Saturday, April 25, 2009

My First Gig as a Reporter (pt2)

This is my second "lessons learned" post about covering the Tax Day Tea Party here in Saint Louis as a PJTV Citizen Reporter. My first post focused on my audio reporting. This post will review my video work.

The seven part YouTube playlist at the top of this post is the whole enchilada. Mostly. All the copyrighted material (music) has been cut from the beginning and I dropped about 15 seconds of Dana (between parts 6 and 7) to avoid having an eight part playlist. Sorry Dana! Note to self, plan your edits better so you don't have to do that in the future.


I only did one video interview at the tea party and it came after the event when I spoke with Rich from Circle of Concern. Circle of Concern is a local food bank. They provided a van to pickup donations of non-perishables. Operation Food Search also participated in the tea party food drive. They had a 26' to 28' foot truck, like the big ones UHaul has.

I don't think I did very well. My questions were sorta ok, but there are three points in the clip where I just can't remember the name of "Operation Food Search". The audio's not very good. I'm too loud and you can barely hear Rich. Perhaps I should've used my MP3 recorder to get a better recording of Rich and then edited the two together. A microphone is probably a better option since I could point that at whoever is currently speaking. The video is blurry, in part, because of the hardware I used. I plan to use a better camera in the future, but that will mean that I'm recording to tape. I hate tape...


It's my experience that conservative voices don't get much airtime. Grainy videos with bad audio are not going to change that. If you're seriously pursuing citizen journalism, then quality has to be a priority. This not only includes quality content (e.g.: an engaging and informative interview, a newsworthy event), but it also includes technical quality (e.g.: 1080i or 1080p video, crisp focus, stereo sound, good levels, etc.)

TV newsrooms across the country have made huge investments so that they can broadcast in high-def. They don't want bouncy cell phone footage except when there's no alternative. They want high quality content: an engaging and informative subject, a sharp picture, and clear audio.

Content is still king. With the economic downturn and the pressures imposed on broadcast news by their competition on radio and the Internet, newsrooms are going to be looking for less expensive sources for their footage. They're going to be looking for open content. Content they can use at little or no cost. Content that you gather with your HD video camera and then make available to them on the Internet.

Open Video Content

Here are my first thoughts on a concise list of parameters that content creators (citizen journalists) are agreeing to when they offer their work as open content:
  1. The content creator must be credited by the content user (TV station, website, etc.)
  2. A link to the complete source material must be provided in the show notes on the content user's website
  3. The content user may edit the content as long as the meaning is not inverted
Obviously, these are my priorities. Editing a video will cause some distortions, if only by omission. The link back to the original should mitigate minor distortions. #3 seems reasonable, but I'm not sure how to put teeth in it. Should a lawyer draft a simple open content license? Perhaps the Creative Commons License would suffice... Please put your thoughts in a comment below!

Video Distribution

Source audio and video content is only the beginning. After the video is shot, it has to be made available online. YouTube is one option, but it has significant drawbacks. Video uploads to YouTube must not exceed 1G and they cannot be longer than 10 minutes. Those factors limit quality and increase the workload for providing video content online.

For the Saint Louis Tax Day Tea Party, I took the time to enter the metadata for each part of my seven part video. That metadata includes name, description, tags, location, etc. Copy and paste helps a lot, but it's still a pain. offers a much better solution. Since you are not limited either by duration or file size, there's less administrative overhead because you don't have to dice-up your video and duct-tape it back together with a YouTube "playlist".

Before I figured out, I exchanged emails with some folks at PJTV. They were helpful, but they really couldn't fix the bottleneck: I did not have enough bandwidth. I discovered this problem before tax day when I tried to upload my interview with Bill Hennessy. It took 12 to 15 hours to upload that interview to PJTV. That's just too long to be useful and it invites data corruption problems and other weirdness. has a tool that manages the upload process. Even though the upload takes just as long, their upload app tool allows you to queue multiple videos. You can also pause all uploads with the app. The pause feature is very helpful when your wife starts complaining that the Internet is slow.

I still don't know if PJTV used any of my footage. I know some of it uploaded successfully. I know that they received a DVD via FedEx Tuesday (4/21) with all of it. This is another reason for #1 and #2 in my list of open content expectations—I'd have a way to find out what they used. Of course, by Tuesday my video footage was, to put it charitably, starting to ripen. News is time critical.

Crowdsourcing the coverage of an event that has over 500 locations on a single day was bound to be messy. If there were other videographers that weren't able to get their content uploaded, I would not be surprised. My conclusion here is that the videographer needs to be responsible for getting their video online. I think that's the view PJTV took. I also think videographers should be independent. It doesn't matter to me who broadcasts my work; I just want it broadcast. If PJTV or someone else, is willing to guarantee that X minutes of my footage will be broadcast or that they're willing to pay me, then that's another story.

If content creators are independent and responsible for getting their footage online, then is the linchpin. It took awhile for me to upload my recording of the tea party to The video was online for a couple of days when it was shutdown for bandwidth/cost. That's not a slight on I love their service! Their prompt and helpful email support was excellent. They offered me the option of keeping the video live by upgrading my account from Premium to Pro, but I was concerned about cost. Their Pro accounts are charged per gigabyte for bandwidth (at a competitive rate)... here's the thing, I couldn't figure out how, in two days, a video of the St Louis Tax Day Tea Party had been viewed 600+ times. Paying for each of those was a scary prospect.

With a Premium account on there's a limit to the number of views/downloads your videos can have. There's no limit to the quality, file size, or duration, so is a nexus between content creators and broadcasters. I grant access to my video content to PJTV, C-SPAN, NPR, Fox, whoever and they download the original hi-def content for broadcast.

What would happen if PJTV (or any outfit, really) had video content from hudreds of locations? If those are live feeds, it's going to be a brutal day herding cats. If it's recorded video like mine, then video editors are going to be working day and night reviewing the content. They might just pitch some of the video, but then they're alienating the people that shot that video for them. The most labor intensive component of video editing is annotating the video, marking the start and end points of various segments, and identifying who or what is on the reel.

Edit Decision List (EDL)

The annotations of a source video are often in edit decision list format or EDL. EDL was originally developed to keep track of all of the edits on multiple source video tapes so that they could be assembled into a movie or show. It's the lingua franca of video editing. It can be the basis of a library of metadata that allows a video editor to quickly find the clip they need.

If you still don't get what EDL is, here's an edit from a recent diavlog between Peter Singer and Tyler Cowen. The edit link captures the "in" (start time) of 48:41 and "out" (end time) of 51:04. You need me to annotate it by telling you that Tyler asks Peter whether it's moral to eat fish and I'd give the clip four out of five stars. has already implemented part of the annotation process: defining the in-out points. Now we need someone to incorporate the other metadata: who, what, where, why, and how good is the clip. The nice thing about that work is that it scales—you can crowdsource it. Here's how that might have been done for the Saint Louis Tax Day Tea Party:
  1. Bill Hennessy announces a phone number and asks for volunteers to text it from their phones
  2. People who texted the number are sent a weblink to access the next day from home
  3. When they follow the link, they're given a ten minute block of the footage to review
  4. They enter in-out points, annotations, and star ratings for the edits in their ten minutes
"Whoa!" you think. "You'll get tons of hits on that account and go over your limit." Yes. If the link is to, but if the source is diced into ten minute blocks and put on YouTube, then there isn't a problem. YouTube versions are used for review and is used for the hi-def version. Ideally, I'd like to see offer some sort of journalism package. I'd want a way for people to annotate the video content. To reduce bandwidth costs, they could provide a low resolution version of the video for annotation.

I think I'll do a third lessons learned post... This doesn't feel done, but I need some sleep.

Related Posts:
Update: Thanks for the Instalanche, Glenn! Part 3 will be up some time Sunday so please check back.


Anonymous said...

A few suggestions about your content license:

1. The content creator must be credited by the content user (TV station, website, etc.)

I don't see any problem there. If the user won't agree to this, you don't want them using your content in the first place.

2. A link to the complete source material must be provided in the show notes on the content user's website.

This is more of a problem. Do you mean if you shoot an hour of video and the content user only uses 5 minutes of it, they have to post the rest on their Web site and link to it? I would say, I'll link to your unused video if YOU host it (and pay the hosting expenses.)

3. The content user may edit the content as long as the meaning is not inverted.

I would prefer "as long as the meaning is not changed," or "as long as the speaker's meaning is conveyed" or "preserved." "Not inverted" means, "as long as you don't make the person appear to be saying the opposite of what he said, editing is OK." I would not be comfortable with that standard.

The Grey Man said...

I have a couple of questions, if you have the time.

I was one of the Cellecast reporters (Baton Rouge, LA). My initial report went well, but I couldn't get any updates to record. Any ideas?

The email is on the blog, thanks.

The Grey Man

dsm said...


Thanks for your comments!

I agree that the burden of hosting the complete content should be borne by me the content creator. Ideally, the content user should link to my video; however, the cost concern means that I'll likely have a YouTube playlist for both for the link back and for promoting the content.

In principal, I agree with you on #3. In practice, I don't think it's realistic. #2 should provide some redress for the minor omissions and slight meaning changes. The link to the original will facilitate a reply to set the record straight. It's clear when someone deliberately inverts the meaning, so proving that a derived work "inverts the meaning" should not get mired in the weeds the way a "changed the meaning" accusation would.

I'm still not sure how to put teeth in it. I guess a take down notice.

dsm said...

The Grey Man,

I've learned too much and know so little... I had a similar problem; however, I never replied to any Cellecasts, so I'm not sure if this will work for you....

As I mentioned just after the main event, my wrap-up report didn't showed up on the page. At some point it did, but I don't know why.

That said, I was able to reconstruct the URL from the confirmation text message sent to my cellphone. Our Tea Party URLs look like this:

I found the cellecast ID code in the SMS message on my phone and altered the bold-italic number in that URL. It's only a partial solution. You'll still have to post the fixed URL to the PJTV page for your Tea Party and/or on your blog, but it's better than nothing.

The Grey Man said...


Robert said...

Good lord! You're a genius and a video god! I grasp the hem of your garmet, and all that.
I used the caveman approach and just put my Coolpix digital cam on "movie". I panned the signs and asked one simple question: Why are you at the Rally?" and let people talk. Editied to nine minutes and posted on Youtube. Most hits of any of my videos at this stage, and I have 93 up.

My great idea for improvement was to change my question for next time to: "What do you have to say to the President and Congress?" and let people talk.

My video is on the Blackfork6 Channel on Youtube, by the way. I rarely talk or appear in my own videos. It was easy. Used Imovie. Never read any instructions. Ugh. Youtube good!

dsm said...


Thanks! I'm not really a video god or a genius, but you may continue to think so ;-)

I like your 9 minute spot. The reality is the shorter content will get more play, because who wants to sit through an hour long protest video, so keep up the good work!

The longer content is valuable because you can edit it down to the nugget you need. My most watched video includes such a nugget and is only one minute long. Some of the reports for the St Louis Tea Party put the attendance at only 5k. It was more than that and I make that point here.

JayDee said...

For what it's worth, on the pro level we make our editing decisions for episodic shows on low resolution versions of the raw footage, then go to high resolution only for the final edit. This helps get us around the massive amount of storage hi-res video uses.

Normally, we have matching time code that's accurate to the frame on both the low res and high res footage. In this case, you can probably just get away with doing a quicktime version of the low-res version of the original, which would enable the user to read from the viewer the rough times into the original roll. You can then post the original footage in pieces of roughly five minutes each, preferably with a little overlap.

Because the low-res version doesn't use a lot of bandwidth, you can get it up fairly quickly, the prospective user can make edit decisions while you're uploading the full-res, then they can just download the chunks they actually intend to use, reducing bandwidth on the hosting service.

You might also do a little exploring on hi-def codecs. There's a few that are in fairly common use and allow quite a bit of compression (and therefore bandwidth reduction) without a major quality hit. See what your potential customers are using and supply footage in that format.

On another vein, doing pre-logging may be of use to some, but the average broadcast editor wants to view the footage while making the choices, so I'm not sure how much value your proposed self-logs would have to the pro user. I can generally scan the low-res and pick out what I want faster than I can read the log and then go see if the interesting sounding stuff is any good.

And... because I'm getting long-winded tonight... the spot-news oriented users are probably more likely to use a two minute or so edited piece than they are to sort through lots of footage, so that's what I'd probably put up first. It's a pretty rote format, start with an establishing shot, throw in some interesting people and signs, then short sound bits from the speakers, interspersed with crowd shots as cut-aways, then a wide shot to end it, and you're there. Get that up asap after the event and it'll get the most use.