Digitizing Nagra IV-S TC with time code

[cedi]

Hello everyone--hoping you can help. We're looking for a very particular analog-to-digital converter (if it exists).

We have an old Nagra IV-S TC and plenty of tape that was recorded on it. The tape is two-track stereo with a center track of time code. We're looking for a converter that will take the stereo signal AND the time code signal and export something we can import it into a video editing platform.

The ideal box is something with stereo input and a dedicated time code input that will export a single digital file with the time code as metadata in that file.

I understand that there are shops and vendors who perform this surface, but the production has a strong preference to do this in-house, hence the search for a hardware solution.

Many, many thanks in advance.

[Tim Gillett]

I'm no expert on time code but imagine the first step would be to digitize all three tracks in parallel as if audio into a suitable recorder with at least 3 tracks.

[James Perrett]

Are you looking to have the sample rate of the audio matching the timecode? If so you would need a box that can generate word clock from timecode. I know that the MOTU Digital Timepiece can do this but there are probably others too.

However, this may not be the best way to do things as the sound quality is likely to be better with the analogue to digital convertor running on its own stable internal word clock so I'd be tempted to just digitise all 3 channels and then decode the timecode and match the timing in software.

[Hugh Robjohns]

We're looking for a converter that will take the stereo signal AND the time code signal and export something we can import it into a video editing platform.

The ideal box is something with stereo input and a dedicated time code input that will export a single digital file with the time code as metadata in that file.

You probably need to hire a time-code synchronisable tape machine -- a Nagra T would be ideal -- and a master word-clock source with a time code generator.

The reason is that if working with digital video, there must be a whole number of audio samples in each video frame, and timecode counts video frames... So you need a timecode generator thats locked to a master word-clock, and a tape machine that can be synchronised to that timecode. You can then digitise the audio from the synchronised tape machine with a converter locked to the master word-clock, and the resulting audio file will be locked to the video.

Complicated... but necessary.

There are variations on the theme, but that's the basic concept and requirement.

[cedi]

Thank you for all the responses--I very much appreciate you lending your expertise on this.

The producers of the project requested an exact digital copy of the Nagra tape that can then be input into the video editing platform. I doubted that there was a simple "A/D converter box" that bridged 35 years of technological development with a couple cables, but I was open to looking.

After reading your responses and others I've been gathering, I'm seeing that my initial thoughts seem accurate. Thanks again for your time. I may have follow-ups in terms of the actual implementation.

[Hugh Robjohns]

The producers of the project requested an exact digital copy of the Nagra tape that can then be input into the video editing platform.

I've been looking into this a bit... and I think there is a solution in the form of a program called DaVinci Resolve:

https://www.blackmagicdesign.com/products/davinciresolve/media

It's possible a similar 'sync to audio timecode track' is built in to some video editing platforms -- not my sphere of knowledge anymore I'm afraid.

Anyway, DaVinci appears to have a function whereby it can re-synchronise digitised audio tracks using an LTC timecode captured simultaneously as another audio track. The intention idea is for video shot on DSLRs which don't have an internal timecode sync function, but an external reference timecode can be recorded on the audio track. In that situation, there will be no steady relationship between the timecode and the audio sample wordclock -- which is exactly the same as your Nagra tape problem.

There's a video here which mentions the re-sync process in passing...

https://www.youtube.com/watch?v=50VFZ6Of5Ro

So, if DaVinci (or some similar function in the video editing platform your producers are using) does what it seems, -- and please note, I've not used it so I'm just going on a bit of goggle research here -- then I think all you need to do is a simple digitisation of the analogue audio and its associated LTC timecode track from your (free-running) Nagra tapes.

In addition to a tape machine that can replay the three tracks, you'll need a four-channel A-D converter (or two identical stereo converters) -- but there are plenty of those around, either freestanding or built into four-channel mic preamps.

Note, it is important that you use a four-channel (or two identical stereo) converter(s) because different brands of converter have different processing delays. So if you use different types for the Audio and timecode tracks you'll build a small timecode offset into your digitised tracks which won't help when you come to sync everything together!

With the audio and LTC timecode digitised (at 48kHz, of course), you can load the material into DaVinci and then use it's 'sync from Audio timecode track' function to effectively re-sync the wanted audio, time-stamping the resulting file in the right way to enable it to be imported into the video timeline like any ordinary digital video/audio file. ... I think!

It's certainly an avenue worth exploring, anyway...

Digitising the audio and with the LTC timecode captured as another audio track is very simple, so it should be easy to experiment. I hope that helps.

H

[Hugh Robjohns]

From the DaVinci Resolve manual (page 213)

https://documents.blackmagicdesign.com/UserManuals/DaVinci_Resolve_12_Reference_Manual.pdf

It makes no mention of how it deals with a free-running audio-track timecode that is inherently non-synchronous with the sample rate... but it's clearly a function designed to deal with that specific situation, so I guess it works...

[cedi]

From the DaVinci Resolve manual (page 213)

DaVinci Resolve.png

https://documents.blackmagicdesign.com/UserManuals/DaVinci_Resolve_12_Reference_Manual.pdf

It makes no mention of how it deals with a free-running audio-track timecode that is inherently non-synchronous with the sample rate... but it's clearly a function designed to deal with that specific situation, so I guess it works...

Hahaha, yes the old "hope and pray" methodology Thank you for taking a deep dive on this, it's actually very helpful information. The video editor is actually already using Resolve, so this all sounds very promising.

I will update you on how it works out!

[Hugh Robjohns]

[cedi]

Hi Hugh and others,

I thought I'd circle back for an update about the current state of things.

Getting the tape into the video platform--conveniently being Resolve anyway--was one issue to contend with. Making that "exact digital copy" that the producers want was another. After serious cogitation and research and conversations with much bigger-brained audio people than myself (like you), here is the current implementation intended to address both issues:

Step one: reach back in time for a device called a CSS Time Code Resolver that was specifically made for the Nagra IV TC.

http://www.mikesfilmsound.com/used_item84.htm

Step two: The 1/4" tape gets resolved against the internal crystal and is then re-recorded using a digital converter (on the table is a Zoom F4) that is itself locked to a TC generator (Timecode Systems UltraSync ONE). My rudimentary understanding is that it this satisfies "TC generator locked to a master word-clock + tape machine synced to that TC" formula that you laid out earlier.

Step three: into Resolve it goes.

Gear hasn't yet been ordered, so do let me know your thoughts about this implementation. Thanks again for all your help with this!

[James Perrett]

I'm not as much of an expert as Hugh is with broadcast gear but as far as I can see the F4 has its own built-in timecode generator so all you would need is the CSS timecode resolver to read the F4's timecode. As a general rule it is better to have your main analogue to digital convertor running from its own internal clock which would be the case here.

Of course, you may already have a timecode generator installed so it may suit your workflow to work in a different way but I would try the Zoom on both internal and external clock to make sure that the quality isn't degraded when feeding it timecode.

[cedi]

I'm not as much of an expert as Hugh is with broadcast gear but as far as I can see the F4 has its own built-in timecode generator so all you would need is the CSS timecode resolver to read the F4's timecode. As a general rule it is better to have your main analogue to digital convertor running from its own internal clock which would be the case here.

This is a great point--thank you.

[Dela]

I would guess that the easiest way to digitize the material from the IV-S TC would be to download a trial version of Avid Media Composer; it has a 30 day fully featured trial period. With that, you can record the Nagra output on three tracks: The stereo tracks on 1 and 2, and LTC on track 3. After recording, the Media Composer can read the audio TC and convert it to TC metadata (Auxillary TC). You can then export the files out in the format you prefer.

The recordings on the tapes will probably have time code breaks, which will confuse the TC conversion, if you try to convert across breaks. In that case, you can make subclips (3 track) and then read the audio TC after citing the recordings up in subclips.

If it is to be imported into a video editing system, you might export AAF files from the Media Composer. These will be read by the editing system directly as clips, making the import a lot easier, depending on the editing system (if they edit in Media Composer it will be really easy).