Remake the dialogue in Post!?
 

Hello,


I was considering to start dubbing short movies in post (the hollywood way). I meen recording all dialouge in a studio in post-production. This may be the only way to get a real proffessional soundtrack to movies. Especialy on exterior shots.

I figured that I'd use a program like, let's say Cubase 5.0. And input video to the program. Then let the actors synq their voices real-time along with the picture. And do this over and over again, until it fits perfect.


Am I thinking the right way? Is there some special program/method for this. Has anyone experience with this?
Also, if I synq sound to picture, then I'd really need to output the picture to a tv-monitor, allowing the actors to see detail for good lip-synq. Is this possible with realtime audio recording? Last, any special suggestions for microphones in studio enviroment? should I go with a dynamic or a shotgun mike?


Thanks a lot in advance,



Regards,


Erik

Hello Erik,

I have tried dubbing using a minidisc recorder and Premiere 6

Basically we had shot footage in a car and the sound was completely horrible, and so decided to give dubbing a go.

What we done:
Turned down the volume on the PC speakers, played the movie in front of the actor and placed a microphone in front of him/her, got them to 'read the picture' for a couple of goes and then recorded there line on minidisk.

Captured the audio in premiere, dumped it into the timeline, zoomed into to about 12 frames and aligned the audio.

This worked really well, although I would recommend recording the voice in a separate room to the computer (there was a bit of humming which was not pleasant).

There is a little bit more about this story at this link:

http://www.dvinfo.net/conf/showthread.php?s=&postid=7152#post7152

Hope this helps, if you plan to go down the gorilla style route,

All the best,

Ed Smith

This process is called "looping" or "ADR" (for Automatic Dialogue Replacement). I've never actually tried it, but will need to be soon, in the best prosumer way I know how. Here's the recipe I'll be cooking up.

INGREDIENTS:
1 MiniDV tape source footage with dialogue to be replaced
1 Blank MiniDV tape
1 Video editing workstation (Canopus DVStorm-based in my case)
1 DV-standard video camera (XL1s in my case)
1 Good microphone
2 Pairs good isolative headphones (I like SONY MDR-D77)
1 Microphone mixer (My favorite is a cheap Radio Shack 32-1106 4 channel stereo microphone mixer)
1 Acoustically treated room (recording booth)
1 Silent video monitor, e.g., small LCD monitor
1 Bottle water
Assorted cables (Firewire, XLR

PHASE ONE:
Load source footage into video editing workstation with Firewire cable. Cut up source footage into chunks consisting of one or two lines of dialogue. Save each chunk as a separate project file in editing software. Set timeline to "loop." Place good microphone, silent video monitor, and 1 pair headphones into recording booth. Run video output from video editing workstation to input of silent video monitor. Run audio output from video editing workstation into headphones inside recording booth. Run audio output of micrphone to audio input of microphone mixer, and audio output of microphone mixer to audio input of DV camera. Mixer and camera should be outside recording booth. Place fresh tape into DV camera. Run headphone output of DV camera into second pair of headphones. Await arrival of talent.

PHASE TWO:
Escort talent into recording booth. Booth should be acoustically treated for minimum reverberance. (If you can clap and you hear a flutter, your recording space is no good. Recording booth should also contain no sources of electrical noise, e.g., fluorescent lighting. There's more to a good recording space than this, of course, but this is a recipe, not an acoustics course.) Instruct talent to don headphones, speak clearly into microphone, watch monitor, sip water when required. Don second pair of headphones and ask talent to speak clearly into microphone. Adjust audio levels on mixer. Hit playback on video editing workstation. Let video footage with dialogue loop a few times to allow talent to get into the rhythm of the lines. The talent will be able to see the footage of themselves speaking on the video monitor and hear the old dialogue through their headphones. (It's important the headphones are isolative so that the microphone doesn't pick up the audio from the headphones.) After a few cycles, the talent should begin speaking the lines in synchronization with the footage. The talent should be made that at this stage lip synchronization is more critical than dramatic reinterpretation of the dialogue! Hit record on the DV camera, which is acting as the audio recording device. (The DV standard allows for 48 kHz, 16-bit audio, which is plenty good for prosumer work.) Monitor new dialogue over your headphones to insure noise levels are tolerably low--if there's noise, check the signal chain pathway for noise sources. (Dirty AC is often a culprit, so I recommend using battery power for microphone, mixer, and DV camera.) After a few good takes, turn off camera, and load up the next project file with the next line of dialogue to be looped. Repeat looping process until there are several good takes on tape for every line of dialogue. When finished, thank talent and offer him or her the remainder of water in bottle as a present.

PHASE THREE:
Load all audio from dub session tape into video editing workstation and begin tedious process of cutting and lining up replaced dialogue. For more precise audio alignment, use software which permits of sample-accurate audio placement. Premiere and Canopus Editing packages only allow for frame-accurate audio placement.

Thanks a lot for the info guys. Those methods may work good.

I produce own music (as a hobby), using different programs such as Cubase 5.0. Those programs basicly do all the synqing for you. And they show video at the same time. For me it sounds like a more efficent way. I don't know. Do you guys know if any program like Premiere etc. can record audio while playback video and audio?




Regards,



Erik

Hello:

To answer your latest question: I don't know any quality programs that can record sound and sync it with video. If Cubase 5.0 can record live sound as your doing a video playback (and syncs the sound with the video), I'd use that program. I like Cubase for the most part anyways - I just recorded my latest "album" with a Cubase this past winter. Anyways, once you have files of synced dialogue with video, import those files into Premiere and begin your NLE work.

For the mic, to use you might want to go with a ribbon mike that they use for radio broadcasts. I've used some moderate-priced omni-directional, dynamic mics before. They sound alright for a rock-concert, but for studio sound, you want something a little more clean. Just make sure you get a good, low-Z mic. Since you're considering doing voice-overs, perhaps a lav mic would be good. I'm not too sure. Generally, lav mics are used for doing interviews (which is like what you're doing with voice-overs). I wouldn't use a shotgun mic in a studio setting.

As far as editing and handling the actors for the voice-overs, here's my cheap-o way of how I'd handle it (you may think its stupid and backwards but. . .). When you shoot your movie in the field, get lots of b-roll footage. In other words shoot other things other than the actor saying the lines; shoot the other actor listening - reacting. Shoot something from the location. Shoot extra things (like if you're actor is in a car and the dialogue gets tense, get a shot of the actor's handles getting white-knuckled, gripping the steering wheel tighter . . . stuff like that). Go into where you record your sound with Cubase (unplug your speakers and have sound running through quality Sony earphones). Have the actor practice the dialogue a couple of times - record these even though you're saying to the actor that its practice. Then say that you're recording for real. Honestly, you're probably going to get the best, realistic, good-acting diaglogue within either the two or three practice lines, or the first 2 "real" lines. I wouldn't do more than that - the actor and voice probably begin to sound mechanical (like recording a rhythm guitar over and over after messing up).

Ok, you've got the natural, awesome sound, but it probably doesn't sync exactly with the lips in the video. But, it sounds good - like the emotion you're looking for is there. Thats where the editing in Premiere takes place. When you've got the voice synced well with the lips, put that shot in the movie; when the dialogue of the lips begins to become out-of-sync, cut to the other actor, cut to the fingers gripping the steering wheel. Come back to the talking actor when the voice becomes in-sync again.

That may sound like a cheesy-way to edit. But, I'd rather have natural sounding voice-overs and get them quickly so my actor doesn't have to keep going over the same line millions of times, sounding less and less human. Of course, if you're giving the actor money, thats what they're paid to do . . . go over things millions of times and sound good. But, I'm really into shooting fast and getting a natural acting shot. Oh well. Thats my two-cents, you don't have to do it.

Regards,

Kyle "Doc" Mitchell

Thank you very much Kyle. I have a pretty good dynamic mike that I'll use. And I will really remember the acting tip (record at practise), as it's often also a good thing when you shoot the footage. Telling your talents to do a practise shoot while the camera is rolling.


Thanks again Kyle,



Regards,


Erik