Re: Sound Devices 744T timecode questions
- From: Matt Mayer <Matt@xxxxxxxxxxxxxxx>
- Date: Wed, 10 Dec 2008 09:54:43 -0600
Nayeli wrote:
Hi all,
I have a question about the relationship between the frame rate and
the sample rate in the 744T (and in general, actually). I'm a post
person, and am trying to find the best post workflow, but have some
questions regarding the sound recorder the production is going to use
so I can give recommendations based on what we're going to do with
that audio later.
As far as I understand, if you record audio at straight 48K, and have
24fps timecode, you have audio that will sync perfectly with 24fps
video or film. One second of time code will be real time second. So
far, so good. We've actually even successfully had audio recorded
with this frame rate and sample rate combination and synched it to a
23.98 reverse TK of MOS TKed 35mm film, relying on the fact that Final
Cut applies an automatic pulldown to audio imported under a 23.98 or
29.97 sequence preset when the file format is WAV. I'm curious about
the other options the recorder offers, however.
The 744T manual says if you are going to be synching with 23.98 or
29.97 video and require audio pulldown, what you can do is select
48048kF (fostex, faux, etc.), which will be audio that is recorded at
48048Hz but stamped at 48kHz, and this tricks an editing system that
doesn't recognize a 48048 sample rate into playing the sound back with
pulldown already "built in." In this case the frame rate is limited
to 30 ND. (I have yet to do a test with Final Cut Pro to see if you
can avoid a double pulldown in this case, because of the
aforementioned auto pulldown.) In any case, this option makes sense
to me.
My question is, what happens if you record audio at 48K, but choose
the 23.98 or 29.97 frame rates? (If it's possible to choose this
combination). Do you also get a pulldown? Because the manual says you
can choose a 23.98 frame rate to sync audio recorded with the 744T to
video recorded with HD cameras. Can you choose this being at a sample
rate of 48K? If so, isn't that a contradiction? If 48K is really 48K
per second, then doesn't it necessarily match up with 30 or 24fps
video ONLY? Because with a frame rate of 23.98 a second of video
timecode is .1% slower than a realtime second! Does the Sound Devices
recorder add a pulldown to your audio if you choose a fractional frame
rate, and somehow compensate the sample rate, perhaps still stamping
it as 48k but with it really being 48048Hz?
And does that choice end up being the same as choosing a 24fps frame
rate with 48048kHz sample rate (in which case you end up with files
that are stamped 48048kHz, unlike the Fostex option, so that not all
software will be able to play them back)?
I find this confusing because of how I understand timecode: that
there really is no such thing as 29.97 or 23.98 timecode, since you
can't have a fraction of a frame. Rather, you're dealing with a
numbering system that is base 24 or base 30, but that plays .1% slower
to get the rate of 29.97 or 23.98. I think the key with the Sound
Devices is the menu option is FRAME RATE vs. timecode format. What
I'm confused about is how this interacts with the sample rate when you
choose different combinations of the two.
Has anyone done tests to find out what happens with each combination?
I want to do them myself, also to see how they interact with Final
Cut, but have limited access to the recorder at this time.
Thanks to anyone who can shed some light on this! I hope I've been
clear enough about what I'm confused about....
Nayeli
Wow, there's a lot of different questions in there. I will try to cover them as best I can. Wolf's books are an excellent starting point.
First we need to differentiate between video and film that has been telecined to video. If a project is straight video (HD or SD), there is no need for any kind of pull down. You can use 48k and whatever TC frame rate the video is using. The thing to know is that Panasonic HD uses 29.97TC externally to the camera, Sony HD uses 23.976. Standard Def video will always use 29.97.
If the project uses film that has been telecined to video, there most likely will be pull down to make the film fit a video paradigm. In the US, it is still common practice to use 30ND for film audio recording. In the transfer process, the audio is slowed .1% -- the problem being is that this makes 48k audio 47.952 audio and makes a direct digital-digital transfer impossible. This is where 48.048 comes in. 48.048 played back at 48 has the .1% slowdown built in and allows a digital-digital transfer.
I was not aware that FinalCut would apply a pull down to audio imported in 23.98 and 29.97 sessions. If that is the case, obviously, you would not want to use the 48.048F mode at all.
48.048F and 47.952F are the only two sample rate setting in the 7 series machines that force a specific frame rate. All other sample rates allow the use of any available frame rates. These are also the only two modes that stamp the file metadata with something other than the actual record sample rate, effectively pulling down the audio when playing back at 48k.
You are correct that there is no such thing as a fraction of a frame. The .03 of a frame doesn't just disappear, it carries over into the next second of time. Because of this, 29.97 time code will quickly get out of sync with 30ND code. At the end of one minute of video, the offset would be 1.8 frames (60 seconds x .03 frames). Drop frame Time code compensates for that, but causes it's own batch of problems.
The bottom line is that the best workflow for you depends on:
1)the original record format (Film, HD-Sony or Panasonic, SD)
2) the transfer process (Telecine--HD intermediate or SD intermediate, Direct computer ingest)
3) The audio equipment used to transfer the audio to the Edit system (Fostex, Aaton, Direct ingest)
4) the editing platform used (Avid--Which platform and software revision, FinalCut, etc)
5) the output format (Film, HD, SD)
Each one of these steps has the potential to "force" you into a specific workflow to conform to the needs of the system chosen.
Hope this helps,
---Matt
.
- Follow-Ups:
- Re: Sound Devices 744T timecode questions
- From: Peter
- Re: Sound Devices 744T timecode questions
- References:
- Sound Devices 744T timecode questions
- From: Nayeli
- Sound Devices 744T timecode questions
- Prev by Date: Re: New Sanken COS-11xR
- Next by Date: Re: Kata Bag Question
- Previous by thread: Re: Sound Devices 744T timecode questions
- Next by thread: Re: Sound Devices 744T timecode questions
- Index(es):
Relevant Pages
|
