[Jimmy See]

Another thing you could try (though it’s pretty experimental and you’d want to test this thoroughly first), is looking for solutions other than Furnacecore that don’t cost as much as nuke. Then you can update the hardware which will likely be a useful investment anyway.

If the work is mainly revolving around the film restoration then it sounds like you don’t have as much to consider as you would around choosing editing software for example, where you’d have to worry about wide support for the platform and the ability to collaborate and many other factors. It seems like you can be pretty free to learn how to do the specific tasks you need to achieve the result you want and then mostly repeat the same task, making learning the new software a less burdensome task.

With that in mind, Natron https://natron.fr/features/ is free and open source and makes particular note of it’s ability to leverage available system resources. It also relies on the OFX framework for its effects, along with Nuke and Resolve. The nice thing about that is that users develop OFX plugins for the community for free. As they’re OFX they often will work with on other OFX supporting software as well.

I had trouble finding OFX film restoration plugins but I did find one some Engineering students put together for Natron. They did it as a thesis so I don’t know if they’ll ever develop it any more, and I also don’t know how good a job it does, but the thing is you can borrow a decent computer, install natron for free and use the plugin and test if it will achieve results as good as what you were using and with the improved speeds you’re hoping for. If it does this well enough, then it won’t matter if its ever updated.

Link to forum post announcing the plugin below and below that the link to the github page for the plugin below that.

https://forum.natron.fr/t/film-restoration-plugins-for-natron/1576
https://github.com/SebastianBugna/PARP

I believe they did not make any binaries, so you’d have to compile it yourself, but that’s relatively easily done on a mac, just a bit frustrating when you’re not used to doing that, but you’d only ever have to do it once.

I know this seems a bit out of left field, but the crux of it is, you can test it entirely for free, if it’s any good you won’t be chained to old hardware, the software it was written for specifically boasts particularly good optimisation for getting the best out of a powerful CPU and as it’s an OFX plugin there’s a chance it could work in other software like Resolve which is also free (for the ‘lite’ version).

[Jimmy See]

An SSD if you don’t have one already could give you a little boost launching FCP, and saving (if using the SSD to store the project file). If you can afford a large enough one, perhaps you could use it as a scratch drive while you work on a project and store your working media on there, that would increase the speed at which it could read the media which might offer some small improvements in some aspects of the process.

[Jimmy See]

This intrigued me because in the years that I used Final Cut I never heard of that happening. I couldn’t find you a nice single easy to understand and test answer, but here’s some suggestions.

[Jimmy See]

Would probably need more details to help with this. A lot depends on what ‘Dropping frames’ means. When I come across the term is usually referring to frames that exist and are for whatever reason not playing back correctly, ie. ‘dropped’. In circumstances like that, it’s unlikely the file itself that has something wrong with it, or at least, isn’t missing any frames. In such cases it would be an issue with the system attempting to play it. That would mean the file was okay in that it is not corrupted or anything, but it doesn’t necessarily mean that it is ‘correct’ in terms of meeting the requirements for the equipment and software the newsroom is using, so again, more details would help because if they gave you any specific requirements for what they wanted and you haven’t met them then that might be the problem right there.

Then again, they may be using the phrase ‘dropping frames’ in a less specific way, the file itself might look ‘jerky’, or have black frames, or jump unexpectedly no matter where or who is playing it which would mean something went wrong in the export process. Have you played the mp4 file on your computer, does it look normal to you? What about the mov file? If they look normal to you, that indicates it’s more likely something going wrong with their equipment or setup, or that you’ve delivered something out of their specifications, if they have any, and that’s why it isn’t playing properly.

To help with this you’d need to provide some of this information as a start:

[Jimmy See]

Hey Dave just realised I never responded to this. We never did find a solution that I can remember. The macs didn’t have any equivalent to dynamically assigned drive letters so there was no potential point of failure from that. I think what I remember was having to to the import again on each machine for which we intended to use the material. Really annoying. I hope Adobe are watching because C300 is a popular camera, shared storage environments are common and the need to collaborate on work is important so if you can’t open a project on different machines connected to the same shared storage and expect everything to be online then you really lose the ability to effectively collaborate.

[Jimmy See]

Hey Travis

I found your predicament interesting because it does sound like it ought to be something you can fix and it’s frustrating that such a little thing is standing in the way.

I’ve found a solution for you, but if Brent’s method works, I’d advise using his instead, because his will work with existing edits and mine will not.

However, if you don’t have any luck with Brent’s method, I have tested this and it worked. Since you’ve already had to create entirely new media, and what you’re afraid of losing is in fact your logging work, not necessarily instances of the clips or their use in timelines, then really relinking isn’t actually necessary and Avid finds it problematic anyway since as far as it’s concerned the difference in number of tracks means that there is no relationship between your original transcodes and the new ones and you’re asking it to relink one piece of media to something totally unrelated. What you actually want to preserve is your metadata, not the relationship between clips in bins and locations on disk.

You can do this using ALE files which are just text based versions of the metadata in a given bin.

• 1. Make a new bin, and populate with your newly created MXF media that has all the tracks. During this process you should put only the same clips in each new bin as were in your old bins, so if you had a bin originally that contained clips 1 to 8, the new bin should also contain only clips 1 to 8 (but of course with the NEW versions of those clips not the old),
• 2. Save the bin layout of your old bins and choose it as the layout for each new bin as well.
• 3. Right-click inside one of your old bins and choose ‘export bin’ an export window will open up with Avid Log Exchange as the default option. Choose a location and save.
• 4. Having now made an ALE version of your old bin on disk, you want the metadata from that bin to attach itself to master clips in the corresponding new bin, specifically, we only want the media that you entered manually attached to those clips and don’t want anything else that might be there by default ruined (especially track data). So first, select all the clips in the bin and then right-click on one them.
• 5. Select input > import media.
• 6. Click ‘options’, in the ‘shot log’ tab of the import settings window select ‘Merge events with known master clips’ from the ‘Events’ list. Click OK.
• 6. Choose the correct .ALE file that you previously saved on disk, and then click ‘open’
• 7. The columns in your bin will populate with the metadata you previously entered.

I tested this with your exact scenario using resolve for transcoding and two sets of separate transcodes from the same source media one with and one without mixing down audio tracks. It works flawlessly.

The biggest drawback with this that anything you’ve done with the clips other than logging is still lost because the new media is considered just that, new media. But if you’ve only done logging then you should be home free.

[Jimmy See]

I would add a few things to widen the potential options you can pursue, though the answers you have are probably best practice.

One thing I’d investigate though is, if you do end up going with a workflow where you transcode anyway, then do you need to shoot at a DNX codec? It would depend on the camera, because there might be an argument in favour based on quality regardless of whether you save any time in transcoding, but that’s something worth checking. If your specific setup confers no advantage in quality over a different acquisition format, and if you find yourself having to transcode anyway for better performance with groups then you may as well investigate what other acquisition formats the camera has and see if there are options of equivalent quality for less storage consumption, so only your edit server has to store the comparatively storage heavy DNX codecs, and you don’t need as many G-drives.

You may also want to test a workflow where you shoot DNX, consolidate DNX and group the consolidated clips and see what kind of performance you can expect with your machines and server and get the people who’ll also be cutting to test as well. Often the performance sacrifice, when the media is consolidated and re-wrapped is in my personal experience usually pretty minor and if it’s subjectively acceptable to the cutters then you can potentially skip and entire offline/online workflow step and don’t have to worry about potential conform issues later down the track.

Relinking in Avid, to anyone used to the methods used with other systems, is a pain but if all the planets are aligned then relinking to AMA clips in a bin does work and is the closest approximation to the approach of specifically designating a filepath as would be done on other software. It’s very picky about this, sometimes it’s like you’re in court and Avid is the prosecution looking for technicalities and loopholes in your approach in order to deny you your request to relink. It’s like it will find ANY excuse to come back with the old ‘no clips were relinked’ message. But I’ve found though that when using a workflow which involves offline to online conform, having DaVinci Resolve take care of the transcoding and using the steps Brent posted works very well. I may also just have been lucky that in the circumstances where I’ve had to do conforms, the crew were diligent with best practice and the camera was one that worked well with professional workflows so each filename was absolutely unique, reel numbers had been set to increment between shoots and timecode was free-run so in assembling my ‘case’ against the ‘prosecution’ there was lots of evidence. Therefore if you have sway over choices in production and if there is flexibility in camera settings and options, try to ensure that it isn’t possible to get duplicate timecodes on a given shoot or duplicate names over the whole project.

Another thing to consider is what formats can your hardware work with at an acceptable level of performance, including with groups? I mention this because, sometimes it’s possible to transcode to an Avid editing codec that is of acceptable quality for the deliverable specifications, and also at a level of quality as to be essentially no quality loss from the originals. In such a case your transcodes can be your online media and there’s no need to conform. This is good in situations where the camera only records compressed formats like h264 and where it’s not particularly high spec anyway since in such cases you have no choice but to absorb the time and storage costs of transcoding anyway, but you can at least avoid the conforming stage. Hopefully in your case, your testing would find that the proposed flavour of DNX for acquisition is one the your hardware can work well with for your circumstances because then a simple consolidate workflow is your best choice, and there’ll be no need of a conform.

[Jimmy See]

Hello Rory if you’re still around on here and anyone else who might stumble in to this old thread. 8 years after the fact, and this is still something some folks are struggling with.

In this instance, RTFM actually was the best course of action, https://resources.avid.com/SupportFiles/attach/Media_Composer_Editing_Guide_8.8.pdf#page=89

For current version of Avid as of 2017, to change the project window font size, click on the project window so it’s focused, then click the ‘window’ menu from the application top row in windows and os x menu bar in os x and then select ‘set font’ here you can choose both what font and what size.

EDIT: I tried to embed the page number for the PDF in to the link but it doesn’t seem to work. It’s Page 89.

[Jimmy See]

Hi Eric

I realise this post is several years old now and also that you do not represent Blackmagic, but this is one of the few places around where the fact you pointed out, that blackmagic products do not support audio input without accompanying video input, is actually clearly stated.

I ask you, but I guess the internet in general and hopefully blackmagic can see this, WHY!? Is this a technical limitation of how the cards work or some kind of decision to deliberately restrict users in this fashion? It is actually really not made clear at all to people who are buying this product in any obvious way. I’ve actually looked and forum posts are the only place where I’ve seen this spelled out in black and white. If it is documented somewhere, it’s obscure enough that people are unlikely to discover it by chance and I could not find it even by looking. I can’t see the justification for making the products behave in this manner. I’m not aware of any blackmagic audio-only capture devices so it doesn’t seem like a cutthroat strategy to force people to buy two products, I can’t see it making much sense from a piracy prevention standpoint either if it’s somehow related to that. I thought perhaps they were trying to force people to purchase higher end models with the capability but this post https://forum.blackmagicdesign.com/viewtopic.php?f=3&t=32050 seems to imply that no blackmagic products will do it. Also if they did, surely they would be making sure people were well aware of the limitation/feature so that they could persuade people to buy their more expensive models which did have the functionality.

I discovered this limitation and stumbled on this thread trying to help a TV editor with their system. They wanted to record temporary narration with a microphone plugged in to a mixing deck going in their intensity pro 4k card. This, I suspect is a very common setup for broadcast editors, certainly it is the standard arrangement I’ve seen on older systems which used matrox hardware for I/O prior to blackmagic entering the market. We were left scratching our heads why this couldn’t work until finding this and some other forum posts reiterating this limitation, I found it nowhere on Blackmagic’s website or in the Blackmagic desktop video manual which appeared to be the only form of manual they offered for this hardware. It makes very little sense and I think if things HAVE to be this way, they could at least signpost this a lot earlier given many of their customers would likely be film and TV editors who will likely encounter this very situation.

[Jimmy See]

Thanks John.

The sequence has always been free of any video so I’m afraid that wasn’t the issue. I have in the interim figured something out that I’m pretty confident is the solution but I’ll have to hear from the sound designer before I can proclaim this resolved.

What seems to have happened and what I’ve done are:

I manually replaced all clips with their unmerged counterparts as explained before. I noticed that the sound designer seemed to be getting only the odd numbered tracks and I realised it was related to the unusual way I setup Premiere. I migrated to using Premiere from Final Cut Pro 7. I found that I did not like the way Premiere handled audio whereby single tracks can contain multiple channels, I was used to each channel being a track and it confused me and annoyed me so often I decided to set up Premiere to behave as I thought FCP 7 would have. So I had my audio set to be interpreted as separate tracks on import, which has been working great for me, as when I put a clip with 5 channels on to a timeline, I see 5 tracks one underneath the other just how I like it. However I don’t come from an audio background, and Premiere’s sound handling is much deeper and richer than FCP 7 was and Avid too, sometimes I didn’t grasp the implications of things. I had my tracks for my sequence setup as ‘standard’ and I think that was the issue because even though there oughtn’t have been any stereo sound that wasn’t split up to individual tracks (as it was all set to interpret to mono tracks anyway), what seemed to be happening on exporting OMFs was that Premiere would anaylse the timeline thusly: Timeline Track 1 = Standard type, so must be dual channel stereo, channel 1 will be one track, channel 2 will be track 2. Given that channel 2, which oughtn’t even have existed, would always be empty I think Premiere concluded that ‘track 2’ of the OMF was empty and thus it discarded it as it does with any empty tracks when creating an OMF. So for every single mono audio track of every clip, it seemed to create 1 extra track, then delete it. After finishing with timeline track 1 it then looks at Track 2 and does the same thing, erroneously considering it to be dual channel, splitting the channels in to OMF tracks, then discarding the empty one. I believe this process resulted in my timeline tracks 1,2,3,4,5,6,7,8,9,10 turning into 1,3,5,7,9,11,13,15,17,19,21. By this theory I assume that despite the confusion there shouldn’t actually have been anything missing in the sound designer’s DAW but it definitely isn’t a good way of doing things and of course one would naturally assume something was missing if you can see track numbers getting skipped.

My solution to the track numbering problem then was that I needed my track types set as ‘mono’ but you can’t do that after you’ve already created a sequence so I had replace every clip on the timeline yet AGAIN, but this time first creating a new timeline with the requisite number of ‘mono’ tracks. I also put my reference export on the video layer temporarily to check sync and make sure I’d not made any mistakes (deleted it again before any more OMF exports).

This logically fixed things but actually I was STILL experiencing issues with embedded OMFs and referenced OMFs just as before when getting them in to Audition (opening as sound files not sessions, failing to link to media except one at a time). I don’t know if this was just audition but I didn’t want to take any chances that the same would happen to the sound designer so I figured I had to be able to produce something I knew worked for me and my tools at the very least. Then I realised what I should have realised from the start, you can use Adobe’s dynamic link to edit a sequence from Premiere in Audition and you can use audition to export an OMF.

I clicked edit > edit in adobe audition and everything worked beautifully, it uses an XML process and copies media over to a directory of your choosing to achieve this so it doesn’t appear to work as smoothly as other adobe dynamic link applications but it does it all automatically at least. My timeline was replicated beautifully in to Audition and I was then able to export the OMF using audition which is actually dedicated audio software for these kind of tasks and that seemed to make all the difference because the resulting embedded media omf was openable no problem in audition after that.

So the problem is partly my fault but a lot Premiere’s fault for some really weird implementations of standard professional workflows and for the ABSOLUTELY CRAZY way it handles merged clips coupled with there being NO WAY to unmerge them. In fact, if anyone needs to know how I even figured out which sound clips belonged to which merged clips, I had to export multiple EDLs (as each can have only 4 tracks) because that was the ONLY method I could find that revealed the original filenames of constituent media of merged clips.