I Don’t Understand the Amazon Video Algorithm

I have no idea how Amazon Video chooses what videos to show people.

I mean, I know that you can impact things by having more reviews, or sharing the links to the video and generating more traffic. But when that’s not happening, what makes Amazon Video show content to people?

Recently I saw a spike in older content that’s on Amazon Video, which I placed their through Amazon Video Direct. I’ve done zero promotion for this in the last year. It just sits there.

I assumed that people who are searching for religious video content on Amazon might stumble across it. And I’ve seen a steady trickle of payments that reflect that. I’m getting enough to pay for my lunch once a month. Not bad for content that was just sitting on a hard drive. AVD payments are always more than my similar content on Youtube. Always.

Then October happens. Suddenly there is a spike in minutes viewed and number of individual streams. And almost all of it is for the series. And based the a couple of reviews, the people seeing it aren’t looking for religious content. (It’s always a fun adventure to see how a non Christian reviewer sees content that is meant specifically for a Christian audience.)

So I don’t get how Amazon is populating the series to viewers. I search through Google to see if the link had been shared somewhere, but didn’t find anything other than the normal listings. Its actually easier to find the series on Youtube or Parables TV if you search on Google.

Overall the numbers of views aren’t huge, but its about 5 times what I normally get. Now, in November, the views have dropped back to the normal rate. I wish I understood this better so I can capitalize on it more when my doc film releases next year.

Advertisements

Technically Superior Internet Syndrome

There’s a virus infecting the internet. Really, it’s infecting internet users. Luckily there are steps you can take to protect yourself.

Symptoms of this infection are most easily seen in the comment sections of popular sites like Youtube or social media platforms. While it may show up anywhere, it’s rampant in tech communities.

The affliction comes on you like this: You do a Youtube search to find out more about a piece of gear, maybe a drone or camera. You watch a video, and realize that the video didn’t include some very critical information. So you have the overwhelming urge to comment, to share you’re superior understanding of the issue.

Maybe someone has asked a question in a group you frequent. But you know that the question, at it’s base level is fundamentally flawed. So, rather than answer the simple question, you proceed to explain why the original post is wrong headed and tell the world the correct approach, provide the correct information. At no time do you even consider answering the question as asked.

It can strike at any time, on any issue.

Here’s an example about van graphics:

Original post is looking for design inspiration. The two commenters were stricken with the need to share vital information, unrelated to the original question. One even correcting the 1st commenter.

I recently saw a comment on a professional videographer’s video about a drone flight. The commenter was quick to point out that the video has been taken illegally, since the location required a permit to fly drones. At no point did the commenter consider that the person flying the drone may have actually acquired a permit, and been flying with permission.

Once a member of a FB group posted an innocent questions about showing videos in their church. Rather than answer that question, several responders pointed out that showing the clips would be illegal without the proper licensing. At no time did the infected consider that the person asking the question might already know that. They assumed they knew better, and told everyone so.

Technically Superior Internet Syndrome (or T-SIS) is rampant.

There is at this time, no cure. But you can fight the symptoms by being aware of them, and taking a moment before you answer a question to make sue you’re actually answering a question. Pause before you tell the world how wrong someone is to make sure you know all the facts. Generally, if you can’t say something nice, don’t say anything. If you aren’t answering the question, don’t post.

Just a Few Hours Left in the Crowdfunding Campaign

Less than 7 hours until my crowdfunding campaign ends.

To be frank this one has been tough. I put in more work setting this one up than any of the previous ones. I had the “large” donors set up to drop their donations in the first few days. And I had built relationships in communities that will be the target audience for the finished movie.

When the campaign launched, I had several large donations come in. But almost zero small donations. When I did the campaigns for my TV show I had lots of small donations and almost no large ones. A couple of the communities I was in were a bust. I don’t know what happened to the rest.

But a few larger donations have come in outside the campaign. I am under $500 away from reaching the goal.

It’s not likely that I will reach it, but I will have enough to make the film, and tell these stories. Later, I will try to figure out how I misread my audience so much.

New Project!

I can’t believe I haven’t posted about this here, yet. I’ve been so busy with everything else, I’ve let you down. Here’s the deal. Some major news on the filmmaking front!

I’ve got a new major project. It’s a documentary film about competitive speech and debate. Here’s the pitch:

If My Judges Are Ready? is a documentary film project following a group of homeschool speech and debate competitors from Texas through the 2017-2018 season of competition as they work to reach and win the national tournament. The story is the personal struggle of each competitor to overcome their own limitations, perfect their pieces and navigate the tournament schedule in order to compete in the national tournament at the end of the season. More important than how they place is how they will grow and mature through the process. The film will give the audience a glimpse into the life of these competitors, show what it’s like to be in a club, to select content, practice and practice, be judged on performances, win, lose, and learn.

It’s a huge project. It will take a year to shoot and months to edit. But I’m excited to tell the story. Follow the link to the website to see production updates. Or sign up for my email list and get them right in your inbox!

Sigma 19mm DN Art f2.8 Amazing Lens with a Video Auto Focus Problem.

The very first lens I bought for my Sony a6000 was the 19mm Sigma DN Art f2.8. For under $200 you get a lens that is super sharp. According to most reviewers, dollar for dollar, it outperforms the competition.

Flag over Washington (Half Mast)I have been very happy with it. I’ve taken some great pictures. On an APS-C sensor, the 19mm is a handy focal length for catching pictures of kids inside the house. It’s not too bad for wider landscapes. Or shots like the one above.

There is a problem, however, when you use it with AF turned on for video recording.

I didn’t notice it for a long time. If you’re running hand held, you might never notice it. On Youtube it’s not easy to see, unless you’re looking for it. And I wasn’t. The I pulled up the footage on my computer. How could I miss this?

But now that I’ve seen it, I can’t look at any footage on a tripod or slider without seeing it. What is it?

Here’s a video that shows the issue very clearly. Watch the edges:

Slowed down like this, the jitter on the edges of the frame is very visible. I pulled down a few videos I shot with it. (But not everything.) But every video I’ve checked has the issue. I’ve used the lens for several videos, but none were reviews of the lens. All were about something else, so I didn’t notice. they look fine where the center of focus is, where your attention is drawn. I don’t “pixel peep” with most of my gear. The center of the lens is sharp, and looks great. But with the autofocus on, the edges shake and jitter.

When I was researching this lens, every review was positive. I didn’t find one mention of this issue. Now there are a few posts about it. Some videos like the one above. I’d hate for someone who wanted this lens for video to not know about this issue, so I’m doing my part.

I love this lens for pictures. But I can’t recommend it for video.

Captioning for Amazon Video Direct (using Adobe Premiere)

I have previously written about Amazon Video Direct (AVD). It’s an awesome opportunity for indie filmmakers to get your content in front of a large potential audience, and it pays better than Youtube. For stand alone or episodic content, it’s a great outlet.

One thing might slow you down as you start to publish your videos on Amazon: Captioning.

Amazon Video requires that all content be captioned before they will publish it. Period. That can be a bit scary. A few years ago I paid about $2500 to a captioning/delivery house to caption and deliver 10 episodes (22:30 each) to a TV network. Now, they captioned the shows in both 708 and 608 captions, and delivered the files in HD to the network and gave me copies of the .scc 608 files so I could use them later. But still, $250 per episode. I’m making indie films with budgets less than that.

Luckily, Amazon suggests a few online captioning services which are much less expensive. One, Rev.com, offers captioning for $1 per minute and delivers in various formats. They can provide captions that are AVD compliant. They even have a free caption converter, should you need one. That means my 22:30 shows would cost about $23 for captioning for AVD.

Still, $23 is money you may not want to spend. What if you want to make your own? You can, but Amazon is very finicky about their files. I will share what I have learned. I have 1 season (10 episodes) of a show and 2 short films available right now, with 1 more short film in review. (Now published)

Adobe Premiere has the ability to create and export closed caption files. But getting a caption file that AVD likes is not simple. Here’s what I’ve learned.

Create a 608 Caption file. Premiere will do 708 files, but Amazon does not seem to like these 708 files. I have only had success with 608 files exported from Adobe for Amazon. Premiere can even import existing .scc files, allowing you to edit them.

Export .scc or .srt. When exporting your video file for upload, export a sidecar caption file as either an .scc or .srt. If you have content that is 29.97 use the .scc format, otherwise use the .srt. AVD says they will take an .stl file, which Premiere will export. But I’ve not had any luck using that format.

If you’re lucky, that’s all you need to do. Just upload and publish.

For my last short film I was not lucky. I was exporting a 23.976 fps file using .srt, and I could not get AVD to accept it. It was exactly like a previously accepted caption file for a previous short film. What was the problem? After trying multiple files over multiple days, I was frustrated. I turned up this post in the Adobe Community Forums. Scrolling through I found 2 solid things to try.

1. The timecode of your captions cannot overlap the same frame.

In Premiere you can see where one caption ends and another begins. Here’s a screen shot from premiere of my latest short film:

If the 1st caption you see ends at 00:00:20:08 and the next captions starts at 00:00:20:08, AVD has a problem with that file. So you need to go through all your captions and make sure none of them overlap.

2. Remove extra content.

During the exchange in the post in the Adobe Forum “Joshb88988268” says, “open the .SRT file with notepad and do a search for this: or the word font color. Delete any that pop up.”

As a mac user I found a free program called Brackets and was able to open the .srt file. Sure enough there were 2 lines with the tag and some extra info about “font color”. I deleted those lines and hit save. My captions in the code editor looked like this:

No extra tags or words. Just number of caption, timecode, and caption content. Brackets should also be able to open a .scc file.

So far that seems to have worked. At this point I have to ask myself, if I’m uploading a 4 minute short film, is it worth spending $4 to bypass all this effort? It might be. But since I have the captions done, I would like to be able to use them.

[Update: while I was typing this post, Amazon has begun approving my video. Looks like the latest captions with these changes worked.]

What to Do When a Copyright Claim is Initated on Your Youtube Video For Music You Have the Right to Use

If you use music from royalty free libraries or websites on videos for your Youtube channel, there is a high likelihood that you will eventually get an email with the subject “[Youtube] A copyright claim was submitted for content in…”

Don’t panic. It’s not personal. No one is targeting you. And while it’s annoying to be accused of stealing music, the Content ID system is automated.

Here are the steps to fixing this.

Read the email. In most cases you are in no danger of having your channel shut down, or even seeing the video take down. Most of the time the claimant just places ads on your video and has the money sent to their account. That’s an annoyance, but any money that is collected while the claim is in place will revert back to your account once it’s removed. Still, I normally take the video out of public view, setting it to “unlisted” so only people with a link can see it.

Note the content that has been claimed. Who is making the claim, etc… Is it AdRev? If so, then your dispute process is very easy.

Check your content. Do you in fact have the rights to use it? I’m sure you do. But dig out the library you took the music from, find the name of the track. If you can find the receipt, even better.

Research the claimant and the company making the claim. Every single time I have been flagged for copyright on my monetized videos it has been an individual making a fraudulent claim through a 3rd party. There is a guy in Denmark or somewhere that has claimed dozens of tracks from Digital Juice‘s library, and he uses AdRev to make copyright claims online. He does not own the content.

The last claim I had was from a Canadian using a French Canadian company, so much of the communication was in French. (Thank goodness for Google Translate.) This gentlemen had exported a Digital Juice audio track and added ambient waterfall and bird noise. Then he put that track on an album that is available through iTunes and Spotify. His record company was policing the content they think he owns. So a video using the same track (which I purchased from Digital Juice just like he did.) was claimed.

Contact the company making the claim. You’re first instance will be to start a dispute through the YouTube process. That process can take 30 days. And YouTube is just acting as the intermediary between you and the company. I normally go to the company first, and then start a dispute in a few days. I find that the claims are released wishing a few days.

If the claim was made by AdRev, just go to their website, and scroll to the bottom and click “contact us.” They get so many emails about this they have their contact page set up to hear your dispute. They have now added a place to upload a copy of the license for the music in question. in the past, regarding the Digital Juice tracks, I have just explained the origin of the music and reminded them that the person making the claim does not in fact own the copyright. They know this, they have heard it many times. But they still keep him as a customer and make fraudulent claims against legally used music. In every case so far (so far) the claim has been released within 24 hours. I have had to dispute music like this over a dozen times with AdRev. (One caveat, if you are using a MusicBed track, the account rep from MusicBed must contact them, otherwise they won’t release the claim.)

In other cases, with different companies, I have simply emailed and explained the mistake. I outlined where the music in question was taken from, what library and what track title. I explained the rights to use the music as it was in the video had been purchased from the company owning the library. And pointed out that the person making the claim did not in fact own the rights or have the legal right to claim the music as his own work. And asked for the claim to be released.

In the case of the Canadian, I also pointed out the differences in the audio. I had not used all of the stems from the Digital Juice tracks, and had not added the ambient noise. It was literally impossible for me to have stolen his recording and used it in my idea. the audio in my video did not actually match his. The claim was released the same day.

File a dispute through Youtube. I almost always file a dispute through the Youtube process if the claim is not released after the initial contact. Once you file a dispute, the company making the claim has 30 days to respond. If they don’t respond, the claim is removed. If they do respond and do not release the claim… well, I’ve never had that happen.

Every single time I have ever contacted a company and explained where I got the music and why I legally have the right to use it, the claim has been released. According to Youtube, if the claim is not released, you can appeal. There is another 30 day period. If the appeal is rejected a 2nd time, and the claimant requests a takedown of the video your account will get a copyright strike.

Most of the time you won’t get to this level. I never have. You can find horror stories online, but normally companies are not that difficult to deal with.

For whatever reason, some people think that buying Royalty Free music and using it in some sort of creative work means they have the exclusive right to use the track. They don’t, but their misunderstanding means you may have annoying claims on your videos. Stay calm, and politely state your case. Most of the time you can get them released without too much trouble.

Why I Entered the Rode Reel Competition Even Though I Don’t Expect to Win- And Why You Should, Too

The Rode Reel short film competition is one of the largest in the world. Entries from 88 countries are all under 3 minutes long and must have been shot using a Rode microphone. In 2017 the prizes total over $500,000. If you watch finalists from previous years, many of them are just amazing looking, amazing sounding.

How can you or I, average independent filmmakers, compete? Why should we enter if we probably won’t win?

Perfecting your craft. Experience always teaches you. I made my first actual short documentary film. I learned a ton in the process and got to experiment with a new genre. Every project you complete has the potential to help you learn and improve. Do you think those Rode Reel finalists just woke up and magically were amazing filmmakers? No, they worked and worked. This is a chance for you to become a better filmmaker.

Exposure. We all have a sphere of influence. We have an existing audience, whether it’s just family and friends or something larger. But entering the Rode Competition will expose your work to potentially thousands of new viewers. Viewers who will meet you for the first time, who might find your social media contacts, who might subscribe to your channels. Viewers who could be fans of your work. And those viewers are available for free.

Free T shirt. And maybe more. If you’re among the first 1500(?) entries Rode will send you a nifty Rode Reel T shirt. Sometimes they throw in some of their small products. Who doesn’t like free stuff?

Deadline. Most of all, committing to enter places a real deadline in front of you. Talk is cheap. If you are actually a real filmmaker, what films are you making? A deadline puts a real goal in place. I wanted to enter last year, but I never committed. So I never entered.

So, want to see my entry?

You can watch it here: https://www.rode.com/myrodereel/watch/entry/3102 Hope you enjoy it. If you did, please take a minute and put in a vote for the People’s Choice award.

Before I submitted my film, I watched some of the finalists for that category in 2016. They were awesome. None of them were telling a story of an event. They were more like showcases, testimonies with nice B Roll. After completing my Rode Reel entry, I know why.

Trying to tell an actual story in 3 minutes, a non scripted story, is extremely hard. My film has a beginning, middle and an end. (Spoilers) There’s a mid point crisis and turn into the 3rd Act. But it all happens in 3 minutes. So it’s fast. I cut so much good stuff out I’m seriously considering an expanded version at a later date.

It’s not perfect, but it isn’t terrible either. And I can guarantee my next documentary will be better because of what I learned doing this one.

[Image courtesy of Greenleaf Designs at FreeDigitalPhotos.net]

6 things I Learned Shooting My First Short Documentary Film

I’ve been working on a short documentary about my son’s last speech and debate tournament, specifically focused on the Team Policy debates in which he competed. It was a guerrilla style shoot. I had permission to shoot his teammate, but no one else. I could not disrupt the competition any more than any parent with a camera might. No extra lights. No extra people. Just capturing the event in real time with my Sony a6000, 3 prime lenses, and a Rode Smartlav+ microphone recorded into my phone. It was a true Run & Gun situation. Here are a few things I learned…

Story. Doing an actual documentary is different than most of the work I’ve done. I know how to shoot and edit a testimony video, but that’s not a documentary. Before the tournament, I spent time mapping out the structure of the short film. While I didn’t know what would happen, I did know the sequence of events, so I laid out the possible plan and tried to capture the actual events as they happened. As the tournament progressed, I could see how things would fit into my traditional story structure.

Pack Light. Because I was a one man crew, everything I needed was with me, all the time. I had gone through my gear, and left much of it at home. But I was still carrying around a medium sized camera backpack. And I still had gear I didn’t need. In order to grab my camera for a quick shot, I had to take off the backpack lay it down somewhere, open it up and pull out the camera. To downsize a bit more, and make access to gear a bit faster, I just ordered a camera sling bag. It’s large enough to carry a camera and a couple of lenses, etc… But smaller than a back pack and you can sling the bag around to the front, and access the gear on the run.

Invest in a zoom. Lens swapping is a pain. And real life doesn’t wait.

When shooting on a set, there is always time to swap out a lens. In between takes, you can switch over to a different focal length of the super fast prime you have. But in a documentary shoot, people aren’t waiting. Life is happening, the event is going on. Not only do you have a chance to miss the shot, but you might also disrupt the very event you’re trying to capture. During one debate round I was using my 19mm lens, and wanted a tighter shot. I was so nervous that opening my camera bag would be noticed by the competitors. I hope that didn’t happen, I tried to be so quiet. With a zoom, this wouldn’t be an issue.

Which zoom? On the Sony E Mount system, the reach and quality of the 18-105 F4 G series (SELP18105G) would seem to be a good fit. The longest lens I had with me was a 50mm, and I was wishing for longer options. It’s a constant aperture. I wish it was a bit faster, but it would only be a problem in the most dim rooms. I found that most of the time I was shooting f3.5 to 5.6. Of course the ISO was almost always at 1600 in the classrooms. Assuming I can continue to push the ISO that high, losing a couple of stops of light might be a decent trade off for the extra length. But at $500+, it’s out of reach for now.

Another option would be to adapt an older zoom of similar reach. You can often find vintage 35-105mm zooms for cheap. Just read the reviews on each one and make sure you have the proper adapter. Of course, you give up all automatic functions with these. I just ordered a Vivitar (Made by Koburi) 35-105mm f3.2-4 Macro lens for $26, shipped. I already own the right camera mount adapter. It won’t be as sharp or easy to use as the Sony 18-105mm. And I wish it was a constant aperture, but I’m hopeful it can fill the gap until I can swing the money. I’m sure I will still carry the 19mm and 35mm primes I have, but the 35-105mm could be my go to glass for future shoots.

A shotgun mic would help. Prior to the event I though I had worked out how to use a small shotgun (Rode VideoMicro) and record it into my phone. My goal was small footprint. I did not want to call attention to myself. I didn’t want to set up a full size shotgun with an external recorder. I tested the small shotgun, and would have sworn that I had the cabling worked out. But the day before the event I was charging batteries, and set up the mic to test it once more, and discovered that it was not passing signal. I needed a special cable to convert the TRS connection to a TRRS for the phone input. (Rode sells one: the SC7). I didn’t have time to get the proper adapter, so I punted. I ended up using the omni directional Smartlav+ to record audio. And, while it’s not as good as… pretty much any directional microphone at a distance, it was a lot better than the on camera mic. With some post work, some of the audio will be usable. But a shotgun mic would have been a huge help.

A camera with an audio input would help. My a6000 is a solid mirrorless camera. But it isn’t perfect, and one of the flaws is that it lacks an external audio input jack. While I would probably still use the Smartlav+ with my phone, having an on camera shotgun, recording directly into the camera would be good. Even if the small shotgun had worked, mounting the mic to my camera and then extending the cable to my phone would have been awkward at best. A much simpler solution would be to shoot on a camera that actually has the ability to record external audio. Of course the simple solution costs hundreds of dollars.

Get permission. I mentioned that this was a guerrilla style shoot. I got verbal permission from the judges in the room, and competitors. But the competitors are minors. So in order to actually use the footage I shot I cannot show any faces of minors since I don’t have permission from parents. They cannot be recognizable. I won’t identify the location, or even the organization. I knew that going in, so I shot accordingly. It would have been infinitely better to have the written permission from the event organizers, the location, and every parent of every student in each round. That wasn’t feasible for this project. In the future, I want to do more to get permissions, so I won’t be as constrained on the shoot.

As I’m closing in on the final edits of the project, I’m fairly well satisfied with it. Assuming I do similar projects later what I’ve learned with help make them even better.

[Image courtesy of Greenleaf Designs at FreeDigitalPhotos.net]