TV monitoring service is fair use, judge rules | Hacker News Reader

TV monitoring service is fair use, judge rules

arstechnica.com

131 pointssehrope34 comments

I built and ran a broadcast monitoring business for 5 years from 2003 to 2007, competing with services such as TVEyes and Critical Mention. It was called RooseveltMedia.com. At peak, it recorded and catalogued about 400+ shows per day in 15 media markets and had about 75 customers.

As we grew, I knew we needed to expand nationwide (which would require raising VC). My saying was "We need Johnson and Johnson, not Congressman Johnson's office." but this question of legality loomed very large. Legal or illegal? If illegal, no investor in his right mind would give us money. Unfortunately, the evidence I had at the time pointed towards "illegal" rather than "legal":

1. There were lots of lawsuits. This Fox vs. TVEyes is not the first. We definitely tiptoed around TV groups so as not to draw attention.

2. The International Association of Broadcast Monitors (IABM, a group of regional services and some single-market self-employed folks) spent several hundred thousand dollars to try to get the law rewritten so that royalties would be set in stone, much like jukeboxes. Why would they do this if they didn't all also believe it was illegal? In any case, the effort failed.

With this evidence in hand, I tried to grow the company organically, but it became untenable. We weren't getting enough money from the single-market (or few-market) customers to fund the expansion. By contrast, Critical Mention was self-funded by a dot-commer, and completely ate our lunch. Even if we'd attempted to raise money, the fundraising sidetrack would have set us back by a few more miles and I had zero additional bandwidth. There really wasn't a good answer to the question.

Ultimately, despite finding product-market fit almost immediately with this freshman business effort, I decided to pivot into a more "legitimate" business. I altered the tech to provide archiving systems for TV stations... and it went over like a lead balloon, cratering the businesses and leaving me with a pile of personal debt which took a couple of years of "real world" work to pay off.

Since then I've tried to start 2 more companies (huzon.tv and words4chrome.com) and while my execution and engineering skills have become vastly better, neither found product-market fit like the first. Beginner's luck, I guess.

Sounds familiar, did you colo in Pennsauken, NJ by any chance?

I once built a system that pulled multiple TV channels, encoded them on the fly and rebroadcast via multicast over the office LAN to reporter's desks. This was for a newsroom and would have otherwise required dozens of DVB cable drops (and subscriptions). The DirecTV guy was mighty suspicious that I had a pile of Linux boxes set up next to the receivers, and no TVs in sight.

I've heard a similar setup is similar in hotels. The hotel gets a box per channel, re-broadcasts them onto coax (same as VCRs used to put eveything on channel 3, but with more channels), and then the hotel TVs can use their built-in tuners.

It's even moving past that: the last hotel I was at used the TV as a thin client that transmitted incoming IR signals back to its box in the server room.

Needless to say, the latency was atrocious. I even saw a Windows desktop flicker during a channel change.

So that's why hotel TVs always take so long to change channels. I thought they were just really cheaply made tuners.

It's infuriating when you just want to channel surf before bed and ever button press takes 5 seconds to register.

Were you just using HDMI capture cards on each receiver and tuning each to the channel you wanted to capture?

>The company has more than 2,200 subscribers, including the White House, 100 members of Congress, the Department of Defense, as well as big news organizations like Bloomberg, Reuters, ABC, and the Associated Press.

Yeah I don't think they were ever in any real danger. This isn't something average joe consumers like us use, this is something big corps, media companies, and politicians use.

The trick with all disruptive businesses is to generate broad consensus of value early, making adverse judicial decisions unlikely.

On the one hand ... TVEyes sounds like a great, useful service.

On the other hand, once all of the legal challenges are out of the way, I can't imagine there won't be a bunch of other companies swooping in to undercut them. $500/month for a service like this? I'm thinking that has the potential to go down significantly and quickly, with competition.

For instance, I can imagine a cut-rate service that doesn't do speech-to-text but instead relies solely on closed captioning. Or a plan that monitors only a limited subset of channels. It might not have all the features of TVEyes, but may have enough to hurt TVEyes' business if it doesn't lower prices and offer a limited tier itself.

It might not make sense for such a service to pop up now, while the legal issues are unsettled, but if TVEyes bears the litigation burden to get those issues settled, I can't imagine it wouldn't face stiffer competition down the line.

Are you saying you can set up an infrastructure to record every TV station and then distribute clips via multiple platforms, and charge a small amount of money for the service?

> $500/month for a service like this?

$500/mo sounds like peanuts for what a service like this provides. For a business that would use this type of information (e.g. TV shows like Daily Show or Colbert or a political campaign) it's both worth more and cheaper than doing it in house. The power bill alone for all those TV-tuners/DVRs would be more than that!

I'm surprised that it's not more expensive.

I'm fairly certain this is what the daily show uses:

http://www.snapstream.com/

... or at least something similar. In other words, they do it themselves. That's the only way I can explain their uncanny ability to find a dozen clips containing a certain word or phrase across very long time frames.

A tangent, but I thought I remembered The Colbert Report staff developing something like this in-house. After a bit of searching, it turns out that staff members had actually developed and spun-off a segment writing and producing tool [1].

[1] http://www.businessweek.com/articles/2014-06-17/stephen-colb...

This market is extremely mature now. Between TVEyes and Critical Mention, it would be very difficult to get a foothold. The disruption occurred in the 2002-2005 timeframe.

IMO, you should never attempt to compete on price alone.

>>It's a significant digital-age fair use ruling, one that's especially important for people and organizations who want to comment on or criticize news coverage.

This is why FOX News sued.

>>The company has more than 2,200 subscribers, including the White House, 100 members of Congress, the Department of Defense, as well as big news organizations like Bloomberg, Reuters, ABC, and the Associated Press.

This is why FOX News lost.

Can't they work around this with licensing?

Basically Fox add a clause to their sales contract that says 'use of this type of thing is not allowed with our service'. That gets pushed down the line to those buying the service. TVEyes are buying a domestic service, contractually they can't get that service without committing a tort (breach of contract). Et voila?

Fox surely aren't obliged to provide the service to anyone they do not wish to supply it to. The inclusion of the Fox media on the service is proof, on the balance of probabilities, that the contract was broken ...

The judge is ruling that this sort of copyright infringement is not disallowed by law; but that doesn't surely mean it can't be made impossible by contract. There is no right to rip down TV broadcasts, surely?

It does seem now that a public facing service offering clips of up to 10 minutes of TV is allowed under this ruling. Presumably that's a new thing in USA? Also how is this different to offering digests of news websites - so I can sell NYT stories now as long as I sell lots of other peoples stories too, and so long as it's recent news, and so long as I don't let them see the whole NYT website. Clipping services bought the rights to the content.

I think it's a great ruling but I don't think the breadth of it will stand as I don't consider it to match with the general impetus of copyright law. Creating a video index using CC and/or speech-to-text (or human transcribers) doesn't seem particularly transformative to me, especially as the service Fox are using already is doing similar indexing. The only transformation is in the utility of the colocation of many channels - since when (other than for Google!) was "but if we sell everyone's content then it's easier" been a defence against copyright infringement.

They can't add a licensing clause if they broadcast it over the air. No one has to sign a license to receive and record broadcast TV.

If they were solely a cable or internet network, then possibly.

Ah, right, so they only way this would work would be to not off FTA broadcasts. Who FTA broadcasts Fox, themselves?

Does this ruling have any implications for Aerio?

No. They are very different services

2200 subscribers = $1.1M monthly revenue. Not bad considering the costs don't grow that much with additional subscribers. Some for serving the content on request, but I imagine the ingestion and storage is the big ticket item.

I wonder how long they archive for?

Per the article, 32 days.

So, for research farther back, LexisNexis will be needed to find transcripts only.

Or for television news content, Internet Archive https://tv.archive.org has news from 2009 searchable by the closed captioning text, only Bay Area media market, national and international. Just added Philadelphia news too.

Seems like an interesting technical problem. Does anyone know a good way to record many TV or radio stations at once, either for storage or realtime processing?

MythTV allows you to record multiple channels on the same multiplex in the DVB world.

http://www.snapstream.com/ (I assume it's not cheap.)

Wow. Click the prices link. That site is extremely annoying.

Lots of storage + Lots of I/O throughput + Lots of software defined radios

It's actually a fun and interesting real-world distributed computing problem. You need computers w/ TV tuners in every major market across the country in order to get the _local_ broadcast news. The computers then FTP the data back to the central servers.

But you can't easily/cheaply get server racks in 100 cities that have cable TV lines going into the cages.

So you have to find reliable places to put all your distributed computers that have good TV signals, good upstream bandwidth, a person who can help reboot it when it crashes, and reasonable cost.

Back when I ran a business that competed with TVEyes and Critical Mention, I used up to 8 analog tuners in a single box to record channels. Happauge PVR-150 and PVR-500 (2 tuners in one card).

Nowadays, it's harder because you can't get many of the necessary channels via analog. You need the decryption of a digital box (imagine several cable boxes stacked on top of each other) to grab the video and even then, it's difficult to get the closed captioning. This makes the colocation complexity more expensive and many facilities would just say "um, no thanks, we're not running a cable line in here for that."

Silicondust sells some network tv tuner devices that are very usefull, I've used the home version, HDHomeRun device a few times, and it looks like thay have some big rack mounted ones that would be well suited for this sort of purpose.

http://www.silicondust.com/products/models/tech3-8us-2x4/

As someone who used to work at Critical Mention I can speak a little bit to this. The infrastructure was set up about 9-10 years ago and initially involved one very dedicated engineer driving around the country installing DVR's at the different affiliate stations in almost all (if not all) 210 DMA's. These DVR's would "phone" in with video, as well as close captioning text to the datacenters. Over the course of the last 10 years the stack has changed a few times, but more recently the real time processing occurred as a series of Python lookups queued in Rabbit MQ. The resultant data (Close captioning text as well as Schedule, Demographic, Affiliate) information was indexed in a full text search cluster. Creating a searchable set of all TV video across 90 days is an engineering feat in and of itself. That being said I don't think any companies have been able to mine the data, summarize or aggregate it in any way that can be truly beneficial to a brand.

Reply on news.ycombinator.com