Jump to content

35mm to SR HDcam sync - .1% pull up is correct, right?


Izen Ears

Recommended Posts

  So it's a 35mm job, but differently posted:

  The lab (Fotokem) will be 1- pass tranferring the film to SR HD Cam tapes (23.976), which are then being thrown onto an Avid 48k/23.976 timeline, and THEN the sound will be imported into this timeline and synced.

  After some investigation, we're leaning toward recording the sound at 30 ND fps at 48.048, this way both the sound and picture will be similalry slowed down, and sync perfectly.  Right?

  Thanks!

  Dan Izen

Link to comment
Share on other sites

  Thank you Rich this is what this group's all about, yea!  This is huge and gives me a very important question to ask. If the version of Avid works to import the 48.048k, great, if not I'll use that cheat.  Thanks for the great great info.

  Dan Izen

Link to comment
Share on other sites

Dan,

Your post reads that the film will be transfered to tape, and THEN the editor will sync your sound in the Avid.  If that's the case, then shouldn't your sound be presented in the same format that will already be in the Avid (23.976)?

If you were giving your sound to the transfer house, then I would agree 30ND/48.048k - But if you are giving your sound to editorial to sync after transfer, then I think you should be giving them 23.976 or 29.97 at 48k.

Anyone else?

Robert

Link to comment
Share on other sites

Robert,

No.  Because the film will have been pulled down but the sound won't be.  For short takes most people won't notice the difference, but it wouldn't hold accurate sync for long.

Maybe 23.976/48.048k if it's the version of Avid that will ingest and play it back as if it were 48k.

Someone correct me if I'm suffering from the late night muddles.

John B.

Link to comment
Share on other sites

Yes, right.  Late night muddles for me.  I believe you are right, John.  So it must be imported into the Avid and then pulled down before syncing.

I spoke with someone in post recently, maybe it was Marc W., who believed that soon we will all be recording 30/48.048 for film.  He believed that all sessions will become 23.976.

Robert

Link to comment
Share on other sites

The newer Avids allow you to do a sample rate conversion of the audio upon import.

That being said, if you record at 48.048/30,and stamp at 48k.  No sample rate conversion is necessary.  The avid will play back the files as if they were recorded at 48 and this will pull them down the necessary amount to match the pull down being done to picture when it is pulled down to 23.976.  This pull down will also pull your 30ND T/C to 29.97.

Darren

Link to comment
Share on other sites

Dan,

Although this might seem stupid, perhaps you can run your two machines independently, for a day or two anyway.  Deva 5.8 and PD6 right?  Run the Deva 5.8 at 30ND/48048 and run the PD6 at 30ND/48 as the back-up.  Either one is ok really, but they will probably want the 48048 to pull down to 48k.  Use the grown-up version on the Deva, not the F version, since you will have the back-up going.  This way everything is stamped properly, etc.  The Deva is what they will probably want, and the failure rate on both of your machines is EXTREMELY low, so you'll be very safe either way.

Once you chat with editorial on day two or three, then you can run both machines at the rate they want.

Robert

Link to comment
Share on other sites

If you stamp at 48.048, you need to ensure that the avid being used has the option to ignore the stamp and still play it back at 48 in order to pull it down, and that the assistant editor knows how to accomplish this.  I wouldn't make the decision to stamp at 48.048 just because it is the "grown up version", whatever that means.

Link to comment
Share on other sites

If you stamp at 48.048, you need to ensure that the avid being used has the option to ignore the stamp and still play it back at 48 in order to pull it down, and that the assistant editor knows how to accomplish this.  I wouldn't make the decision to stamp at 48.048 just because it is the "grown up version", whatever that means.

I just meant to not "trick" an older version of the Avid when it might not be necessary.  I like the idea that if the file is 48048, then it is stamped 48048.  If it's not, then it's not. 

Link to comment
Share on other sites

  Yah I agree Robert, 48.048k should have the same stamp.  I actually proposed the idea of running two different sample rates on the two decks like you suggest, then we realized it might mess up the timecode numbers and end up with the timecode not being consistent.  Or something...  Anyhow I like my backup to be identical to the original... 

  Anyhow they're going to make sure it's the right version of Avid and it should be all fine...  (gulp)

  And by the way - the Deva DIED in the middle of a take the other day!!!!!!!!!!!!!  I heard a digital crunch, then the screen slowly went white, from the outside in like a wacky reverse "that's all folks" wipe...  Only lost the last line and it wasn't a take they liked anyway.  But I DID LOSE SOME SOUND when that thing died.  I power cycled (that's what Zax likes to call it) and it was fine for the rest of the day...

  Let me say it again: THE DEVA FROZE IN THE MIDDLE OF A TAKE AND I LOST A LINE OF DIALOG!!!  And I just updated to software version 5.440 the day before...

  Dan Izen

Link to comment
Share on other sites

I am by no means an AVID wizard, but it is my impression from talking to those who are,

that if the AVID Project is set up as a 24 Frame film project. (Not HD 24P which is 23.976 native and 48k audio)  The audio and video both play at the pulled down rate.  (audio runs at 47.952Khz and video is 23.976)  So if the audio was 24 fps TC (a rarity)  at 48K TC it should sync up if imported to the 47952 timeline and played back at the slower rate and match picture.  Or 30 Frame 48k will work as well although you will have to sync on the even second.  If I am wrong on this please some AVID wizard chime in and set me straight.  This may only work if AVID HD files are created during Telecine and no Tape is used as an intermediary and later digitized using some black-magic card. 

It is just too confusing now with Kludge applied to Kludge applied to Kludge.  After you get 3 Kludges deep nothing can be put forth as the "correct" way to do things.   The answer is always "It Depends".  And the things that the answer depends on are not known yet because they are based on future decisions.

----Courtney

Link to comment
Share on other sites

I am by no means an AVID wizard, but it is my impression from talking to those who are,

that if the AVID Project is set up as a 24 Frame film project. (Not HD 24P which is 23.976 native and 48k audio)  The audio and video both play at the pulled down rate.  (audio runs at 47.952Khz and video is 23.976)

That certainly has been the common method of pulling down.  But  I am currently working on a project that is being shot on film for a film theatrical release.  It was requested by editorial that we record at 48.048, which we are doing.  Apparently, so I'm told, this workflow is becoming more and more common when editing on the newest Avids at full HD resolutions.  Our workflow is to record at 48.048/30.  We stamp at 48.  Avid plays back the files at 48/29.97.  It all seems to work rather well and the telecine process and syncing the dailies in the Avid have gone smoothly.

I use Metacorder and Sound Devices recorders.  I'm not certain that this workflow is possible with B.R.  If anyone knows if it is, please post and let us know as I've had people ask me this but was not able to answer.

Best,

Darren

Link to comment
Share on other sites

After you get 3 Kludges deep nothing can be put forth as the "correct" way to do things.   The answer is always "It Depends".

I couldn't agree more!  There's definitely no longer *a* correct way of doing things.  There are so many different workflows now.

In all but very few cases, it's not our job to choose one, but to give them what they want.  That being said, its my feeling that it is also our job to understand what they want, and why they want it.  And also to ensure that what they "want" is indeed what they "need", and that it is indeed going to work for whomever requested it and everyone else that will be dealing with our sound files down the line.

It would be interesting to compile a list of all of the different workflows that people that are on this forum have used, and what type of projects (film, video, HD, etc) they were used on.  For example, I do know that some editors have been requesting 48.048 with t/c at 24.  I definitely find myself wondering... has anyone actually recorded this way?

Cheers!

Darren

Link to comment
Share on other sites

There are many ways to concoct a workflow. Personally, I am always against any method that involves recording off speed. 

48048  and marked as 48k is recording off speed and marking it as on speed.  So if you were to play that track back on any normal player it would always play slower than real time and give no evidence that what you were hearing was not a faithful playback of the original event. Most devices that record images and or sound are designed to record them and play them back in "real" time.  In other words 30 minutes of recording when played should last exactly 30 minutes.  Disregarding the exceptions of Special effects like slow motion and under-cranking,  I feel that you should record at standard speeds.   Manipulations to accommodate transfer of 24 frame Film to 29.97 fps Video later should be done at the time of the transformation.  i.e. if you slow down the film you should slow down the sound at the same point in the workflow.Then pulled back up at the same point if necessary for a film output.   That way camera original will always match your separate Recorded sound at any point in the workflow.

That future proofs your sound so that in the near future where we have a way of viewing 24FPS film in sync with the original double system sound without resorting to NTSC Video the two media will playback at the standard caputure rates and stay in sync.

-----Courtney

Link to comment
Share on other sites

Courtney,

You make a good point.  Your position that we should "future-proof" with accurate labeling and metadata is well worth consideration.  Despite the short "shelf life" of much that we do, you just never know when materials may be reused down the road.

But, while considering this, I am struck with a another thought.  Stand back... here it comes... 

In terms of speed accuracy I'm amused that we're concerned over something that, even off-speed (one-tenth-of-one-percent), is twice as accurate in respect to the original recorded speed as the reproduce timing accuracy of the venerable Ampex 351 (+ or - .2%).

John B.

Link to comment
Share on other sites

I use Metacorder and Sound Devices recorders.  I'm not certain that this workflow is possible with B.R.  If anyone knows if it is, please post and let us know as I've had people ask me this but was not able to answer.

I use BR and Sound Devices stuff - I have used 48048 workflow often without incident.

Robert

Link to comment
Share on other sites

There are many ways to concoct a workflow. Personally, I am always against any method that involves recording off speed. 

48048  and marked as 48k is recording off speed and marking it as on speed.  So if you were to play that track back on any normal player it would always play slower than real time and give no evidence that what you were hearing was not a faithful playback of the original event. Most devices that record images and or sound are designed to record them and play them at back in "real" time.  In other words 30 minutes of recording when played should last 30 minutes.

-----Courtney

As always, Courtney's take on this is as straightforward and clear as it can be, but it represents a philosophy that is generally not shared by so many of the people with the power and authority to concoct some of these workflows. In the early days of non-linear (video) based editorial for film, these sorts of transfers, pull downs, etc., were always performed only ONCE (as film and audio material was brought into the editing system). Now, these wild and crazy manipulations happen at several stages of the operation, often without documentation and sometimes even without anyone's clear knowledge --- this is really bad and certainly creates NEGATIVE future proofing (what would you call that?). I am in favor of doing all the things that are necessary to insure that our sound, recorded on our devices, will be in sync with the imaging device (dare we call it a camera) and the further manipulations that may be required to accommodate editorial devices be done by editorial.

There are so many jobs I would venture to say, film and video, that would work just fine if we are set at straight 48K and 30 nd, and the "speed" manipulations required by the editorial process be handled entirely in the editorial process --- going in and coming out.

-  Jeff Wexler

Link to comment
Share on other sites

In terms of speed accuracy I'm amused that we're concerned over something that, even off-speed (one-tenth-of-one-percent), is twice as accurate in respect to the original recorded speed as the reproduce timing accuracy of the venerable Ampex 351 (+ or - .2%).

John B.

Of course, the Ampex 351 could not be used to record double system sound because it used a pinch roller and a loosely controlled (AC Line controlled) capstan motor (with a big flywheel) to maintain speed.  Slippage and speed variances due to changing reel tensions prevented more accurate recordings.  That was why the Nagra and Uher methods of recording a servo reference track on the tape for later resolving was so revolutionary.  The Nagra didn't have any better speed accuracy, it just had a way to resolve them on playback.  Once you remove tape movement from the equation accuracy goes way up and is only dependant on the accuracy of your word clock.  Most cheap solid state MP3 / wave digital recorders are far more accurate than any of the most expensive analog tape machines.

----Courtney

Link to comment
Share on other sites

Thanks Courtney,

I can certainly flick switches/set menu items for any flavour of timecode/sample rate post and the producers ask for (and there've been some weird requests), but I'm always vaguely uncomfortable doing so. Your post was very encouraging. Simple works for me.

Chris (old dog) Newton

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...