In a word, PERFECT.

Bob.

Video quality and editing were very good. The narrator's voice kept it from being topnotch overall. He lacked that authoritative, professional spokesman touch that you hear in many spots. Small point.

Also, how does a movie theater play 1920x1080x30p material? Do they have one projector for film and then a Blu-ray projector for video material?

Great work, Laurence - I like it a lot. :-)

I like very much, when non profit org. like this make good presentations, then it in my opinion often proves that it not always is about big $$ but also passion and hard work.

/Ulf

The "narrative's voice" was actually a 16 boy. When we shot the kids, we invited a couple of high school kids, but he was the only one who actually showed up. All the other kids were elementary school age. We had all the kids read all the lines, but the older boy was the only one who didn't fumble that particular line. He begged us not to use his shot because he didn't want to the only teenager amongst all those kids. Since he was the only one who got the line right, I asked if I could just use his voice if put something over his face. He was OK with that. I was hoping to keep the whole project nothing but kids and that's why I used his voice.

The production specs for the theater preshow release are listed on a PDF http://www.widescreenmedia.com/wscs/prodspecs.pdfhere.[/link] The two things throwing me a little are the 30p (specifically not 29.97) and the requirement for highest quality (rather than a bitrate) mp4 video. Rendering it out at exactly 30p seems to look fine since a doubled frame here and there on the talking heads is no big deal and the still picture movement renders out at whatever rate I set.

Since so many people doing this sort of thing use FCP, I tried Quicktime and now have a template with mp4 video at best quality (but I don't know the bitrate) and uncompressed 16 bit 48k audio.

I really don't know what they are talking about when they say "audio normalization at -12db). That seems like an awfully low level.

Laurence, as usual, an excellent job!!

I'm curious though as to why you did this as a chroma key.
IMHO, it would've been much simpler to use a large piece of black cloth behind the kids and avoid all the problems.

Mike

When I shot it I wasn't sure what background I was going to use. It wasn't until I keyed out the green screen that I realized what a great look a black background is. Especially in this day of online data compression, a simple black background compresses spectacularly. Yeah, I think I need a big piece of black cloth in the future. I am really tempted to use plain black for all my interviews.

As far as black cloth goes, It seems like I would need something really dark and really unreflective like maybe black velvet. Has anyone else here shot against a black drop cloth? What do you use or recommend?

When I shot it I wasn't sure what background I was going to use.

I realized that after I made my post :-)

Has anyone else here shot against a black drop cloth? What do you use or recommend?

Duvetyne (aka Commando Cloth) is what I see recommended a lot.
Check your local fabric stores.
If you strike out there, it's available at online sites such as Filmtools.

Mike

One nice thing about going the chroma key route is that I could shoot everyone in the center of the screen and then move and resize them at will during the edit. What I did in this case was take the bits I wanted to use and save them as .mov clips with a transparent alpha layer using lossless png compression. It was really nice being able to move the kids around and fit them in with the text so easily.

Has anyone taken an HDV camera and flipped it on it's side 90 degrees to shoot talent against a greenscreen? It seems like that would give you really high resolution that would let you really play with the framing afterwards. Let's say you shot with the camera sideways and shot from the waist up. You could go anywere from an upper body shot to super closeup on the face with the same footage. I think I'm going to try this the next time I shoot this way.

The original version was exactly 45 seconds for the local movie theater preshows. Here is a longer (and hopefully better) version:

As said before, Laurence, I like it a lot, and yes this one might be better, it can be tough to be limited to a certain lenght as max or min.

/Ulf

Just uploaded very slightly different version of the Give a Kid a Backpack video. Added a bit of movement to the stills, a little better color correction, and the website name at the end holds on a little longer. It is now also exactly 2 minutes and 45 seconds long. The link is the same as it was in my last post (or here again):

Laurence, this video is fantastic. If you're striving for total perfection, and looking for nitpicking feedback, I would say that some of the stills are cut ever so slightly short, especially towards the end. I was wanting them to hold just a split second longer so I could absorb the image. Also at 1:22 you could freeze the girl after she's said "that" rather than seeing her mouth start quickly moving again after you've dropped the volume. This is being totally nitpicking though. You and the kids have done a great job.

Great job all around Laurence. And yes, I do agree with Nick too.

Black backgrounds for a single "on-message" message is perfect - "Less is More".

Grazie

One thing about the simple black background is that the image compresses really well on Youtube. I've noticed that progressive scan shot footage looks way better on Youtube. On a TV, the quality of the HDV interlaced footage looks really good, but it seems to deteriorate much more after a deinterlace followed by Youtube recompression.

Boris FX is a kick ass chroma keyer. Much better than anything else I tried. I just wish it would work with 64 bit Vegas. Next time I think I'll bring a little hairspray. Little hair fringes drove me absolutely mad when I was trying to get a good key.

The Z7 looks way better on Youtube than my HVR-A1. I think that's mostly due to my shooting 30p on it.

I used my LED panels for lighting the kids for chromakey. They are daylight color temperature so I didn't worry about outside light. That particular day it was partly cloudy and the sun kept coming and going. I notice that the shots where the LED lights were augmented by sunlight look fantastic, but the shots where the sun was behind a cloud don't look anywhere near as good, even after my best attempts at color correction. This was even though I was in a room and the ambient sunlight wasn't something I was concerned about at the time.

The classroom kid interviews were shot with a very inexpensive http://www.bhphotovideo.com/c/product/379776-REG/Lowel__Ego_Digital_Imaging_Fluorescent.htmlLowel Ego light kit.[/link] The interview at :55 was shot with just a simple reflector before I bought any lights and looks fine as well.

Nick's suggestions are good. As I was editing, I felt like a lot of the shots should be held out a little longer, but I wanted to keep the whole thing uptempo and moving along quickly. The lips moving after I cut the audio: I have mixed feelings about that. I kind of like the feeling of it being an excerpt rather than a lone statement, but you are probably right.

I put the music together with a beta of the new Propellerheads Record software. I just LOVE that program. Musically, I do the whole Line6 / Properheads thing. The guitar is a Line6 700 faux acoustic, the audio interface is a Line6 Toneport, the software was the beta of Record. What is so nice about Record is that you can adjust the tempo after the fact even with a bunch of acoustic tracks in the mix. Not an artifact to be heard when you do this. It means you can record at a comfortable tempo and still get an exact timing at the end. Way cool.

I like using still pictures in things like this because you can go so much faster between them. With HD footage especially I feel like you need to stay on a shot for a good four seconds or so. With stills you can just keep popping them up there.

Everyone seems to like the video except my wife. I keep trying to explain about how you can only put in things you actually shot or have a still for...

Very nice Laurence

Ok who is the little lady who waves her arms around when she talks?
She really had a strong emotional connection and a honest positive energy, enjoying helping others
They all were but she just made my heart smile

Laurence, perhaps there was a slight element of familiarity-with-the-pics happening, by which I mean you, as the editor, know the pics well and so you don't need that same amount of time to absorb/appreciate them as someone viewing the video for the first time. I'm sure someone's put a buzzword name to that phenomenon but I don't know what it is. It's certainly a trap that I've fallen into myself.

Didn't realise you'd done the music yourself too. It's really perfect for the video. Well done!

As for Propellerheads software, I played around with Reason quite a bit a few years ago and would still love one day to make some of my own music for my videos. Do you know Reason? As someone who is not a musician and wants to make non-vocal tracks from samples and synthesizers, would Reason still be the Propellerheads product of choice for me or does Record have something extra?

Record is now in the public beta stage, but I was working with the private beta as well. They have a deal were if you can point out a new as yet undiscovered bug, you can get the program for free when it is released. As much as I have tried, I am still going to have to buy my own copy when it is released. It is about as stable as Vegas is typically on a "c" or "d" release.

As far as Reason vs Record, it doesn't work that way. What happens is that if Reason is installed when you start Record, all the Reason modules show up in Record as if they are part of the program. If you don't have Reason installed, all that you have instrument wise is a general midi player. When Reason is installed, you can even open a Reason sequence in Record and everything will be assigned and play correctly.

Also, if you use a Line6 Toneport for your audio interface, and have bought any of the extra amp and cabinet simulator models, they can be enabled within Record. As a guitarist, I have done this and it is an important feature.

What does Record add? A wonderful mixer modeled very accurately on a real high end mixer (I can't remember which one, but it's really a nice layout). That and the ability to record and playback audio. It has a way of comping tracks and picking the best phrase from each take to assemble a composite take. For us video guys, the feature that is most cool is that you can change the playback rate of the entire piece (audio and virtual instrument tracks both) to whatever you want, and the resulting mix is artifact free. You can even change tempo according to a tempo map. Really cool stuff. It lets you go for exact timings while recording at whatever tempo feels comfortable.

What it doesn't do is let you pitch and time correct individual notes a la Melodyne. Because of this, what I do is rewire Reason into Melodyne Studio, record all my acoustic parts into Melodyne at 24 bits, correct them, then save as wav files. I import these into Sound Forge and normalize them and dither them down to 16 bits and import the wav files into Record. It sounds like a pain, but because Melodyne lets me perfect a part that is close, I actually spend a whole lot less time than I would if I had to do it the old way: keep recording until I get a perfect take (which for me can take hours).

Perhaps the most glaring thing missing is a way to use the VST plugins that all the other DAW software uses. I don't mind this though because the Propellerheads plugins are really good. The problem with using plugins with a host is that every time you use a plugin, you need to have some kind of pipeline between the host and the plugins. This works well but puts way more demand on the CPU. The problem I kept running into is that by the time you have a good sounding mix happening, your poor little CPU can be sputtering. To get around this, most DAW software has some sort of track freezing (where individual tracks can be temporarily rendered to free up CPU cycles). Also, a fast quad core can handle any VST plugin mix. The problem is that I HATE desktop PCs. I do everything from a laptop and with VSTs in a DAW host, it was challenging. Reason and Record on the other hand do everything within one program and the CPU efficiency is orders of magnitude better. I can do the most complex mix I can think of on my HP Core2duo and the CPU meter barely budges.

Another advantage of built in FX over VST plugins is the patching. Everything in Record and Reason has got patch points galore. You can patch in things like sideband compression so easily. You can make layers with numerous types of synthesis, you can take one note of a drum machine and trigger a synthesizer. It's like having the eighties studio of your dreams, and as complex as you go, even a moderate PC will keep up. I had been using Reason rewired already so I have a pretty amazing Reason library. Now I have access to all that in Record.

Propellerheads has added copy protection to Record. It works in two separate ways. You can use a hardware USB key, or you can have your computer online and sign into your Propellerheads account. Copy protection only matters when you load a file. Without the dongle or signing in, you can try everything out and even save files. You just can't load anything. I don't really like this, but that's the way it is.

What also is missing is a way to sync to video. Apple users have a cute little app that plays back video with a rewire slave, but PC users have to use a program like Acid. I usually end up just going by the timings, plus I usually do the music first. I'll figure out how many themes there are in the video, and write music that will accent that number of things. Of course this isn't always perfect and you can end up rushing certain pictures...

If you are interested in the public beta, go to http://www.recordyou.com/recordyou.com[/link] and apply for an invite code. There is also a public beta forum http://www.propellerheads.se/forum/forumdisplay.php?f=19here.[/link]

Thanks for all the info Laurence. Many Reason users were always wittering on about adding audio capability to to Reason, and I now see that they're achieved it with a totally new product. Together they do sound like an outstanding solution.

In your experienced opinion, would Rewire support in Vegas Pro help your workflow?

>would Rewire support in Vegas Pro help your workflow?

Would rewire in Vegas help my workflow?! I would be dancing and singing up and down my street, I would be kissing complete strangers, I would be grinning from ear to ear even during funerals, I would be ... uh yeah, it would help my workflow.

Laurence,

The video is fantastic - not only content, but quality.

What settings did you use when encoding for upload? Everything I've uploaded looks like cr*p in comparison - artifacting, stuttering, etc. I must be doing something screwy.

Tom

I have a template for Youtube renders that I really like:

Sony AVC mp4 compressor, 1280x720 pixels square pixels, 29.97 progressive scan frame rate, bitrate is 5,000,000 bps, audio bitare 128,000 bps at 44k. Everything else at their default settings.

I've tried wmv but don't like it. The detail is there but the color looks washed out. DivX is better but I like AVC mp4 the best.

Simple stuff compresses best. Faces against a one color background mean that the Youtube compressor can use all it's bits on facial detail. Shallow depth of fields stuff and still pictures that move only slightly compress great. Progressive shot footage compresses better than interlaced even after a deinterlace. Noise drives the compressor nuts. On my Z7 I've started shooting at -3db gain instead of 0 because it has noticeably less noise that way. Noise that you hardly notice on the original will look like crap after it gets rerendered by the Youtube compressor. A lot of times as I'm downrezzing from 1440x1080 to 1280x720 I'll apply a little sharpening. I only ever use sharpening on downrezzes, but I really believe in it in this one instance.

So much of it is in the shooting. Don't use hard pans. Don't use big zooms. The more each frame has in common with the frame that precedes it, the better the video will look online.

I like Youtube better than Vimeo because it shows up on searches more readily and plays back on my phone. For a while it didn't look as good, but I think they are about the same now if you upload a nice source file.