Hi
Cinema Tools from Final Cut Studio is used to conform frame rates - it converts the audio track's run time to remain in sync with the video.

To go from 25 to 23.976fps the audio's pitch is lowered 4% (and vice versa).
It does this in a single mouse click.