Jump to content

Who's had luck with 48.048k 30ND in Boom Recorder?


RPSharman

Recommended Posts

Hello Folks,

I'm trying to nail down what I need to do to turn in 48.048k files with 30ND time code to an NTSC TV show which shoots on film (24fps).  They are accustomed to DEVA V files stamped as 48048 30ND time code.

Previously with my DV824 I set 29.97ND 48k with 1% pull up.  No complaints.  I turned in a back-up from the 744T at 30ND and 48048 - no "F" - but imagine it was never looked at.

Now I am using Boom Recorder and 744T.

The 744T has 48048F, which I believe automatically defaults to 30ND time code, stamps files at 48k forcing DAW to slow files to 48k 29.97ND - But since this is a real show with real equipment, I don't believe I need to trick any systems, and I feel I should set the 744T for 30ND 48048 Hz, and let the post Machines do what they are designed to do.  Correct?

My Fireface 800 has a master clock setting for 48048, and when set for this, the Boom Recorder software offers a "File Sample Rate" for 48048 but a "Hardware Sample Rate" at 48000k - Who has had luck with using RME hardware and Boom Recorder and creating 48048 30ND files?

Thanks,

Robert

Link to comment
Share on other sites

<snip>

The 744T has 48048F, which I believe automatically defaults to 30ND time code, stamps files at 48k forcing DAW to slow files to 48k 29.97ND - But since this is a real show with real equipment, I don't believe I need to trick any systems, and I feel I should set the 744T for 30ND 48048 Hz, and let the post Machines do what they are designed to do.  Correct?

Robert, 

Your 744 settings depend on what machines they are using in post.  As far as I know, Fostex machines will still choke on a 48.048 stamp.  This was most of the reason behind the 48048F mode.  An InDaw won't care about a 48.048 stamp.

This is definitly an area to check with your post people and find out what workflow they are using.

---Matt

Link to comment
Share on other sites

I recorded a file with FF800 set to 48048 and TC sending 30ND - BR settings were 48000 interface sample rate and 48048 file sample rate.  I sent Take Vos an incident report so he could see the file information.  He says...

Hello Robert,

From the incident report Boom Recorder sees a 48000 sample rate, in that case you should use the following settings:

- Boom Recorder hardware sample rate 48000 (Boom Recorder uses this sample rate for timecode calculations)

- Boom Recorder file sample rate 48000 (This is the sample rate the file is set to)

- Boom Recorder frame rate 29.97ND (As Boom Recorder already sees a pulled down sample rate, you should use the pulled down frame rate as well)

- Fireface clock to 48048

- External frame rate 30.00ND

In the future I will add a feature in Boom Recorder to handle all the sample rates of the fireface natively.

Cheers,

    Take

Link to comment
Share on other sites

Robert,

I'm confused by Take's suggestions. I'm hoping Courtney Goodin sees this thread and offers his expertise.

The way I read Take's settings suggestions is that BR is doing the "pull down" while recording.

The 48.048 scenario can either be done at the telecine end or at the record end. I prefer to leave the pull down to be done at the telecine end. So I record 30FrTC at 48.048FS - marking the discs and all supporting logs with those specs. The TC slate is also reading 30NDF. There might be a sync issue in telecine when trying to lock your recorded TC at 29.97 from the discs to the numbers on the TC slate reading 30NDF.

As I noted I did record on BR (and the Deva 5 simultaneously) The Fireface was my master clock sending 48.048 to BR.

I set BR to 30FrTC (that's what it was receiving from my Denecke GR-1) I set BR to use the Fireface (Hardware) to get the external FS. (The current BR software does not have a checkbox for 48.048 from the Fireface - Take indicates he will address this in a future update).

I took my test discs over to Eric Benton at Modern Video, who gladly ran them for me to see if pull down was successful.

I now follow Eric's suggestion when I do my head ID and tone:  I audibly read out the timecode I am recording for 15 seconds.

When we played back the test discs at Modern Video, I could hear the timecode being read and exactly match the TC readout from the discs -- pulled down to 29.97.

That's my way of having "luck with 40.048K 30ND in Boom Recorder".

Happy 4th and regards,

Richard

Link to comment
Share on other sites

Thanks, Richard.  Helpful as always.  I plan to set FF800 at 48.048k, run TC into BR via analog input at 30ND, set BR to 30ND and 48k for the "Hardware" sample rate.  Did you set the "File" sample rate in BR at 48k or 48.048k?  There is that option in that drop-down menu.

Thanks again,

Robert

Link to comment
Share on other sites

Robert,

I set the file sample rate at 48.048. I was initialy concerned that the BR Hardware settings for the FF400 didn't show 48.048 - but I realized that it was just that Take had not set up that ability. The Fireface software confirmed that I was at 48.048 and thus set BR's file sample rate at 48.048. I do the same as you, TC via an audio input at 30NDF.

Before I went over to Modern Video, Robert Kennedy at Coffey sound checked the files with BWF-Widget and it showed that it had 48.048, 30NDf flags.

I suggest you do your test disc and then call Eric Benton at Modern Video 818-840-1759. I'm sure he'll be more than happy to check it with you. (I'm assuming you are in L.A?).

Do the audible TC count - when it's played back and you hear your T.C count perfectly in sync with the TC read you know that everything is "A-OK'.

Good luck,

Richard

Link to comment
Share on other sites

When doing a test by reading the timecode as the machine is recording.  You might want to set the Time Code to a high hour number.

(Greater than 12 hours).  This will make any misinterpretation of the time stamp very evident and easy to spot by the verbal sync method described.  If you are at 1 hour or less, the calculation error of the time stamp may be too close to the displayed time to denote a difference by simply reading the changing seconds.  At that point the error would only be a few frames and you can't observe that without putting the interpreted time code and the intended time code (like the slate numbers) against each other in a freeze frame.  Many slates have an offset of a couple of frames from decoding then display latency so it would be hard to gauge if the settings are being interpreted correctly at low hour counts.

I am not sure what Take is talking about when he talks about "file Sample Rate" and the "hardware Sample Rate"  Does this mean that boomrecorder allows you to record at a hardware sample rate of 48048 Hz and set the "file Sample Rate" to 48,000 Hz? (I guess that means the rate stored in the format chunk of the file will be 48,000.  Not sure what he puts in the coding history or iXML chunk.

This would effectively be the same as the "48048F" mode on the 744t or the Deva. This would also mean that the Time Code Generator should be running at 30.00fps and the file should be marked as "29.97" (or 23.976)

I haven't played enough with Boom  Recorder and Metacorder enough to be well versed in their inner workings. Maybe we will have more light shed on this tomorrow in the test setup we are doing.

-----Courtney

Link to comment
Share on other sites

  • 6 months later...

Hi, new poster here.  Awesome forum!!!! 

I'm about to use boom recorder on a 35mm feature and got a few questions.  I'm using a MOTU Traveler along with a Macbook Pro and a Tascam HD-P2 for backup.  I'll be syncing with a Denecke timecode slate and a SB2a sync box.  I'll be using a I'm trying to figure out what's the best way to deliver 48.048kHz 30ND fps to post.  I emailed Take Vos about this and he suggested that I try tricking the system by setting the Traveler to clock to SMPTE and setting it to 29.97ND.  Then feed 30ND timecode into the Traveler, setting Boom Recorder to 48k Hardware and 48k file.  The Traveler will then supposedly clock it at a pulled up rate?  Then upon playback during Telecine at 48k it  will already be pulled down. 

Has anyone tried this method successfully?  It seems this whole pull up/down issue is being discussed a lot and there's still a lot of inconsistencies about how it's supposed to be done.  Any advice is greatly appreciated.  Phillip Perkins already made a few comments about this on another forum.  I'm looking for more input.  Thanks   

Link to comment
Share on other sites

...Phillip Perkins already made a few comments about this on another forum.  I'm looking for more input.  Thanks   

Philip is an excellent person to listen to.  1) He's knowledgeable.  2) He does both location sound and post production. 

The question you're asking is all about the interfacing of location sound and post production.

(Of course, there are numerous others here who are also eminently qualified.)

JB

Link to comment
Share on other sites

Ok, I just did some tests on this.  This is probably not the most accurate test out there but it does shed some light.  Sorry but this is kinda long....

I set both Boom Recorder and the Tascam HD-P2 to receive LTC from the same source (a Tascam DA-88 with the SY-88 time code card option).  I spilt the timecode output of the DA-88 feeding both the Traveler and the HD-P2 simultaneously.  I set the DA-88's timecode to generate at past 19hrs so timecode issues would be more noticeable if it occurs as suggested by Courtney?  I then feed the same audio to both systems simultaneously from my mixer. 

First test:  Both systems clock were set to SMPTE, 30ND and 48k throughout. I recorded a vocal slate and a series of claps at the beginning, then let it run for about 5 minutes, then record a series of claps again at the end. 

Second test:  The Tascam HD-P2 remained the same settings.  However, I set the Traveler's Frame Rate to 29.97ND in the MOTU SMPTE Console.  In Boom Recorder, I set the Frame Rate to 30ND, Hardware Rate to 48k and File Sample Rate to 48k.  I once again recorded a vocal slate, a series of claps at the beginning, let it run about 5 mins, and then claps at the end. 

Then the files goes into my studio DAW (Samplitude Professional).  I set the project frame Rate to 29.97ND, 48kHz.  I set it to import the WAV files into the respective timestamp position.

First recordings:  Both files lined up near perfectly and plays in sync throughout.  I zoomed in all the way to sample level looking at the waveforms and the 2 files were offsetted approximately 15 samples apart consistently at the beginning and at the end (I guess this would be equivalent of around 0.0093 of a frame?)  This should be acceptable.  I'm using this first recording as a baseline.

Second Recording:  From the beginning I can hear slight flanging and towards the end they were obviously way off sync.  I then take the file recorded by the HD-P2 and slowed it down 0.1% (I used a stretch factor of 1.001; resample mode).  I visually line up the claps in the beginning of the files, then go to the end.  They now seem to sync.  I zoomed in to sample level noticed that they were approximately 13 samples apart at the end of 5 minutes.

Based on this test, it seems Take Vos' suggested method works.  The Traveler clocking to SMPTE essentially sped up doing the equivalent of recording the file at 48.048k.  Upon playback at 48k, the recording automatically get's pulled down.  Slowing down the HD-P2 recording by 0.1% would be the equivalent of a telecine analog transfer pull down?  I have been racking my brain trying to figure this whole thing out.  Please let me know if there are any flaws in this test.  Please point it out to me if my math is off.  Ultimately, the true test is when the files goes to post after the shoot.  However I would like everyone's opinion on this.  Thanks.....sorry for the long post.

Link to comment
Share on other sites

Just to be explicit about some details:

Where will sync take place?  During telecine or in editorial?

You didn't say specifically what LTC rate you were outputting from the DA-88 for your tests....?  I assume 30ND, correct me if wrong.

In your test, what was the clock reference for Samplitude?

If sync'ing in telecine, I'd make sure telecine gets those files and plays them back in their machine.  A DV40 or 824 (the most common these days, I think) may react differently to the "tricked" TC and sample rate than your DAW.

I don't know enough about the "F" flag that Fostex machines use to deal with 48.048 to say if that will have any influence on proper playback speed.

If you look at the metadata chunks for the files, are there any values which indicate a 48.048 sample rate?

If you go for it with this method, get an email confirmation that everyone has agreed on the workflow parameters (sample rate, frame rate).  The fingers will inevitably point to the source of the recordings if something goes wrong.

Brian

Link to comment
Share on other sites

I don't know how this will work with your TC source being a Denecke box feeding 30F TC to your recorder and your FW interface.  I don't think the MOTU has the pull up option to change the sample rate to 48.048k, like the RME FF400.  If you set your HD-P2 for 29.97/48k and send it 30F TC it may pull up, I don't know.  Or does it have a 30/48.048 option?  Boom Recorder will look for the 48048 sample rate from it's TC source, your MOTU, but if the MOTU  is set for 48k, then that's what BR will see, regardless of the TC the MOTU is receiving from the Denecke box.  You can check this by clicking on the TC readout in BR.  It will show the sample rate.

The only way I have known this to work for people with the MOTU is to get the TC from a 744T set to 30/48048, which tricks the interface and Boom Recorder.

Robert

 

Link to comment
Share on other sites

I don't know how this will work with your TC source being a Denecke box feeding 30F TC to your recorder and your FW interface.  I don't think the MOTU has the pull up option to change the sample rate to 48.048k, like the RME FF400.  If you set your HD-P2 for 29.97/48k and send it 30F TC it may pull up, I don't know.  Or does it have a 30/48.048 option?  Boom Recorder will look for the 48048 sample rate from it's TC source, your MOTU, but if the MOTU  is set for 48k, then that's what BR will see, regardless of the TC the MOTU is receiving from the Denecke box.  You can check this by clicking on the TC readout in BR.  It will show the sample rate.

The only way I have known this to work for people with the MOTU is to get the TC from a 744T set to 30/48048, which tricks the interface and Boom Recorder.

Robert

 

I'd recommend getting TC AND WC from the SD machine--this works well for me w/ Travelers, and saves an audio input on the Traveler.

Philip Perkins

Link to comment
Share on other sites

Just to be explicit about some details:

Where will sync take place?  During telecine or in editorial?

You didn't say specifically what LTC rate you were outputting from the DA-88 for your tests....?  I assume 30ND, correct me if wrong.

Yes the DA-88 was outputting 30ND timecode.  I will be contacting both telecine and editorial to get all the details for this project.

In your test, what was the clock reference for Samplitude?

I used the Traveler for playback with Samplitude.  I figured since we're comparing 2 files playing back simultaneously in the digital domain, the external clock shouldn't affect it?

If sync'ing in telecine, I'd make sure telecine gets those files and plays them back in their machine.  A DV40 or 824 (the most common these days, I think) may react differently to the "tricked" TC and sample rate than your DAW.

I don't know enough about the "F" flag that Fostex machines use to deal with 48.048 to say if that will have any influence on proper playback speed.

How do you deal with 35mm features?  I know you're using BR and Traveler.  Are you using the Fostex PD-4 to output 30ND timecode and also s/pdif out at 48.048k?  Are you setting BR's files as 48 or 48.048?

 

If you look at the metadata chunks for the files, are there any values which indicate a 48.048 sample rate?

Samplitude sees some of the metadata upon file import.  It sees the project name, timecode, file originator....in this case Tascam HD-P2 and Boom recorder was recognized.  And finally both set of files shows 48k.  There is no indication of 48.048 anywhere and I did not set that rate on any of the recorders. 

Brian, thanks so much for the response.  Please elaborate a bit more on how you achieve this and how your BR files are delivered.   

Link to comment
Share on other sites

I don't know how this will work with your TC source being a Denecke box feeding 30F TC to your recorder and your FW interface.  I don't think the MOTU has the pull up option to change the sample rate to 48.048k, like the RME FF400.  If you set your HD-P2 for 29.97/48k and send it 30F TC it may pull up, I don't know.  Or does it have a 30/48.048 option?  Boom Recorder will look for the 48048 sample rate from it's TC source, your MOTU, but if the MOTU  is set for 48k, then that's what BR will see, regardless of the TC the MOTU is receiving from the Denecke box.  You can check this by clicking on the TC readout in BR.  It will show the sample rate.

The only way I have known this to work for people with the MOTU is to get the TC from a 744T set to 30/48048, which tricks the interface and Boom Recorder.

Robert

 

The MOTU has no 48.048k option.  Take Vos suggested I try the method i did the test with.  He never said that it's a proven method, just something i can try.  The HD-P2 does have a 30ND pull upsetting.  I have tried recording with that setting and it does NOT stamp the files as 48.048k.  So is it doing the "f" mode by recording it at a higher rate and stamping it at 48k?  This is what the HD-P2's 1.02 revision says

----------------------------------------------------------------------------

Pull Up/Down Sample Rate Settings

A pull up/down sample rate is always relative to the chosen time code rate.

For example, when it is required to chase 30fps LTC but the project will eventually be pulled down to 29.97, a 30 Pull Up setting would be used providing an operating sample rate of the selected sample rate plus 0.1% while chasing 30fps LTC. Later, when chasing 29.97, the sample rate will be normal.

The following 0.1% pull up/down rates are supported by the HD-P2:

23.976 Pull Down

24.00 Pull Up

24.975 Pull Down

25.00 Pull Up

29.97 Pull Down

30.00 Pull Up

--------------------------------------------------------------------------------------

The original HD-P2 manual did not make sense to me, but the revision notes seems clear.  I will try another test where I will record the files using the HD-P2 30ND pull up mode and at the same time having BR record clocked to the HD-P2 via S/PDIF.

Link to comment
Share on other sites

So the HD-P2 will work, but you are still stuck not being able to send 30/48048 to you MOTU in order to have Boom Recorder behave properly.

My suggestion is to buy the FF400.  You can send it 30ND from the Denecke box, set the FF400 for 48048, which in turn will give BR the settings you need.

Link to comment
Share on other sites

Syncing the MOTU to timecode at its 29.97 fps setting while sending it 30 fps seems to me like a brilliantly simple way to get it to record at 48.048. When I first got the MOTU I experimented with this too, by recording pieces of audio of known length in this manner and then comparing the recorded version's length to the original. As far as I could tell, it was a perfect .1% change. Since I just do commercials and video, no one's ever asked me for 48.048, so I've never gotten to really test it either.

I know this doesn't help you with the other settings you'll need to change (or not change) in Boom Recorder, but if you're going to do tests with a post house, I would be very interested if you would test this method. Respectfully, it seems as if people who have never tried it are saying that it won't work.

Link to comment
Share on other sites

I would defer to Robert and Philip for their real experience delivering 48.048k, because, so far, I haven't actually done that.  I've stuck to convention, delivering 48k.

As far as my system, the PD-4 is about to go into cold storage as I picked up a 744T recently to implement as the backup/portable recorder.  So, in the next few weeks, I will also be revisiting how the machines talk to each other.

That said, I used the PD-4 as a backup recorder and timecode source only.  No SPDIF, no AES output.  The only connection between the PD-4 and the Traveler was 30 NDF piped out of the PD-4's TC OUT, routed to a Traveler analog input.  BR clocked to LTC/SMPTE.  48k across the board.

I think JBucher is right, and I don't want to be an armchair skeptic.  My biggest concern is the 48k stamp that your pulled-up files will carry with them, especially with regard to how your recordings will be dealt with down the line.  Here's some scenarios I can imagine (and a lot of unanswered questions...sorry):

You record files with the "trick" pull-up, yielding 30/48k files.  At what frame rate will telecine play those files back?  If 29.97, will this slow down the recordings to sync appropriately (in essence, the "reverse trick" of how they were recorded), and if so, what sample rate will go to the digital output?

If telecine plays back at 30, we'll assume they will need to engage a pull-down mode.  According to the DV40/DV824 deliverables guide found at <a href="http://www.locationsound.com/PDF/DV40Dv824%20Deliverables.pdf">this link</a>, if they engage 0.1% pull-down, it will digitally output 47952 kHz.  Most likely not compatible with the transfer medium, or the editing software.

Assuming recordings at 30NDF/48kHz and a telecine house using a DV40 or DV824, it sounds like they need to sample-rate convert to achieve 48K audio going to the transfer medium, IF they transfer the recordings in the digital domain.  I guess I don't like the idea that the first stage in the new life of our precious recordings is to go thru a sample-rate conversion in a telecine house.  I think it happens all the time.

If the material is sync'ed in telecine, we might guess that the editors will be working with that sound.  But if they have to bring in anything directly from your recordings, how will the editing software react to the "trick" files?

I read this from RAMPS regarding 48.048 workflow:

"The BWF poly files HAVE to be at 48048 and flagged

this way in order for AVID to actually import the files. To sync up,

your files must be clocked at 48048 and flagged as 48048. "

source = <a href="http://groups.google.com/group/rec.arts.movies.production.sound/browse_thread/thread/25ac77256310dd0a/55d5291349b958dc?hl=en&lnk=gst&q=clocked+at+48048#55d5291349b958dc">this link</a>

I'm not sure I've really given you a solution, but there are so many variables that you have to get firm answers from production as to the workflow.  I tend to think one of the most important parties to be involved is someone from post-sound.  Of keen interest to post-sound is:  what will be the source of the sound masters?  An OMF or AAF transfer from picture editorial?  Or will they go back to the original recordings and conform?

Considering that you're not really delivering 48.048k files, I'll urge you again to make sure they'll be compatible with all the machines during the various stages of telecine, editorial, and sound.

And I would also be keenly interested if you test it and this workflow is viable.

Brian

PS - I don't know how Samplitude displays metadata.  I usually use the free iXML reader or BWAV Reader to be able to see all the fields in the BEXT or iXML chunks (such as File_Sample_Rate, Hardware_Sample_Rate, etc.)

Link to comment
Share on other sites

Brian, thanks for the insight!  You said you stick to convention and deliver 48k files, so does that mean they pull-down via an analog transfer in telecine?  So far everything I've read about this whole 48.048k workflow is inconclusive.  :)  The only thing I can be sure of is to check with the people doing Telecine, Editorial, and Post Sound to ensure we're all on the same page!  I wanted to do some tests myself first so when I talk to them I can tell them the options of how I can deliver the files.  Once again, it is great that everyone has been so helpful.  I'm learning so much here. 

Link to comment
Share on other sites

does that mean they pull-down via an analog transfer in telecine?

Not necessarily.  It's a question worth asking, though.  On the first feature I did with Boom Recorder, I called the telecine house to make sure things were going smoothly.  I found out the first thing they did with the recordings was to transfer them to a DA-88, because that was the machine they used for sync in the telecine suite.  I don't know if the conversion was done digitally or analog, and I didn't pry into it, because I already knew that sound editorial would ultimately conform from the original recordings.  So whatever happened between telecine and editorial essentially "didn't matter" to the sound quality that sound editorial would eventually use.

Link to comment
Share on other sites

Click on the TC read out window in BR and it will tell you if sample rate is 48048.

Wow I didn't know about that feature in BR.  I Just checked with the following settings:

-30ND timecode from the DA-88 into the traveler

-MOTU SMPTE Console at 29.97ND,

-Boom Recorder frame rate at 30ND

-Clock Source to SMPTE

-Hardware Sample Rate 48000

-File Sample Rate to 48000

Click on the timecode window the rate drifts between 48045 to 48051 but stays in the 48048 range mostly!!  So can I say it's being clocked to 48048? 

Now I set the file sample rate to 48048.  Record a file, then brought it into Samplitude again.  Samplitude sees it as a 48048Hz file.  I try  Import into Pro Tools, it also sees it as 48048 and a warning comes up saying the file will be played back at the wrong speed.  (Session is at 48000). 

So, can I say that I just recorded a legitimate 48048 file that was clocked at 48048 and tagged as 48048 using a MOTU Traveler and Boom Recorder?  Those who have this setup, can you try it and confirm?   

Edit:  Well I just checked back and noticed that after setting the file rate to 48048, the timecode readout doesn't match :(  Sigh....and the tests continue.  :)

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...