Tips for scoring/soundscaping video?

[swishniak]

one tip - pay attention to sample rates! video people tend to capture + export material using 48000 instead of 44100.

[sporkles]

one tip - pay attention to sample rates! video people tend to capture + export material using 48000 instead of 44100.

Heh... Sample rates are the least of my worries; the project I'm working on is a nightmare: the entire soundtrack was recorded with the camera's on-board microphone (a real piece of junk). Fair enough - I'm in the process of replacing the entire soundtrack with my own recordings, but because the actors are not available, the dialogue is, save for one sequence which was re-recorded, absolutely abysmal. I'm doing the best I can, but I'm polishing a turd and it really vexes me.

But, yeah: 48k and 24-bit for film.

[swishniak]

one tip - pay attention to sample rates! video people tend to capture + export material using 48000 instead of 44100.

Heh... Sample rates are the least of my worries; the project I'm working on is a nightmare: the entire soundtrack was recorded with the camera's on-board microphone (a real piece of junk). Fair enough - I'm in the process of replacing the entire soundtrack with my own recordings, but because the actors are not available, the dialogue is, save for one sequence which was re-recorded, absolutely abysmal. I'm doing the best I can, but I'm polishing a turd and it really vexes me.

But, yeah: 48k and 24-bit for film.

ugh - i feel for you; thats never fun. one other thing ive found helpful is to take on a job either as a music composer or sounddesigner (responsible for audio / foley / mixing). doing both of these jobs can be overwhelming, especially if the live sound capture was shoddy.

of course we all take on projects for friends / experience / etc. . but after doing a couple projects like the one youre describing ive tried to avoid it.

[ttilberg]

I'm still on about this a little bit. Been working on various projects for a friend as I mentioned a few months back...

There are youtube videos for everything in the universe, but I'm having a hard time finding one of someone doing some AV work in Live.

So, if you know of one you've seen, please link it, or if you wouldn't mind showing your workflow a little bit and recording your own, that would be awesome.

Specifically related to things like cueing audio to video triggers, and managing tempo in this fashion. I've tried a few things, and it's driving me nuts I feel like I get stuck in "whatever tempo is convenient" but then it's like... "What I'm really feeling here is 100 bpm, but my video cues are requiring me to set 140 (or 70)." And then when doing it in terms of setting to 240 BPM -- I find it hard to get around and keep sanity.

Also if you have other forum resources, or generally decent sites that might help me with this type of project in general, that would be awesome. I have lots of experience making dance music, and lots of orchestral experience, and a small amount of orchestral composition experience... but this feels much more difficult than I assumed it would be!

[ttilberg]

To anybody who digs this thread up in a search, I found this guy on Youtube, who uses Live for video. He started a v-blog discussing how he was organizing a certain project almost from start to finish, posting 10 minutes every day. He also had a video where he shows his technique for getting audio to sync to video for impact.

Sync audio to video with Live: http://www.youtube.com/watch?v=TOgxUevhwh0
Video blog part 1: http://www.youtube.com/watch?v=eqhb9Kg5st0

He rambles an awful lot, but hidden within the rambling are some excellent points he brings up that might help inspire you.

If anyone has other resources similar to this, please post them here!

[Angstrom]

I simply wouldn't attempt to assemble anything in Live. I just use live as a dedicated construction platform. Make the music and foley while watching the video, but nothing more than that. Then assemble video and audio in a dedicated video app.

All this business about running live at a BPM multiple of the frame rate is pretty nasty, it's not going to provide reliable sync and is destroying your only advantage of working in a music app.

Really: it's 1000x easier to use the correct tools for the job.

Compose your music to suit (not match) a rough video edit, and use all the musical tools to make music and audio effects. Now render the music, and the individual foley items to separate audio files and take them into your choice of NLE, into the project with the rough video edit in.
Now, tighten up that video edit to match the music (if required) and put the foley effects on the NLE timeline.

Doing it all in Live is like trying to make a song in FinalCut. Yeah, its possible ... but it's craziness!

[chris vine]

What Angstrom said. The problem of the Live sequencing grid and the timestretching on/off vortex can open a massive can of worms.

[ttilberg]

Doing it all in Live is like trying to make a song in FinalCut. Yeah, its possible ... but it's craziness!

Lol

[sporkles]

That analogy is a little too simplistic, Angstrom. If you zoom in enough on the timeline, Live does a pretty decent job for foley, too. Not everyone has access to proper video editing tools in addition to their audio software; I'm sure that anyone who'd want to attempt to do it all in Live is in the same boat as I: a) they don't normally do video, i.e. don't have access to Final Cut or Premiere. b) they know Live, see the foley/soundscape/soundtrack as a challenge and want to have a go without having to learn new tools.

I've come up with a pretty good workflow since my last post in this thread:

1) I have the foley stuff and the music in separate sets.
2) The foley set is, as mentioned set to 225 BPM, just for simplicity, and I keep all warping off.
3) When I had done most of the foley work, I exported the entire thing as a single, roughly mixed audio file, which I dropped into a new set, and put the video in there, too, both with warping turned off.
4) I then started playing around with instruments, and landed on a 100 BPM for the music.
5) I'm now in the final stages of the music composition, and when I've mixed the music, I'm going to do the inverse of what I did in step 3 - export the music to a single file and import it into the foley set, where I'll finalise the foley and mix it alongside the music.

[ttilberg]

Sporkles, that fits my use case exactly, actually.

The one thing that the guy in those videos I linked to above did regarding the tempo was:
1) Separate each musical scene so each piece can be handled on it's own
2) Load the video into Live and mark key frames
3) Enter the times into a tool to find appropriate BPMs ( http://www.fransabsil.nl/htm/eventhit.htm ), and put your master tempo to that
4) Utilize time signature changes to make sure things are hitting on the proper downbeats, instead of the "2" of the measure. Put the measure before it at 5/4 (or where you want to make up that time).

I've been playing with this a little bit, and once I came the realization that I can use time signature changes to shuffle around audio that isn't to a straight 4/4, or 3/4 -- I'm finding a little more success...

[sporkles]

Yeah, doing it scene-by-scene makes sense for longer projects, where the score complements the rest of the audio. In my current project, though, a 20-ish short, the music is an integral part of the overall soundscape, so I don't want to ruin the fluidity of it by splitting it into smaller chunks.

Thanks for the links - I'll check out the videos, and the tempo calculator looks very useful, although I haven't bothered with getting my clips to start on-beat yet; I just don't see the value in it right now; my objective is to sync the music to the picture, and I can do that by nudging and aligning clips relative to each other.

[ttilberg]

My first proper attempt at writing music for video! The tips I gathered from the above video blog I referenced helped a lot.

http://www.youtube.com/watch?v=GLefrwi2 ... Uv708e9v3A

[stringtapper]

Heh... Sample rates are the least of my worries; the project I'm working on is a nightmare: the entire soundtrack was recorded with the camera's on-board microphone (a real piece of junk). Fair enough - I'm in the process of replacing the entire soundtrack with my own recordings, but because the actors are not available, the dialogue is, save for one sequence which was re-recorded, absolutely abysmal. I'm doing the best I can, but I'm polishing a turd and it really vexes me.

But, yeah: 48k and 24-bit for film.

Wish I had seen this earlier. I had a similar situation on a full-length film I did post on a few years back. Audio recorded in different environments; air conditioning sounds; car sounds; fucking bird sounds at one point

Izotope RX saved my life on that project. Being able to remove background noise and even out the dialog helped immensely. The Spectral Repair tool in particular performed a few miracles.

I use that app every day in my work and every day it amazes me with what it can do.

[sporkles]

Wish I had seen this earlier. I had a similar situation on a full-length film I did post on a few years back. Audio recorded in different environments; air conditioning sounds; car sounds; fucking bird sounds at one point

Izotope RX saved my life on that project. Being able to remove background noise and even out the dialog helped immensely. The Spectral Repair tool in particular performed a few miracles.

I use that app every day in my work and every day it amazes me with what it can do.

Thanks for the heads up! I downloaded the demo, but I couldn't immediately wrap my head around it; I've achieved arguably better results with the Noise Removal tool in Audacity.

[stringtapper]

Thanks for the heads up! I downloaded the demo, but I couldn't immediately wrap my head around it; I've achieved arguably better results with the Noise Removal tool in Audacity.

Denoiser Tool. Find a piece of silence within the recording and select it, doesn't have to be too long. Then Click "Learn" on the Denoiser. Then select the whole document and click "Process". That will remove the background noise. You can also adjust how much it attenuates. There's a learning curve but I guarantee it can do better than Audacity (or anything else).