Suggestions for specific recording situation

[Tommy2013]

Hi all,

Thanks for your help.

I have a bit of an interesting recording situation and I'm trying to figure out the most effective way to manage it.

Basically we want to record a tutorial where a teacher shows how to do things on a smartphone. The format we envision is a picture-in-picture style, with the talking head on the left side of the screen and the phone on the right.

To accomplish this we will need a video camera focused on the teachers head and a video camera focused on the smartphone. We need to actually see the finger moving things on the screen so trying to capture some sort of video-out from the phone is not an option.

One element we are trying to accomplish is to make this a simplified setup (if possible) so if I am not around the local staff can more or less press a couple of buttons and start recording. I would eventually be in charge of putting the final product together. If that's not feasible than I would be on-site the whole time (2nd best option).

What seems easiest would be to simply have the two camcorders setup and in position. Press "record" on both and just leave them running for as long as it takes to finish the tutorial (i.e. not stopping for edits or corrections, etc.) so I would have two long videos which I could bring into Adobe Premiere, and as long as I synch the start points, they would be in synch for the entire storyboard. As long as I cut both streams as I mix I would be able to edit it fairly easily.

Another option we discussed would be to run the hdmi out from both camcorders into a video mixer, generate the picture-in-picture from there on the fly, and run a single hdmi out to a laptop that captures the resultant live stream. This would potentially give the videographer (assuming I am not there) a more realistic view of what the final product would look like as they would have sort of a "live" monitor.

The second option is more complicated obviously, so runs against the simplified approach (and would require two more components, the video mixer and laptop). However, assuming I am on site, maybe this would potentially be a preferred approach?

Just wondering if anyone has any thoughts or suggestions?

Thanks,

Tom

[DB83]

The second option requires more than two more components as direct capture over hdmi to a laptop is a non-starter. You would need a capture device between them. But maybe your terminology is wrong and you actaully meant DV which could then, theoretically, capture the video stream. I say 'theoretically' since it is not so easy to do any form of capture direct to a laptop.

The easiest method is the two camera aproach.

[budwzr]

Why the talking head?

[Tommy2013]

The second option requires more than two more components as direct capture over hdmi to a laptop is a non-starter. You would need a capture device between them. But maybe your terminology is wrong and you actaully meant DV which could then, theoretically, capture the video stream. I say 'theoretically' since it is not so easy to do any form of capture direct to a laptop.

The easiest method is the two camera aproach.

Thanks DB. I tend to agree about the two camera approach.

I was actually considering the Blackmagic HDMI to USB 3 converter, which looks pretty interesting and at least the Thunderbolt version has gotten good reviews: http://www.blackmagicdesign.com/products/intensity/models/

[Tommy2013]

Why the talking head?

Unfortunately the spec from the client...I'm trying to talk him out of it.

[DB83]

Even with the Blackmagic you still have the issue with the laptop to consider.

You have one HDD and anything, particularily the OS, that decides to write to the drive at the same time as you are doing a capture can screw that up. At least with the cameras you have an independant recording and an opportunity to do multiple captures.

[Cornucopia]

A third, and even more complicated but much higher rez, option would be to:
1. Perform the tutorial once, live with 2 cameras (1 shooting the talking head/torso, 1 shooting the smartphone). Always remember to slate/clapper each take for better/easier sync.
2. Afterward, in sync with the tutorial playback, take phone screencaps (there are apps for this) of the main shots/highlighted moments and afterward transfer the pix to your edit system
3. Later, also in sync with the tutorial playback, shoot a "finger body-double" of the finger action, using a solid green as the phone's screen (there are apps for this if you don't already have it). I would use some semi-hidden vise grip to "hold the phone" steady so that the finger is isolated nicely (and later do a crop around the edge of the phone so the grip doesn't show). This will be your "green screen" foreground.
4. Then in editing, use your original "phone shot" as a reference with which to replace with your screencap shots, then overlay the "finger-motion-greenscreen" shots on top of that (you could even do a corner-pinned PIP replacement of the greenscreen section to make it exactly match perspective). Use the correct compositing mode to include "finger shadow".
5. Then composite your Talking Head shot with your Screen+Finger composite and add in SFX as desired and you will have a finely-tuned master composed of the highest resolution elements possible.

Like I said: more work, but you couldn't get better quality.

Scott

[Tommy2013]

A third, and even more complicated but much higher rez, option would be to:
1. Perform the tutorial once, live with 2 cameras (1 shooting the talking head/torso, 1 shooting the smartphone). Always remember to slate/clapper each take for better/easier sync.
2. Afterward, in sync with the tutorial playback, take phone screencaps (there are apps for this) of the main shots/highlighted moments and afterward transfer the pix to your edit system
3. Later, also in sync with the tutorial playback, shoot a "finger body-double" of the finger action, using a solid green as the phone's screen (there are apps for this if you don't already have it). I would use some semi-hidden vise grip to "hold the phone" steady so that the finger is isolated nicely (and later do a crop around the edge of the phone so the grip doesn't show). This will be your "green screen" foreground.
4. Then in editing, use your original "phone shot" as a reference with which to replace with your screencap shots, then overlay the "finger-motion-greenscreen" shots on top of that (you could even do a corner-pinned PIP replacement of the greenscreen section to make it exactly match perspective). Use the correct compositing mode to include "finger shadow".
5. Then composite your Talking Head shot with your Screen+Finger composite and add in SFX as desired and you will have a finely-tuned master composed of the highest resolution elements possible.

Like I said: more work, but you couldn't get better quality.

Scott

Thanks Scott,

I was afraid I would potentially have to go that route, which would admittedly have been too much work for the client. Luckily I put my HD camera (Canon HF G10) on a tripod and the phone on a black stand (still need to tape it down or something similar as you mentioned) and got really high quality, much better than I was expecting, so hopefully I'll be able to go that route.