OT - filming a solo performer demo

vicmilt schrieb am 03.01.2007 um 11:04 Uhr

I recently got a very nice letter asking me for lighting and shooting tips. This particular person wanted to film a solo guitarist/singer for a "live demo" to submit to a music club.

Fancy, schmancy video effects were uncalled for.

Here are my suggestions - perhaps they will inspire one of you to take on the same or similar challenges:

Plan to shoot in a simple setting (i.e. don't aggravate yourself with complex lighting setups) - I'd think in terms of a simple roll of
dark grey or dark brown seamless paper, if possible (about $30 bucks US0 or even a 12'x12' painters dropcloth (dye it a dark color if you're ambitious) or (best) see about using a local stage, either in the local school or a club. The idea is to keep the background dark, and the subject the central item in the shoot.

I'd tell the gal you want to shoot each song three times, and that she should try to keep each performance similar. With only one guitar,intercutting between the three performances won't be a big deal.

Ideally I'd mike her with two microphones, one on the guitar, and one on her voice, but I've had great results with the built-in stereo mike on the Z1, IF the local acoustics are good.

I'd light her once with a soft light on her face and a spot slash across her guitar "strum:" hand. Dim the spotlight, so it doesn't overwhelm the softlight on her face.

Since you have two cameras, I'd set the A1 as a close-up camera on her face, and the Z1 as a more general half or even full length shot. Each time she's done with a performance, I'd move each of the cameras about 15 degrees and raise or lower it, not missing
"hawk-eye" and "worms eye" views. Now you'll effectively have six perfomance angles to cut from. Tell her NOT TO STOP if she goofs up - that you'll cut the cameras to get the best amalgam. If she's really good, you can do it in two takes.

The low light sensitivity of the Z1 is not nearly as bad as I had been led to believe (before I bought my own), and for this kind of performance I'd have NO problem with a 6db or even a 9db gain boost. I feel you can get acceptable footage even with a 12db boost, although it is a little noisy. But this is in a "nightclub" so don't be afraid of a little noise - it just lends itself to the whole effect.

The A1 is even less sensative so I'd use that for the head close-ups.

If you only have one camera, then definitely shoot three times.

BTW - you can also do something like this, with lip syncing to a pre-recorded playback, but if she's any good, the live performance will be easiest. With a multi-person band, I'd be much more adament about shooting to playback.

Hope this helps any of you music-video folks out there.

v

Kommentare

farss schrieb am 03.01.2007 um 11:56 Uhr

From my experience shooting a solo piano player recording the piano first and then having them perform to that while recording both instrument and vocals is much easier when it comes to editing multiple takes from different angles.
I was able to keep the replay low enough so the vocal mic didn't pick it up. Keeping every take in some form of sync makes cutting so much easier, you can then sync everything up and do a multicam edit. Plus as I found the performer can then concentrate their attentions on the vocals.

The other tip I've had to give performaers who've never been on camera before is not to jump up as soon as they hit the last note, to let the note fade for a few seconds and then ask if they were OK.

To mic an acoustic guitar (which I've never done) two mics seems to be the norm. I've miced a sarod and wished I'd had more than one mic on it and more isolation from the other instruments as well.
Good acoustic instruments sounds fantastic when miced well and the sound of them can really lift a performance, I sure wouldn't be trusting them to the mics on a Z1, maybe, just maybe use them for ambient to add into the mix but they'll not get right into the guitar and pickup the direct sound from the strings. If you really want to get good ambient then a X-Y mic in the right location would be the way to go if you want to place the sound in the venue.

As for lighting and setting, I agree, keep it simple. If it's a demo reel remember the reel is there to show off the performer, not the video of the performer.

mountainman schrieb am 03.01.2007 um 15:39 Uhr

I’ve been casually talking to a local band about this very subject. And of course the biggest single thing is the audio. There are of course 2 ways to go, live through a mixer or syncing to playback.. The sync method would be easiest for me for sure but… the guitar player is one of the best I’ve ever seen. This guy plays Hendricks better than Hendricks did. If I use the sync method, I’m afraid I’ll lose the nuances of these great performances.
Should I shoot to playback, or bite the bullet and shoot them live? I was planning to shoot each song at least 5 times, one master and the rest iso. With that many takes am I asking for trouble? Is there a better way?
Sync or live, that is the question. JM

farss schrieb am 03.01.2007 um 19:24 Uhr

Shoot them live into a multitrack recorder with plenty of cameras.
Any part of the audio recording that's in need of repair can be overdubbed. This does quickly become a pretty serious effort.

You want to know why DVDs of bands such as U2 live look and sound so good, every concert it recorded multitrack, what you end up hearing is bits from everywhere.

mountainman schrieb am 03.01.2007 um 20:06 Uhr

Farss, your right. Multi-cam, multi-track. But that moves the project from affordable to expensive.
My thought process involves the band, me, an audio guy to mix down to 2 channels recorded to cam and tape, and multiple takes. Have you or anyone else tried it this way?
Lots of editing but that involves only me. Thanks, JM.

farss schrieb am 03.01.2007 um 20:22 Uhr

I've posted one such effort BUT they had a 3 camera shoot, one cam locked wide and two roving cams.
I've also done this myself with one cam and a mixdown from a board. I recorded into the Edirol R-4 so I had instant playback back into the desk. I'd then got the individual musos to play in sync while I shot closeups. Hard as some of them only improvised but overall it seems to have worked.

However many things have caused me pain. Even a multicam shoot of a dance performance has it's issues without genlock. Half a frame is a heck of a long time and that's as close as you can get without genlock. Dissolves look wierd when you can see a solo performers arms in different positions through the dissolve and cuts from XCU to MCU jump. A drummer or guitarist could be even more of a problem. Need to only cut between very wide and XCU shots or you can give the game away or else choose the cuts very carefully.

One issue I was pretty amazed at was the number of audio guys who freak when you say "Stereo", they're sound reinforcement mix guys and not much of that is done stereo.

Bob.

vicmilt schrieb am 04.01.2007 um 00:08 Uhr

Here's another thought...
shoot the band in sync, to multitrack, but w/o the guitarist. Then shoot the guitarist LIVE to the multitrack.

The best of both worlds.

v

mountainman schrieb am 04.01.2007 um 00:09 Uhr

Ohh, genlock. Excellent point. That smoke you smell is me thinking. Genlock multi-cam would be the way to go. Thanks for the tip. JM

GaryAshorn schrieb am 05.01.2007 um 02:46 Uhr

You can do this both ways. I have done several of an artist live on tour in Scotland and Nova Scotia and locally. I have done single camera setups at multuple locations and mixed all the footage back to one audio track from one location and blended them. However, he is both a studio record artist and live performer so all of his songs played on different days all are almost in sync so it was easy. I could match guitar note for note between two locations over a full minute of playing. That is for the live type videos. I have done multiple cameras at one location and used on audio one stereo track of the live performance. I even do multiple cameras at several locations and put them all together. But this takes a performer you can do this with on live shoots that is very constant. Most are not that easy to do if they can not play the same song the same way and timed the same from each session.

The scripted type, then using a standard copy of the song, lip synced etc and then mixed is the better way for that type. You are talking apples to oranges between live and scripted projects. So pick one and then decide. Understand the purpose of the project and what it is intended for the viewer. Then pick the way. Of course if to be done on the cheap then you have boundaries that may limit what you what for what you can do. You can do good work for very little equipment if you plan ahead and understand the final product first.

Gary

PeterWright schrieb am 16.01.2007 um 08:26 Uhr

Time to "out" myself - it was I that asked Vic for some advice, and as always, he responded generously.

I shot the performer singing the song three times, with a Z1 and an A1, so I had 6 angles to choose from.
I recorded the audio via 2 Rode NT1A mics straight into the Z1, and chose the best of 3 performances for audio purposes, although I did "borrow" a few passages from other takes.

The three performances didn't synch up perfectly, of course, but I used Excalibur 5.5 for the multicam edit then went through and trimmed each part back into synch.

I couldn't resist adding bass and string tracks in Acid Pro 6, and I've put a 256Kbps WMV version - about 7Mb Here

Comments welcome.

farss schrieb am 16.01.2007 um 08:58 Uhr

1) Loose the dissolves at the end OR make them organic.
Add GB to the track and keyframe it from 0 to 0.030 in the middle of the dissolve and back to 0.0 at the end. Adjust to taste. Or just make the much dissolves faster.
I get kicked real hard by a mate over dissolves and the more critical I get of my own work the more I realise he's right, seeing the same thing 'double' in the middle of the dissolve is very distracting. If they're two entirely or patially unrelated scenes fine. By the latter I mean they have some visually matched elements, like the horizons match etc.

2)
I would have done something with the back ground, something simple like drapes but the strong vertical from the aluminium window is distracting. If you had the budget get the background to match the mood of the song a little bit more, I know it's probably meant to look like it wasn't a staged performance but...

3) Curious as to why you used the NT1s in endfire.

4) Some of the timing seemed slightly off but that could have been my PC's lousy playback performance.

PeterWright schrieb am 16.01.2007 um 09:12 Uhr

Thanks Bob

I take your point on the dissolves - one reason I had to use them was that I was joining slightly different performances, so a cut would jar, but some I agree could be shorter.
(I'm probably forgetting something obvious, but what's GB?)

The overall thing is that this is a demo of the song and the performer, not of a music video, so I allowed some latitude - it was also low budget - the girl is a friend's son's girlfriend ...

Not sure what you are asking about the NT1s - I used them because they are the best mics I have and I was happy with the overall mix for a live performance - not sure what you mean by "endfire".

Synch could be a tad off here or there - slight variations in performance being one factor, but I'll go through and try a frame + or -.

Thanks for the comments.

farss schrieb am 16.01.2007 um 09:35 Uhr

GB = Gaussian Blur.

Have to laugh!.Your comments are EXACTLY what I keep saying back to my mate (OK he's worked most of his life as an editor) but he just comes back and says. If it looks wrong it's WRONG, the audience doesn't know about or care about WHY.

I'm being harsh, why, cause when people criticise my work I want to know how to make it better. Telling me it's great and I'm really wonderful etc, etc teaches me nothing.

The NT1A is a LDC, it has a gold dot which should face the sound source. That means it's a side fire microphone. Page 10 of the manual shows how to use it to mic a guitar, get a copy from here
Basically the diaphragm is vertical within the microphone so the mic has to be side on to the sound source. The side away from the gold dot is the rejection side, there's two elements in the mic to provide the cardiod pattern.

PeterWright schrieb am 16.01.2007 um 09:47 Uhr

Thanks again Bob.

Yes - I too like to think I accept constructive criticism, especially when I've invited it!

Very interested to read your info on the sidefire topic - must admit I didn't know that, but as I said, the audio sounded ok to me, but next time I'll definitely try the other way - I have always used side-on for audio only recordings without knowing this was the "correct" way!

farss schrieb am 16.01.2007 um 10:49 Uhr

One problemo with the NT1A is by the time you get it into the cats cradle it's frigging BIG. Add the oversized pop filter I use with mine and it's pretty easy to loose a performer behind the thing, strictly side on shots only and then it's really going to have that 'shot in a recording studio look' to it.

Here's a tip from my editor mate but I'll start with my own tip.

Don't spend a week with a 'best mate' directing your unpaid editing, yish, nearly lost a mate of 40 years!

There has to be a reason for a cut, cutting between two similar shots of the same thing doesn't work. Cutting to / from a different part of the same thing is fine, you do it to draw attention to something. But you can be sneaky, you can use it to fill a cut between two shots where you think your only answer is a dissolve.

So in this case cut to a close up of the hand on the frets or the other hand strumming the strings. It doesn't have to be in real sync either. This is the problem with cutting this kind of stuff as multicam. It tends to mentally restrict you. What you might be able to do is forget the multicam edit and just go through the footage to find a few seconds of XCUs that you can use between the wide shots. Ignore the audio of course, ignore that they come from a totally different part of the song. No one will see them for long enough to pick the trick. Even slow down / speed them up in Vegas to get them to cut properly. At the speed the viewer is going to see them one chord is as good as another.

Now, once you've got these totally different shots on the T/L then you can use a dissolve if it adds something.

Bob.

vegemite schrieb am 16.01.2007 um 10:58 Uhr

Hi PeterWright
You asked for feedback, so here goes. It's meant constructively of course! I've had a look at your clip and must say I enjoyed it. She's an attractive young lady and I liked her voice and style.

Now some bits That were not so good; I found the background distracting. If it could have been a plain, dark expanse, it would have been smoother. Still, the exposure of the performer seemed appropriate.

The audio; I hate to tell you this, but you've used the Rode mics the wrong way! These mics have a vertical diaphragm and should have the side of the mic, with the appropriate pattern, pointed to the subject. From memory, they have an indicator mark on the side.

This placement probably acounts for the slightly "thin" sound of her voice -- and she sounds a little "off mic".

Another "flaw" is one you may not be able to avoid. She scrapes her fingernails or rings or whatever rather prominently. It may have been due to the proximity of the RH mic or maybe it's just her technique.

Whatever, congrats on an excellent first attempt, she should be pleased. I'd be putting the hard word for a few more sessions to refine both your techniques. I think there is the makings of something really good.

craftech schrieb am 16.01.2007 um 12:13 Uhr

I shoot strictly stage productions, live performances, and demos.

For a demo the last thing an agent, cd, or producer wants to see is editing or processing or anything that could even SUGGEST electronic tampering. They want to see what they would see if the person were performing in front of them personally at an audition. No close-ups of fingers on keyboards or guitar strings.

In terms of the video I would suggest ONE camera with decent lighting any place that is suitable and DON'T try to "create" a nightclub. If you shoot in a nightclub - then fine. If you shoot somewhere else, just avoid a distracting background, but don't fake anything.

In terms of the audio, THAT is where you need to really concentrate on getting the best "natural" sound. Many of the suggestions above were great suggestions. However DO NOT lip synch anything or add tracks of any sounds or instruments that aren't actually there. The artist may never get a call to come in for an audition from anyone "worthwhile" if they suspect that the sound was "processed" in any way at all. They want to be convinced that what they are hearing is what they would hear if the person were playing at an audition right in front of them. And practically speaking, they may very well play the demo on a computer and output the sound to reasonably good computer speakers instead of a state of the art audio system.

Now if you are splicing in clips from something they did that was on television, for example, that's different. The artist had no control over how it was produced so they cut them slack.

John

farss schrieb am 16.01.2007 um 12:41 Uhr

John,
you make some very good points. I did a demo reel for a cover band that did high paying OS gigs and that was their requirement too, pleny of closeups of the band members, the clients wants to see what they're getting. As it was a band the audio could be 'mixed' obviously but the main thing was to see what they're getting.

However my comments were based on the fact that Peter has already shot this across multiple takes with multiple cams and then cut that.

Sometime ago I did a another demo reel and it's exactly what you're saying, no (obvious) lip syncing, single camera dead boring on a plain background. As it was 20 mins long (client insisted and he's got lots of work out of it and sold many copies) we did the odd cutaway but not in the first few numbers.

craftech schrieb am 16.01.2007 um 12:47 Uhr

Bob,

20 minutes is really long. 3-5 minutes is really what the target time should be. If the producer or CD or agent has lots of prospective talent to review thay don't have the patience to sit through long video. My guess is that the first 5 minutes of the 20 minute video was convincing enough to get them interested. When you say "sold" do you mean that he was actually "selling" his demos? I have never heard of that.

John

farss schrieb am 16.01.2007 um 19:33 Uhr

3-5 minutes?

I'd say you've got even less time than that. If you can get them in the first minute (or less, like if the talent really sucks) then you need more material. Depends on what the performer does though. This guy plays piano bars and he'll probably be the only act for the night and he'll play there 3 or 4 nights a week on a 3 month minimum contract. So he also needs to show that he's got a variety of material and that he can write new material.