ADR sanity check at DVinfo.net

Shaughan Flynn · January 20th, 2010, 01:13 PM

Has anyone done ADR here?

I have to do it for a scene in a film that my wife is posting. My approach is this:

For a given shot, I use FCP to repeat the shot 3 times. The first 2 with a guide track hard panned to the left. I recorded a cue tone audio file that consists of 1 frame of 1Khz tone burst. I use 3 bursts separated by 20 frames. Then 20 frames following the last burst, the line to be recorded occurs. I also burn TC into the shot at the bottom of the frame to match the cues in the ADR script.

The guide track and cue tone occurs on the first two iterations of the scene followed by the 3rd iteration that has no guide track. This way the talent sees the shot with some lead in along with the tone cue twice and then can attempt a take on the 3rd iteration.

I output this from FCP and load it into Logic Studio 9 as a movie into the video global track. I then have an audio track for the talent to record the ADR onto and this is panned hard right. I figure I can set a cycle region over this set of 3 video clips, use take folders and just let the talent do as many takes as they need to get it right.

Any suggestions here to improve this or is this a reasonable approach?

TIA!

Jon Fairhurst · January 20th, 2010, 01:55 PM

In Vegas, the process is really simple. It's based on punch-ins for multi-track recording. You set a loop that will repeat. You set in and out points for the recording. That gives the actor/musician some lead-in of previous lines/sounds/video before the recording starts, and a second or two after the recording ends, before it loops to keep things from feeling too rushed.

When you record, Vegas keeps each take. You just keep looping and recording over and over, until the actor/director/soundguy are satisfied. On playback, you can quickly switch between the various takes while looping for quick comparisons. When satisfied, just leave the best take as the selected one and save the file.

You can either work phrase by phrase or with longer sections. It's up to you to set the loop and in/out points.

I'm not inferring that you should switch from FCP to Vegas. I'm just saying that if you can use a punch-in process right in your editing tool, it's really simple.

Also, I'm not sure why you would pan the audio. I prefer mono to the headphones. It sounds more natural than a hard pan in a can. You should be able to control the playback levels and panning with the actor, based on the actor's feedback during the session.

Robert Turchick · January 20th, 2010, 02:16 PM

Back when I did this on a regular basis, the tones were always heard. Sounds like you're on the right track (pun intended) Last time I did ADR though, I was using BetaSP as a master driving pro tools and sending TC and audio over ISDN to Los Angeles where it drove the client's BetaSP and Pro Tools. ridiculous setup but they wanted the redundancy as it was Dom Delouise!
I have always used Pro Tools which allows multitake looping which is convenient. also made it really easy to stretch or compress even syllables to match exactly. Logic will give you the same flexibility. Record away!

Gary Nattrass · January 20th, 2010, 02:41 PM

The voice over utility in FCP has cue tones available and is really easy to use as it gives a pre roll on the clip and you can just use the markers to highlight your record window, it also allows multiple takes.

Ive done lots of ADR at pro levels and have used some very good visual wipe systems but have found with people not used to it the best way os to play them the line to replace and try to get them to reproduce it parrot fashion. I then use cut and paste to match although I now have pro tools so have a software waveform matching package.

Shaughan Flynn · January 20th, 2010, 03:59 PM

Quote:

Also, I'm not sure why you would pan the audio. I prefer mono to the headphones. It sounds more natural than a hard pan in a can. You should be able to control the playback levels and panning with the actor, based on the actor's feedback during the session.

I asked a guy at Glen Glen ADR studio in Santa Monica about this process and he described the panning config to me so I followed that in an effort to appear as though I know what I am doing :)

Vito DeFilippo · January 20th, 2010, 09:28 PM

As an actor I've done ADR several times. Some with the visual wipe that Gary describes, other with beeps for cues, sometimes just by hearing the line a few times, then punching it at the right moment during the final playback. The latter seems more common for voice gigs.

The easiest for me were the beeps, or the punching in. The visual wipe worked, but was hard to get relaxed with. I would avoid the latter...

Shaughan Flynn · January 21st, 2010, 01:16 AM

Yeah the wipes (called streamers) I think are overkill for what I will do here. There is a nice, inexpensive software option for generating them: Figure 53 | Streamers | Visual Cues for Film Scoring, ADR, and Foley

Usually they are generated by an expensive piece of hardware. You can see them in use in this video here: YouTube - Sound Design for King Kong (Post/production) 3 of 7

Jay Massengill · January 21st, 2010, 02:12 PM

A side-note on the mic technique rather than the software procedure.
I often find the ADR audio is so noticeably different from what is before and after it, that the fix becomes disruptive even if you've matched sync perfectly.
I would strive to record it in the same manner and same type of room if it was an interior scene as the original material (minus the problems that are making ADR necessary of course).
In other words I'm willing to get 90% good audio that fits the scene perfectly rather than 100% "perfect" audio that is noticeably out of character.

Forgot to mention:
Use sealed headphones or in-ear monitors for the talent.
If the talent is monitoring their own live voice, make sure you have a zero-latency return path for them. This is easiest to do if you use an external mixer. The mixer receives their mic signal from the talent as well as the output from your playback audio track. The return to the talent contains the playback audio and their live voice returned from the mixer directly, but it would not contain the return from their recording track on the computer which may have latency. Some people do better with none of their live voice returning in the headphones at all, it will be up to you and your talent as to what is working best.

Shaughan Flynn · January 21st, 2010, 05:51 PM

Thanks Jay. I am lucky in that the audio for the entire scene is not usable or is just missing completely. It takes place in a noisy market and the rest of the film is quiet interiors all with good production audio. So I can fake the market sound and I won't have a mismatch.

Robert Gordon · February 1st, 2010, 03:25 PM

Soundtrack Pro has the ability to loop and set up in and out points. Then there is the multi-take editor which makes dialog editing a (relative) breeze. So I'm told!

January 20th, 2010, 01:13 PM	#1
Shaughan Flynn Major Player Join Date: Aug 2005 Posts: 272	ADR sanity check Has anyone done ADR here? I have to do it for a scene in a film that my wife is posting. My approach is this: For a given shot, I use FCP to repeat the shot 3 times. The first 2 with a guide track hard panned to the left. I recorded a cue tone audio file that consists of 1 frame of 1Khz tone burst. I use 3 bursts separated by 20 frames. Then 20 frames following the last burst, the line to be recorded occurs. I also burn TC into the shot at the bottom of the frame to match the cues in the ADR script. The guide track and cue tone occurs on the first two iterations of the scene followed by the 3rd iteration that has no guide track. This way the talent sees the shot with some lead in along with the tone cue twice and then can attempt a take on the 3rd iteration. I output this from FCP and load it into Logic Studio 9 as a movie into the video global track. I then have an audio track for the talent to record the ADR onto and this is panned hard right. I figure I can set a cycle region over this set of 3 video clips, use take folders and just let the talent do as many takes as they need to get it right. Any suggestions here to improve this or is this a reasonable approach? TIA! __________________ B-Scene Films

January 20th, 2010, 01:55 PM	#2
Jon Fairhurst Inner Circle Join Date: May 2006 Location: Camas, WA, USA Posts: 5,513	In Vegas, the process is really simple. It's based on punch-ins for multi-track recording. You set a loop that will repeat. You set in and out points for the recording. That gives the actor/musician some lead-in of previous lines/sounds/video before the recording starts, and a second or two after the recording ends, before it loops to keep things from feeling too rushed. When you record, Vegas keeps each take. You just keep looping and recording over and over, until the actor/director/soundguy are satisfied. On playback, you can quickly switch between the various takes while looping for quick comparisons. When satisfied, just leave the best take as the selected one and save the file. You can either work phrase by phrase or with longer sections. It's up to you to set the loop and in/out points. I'm not inferring that you should switch from FCP to Vegas. I'm just saying that if you can use a punch-in process right in your editing tool, it's really simple. Also, I'm not sure why you would pan the audio. I prefer mono to the headphones. It sounds more natural than a hard pan in a can. You should be able to control the playback levels and panning with the actor, based on the actor's feedback during the session. __________________ Jon Fairhurst

January 20th, 2010, 02:16 PM	#3
Robert Turchick Trustee Join Date: Oct 2009 Location: Mesa, AZ Posts: 1,389	Back when I did this on a regular basis, the tones were always heard. Sounds like you're on the right track (pun intended) Last time I did ADR though, I was using BetaSP as a master driving pro tools and sending TC and audio over ISDN to Los Angeles where it drove the client's BetaSP and Pro Tools. ridiculous setup but they wanted the redundancy as it was Dom Delouise! I have always used Pro Tools which allows multitake looping which is convenient. also made it really easy to stretch or compress even syllables to match exactly. Logic will give you the same flexibility. Record away! __________________ The older I get, the better I was!

January 20th, 2010, 02:41 PM	#4
Gary Nattrass Trustee Join Date: Apr 2008 Location: Cornsay Durham UK Posts: 1,992	The voice over utility in FCP has cue tones available and is really easy to use as it gives a pre roll on the clip and you can just use the markers to highlight your record window, it also allows multiple takes. Ive done lots of ADR at pro levels and have used some very good visual wipe systems but have found with people not used to it the best way os to play them the line to replace and try to get them to reproduce it parrot fashion. I then use cut and paste to match although I now have pro tools so have a software waveform matching package. __________________ Over 15 minutes in Broadcast Film and TV production: http://www.imdb.com/name/nm1044352/

January 20th, 2010, 09:28 PM	#6
Vito DeFilippo Trustee Join Date: Mar 2006 Location: Montreal, Quebec Posts: 1,585	As an actor I've done ADR several times. Some with the visual wipe that Gary describes, other with beeps for cues, sometimes just by hearing the line a few times, then punching it at the right moment during the final playback. The latter seems more common for voice gigs. The easiest for me were the beeps, or the punching in. The visual wipe worked, but was hard to get relaxed with. I would avoid the latter... __________________ . http://www.nosmallroles.com

January 21st, 2010, 01:16 AM	#7
Shaughan Flynn Major Player Join Date: Aug 2005 Posts: 272	Yeah the wipes (called streamers) I think are overkill for what I will do here. There is a nice, inexpensive software option for generating them: Figure 53 \| Streamers \| Visual Cues for Film Scoring, ADR, and Foley Usually they are generated by an expensive piece of hardware. You can see them in use in this video here: YouTube - Sound Design for King Kong (Post/production) 3 of 7 __________________ B-Scene Films

January 21st, 2010, 02:12 PM	#8
Jay Massengill Trustee Join Date: Jul 2003 Location: Burlington Posts: 1,976	A side-note on the mic technique rather than the software procedure. I often find the ADR audio is so noticeably different from what is before and after it, that the fix becomes disruptive even if you've matched sync perfectly. I would strive to record it in the same manner and same type of room if it was an interior scene as the original material (minus the problems that are making ADR necessary of course). In other words I'm willing to get 90% good audio that fits the scene perfectly rather than 100% "perfect" audio that is noticeably out of character. Forgot to mention: Use sealed headphones or in-ear monitors for the talent. If the talent is monitoring their own live voice, make sure you have a zero-latency return path for them. This is easiest to do if you use an external mixer. The mixer receives their mic signal from the talent as well as the output from your playback audio track. The return to the talent contains the playback audio and their live voice returned from the mixer directly, but it would not contain the return from their recording track on the computer which may have latency. Some people do better with none of their live voice returning in the headphones at all, it will be up to you and your talent as to what is working best. Last edited by Jay Massengill; January 21st, 2010 at 02:25 PM. Reason: Additional material.

January 21st, 2010, 05:51 PM	#9
Shaughan Flynn Major Player Join Date: Aug 2005 Posts: 272	Thanks Jay. I am lucky in that the audio for the entire scene is not usable or is just missing completely. It takes place in a noisy market and the rest of the film is quiet interiors all with good production audio. So I can fake the market sound and I won't have a mismatch. __________________ B-Scene Films

February 1st, 2010, 03:25 PM	#10
Robert Gordon New Boot Join Date: Aug 2009 Location: Raleigh, NC Posts: 19	Soundtrack Pro has the ability to loop and set up in and out points. Then there is the multi-take editor which makes dialog editing a (relative) breeze. So I'm told!