Post production workflows. What works best for you?

I've been spending a lot of time researching workflows online, but there isn't a lot of solid information so I thought I would start a discussion here.

Over the next year I will be working on a web series with my classmates from school.

Sound design will take place in ProTools. Color correction will take place in Davinci Resolve or Adobe Speed Grade (what do you suggest?). I will use PluralEyes3 to sync the audio.

The confusing part for me is how all of these work together. I used the syncing feature in FCPX for my last project and I hate how it creates a compound clip that loses all of the original video metadata and other info. I would much rather just have the camera audio replaced with the audio from the external recorder. PluralEyes3 can export a file that drops the camera audio and replaces it with the external recorder, but if I do this, will I lose all the original video metadata? We might get the BMCC and it would suck to lose that information.

Then there is the color correction part. I've read about Resolve and how it works with FCPXml and that I could import directly from the camera into Resolve and then import the Resolve XML into FCPX to edit my timeline, then send the FCPXml back to resolve for color grading.

I'll stop here because I'm getting super confused which means you are too.

I'm just having a lot of trouble figuring out workflows because I'm not learning it in school and the training videos I've seen on MacProVideo or other places don't get into specific workflows because everyone will have their own.

Any help would be appreciated because this is super confusing.

Thanks!

What color grading are you doing that FCP X can't do internally?

What metadata do you need to keep when you make a Sync Clip in FCP X? No metadata is lost, it still belongs to the original clips, they're just inside a new clip. You can assign all sorts of metadata to that Sync Clip. So, specifically what do you loose that you really need?

What camera are you shooting with?

Unless you are in a situation where you need to provide daily previews to a third party I'd skip Resolve at the beginning of your workflow. Import into FCPX. Keyword. Edit. Then export to Resolve if you prefer using it instead of FCPX to do your color grading. When finished with Resolve render your clips and export the XML. Import the Resolve version into FCPX and do final cleanup.

One note: Even when you tell Resolve to export your audio it will NOT show up in the exported event. It is, however, in the exported clips. I detach and copy the audio from the original project and paste it into the resolve version. Then I tweak the audio in FCPX.

NB: I'm an amateur playing with home movies. The professionals probably have better info.

Well, isn't resolve just better at grading than FCPX? I mean if I can do everything in FCPX that's fine with me, but I just hear great things about Resolve and the free version is basically as good as the full version, just no 3D.

Maybe its not the metadata that gets lost, but I think one issue is that any kind of organization done automatically on import is lost when the clips are synced. While the syncing feature in final cut is great, I just don't like having to sync the clips one by one. Doing it this way I have to go through each audio and video clip and properly label each of them so I know which audio clips go with the video clips. Is there any way I can just batch process all audio and video clips in fcpx like pluraleyes does, or am I stuck manually syncing one by one? Although, I do trust manual syncing like that more because I can fix any issues that come up, like the boom operator hitting pause instead of stop and causing four different takes to be on all the same audio track.

Right now I'm shooting with the GH2 which really doesn't provide any metadata. But in the future when I upgrade to a camera like the BMCC I just want to make sure that info stays.

I think I just need to get a one-to-one membership so I can really nail down my workflow. I'm really systematic and process orientated mentally and that's what really attracted me to editing, but until I know every possible way to do something I can't figure out what workflow will work best for me.

Right now I'm thinking that my workflow should be something like this.

1. Import footage: create optimized and proxy media ( use proxy for quickly assembling project, optimized for final edits and color correction), move files to event folder.

2. Organize footage. First rename each audio and video clip reel to "original audio" "original video", then rename each clip "scene 1 take 1" etc...

3. Sync each audio and video clip, then enter metadata into each synced clip. I.e. scene, take, reel etc.

4. Edit until I have a final cut.

5. Send XML through XtoPro for sound designer to work on in protools.

6. Color correct in FCPX.

7. Add any titles or graphics as needed.

Are there any other steps that I should be taking? I just really want to make sure that I have everything organized perfectly before I start editing, because if its not perfect things can get really really messy.

You may want to consider a JuiceLink or BeachTek unit and record audio directly into the camera, and save yourself the sync headaches. I've no clue why more DSLR shooters don't do this more often.

PluralEyes I'm not that familiar with, don't know if it can automate things, but you're still stuck with "new clips" that have to then be organized inside FCP X. In FCP X when you create a Sync Clip, it is a "new clip". But you create a Smart Collection based on Clip Type that gathers ONLY Sync Clips together very quickly. You'll be working on those only, not the original un-synced stuff anyway. Become very familiar with organizing media in FCP X, as it is super easy and fast.

You never lose metadata, it stays with the original clips always in FCP X. The "Sync Clip" is a whole new clip that you can assign its own metadata to. At that point you're not using the originals. I'd as what SPECIFIC metadata are you afraid of "loosing" that you really need when you start to edit?

I also tend to think Sync Clips may not export in XML to PPro or PT very well. I'd check with other users first who've done this.

As for Resolve, yes, it is more powerful than the grading tools in FCP X, but do you REALLY need that? Just like PluralEyes, it may be whiz bang, but do you REALLY need it? Here's an article talking about my point. Be appropriate, not whiz bang.
www.philiphodgetts.com/2013/01/do-we-alw...ve-to-have-the-best/

Finally, how large of a project is this? If it is a school project it is not a feature film. It will be a smaller project that doesn't require all that much. So before I'd ever give too much advice, I'd want to know how large this project really is, or isn't. What are the parameter, what is the purpose, how many editors, etc?

I'd go with a ProRes 422 workflow from start to finish.

And the LITE version of Resolve also has no noise reduction, as well as a couple of other small things. But still a nice app besides the point it has a forced marriage to BMD expense hardware.

You're on the right track, pre-planning your production. I wish more videographers would do this. But then if they did, I could be out of a job.

BenB wrote: You may want to consider a JuiceLink or BeachTek unit and record audio directly into the camera, and save yourself the sync headaches. I've no clue why more DSLR shooters don't do this more often.

I can answer that... noise. The analog to digital converters and preamps in a DSLR are just plain nasty. Even when fed a nice, hot signal there is a clearly audible noise floor added by the DSLR when recording audio directly to the camera card. It's a hiss. And quite frankly, a H4N Zoom is not that much better. It's mitigated by feeding it a loud signal, but that is just a signal-to-noise coverup. That's why we record our DSS to a Sound Devices 702.

I know many shooters locally, season pros, get really great sound out of their DSLRs with a JuiceLink.

BenB wrote: I know many shooters locally, season pros, get really great sound out of their DSLRs with a JuiceLink.

That may be, but I'm telling you why people do it. Google "Canon 5D hiss" and you'll see what I mean. I've done the tests on our own rigs and the difference is undeniable. With the right microphone (and for the right content) it is definitely "acceptable". If you feed a loud signal into a 5D, the difference between the his and the recording is greater, so the hiss is not as apparent. But there is a difference.

Another popular reason for DSS on a Canon 5D is the absence of a headphone jack (without custom firmware) and on-screen audio meters. Not monitoring the recorded audio at the end of the chain is risky.

For all of my school projects none of this is an issue as most are under two minutes long.

Last summer I shot and edited a narrative that ended up at 24 minutes. Basically a TV episode. I edited in 10.0.5 and exported the XML through XtoPro and every clip was an audio synced clip, it worked okay, but the sound designer said he had to spend a lot of time cleaning up the organization in protools. I'm not sure if this was an XML issue, XtoPro issue, or fcpx issue, but he was not stoked. We plan on shooting 8 more episodes, but only about 11 minutes each.

I've been reading about PluralEyes3 and I think I'm really leaning in the direction of having it create a new video that dumps the camera audio and creates a new video with the external audio.

PluralEyes stand alone media files demo

This seems to be the best solution because then I can import the synced audio clips without having all that extra footage in final cut and hopefully have a better XML experience.

Also I'm only a videographer because I needed footage to edit, my main goal is to be an editor, but along the way I've really started to enjoy shooting video. And because I'm editing the footage I shot, I'm really able to get exponentially better with my camera because every time I screw up my camera work I pay for it as an editor. Great way to learn problem solving as an editor and how to improve camera work.

I'd assign Roles, export Stems of the audio, and a proxie of the video. No sync issues, no organization issues. I do this with Logic Pro often and am very happy with it.

But you need to find the workflow that works for you.

No clue what your audio guys is talking about. I'd have to see how the AFF imports into ProTools first.

Julian wrote:
Sound design will take place in ProTools. Color correction will take place in Davinci Resolve or Adobe Speed Grade (what do you suggest?). I will use PluralEyes3 to sync the audio.

I would consider Davinci Resolve for your colour grading and PluralEyes3 FInal Cut Pro X is just horrible when it comes to the syncing audio!

Julian wrote: Right now I'm shooting with the GH2 which really doesn't provide any metadata. But in the future when I upgrade to a camera like the BMCC I just want to make sure that info stays.

Also you can consider a external digital recorder like a AJA Ki Pro or Blackmagic Design HyperDeck Shuttle 2.
I checked out the Atomos Ninja-2 at the IBC and it was pretty nice, I really love the feature that the metadata goes right into FCPX. They even have a Reject and Favourite feature.
Do your media management right on the set.

I've actually used the Ninja and the Ki Pro, love them both very much.

This seems like its going to be the way to go. I'm just going to have to do some practice runs on rehearsal footage.

I also think ill try and do the color correction myself. This will be a fun challenge because as I stated in another thread, I'm red green colorblind so ill have to really get to know how to work with scopes. Do you know of any good books or online resources that go into pretty good detail about scopes?

The main thing I have to be aware of being colorblind is that I like to over saturate and have extreme contrast in my images. I just have to figure out how to make this work for my overall image style.

"The Color Correction Handbook" by Alexis Van Hurkman is the colorist bible.

I have this very comprehensive course online.
www.macprovideo.com/tutorial/finalcutprox205-color-grading

There is also good info in the user manual.
help.apple.com/finalcutpro/mac/10.0.6/#ver761ca98b

Scopes won't tell you everything, you've got a challenge ahead of you. Be sure to have a non-editor/filmmaker review your work. Because your audience are not editors/filmmakers, they are lay folks, and no matter what editors think, if the lay folks don't like it, that's all that matters.

smg669 wrote: I can answer that... noise. The analog to digital converters and preamps in a DSLR are just plain nasty.

OK, you've never used a JuiceLink have you? The JuiceLink DT454 on a Canon 5D is super clear, compatible with any EX or HVX or anything else. It bypasses the camera's built-in preamps. As long as the DSLR allows you to turn the gain down internally, do it. Then pump up with JuiceLink clean preamps. I have a shooter who does this for broadcast all the time and no one can tell he's not using a high end audio recorder. Never any need for sync clips.

I do the same thing with a Sound Devices MixPre going to my Marantz 661 and to the DSLR with an inline Sescom attenuator. You turn the gain down to a tick above zero on the DSLR and feed it a signal tone from the Mix Pre and dial the attenuator back until you are at -12db on the DSLR. Then I have 96/24 audio on the Marantz and 48/16 on the DSLR. If the sound clips on the DSLR, I have a better track to repair with on the Marantz as a backup.

The mic pre's on the MixPre are way better than the Juiced Link or BeachTek and the attenuator lets me get the noise floor way down on the DSLR. And the Marantz is a good security blanket for worrying if the DSLR is really recording. Plus with the Marantz, if there's only a mono mic, I can record the signal on two tracks, one normal and the other 20dB lower to prevent clipping.

Is there documentation about the MixPre having far better preamps than the JuiceLink or BeachTek? Seriously, I'd love to be able to document this for a future article I'm going to be working on.

BenB wrote: Is there documentation about the MixPre having far better preamps than the JuiceLink or BeachTek? Seriously, I'd love to be able to document this for a future article I'm going to be working on.

Here's an interesting link from the guy from Juiced Link. Take it with a grain of salt, but i'd say his product sounds on par with the MixPre... at least when it comes to preamp noise.

I'm not sure there's "documentation", but I've used all three, and I found the ergonomics and the audio quality of the MixPre to be superior. The MixPre is made by Sound Devices who build their stuff like tanks for working sound recordists. The mic pre's in the MixPre are of the same calbre as their higher end models. The JuicedLInk and the BeachTek are recent low priced entries into the DSLR run & gun sound world and I have not seen them used by any professional audio mixers. I realize professional is a loaded word, and I'm not knocking JuicedLink or Beachtek, but the SD stuff is in another league and price point. I realize price doesn't always guarantee quality. Who knows? Maybe I swallowed the snake oil.