Most convincing way to add source/diagetic music to a scene?

[Supertramp]

Hi, I'm starting to edit my first short film and I'd like to add diagetic music to a scene. The dialogue and room tone was recorded with lavs (unfortunately I had no other option). What's the best way to add music to the scene? Should I replace the room tone track with a live recording of the music playing on the same set?

I'd really appreciate any advice, thanks.

[Gregg Juke]

Could you be a little more specific about the scene (setting, any shots that include said "music performance," etc.), and the music in question? Not meaning to be wank-ish, but without further detail, you probably won't get any answers, or any that you do will sound as random and vague as the question seems to... Is it a one-location film? Is it in a bar or a concert hall? Is the music source visible, or are you trying to imply something off-screen? Is the music original, and is it written/arranged yet? What resources (music, audio production tools, and performer and instruments/musical props) are available to you? Are we talking accapella vocals, acoustic guitar, full band, orchestra, or a completely made-up Klingon organ type of thing?

GJ

[Supertramp]

Thanks for replying - that doesn't sound wankerish at all, I was really unclear and non-specific. I think I just assumed for a moment that people could read my mind. It's a scene with two people having a romantic dinner in a modest kitcken/dining room. Some old school jazz music is playing on a music player that is not visible. The music player should have been incorporated into the set at the time of shooting, but that's one of many mistakes that I've learned from. The film itself takes place in one apartment. Exterior, bathroom and kitchen/dining room.

[paulears]

That's not digetic music - this is when the source of the music can be seen - as in the bloke playing the piano in the background, so if the source of the music, as in the hifi system in the room can't be seen, then it's simply background - so pick the music and stick it up! I'd suggest eqing it to remove quite a bit of the HF end - leaving the mid and low end , and the ear will detect the lack of HF and dismiss it as what you hope they will - a room stereo system.

[Noiz2]

Thanks for replying - that doesn't sound wankerish at all, I was really unclear and non-specific. I think I just assumed for a moment that people could read my mind. It's a scene with two people having a romantic dinner in a modest kitcken/dining room. Some old school jazz music is playing on a music player that is not visible. The music player should have been incorporated into the set at the time of shooting, but that's one of many mistakes that I've learned from. The film itself takes place in one apartment. Exterior, bathroom and kitchen/dining room.

NO you should not have had it playing when you shot!!!!! That would totally have hamstrung your editing.

Ignore the "diagetic" part. I don;t know anyone in film post who uses that term. You have "score" and you have music that happens in the scene. Make no difference if it's "seen" or not.

You can rerecord the music in the right space, or one that sounds close. Or you can use a bunch of plugins to try and match the environment. I would go with the first option. NO you don't replace, you add on. Ideally you will want 30 or so tracks of sound. If you are going with a lot less you are not doing a real sound post. Even if you are not doing post you want to add the sound in not replace existing sound.

[Supertramp]

I thought diegetic (I think we all spelt it wrong) referred to anything that exists within the narrative, regardless of whether it is seen or not. Was just trying to be specific about the issue. I tried what Paulears suggested and it does work reasonably well, thanks - if all else fails it's pretty decent.

But yes, my audio is unfinished - thanks Noiz2, I do need to add more tracks to make the environment sound more natural. Especially because I was working with lavs. I should clarify - I meant that the visual of the music player should have been incorporated at the time of shooting, so that the audio would be motivated. But 30 tracks...you mean all on top of each other or occurring throughout a scene? The former is way beyond me. The project won't be going anywhere other than the internet. I'm trying to do the best I can, but I am totally new to sound so I'm definitely not doing a "real sound post". It's a no budget short so I don't have the resources to hire a professional...and I live in the middle of nowhere in Japan. All I've got is myself! The only other people involved in the film were two generous friends with no film or acting experience.

I'm going to try rerecording it on the location so, to see how it compares. Thanks for you help guys, I'll post it when I'm finished.

[Noiz2]

30 all at once, though not all would probably play at the same time. If you read the stickies on post there is a much more detailed breakdown but you should probably have a dialog track per character, and you will need 8-16 for FX, 4-8 stereo for backgrounds, some for Foley, some for music.

It's not beyond you but it might be more than you want to spend (time wise) and it may be more than your film needs.
The basic idea is to break down the sounds and build up the FX. So instead of say finding a "getting in a car and driving away" sound FX that kind of works, you break down you needs and build up the sequence. Car door open and close (that works for you) maybe some getting in Foley, a car start that you like and an away that you like. Add some tire on gravel etc. Add in some BG (background) sounds, maybe some wind, some birds, distant traffic, etc. Now you have a real world for your character to live in.

[Supertramp]

Well I've taken much of that advice to heart and am still working on it, but it's coming together - room tone made a significant difference in making everything sound more natural. I've uploaded a private clip with temp music, I think the effect works and the music sounds like it's within the scene, but I'm not sure if my perspective is warped. Might also insert some more dialogue to clarify things - especially the fact that the hair is actually hair. Sorry, the scene is a minute away from the end of the short and probably makes no sense to anyone watching it in isolation, but I'm trying to work on the technical sound of things. Does it sound "off"? I know I'm missing some spluttering towards the end and perhaps some saliva sound for the piece of hair. I feel like one part that bothers me is the part where the male character stands up and moves to the other side of the table. It feels like there should me more noise, but I'm not sure what I should do? Perhaps that's more to do with a lack of vocal noise from the female, Im not sure, but for a relatively "big moment" the noise sounds a little small. Perhaps I also need more cutlery or food sounds throughout the scene. I probably won't try to include an instrumental layer of audio during the "moment of significant" in the final product.

Actually I do have one very specific question - the video below is in mono, but I have the raw vocals recorded in stereo, his voice is primarily on the left channel, hers on the right (except for the second last shot). My problem is that the stereo is really exaggerated, I love the effect but I want to tone it down. Is there any way I can half blend the left channel with the right channel on the vocal track? By that I mean, move some of the audio from the right track to the left and vice versa, but not so much that they equal each other. Probably a dumb question, I'm really bluffing my way here, but if anyone can help that would be great. I used Premiere Pro.

Sorry the video is incorrectly formatted for vimeo, so it's a bit squashed :/ Password is yum

[Egg Born Son]

There were a few things I thought were off until the punchline, then it all made sense. I like it. I thought the music turned out okay but some finessing of the levels might make more impact, sort of float between scene motivated and score. If the levels go down a touch in during the dialogue and up at the action. You could even try riding the levels through the whole scene according to the emotion, control the intimacy with volume (down more intimate, up less intimate). The interplay between them is quite dynamic once you realise there is double-meanings and talking a cross-purposes at play. If you wanted to make the effort I think you could really take this to the next level. After the blood squirt sound after he says "least expect it" maybe even go into full score mode, drop off all the filters, turn up the volume, maybe even drop all other sound. It could be a powerful effect, just visuals and music. While I was watching it I thought the music selection was close but inappropriate as mood music. But by the punchline the slight disconnect really helped to set up what was to come and by the kill scene works very well. The conversation gets increasingly bizarre requiring more attention from the audience and subtly messing with the sound could be a real conduit to the audience's subconscious done right.

Suggestions aside, for what you're trying to achieve I think this sample is more than acceptable.

[Egg Born Son]

Actually I do have one very specific question - the video below is in mono, but I have the raw vocals recorded in stereo, his voice is primarily on the left channel, hers on the right (except for the second last shot). My problem is that the stereo is really exaggerated, I love the effect but I want to tone it down. Is there any way I can half blend the left channel with the right channel on the vocal track? By that I mean, move some of the audio from the right track to the left and vice versa, but not so much that they equal each other. Probably a dumb question, I'm really bluffing my way here, but if anyone can help that would be great. I used Premiere Pro.

Bounce or copy the stereo track, reduce the volume of the copy and flip the channels. Sorry, can't give you specific instructions for premiere pro but they are very basic functions, the means will be in there. Don't invert the phase, it will cancel out and knock body out of the dialogue. You can then adjust the volume of either stereo track to achieve the balance you want.