Hello,
I’ve been doing a lot of game music, so my longest session has been around 3-4 minutes. I recently worked on a documentary, but it was a nightmare every time the director made edits, as it would throw the entire project off time. What techniques are best to avoid this from happening? Several separate cubase sessions per cue? My rig is capable of scoring in one giant session, just not sure how this is possible musically.
Thanks =)
I’d think it depends on your computer horsepower , how extensive is your orchestration, are you using intensive VIs , do your current versions of each cue sync to picture without any hiccups while all in one file, have you had direction from your director, how does your music editor like to receive your finished cues… answering these will get you closer to an answer for your question I think.
I frequently “save as…” new versions of the project file at certain milestones, .2 .3 etc. Doesn’t take much time or storage as long as you don’t duplicate audio, and makes it easy to recover from a corrupt project file (rare) or a stupid mistake (more frequent).
I’d recommend splitting up the film in several larger sections and making one cubase project for each section. Usually a film contains “chapters” where it feels sort of natural to split. Also, make a master project where you throw in all the bounced cues from each separate project. That way you won’t have to adjust the entire timeline for the whole movie and move parts around etc each time they edit something, and it’ll be easy to keep track of what you’re doing.
And this is very important: ALWAYS burn in time code into the video file and hard-sync your Cubase timeline to that time code. You can use programs like QTSync or similar for that. This way you will always know the exact time position within the full movie when you’re working on individual sections of it. Every time they send you a new edit you can burn in new time code and compare it to your Cubase timeline.
Nice question… this is why DP is so appreciated between film composers. In cubase you could find some workarounds but too far from what DP features have: chunks multi sequence in the same project. I think that adunion answer is the easiest way to get your goal.
I see what you mean, so I would literally load individual templates with locators from 30 - 40min, 50 - 55min, etc? Would I start writing at that time in the template, leaving the other areas empty? I can see this being an issue when consolidating the entire wav for the director at the end, pasting all the separate wav files together could cause problems if even a millisecond is off.
Is there a video on this hard sync to cubase thing? I’m very confused, especially since I’ll be separating it.
There are some good tips in this thread such as save often and save project by the cue name as you go however there are quite a few posts on this forum regarding this topic, so many do what works best for them , having said that I have been doing this for a while so here is what I have learned. I highly recommend running a second PC or Mac to run the video, get it off your primary sequencer , set your sequencer to musical midi mode, and sync with vst system link to your other pc of Mac using the Adat light pipe on your audio cards.
This works excellent for me and gives s perfect sync.
Keep in mind I am working entirely in midi until my score cues are approved by the director and all changes are made and approved, then I do any live recording if needed, batch render audio, do a mix and eq any issues, and FTP audio tracks to the post production house.
Anyhow when you change tempo for cues the video will not loose it’s time location, the music may slide after the tempo change but it’s easy to slid it back, or you can save projects by the cue if you wish, but the video will stay locked to a point in time regardless of tempo changes thus way.
Also I wouldn’t accept a film from a director without time code burned into it. Ask them to burn it.
And wouldn’t do the real scoring without a locked cut ( locked to time) this could be a late rough cut with blank moments that need vfx but you want the time locked.
If your signing a contract to score make the delivery date x weeks after receiving locked cut this way you don’t toil.
If there are some seasoned composers on here doing 5mm or higher budget scores that don’t get locked cuts and don’t get time code burned For you I would like to hear from you out of curiosity. I suspect if your running a dedicated video engine with smpte, running your own time code generation.
It’s pretty simple actually. Say you have a video file ready with timecode visible in it, and you want to work on the part between for example 10:31:00 and 15:25:00. You could do the following:
Now the timecode of the Cubase timeline will be the same as the time code of your video clip. And this way you won’t have to leave lots of empty space on the timeline before you start writing.
When you’re done with this cue, bounce it, and paste it into your MASTER PROJECT containing the full movie, at the correct time code spot. Whenever you need to supply the director with a draft, bounce from 00:00:00 in your master project.
Someone please correct me if I am wrong, but my understanding is that this does not happen with Nuendo.
Not sure about the post above ^ - that’s rather complicated. I guess not once you get used to it…
Here is what I do:
My film scoring template has everything (VIs) loaded in VEPro. Nothing is loaded in Cubase; this makes switching back and forth between projects a cinch. If everything is loaded in Cubase, it would take 15 min to switch projects. With everything loaded in VEPro, it takes about 10 sec. The only thing Cubase contains is the vid file, and the MIDI data.
Let’s say I start from the very beginning of a film. I would load my template (in a folder specifically for this film), import the vid (alternately you can create a new template that contains the vid, but importing the 1 vid file is so quick and easy, I don’t do this), I adjust the tempo to what I feel it needs to be, and I write the opening credits. That goes from the FFOA > 1:45. That’s 1M1, so the Cubase project is 1M1.cpr I take notes, outlining the time that 1M1 runs for. I mix down this cue to audio, then save and close.
The next cue is 1M2. I load my template again (new project), import the vid, adjust the tempo to what I feel it needs to be, and I write. That runs from 2:10 > 3:30. My notes reflect this. Now I have 1M2.cpr I mix down this cue to audio, then save and close.
The next cue is 1M3. I load my template again (new project), import the vid, and I write. And on.
I take the audio files, and send them to the editor, with the time info, so he/she can sync the audio to the pic.
1M1_01-00-08-00.wav
1M2_01-02-10-00.wav
When the director calls me and says 1M2 has changed, I open 1M2.cpr, import the new vid, and make my changes. 1M1 and 1m3 are completely unaffected. Mixdown to audio, and send the editor the new 1M2.
I know that’s a lot to read, but hopefully it helps.
Cheers.
Sorry for bumping such an old thread.
I’m currently upgrading to Ryzen 5 assuming the benchmarks prove accurate. Why not do the whole 1 hour film in a single session?
This way seems like it will cause a major headache, because if the director changes 10:31:00 - 15:25:00, to 10:31:00 - 15:27:00, that will throw everything off afterwards, and possibly before if there are edits before 10:31:00. Now, I will have to open every session, do the math, copy and paste, glue everything together, etc. Since I can’t just see everything, I’ll have to keep a spreadsheet handy.
Apart from horse power, whats the point? I’m I missing something?
Thanks!
I don’t understand how this is possible, if say the opening credits is changed from FFOA > 1:45, FFOA > 1:10, everything after will be occurring 35 seconds sooner. Now imagine multiple scenes are edited to be shorter and longer, aren’t you constantly having to nudge every session/cut you made?
If you look at history because of the nature of actual film the movies were cut into reels, and were worked on in reels as well. So a composer or sound designer or re-recording engineer would work on reels separately. So it’s only “recently” that we’ve had the ability to contain everything within one “super-session”.
I think one reason people still like to work in reels even though they’re “virtual” is that it’s a bit easier to manage large-ish chunks of time. It’s sort of a compromise between a super-session and one session (project) per cue. Less loading like you mentioned, but also less problems with dealing with plugins and automation on the same tracks.
I’m on the re-recording side of things so I don’t really worry about this, but I can sort of see how when you get to really large projects things can tend to get out of hand in super-sessions and it’s just more convenient to manage in virtual reels. And yeah, the more you load your project with heavy plugins and instruments the more you benefit from splitting it up.
And maybe it’s worth considering that we should all charge for our time and that in the past a “locked picture” was actually locked, and if people these days want to keep changing the cut late in the process then they get to pay for our time we spend to address that. In other words switching project files is just part of the process and the client pays for it.
But what do I know?..
Even on “pro” productions, “locked” picture can change. This is where Nuendo’s reconform feature has really saved my a$$.
Oh you poor things , you… lol
I always thought that the sound is the last step after the video timeline is set in stone…lol
