Jump to content

48.048 for the first time


Guest Eric Lamontagne

Recommended Posts

Guest Eric Lamontagne

Yes, I'm new at this so I would like to ask the group...

I'm just starting a small low budget 16mm doc on the FR2 and considering 29.97 48k with the pull up option enabled. I've been told this renders 30ndf at 48.048khz. On import into Final Cut, I'm hoping this works out to a slow down of speed to the NLE timeline. Film shot 24fps will be transfered MOS to 29.97ndf video, imported and then sunk up in FCsP (final cut semi-pro!). Film will then be cut manually from the EDL produced by Final Cut. This film cut will be delivered to a transfer house that is requesting 29.97ndf 48k audio to sync with it. This is where things are weird for me as 29.97ndf audio would have to be then sped back up to sync with the original film. Of course I have to be the post supervisor as well on this because it's low/no budget.Please send me any thoughts you have on this workflow as this is my first time with 48.048.

Eric Lamontagne

Link to comment
Share on other sites

My brain hurts.  Are they doing this so they can have an HD delivery while avoiding the cost of an HD transfer with the dailies?  The transfer house is probably asking for a FINAL SOUND EDIT at 29.97/48 to sync with their 29.97 HD output from the transfer of the negative cut.  My guess is that if there is some sort of magic voodoo, they would want 30/48048 final sound to "transfer" with the 24FPS film to create their video master.  But since I don't think that's possible, they will probably sync it after the transfer.

This begs the questions... are they manually syncing all of your audio in FCP to edit?  Are you using a 16mm camera with TC input?  Are you TC slating?  Why not transfer audio with dailies?  Who is taking care of the sound edit?  What do they suggest?

If I were producing a low-budget project, I would want a workflow that everyone old and new knows and every system can handle... 30/48/16bit sound and HD telecine, either via tape or direct HD storage to import to FCPHD.  Then it could all be done quickly and with less staff and room for error.  I can't imagine this would work out costing more in the end.

Robert

Link to comment
Share on other sites

Guest Eric Lamontagne

timecode slate would run 30ndf regardless.

transfer MOS to save costs, sync audio in FCP at 29.97ndf (I assume this will include slow down). OMF to sound edit in ProTools. deliver audio in 29.97ndf 48k is the request for online to film 24fps. No HD

they are skipping a step at the transfer house and it's me trying to determine 48k or 48.048k. Post audio isn't onboard yet

Link to comment
Share on other sites

I don't get why they want 29.97 if the final output is film.  If there are empty holes in the workflow staff, then I would strongly recommend using a traditional approach during you production recording.  30/48.  It won't be wrong, even if it turns out eventually it wasn't exactly what they wanted.  There are lots of people to go through before this mystery delivery request at the end of the workflow.  Let the non-staffed positions in the middle do their jobs when the time comes.  Good luck, and have fun.

Robert

Link to comment
Share on other sites

This workflow Shooting 24 frame 16 MM but running sound at 29.97 and a +.1%pull up on the Fostex means the sound is stamped as 29.97ND 48k but actually outputting 48048hz files at 30fps timecode.  (your slate should be set to 30.00FPS ND)  This allows them to bring the files into FCP at 29.97 and line them up with MOS film that was slowed down .1% in telecine.  They don't have to re-sample the sound since it was recorded at the faster speed but marked at 48k.  (it will playback at .1%  slowed down rate when played normally on a 48k timeline).  They can do all their post audio at 48k syncing to the 29.97 video of the cut film.

Then they pull up at the end if going back to film.  But only if they are going back to a filmout for release.  If going straight to DVD it will probably stay at 29.97 or be converted to 23.98 for HD.

----Courtney

Link to comment
Share on other sites

Guest Eric Lamontagne

Thanks Courtney!

This was my understanding of the process.

The Producer/Director has been made aware of my question re speed and will inform if it would be advisable for post to use a slow down on set. Tomorrow will be 30f/48k for safety sake, probably moving to 29.97ndf/48.048k after tomorrow.

Eric Lamontagne

Link to comment
Share on other sites

Switching sample rates from pulled up (F mode) to normal in mid production is quite confusing. Since both of the files probably look the same internally.  Especially on an FR2.  One requires pull down and the other dosen't.  Confusing in post.

---Courtney

Link to comment
Share on other sites

Robert is 100% right above. If it's a no/lo-budget, my advice would be to keep it simple: 30NDF 48K, camera rolling at 24.00fps.

In post, film runs at a .1% pulldown to 23.976 (in HD or 525), and sound pulls down to 29.97NDF and 48.952K. Literally 90% of everything we do is done this way. The post house is insane to request 29.97NDF timecode and 48k audio, unless (for some very unusual reason) the film camera is running at 23.98fps, which I have almost never seen happen except by mistake.

My feeling is that SRC is so good nowadays, I'm not sure I see the point of doing 48.048 anymore. The potential risk of things going wrong increase exponentially with 48.048 audio projects. On the other hand, if everybody's on the same page, 48.048 can work fine. I know shows that have used this system for years, and (after the inevitable first-day-of-shooting craziness), they go very smoothly.

BTW, to my knowledge, neither Evertz Tracker nor Aaton Keylink* can capture 24fps audio timecode yet, and the visible audio TC inserted in the picture is unstable (numbers change erratically). There are workarounds, but they're clunky and somewhat time-consuming. We're told that they will be demoing fixes for this at NAB next month, so maybe -- finally -- 24fps audio TC will be a simple, easy option for shows shooting on film. 

--Marc W.

* not 100% certain about Aaton, but this is what I've been told.

Link to comment
Share on other sites

Guest Eric Lamontagne

Thanks for the info guys. I'm glad to report a 30/48k (no fuss no muss) day has gone by with only three 11min takes in the can. An easy SRC if I do say so myself. Just like every other film shoot I've ever done.

Eric

Link to comment
Share on other sites

Guest Eric Lamontagne

Phillip,

Nope! I switched to 29.97ndf/48.048k (rendering practical 30ndf on the set and a slow down to 29.97ndf when importing into Final Cut) after the first day. I will manually apply the slow down the first day's work in PT7.4 (only three long takes). In fact I'm already done the first day and delivered the slowed down audio. I am archiving the original audio until everything is completed for safety.

I have spoken with the online house and they are expecting slowed down audio at 29.97ndf 48k. Offline edit will be in a 29.97ndf environment as well, this is where audio will be sunk to picture. Post audio will also be in 29.97ndf from an OMF output from Final Cut.

I AM the post supervisor on this one!

Thanks for all you input, an invaluable resource this board is!

Eric Lamontagne

Link to comment
Share on other sites

  Just starting a movie next week and everyone has agreed to the 48k/30ND.  This is a heavy music movie and there's tons of playback with live vocals and syncing long shots, and still everyone has decided on 48k, not 48.048k.  I guess it's okay to have to do a pull-down for the sound mix or something.  It's cool because the assistant editor has the machine to import the Deva files, and will be performing the pull-down there.  So by the time the post sound mixers (who are old-school pros) get it all the files will already be conformed to their specs.  I guess.  I didn't really have much to add to the discussion other than "I'll record whatever works best for you guys!"

  For the past 5 film jobs I've had, everyone wanted 48.048k (one guy didn't know so I did 48.048k and he said it synced perfectly) so this will be a switch.  (Plus one guy said I should have used 48.048 when I used 48 a while ago.)

  Now, I was using a PD6 on those jobs, so for all I know "30ND, 48K + .1%" meant 29.97 pulled-up to 30 and 48k pulled-up to 48.048k?  (I know the FR2 doesn't work like that.)

  It's confusing to me that there can be a difference between the frame rate that actually exists and the frame rate "stamp."  Why can't these decks just report what they're recording?!!

  Dan Izen

Link to comment
Share on other sites

Guest Eric Lamontagne

Izen Ears,

I agree. The Fostex method for pull up/down is possibly confusing because it also messes with the frame rate. If I set my FR-2TC recorder to 29.97ndf and 48k with the +0.1% enabled, I get 30ndf, 48.048. This then imports into FCP or Avid as 29.97ndf/48k with a slowdown.

Whats great about the timecode being affected by the +0.1% is that I can compare the recorder's timecode output with my settings. A real failsafe.

Once I wrapped my head around this it all makes sense though. I think it's up to us the recordists to know our own gear and know what it's gonna deliver at the end of the day and why this is needed. Reading the manual is not an optional step for us, we MUST do it and test our knowledge! Assume nothing. If we are on our game we can answer our own questions often before post audio is hired! Plus, there is a reason why we have these features built into our recorders - otherwise anyone with a R-09 could walk in and under bid us! Hehe

Good Luck!

Eric Lamontagne

Link to comment
Share on other sites

I agree. The Fostex method for pull up/down is possibly confusing because it also messes with the frame rate. If I set my FR-2TC recorder to 29.97ndf and 48k with the +0.1% enabled, I get 30ndf, 48.048. This then imports into FCP or Avid as 29.97ndf/48k with a slowdown.

  And why on Earth would Fostex make the FR2 and PD6 different in an already-confusing feature?  I noticed the FR2 does not lose timecode when it powers down (thanks to the menu option), but the PD6 ALWAYS does.  If the PD6 has to reboot or shut off I must re-jam slates, not so with the FR2.

  So once I tried to run the FR2 as both a backup of tracks 1 & 2, as well as a constant timecode jam.  I had both machines at 30ND @ 48k + .1%.  Guess what?  The PD6 declared it was getting a timecode of 29.97 from the FR2!  Apparently I should have set the FR2 like you did, 29.97ND @48k + .1%.  Maybe then it would have output the 30ND.

  Why does the PD6 only + 1% the sample rate, and the FR2 + 1% both the sample and frame rate?  I didn't test it I just gave up, thankfully the PD6 held up...

Whats great about the timecode being affected by the +0.1% is that I can compare the recorder's timecode output with my settings. A real failsafe.

  How do you do this exactly?  Set the slate to 29.97 and see if it gets a code 30?  Or set the slate to 30 and see if there's no tc error?

  Dan Izen

Link to comment
Share on other sites

Guest Eric Lamontagne

Whats great about the timecode being affected by the +0.1% is that I can compare the recorder's timecode output with my settings. A real failsafe.

  How do you do this exactly?  Set the slate to 29.97 and see if it gets a code 30?  Or set the slate to 30 and see if there's no tc error?

  Dan Izen

Well, I carry an Ambient 301 slate that compares incoming timecode to it's own internal settings. It reads 30ndf from the FR-2 in 29.97ndf +0.1%. Ambient 501, Ambient Master Controller and Denecke GR-1 also do the same comparison. Then I know that everything is right. Correct setting for the slate is actually 30ndf in my case.

Check that your PD-6 actually does what you think it does when in +0.1%.  Read the timecode output with a timecode reader/slate.

READING THE PD6 MANUAL SUPPLEMENT: V220e  pg13....

"With the earlier version, the TC frame rate always changes in sync with +/- 0.1% speed

change. With the new version, the TC frame rate does not always change, as shown in the

table below."

Anyone out there with a PD606 to check this also?

Eric Lamontagne

Link to comment
Share on other sites

All Fostex recorders (FR-2,PD6, PD606, PD204, DV824) behave the same way, i.e.- when the Pull Up/Down switch is engaged, both the sample rate and the TC rate are affected by .01%. **Except where doing this would result in a non-standard framerate**

/snip

On all Fostex machines:

29.97ND/48K +.01% = 30ND/48.048 Output (48K Stamp)

30ND/48K -.01% = 29.97ND/47.952 Output (48K Stamp)

Mike

  Hmm, I must not have the latest update for the FR-2, since it was definitely NOT outputting 30ND when I had the + 1% set.  The file WAS 30ND, but the TC out was 29.97.  This was confirmed by the post house to my extreme embarrassment.

  Your statement above seems to hold true with my PD6, but the FR2 was another story.

  So for the future I guess setting the FR2 to 29.97/48k and then +1% will give me what I want.  I'll have to test it though since I just got my Deva 5.8 I doubt I'll ever use it again!

  Dan Izen

Link to comment
Share on other sites

  • 4 weeks later...
Guest Eric Lamontagne

Report back from post, all is well! The pull up functioned as expected and resulted in matching the slow down during film transfer. I have since worked another sound job with the same 48.048 to good result.

I will check out 48.048 my PD606 when I get a chance sometime this week.

Eric Lamontagne

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...