HD at 24p results in serious sound issues!

[Gary Adcock]

[Tarka the Otter] “shot on HD at 24p”

What kind of HD? HDcam, Varicam- other? [Tarka the Otter] “she was only given the option of importing at 23.98fps, not at 24fps.”

correct – HD cameras need to be specially setup to shoot at 24.0 fps, 23.98 is the default

[Tarka the Otter] “Regardless of this, editing is finished now and went smoothly, her problem has arisen from her sound dub, where she gave the sound mix on a Hi8 tape to a Sound House in Athens and they returned Optical Sound (i.e. printed on film).”

she delivered her sound on a 29.97 video tape???? – she should have delivered a sound master in HD or as audio only on disk.

[Tarka the Otter] “It seems the sound house has returned the sound at a speed that would be parallel to 24fps, therefore slightly faster than the video (2% faster, in fact). Roughly 4 minutes into the film, sync is noticably lost. “
nope they worked with what she gave them- she is in error for not making sure that the timing issues were noted.THe post house was expecting that what they got was gold( ready to go) , when what she gave them was not.

[Tarka the Otter] “as some sources have led me to believe that HD at 24p is NOT actually 24p at all but 23.98fps… and yet MORE confusing si that 23.98 is again just another convenient shorthand for 23.975… argh!”

Actually it is 23.976 and YES that is correct – because of SD legacy- Most Film outs are done from 23.98 masters NOT 24fps version. the audio sync may in part be because she told the sound house that it was 24.0 when the Hi 8 tape is 29.97 ( and not a good choice for audio post since it was a digital to analog conversion on her end to get the audio onto that format.

gary adcock
Studio37
HD & Film Consultation
Post and Production Workflows

My DVD’s are available @
https://www.rastervector.com/dvd/dvd.html

[Annaël Beauchemin]

there was a very clear article about 24.0 fps vs 23.98 on 24p.com a while ago but I can’t find it anymore…

anyway, if there is a 2% difference between the video and the audio, it’s something else than the difference between 24.0 and 23.98 (23.976). The 24.0 and 23.98 difference is 0.1 %.

In general, you should always work in 23.98 EXCEPT if you are dealing with film and separately recorded audio. NTSC telecine and HDCAM transfer of film work in 23.98. You should also always output your movie to 23.98 if it’s going to print to film. If your audio is sync in your 23.98 video master, it will stay sync in the final 24.0 fps film transfer. If you need to send your audio and video separately, you’ll need to ask the transfer facility what to give them.

I don’t know why a Hi8 tape was used… MiniDV would be much, MUCH more reliable as far as audio quality and sync goes.

The 2% difference sounds like some sort of PAL 24@25 conversion problem. Which format did she send her audio mix in? PAL tape or NTSC with pulldown? (I understand Greece is probably a PAL country)

[Michael Gissing]

Firstly, Hi8 tapes are used to record the DA-88 multitrack PCM format and is a standard way to deliver a 5.1 plus stereo Lt Rt mix for optical sound.

Secondly a sync drift that is noticed after 4 minutes is neither the .1% drift of 24 to 23.976 nor 24 to 25 frames per second. That drift is 4% and would be noticed within two seconds.

My guess is that there was not a proper relationship between timecode on tape and the digital wordclock signal. This leads to drift between the digital audio and the timecode. Imagine if you didn’t have timecode synced to video. Same deal with digital audio.

This problem always happens when a DA-88 is on internal sync and the audio is fed in via analogue. Feeding in a digital signal with the DA-88 locking to the external AES audio would solve this.

[Tarka The otter]

Hi guys,

It’s been resolved happily, she did as you suggested and the sound was squashed slightly (by 19frames, in the end) and a new print was done and it’s apparantly all fine and in sync. Thanks very much for your help, especially all the stuff about 24/23.98 stuff which I found very interesting!

I apologise as well for a mistake I madetyping up the problem I had – at one point I said that the sound was 2% faster than the video, which some people flagged up as being a whole other problem. My bad, I don’t know why I wrote that, it was 0.084% faster (or 0.02fps, which is probably sort of where my random typo originated) so sorry for that.

Cheers again everyone!