Software-Based TIVO?

HBergeron asks: "Tivo and her various competitors, including the new Sony box, are intriguing tools for the broadcast media consumer, but devices themselves are nothing more than a variation on the special purpose PC, and you pay a substantial premium. Why don't we see a software based system that will run on our own PCs (Linux, Windows, Mac, etc.) rather than requiring another duplication of hardware we have heavily invested in? It may not work for everybody, but there are enough people with high-end, TV capable machines to make a market. After all, we're the ones that adopt TIVO-like technologies first. Is it access to programming information from the networks? Are there legal issues?" I think the only thing standing in the way of applications to do this are drivers for TV-capable video cards under the OS of choice for the specific user. Does Linux have support for the variety of TV video cards that Windows does? How about for BSD and Macintosh?

Comments

Cost Effectiveness

[Detritus]

How much money are you going to really save?

I have a ReplayTV box made by Panasonic. It includes a 30 GB hard drive, MPEG encoder/decoder, modem, TV tuner, systems software and a lifetime subscription to the program guide service. I'm not sure what CPU is used or how much RAM is in the box.

Without the program guide service, these boxes sell for about $400. It can be less if you get it on sale or with a rebate. $400 is not going to buy much hardware for a PC.

Much of the usefulness of the box is dependent on the program guide service. Are you going to type in the contents of TV Guide every week?

Re:Cost Effectiveness

[HBergeron]

I appreciate the thoughtful reply, and I don't know the etiquette of replying to ones own submission, but I thought I would elaborate on the idea.

This time next year I would expect 1Ghz level cpus with TV capable graphic cards and 40 Gig drives should have penetrated at least 10% of the home market. (This is 9+% higher than the current penetration of Tivo like devices.)

Stats show that this edge of the market tends to own 2 or more computers, (rising with the number of kids) and has at least one in the "living room". That doesn't really take into account us bleeding edge adopters (who are a substantial fraction of current Tivo customers) who may have one (older) machine as an MP3 jukebox, a home server, and several linux boxen running, well, stuff...

regardless, we appear to have the market penetration of hardware to make us of this kind of service, so the investment to launch is much lower - no hardware design, contract gaurentees or distribution. When you look at the difficulty of getting hardware into the home, the idea of using a pre-installed base should be an appealing business model. Heck, the software and weekly programming can be downloaded from a web site.

I'm not saying this is superior to a stand alone device, just that it should be a viable and desirable option to those who already have the hardware capability at home. It would be hard to recreate the device for the same price in a pc case, but if you already have the tools... After all, I don't need three power saws - or do I....

Lots of reasons

[funkman]

TV and home computer don't look together. That may sound petty, but that is reality for the masses, think about it. (Something to digest: The iMac had nothing to do with its technical features, it looked cool)

The family PC, would be need to be near the computer, very distracting for homework.

If the TV and PC are a distance apart, networking will be involved. That will eliminate 95% (or more) of the population. (Face it, the /. readers are an extremely tiney minority)

For software makers, the market share is not great enough to show me the money. Even those who had the ability, would need to perform one of the above.

Support, support support. See previous statement about market share.

Re:Lots of reasons

[maggard]

I agree that *right* *now* it's a better idea to simply buy a dedicated Digital TV Recorder then to retrofit one's PC - I just think funkman got it all wrong. Here's his points, mine, and some more:

• TV and home computer don't look together. That may sound petty, but that is reality for the masses, think about it. (Something to digest: The iMac had nothing to do with its technical features, it looked cool) Don't look good together? Sorry, but even amongst my most decorator-obsessed friends I don't know anyone who'd not find a way to work around this if they really cared. As to why iMacs sold: hint - it wasn't the colors. Yes they're a break away from the traditionial beige box and that made them trendy in certain circles but it was much more that they were the first resonably priced Macs that came out in a long time, they were agressively marketed, technically were good for their market and Apple didn't look like it was about to slip below the waves leaving everyone who bought one with an orphan.
• The family PC, would be need to be near the computer, very distracting for homework.First many of us don't have "homework" in our houses (the most likely spot for a child to be in my house would be on the dining room table with an orange/ginger glaze and an apple in their mouth.) Of my acquaintences who do have progeny many of them seem to do quite well with the TV and the computer sharing the same room. Indeed I expect if one were to survey families with PCs most would be found in the same room as the main TV set, some because it's the common room in the house, others so they can kep an eye on their kids computer use.
• If the TV and PC are a distance apart, networking will be involved. That will eliminate 95% (or more) of the population. (Face it, the /. readers are an extremely tiney minority) W-i-r-e-l-e-s-s. H-o-m-e-l-a-n. Both are happening now - not just amongst geeks but out in middle-class suburbia. Furthermore this has been going on since cable-modems were developed (y'see the computer gets connected to the cable-TV which usually comes into the house near the TV set even though they look just awful near eachother...)
• For software makers, the market share is not great enough to show me the money. Even those who had the ability, would need to perform one of the above. This isn't a software issue: this is a hardware one. A bunch of companies (video-card folks, disk-drive makers, even PC vendors) would love this to happen. It wouuld set a new spec for PCs, a high-end one. The software would get written and tossed out there just to support the hardware, much like CD burner software is bundled with CDRW drives. There would probably be a similar market with the low-end version included in the box and a more fully-featured one available, or the listings able to be subscribed to. As to your second sentence - left over from a bad edit?
• Support, support support. See previous statement about market share. Yeah, Dell, Apple, Compaq, Nvidia, ATI, Seagate, Maxtor, they don't have any market share between them. Just 'cause this could be big sales for them doesn't mean that they'll go after it...

So what are the other points?

• You'll see PCs capable of doing this stuff availiable off of the shelf in the next 12 months. The technology is already there in bits & pieces, someone just has to put it all together. iMacs DV's already do much of it, MS WinME also includes most of the same technology.
• Of course this will require folks keep their PCs on 24/7, the PCs be quiet enough to not be intrusive, the PC's 'sleep' & 'wake up' (power-conserve) to do the recording, and of course the recording not interfere with the computing or vice-versa.
• There will need to be a way to control the computer's recording & playback from the couch. This will reqire some sort of custom remote (another one!) or better yet a USB-tethered IR 'eye' plugged into ones PC with some smart backend software that could work with many types of remotes. This would actually be almost the easiest part of the system as the hobbyists would all rush to make every remote they've ever owned worked with this plus their Palm Pilots and everything else.
• There would need to be a source of TV listings avaliable in a format one's software can digest (TV-XML?) This could come from the Cable vendor or third parties but there's likely to be some sort of revenue recovery involved - this can't be cheap to assemble and I'm sure there are Intellectual Property issues involved somewhere.
• Dedicated boxes will prove popular since they won't interfere with the computing side of things. Additionially they can be loaded with other features like set-top gaming, MP3s storage & playing, web-browsing, etc. They could even be sold as 'Home Servers' offering firewall, censorware, routing, and other features along with their basic TV functionality. You'd end up with a desk-based PC for keyboard-type things such as MS Office and an appliance for couch-potato stuff like TV watching and casual web-browsing (I want Brittany's skirt! Click on it and a brower pops up pointig to her website with the skirt in the video already loaded into the page)

  Look, this is about the merging of the TV and the computer. It's big, it's happening, and companies like MS are rooting for it in so many ways. There's no need to go into the details, convergance has been discussed for years and these are the devices it's happening through.

PowerVCR may be a step in the right direction...

[Hobart]

CyberLink, makers of PowerDVD have a product called PowerVCR -- however I don't think this has the TV guide and record-what-you-might-like features of TiVo.

Wave Systems' joint venture with Sarnoff

[Anonymous Coward]

is doing exactly what you want. BSD using Microsoft TV for PCs.

Wavexpress and Microsoft Demonstrate Advanced Windows Broadcast Software at IBC 2000
http://www.wavexpress.com/00.09.08.html

Amsterdam, Netherlands September 8, 2000 - Wavexpress, a new technology company providing broadcast commerce for terrestrial, cable and satellite broadcasters, today announced they are working with Microsoft Corp. to develop advanced digital television software solutions for the PC industry. Wavexpress will provide a live demonstration at the International Broadcasting Convention 2000. The demonstration will couple the Wavexpress datacasting application with Microsoft TV Technologies, yielding a fully integrated end-to-end solution for delivering interactive content and commerce to consumers. The Microsoft TV Technologies is part of the extensive television support being provided in future versions of Microsoft Windows.

Microsoft is providing a next generation driver architecture targeted at digital television applications and is providing critical API functions to support key operations such as network selection, channel tuning, stream de-multiplexing, audio/video decoding, and IP packet extraction. Providing these functions within the core operating system will yield multiple benefits, including a common tuning and stream control model, a modular component architecture, enhanced datacasting support, and the flexibility to seamlessly utilize either hardware or software based transport demultiplex and MPEG decoder solutions.

"A new class of services is emerging from the rapid deployment of digital networks, and Wavexpress provides a compelling multimedia experience uniquely enabled with broadcast commerce," stated Harish Naidu, general manager of the Video and Audio Division at Microsoft. "By exploiting the power and flexibility of Microsoft TV Technologies along with key Microsoft solutions such as Internet Explorer and Windows Media Player, Wavexpress has created an application
that showcases the benefits of datacasting in a digital broadcast network."

Wavexpress has developed an innovative broadcast commerce service that combines the power of broadband datacasting with Wave Systems' EMBASSY distributed transaction and rights management system. In leveraging EMBASSY's Trust @ the Edge capabilities with the benefits of a high speed digital broadcast medium, provided via terrestrial, cable or satellite systems, rich multimedia content and a broad range of e-commerce merchandising models will enhance the delivery of the Mass Media Internet.

"By addressing critical system challenges, Microsoft TV Technologies represents a significant step forward in enabling digital television applications on the personal computer," according to Cliff Jenks, CEO of Wavexpress. " The Microsoft TV Technologies initiative provides the key building blocks needed by a PC to support broadcast applications and extends the completeness of our solution by minimizing the complexities associated with processing a high speed data stream."

The Microsoft booth at IBC2000 (321) will include a demonstration of the receipt and processing of a live satellite feed, yielding standard digital television and the Wavexpress enhanced Internet service.

About Wavexpress, Inc.
Wavexpress is establishing the digital marketplace for the 21st Century. A joint venture of Wave Systems Corp. (Nasdaq: WAVX) and Sarnoff Corporation, Wavexpress delivers rich content-movies, music, sports, news, computer games, software-over a digital broadcast signal direct to any DTV-enabled PC. Wavexpress provides a full range of secure data broadcast architecture, infrastructure, and services to broadcasters, content providers and consumers. Unlike competitors, Wavexpress' broadcast e-commerce system offers a variety of business models, including ad-supported content distribution, pay-per-use, purchase transactions, and subscription services. Integrating over $100 million in research and development, Wavexpress capitalizes on the broadcast engineering expertise of Sarnoff, and Wave Systems' Embassy technology and back-office systems that handle distributed content protection and transaction reporting functions. For more information, visit the Wavexpress web site at: www.wavexpress.com.

About Wave Systems
Wave Systems' goal is to build a worldwide network of users based on trusted electronic relationships. Trust @ the Edge defines a new architectural model for the Internet, which embeds trust and security in every user device. Wave Systems is developing, deploying and licensing its Embassy Trusted Client technology for the mass adoption of this revolutionary model. Integrating industry standard functions from a wide range of partners that enable reliable, secure digital exchange and commerce over the Internet. At its core, Wave Systems is building the services, and enabling 3rd parties to build services that will take advantage of this new open Trust @ the Edge model. For more information about Wave Systems and Trust @ the Edge, visit www.wave.com.

About Trust@ the Edge
Current solutions involving, privacy, security and commerce are based on the centralized network portal model. This model takes advantage of PC's for processing, the Internet for connectivity, and the World Wide Web for browser navigation, but lacks the one component that will enable the true explosion of digital exchange and commerce, as well as enhance privacy - embedded trust and security in Internet user devices. Trust @ the Edge is revolutionary in that it turns the Internet inside out, moving core security and e-commerce functions out to the edge and places them in the user's Internet device. Trust @ the Edge is the integration of strong security in every user device which provides for the creation of trusted relationships and enables reliable digital exchange and commerce over the Internet. For an overview of Trust @ the Edge, visit www.wave.com.

its being worked on.

[Jose]

There is mention of it every now and then on the V4L list.
You really need V4L2 to actually do it (V4L1 doesn't support two programs accessing the tv tuner at the same time)..but it should be ready from prime time rsn.
For the mpeg encoding, the best tool I've seen so far is mp1e (mirror here ...the real site seems to be down..)
For grabbing the listings, have a look at xmltv. There should be a new release coming out soon, with support for more countries...

ccdecoder (search freshmeat) also looks promising...it has the potential for grabbing listings right off your cable tv line..

TiVo is a service

[digitalbeing]

Fundamentally, TiVo is providing a service. The fact that the first iterations of their product involved a box they put together and designed is so that they can keep the cost to the consumer low by subsidizing the initial hardware purchase, and enforce copyright restrictions, among other reasons. Long term they are not in the business of selling inexpensive mpeg encoders.

ATI probably has the closest to a drop-in, TiVo on a PC offering, the Radeon All-in-wonder.

It includes an interactive program guide, record on demand, and can pause live tv.
http://www.ati.com/na/pages/products/pc/aiw_rade on/index.html

A company called gemstar has been vigorously enforcing their patents on electronic program guides, which is why it is doubtful you will see a completely freeware version of an interactive program guide service anytime soon.

Re:TV guide as HTML

[epeus]

tv.yahoo.com gives a readily parsable HTML version for every zip code.

However, TiVo believe that the TV Guide and correlation subscription is their real business, and the boxes are just a way to sell it. They might support a software version if you used their service

realtime MPEG encoding

[ikekrull]

Realtime MPEG encoding in software is not really an option. You might be able to do it on a 1GHz PC or something, but thats an expensive MPEG encoder...

What would be really cool is if a device like a Tivo supported streaming of the MPEG it was playing back over ethernet through, say, a Java based (for cross-platform considerations) client. Something like the Video4LAN application.. You could control and view your Tivo's playback stream through a wireless network so you could watch TV on your laptop or something like that.

I guess you could build something like that out of a linux box without too much trouble, given hardware support for the MPEG encoder card..

You could then archive the stream to your local hard drive easily, without needing an MPEG encoder in that machine.

The Tivo-like device is your local TV-server. Just log on and watch what you feel like watching out of the 30GB buffer.

This of course, would probably make the MPAA, RIAA and probably various other dumbass corporations throw a squealing fit on the floor, since the possibilities with regard to then broadcasting or archiving the content on the internet are obvious and tempting ;)

Re:realtime MPEG encoding

[clackamas]

I don't believe you are right. The windows media encoder for one thing does not do MPEG2 the next thing is that it requires a file as the input source not a live stream. Anyway even if it did it can't match the quality of hardware specific MPEG2 encoding.

Re:realtime MPEG encoding

[ikekrull]

Well, i have yet to see a software solution that will even encode fullscreen MPEG-1 in realtime. By fullscreen realtime i mean taking a SIF (720 x 576 x60fps PAL) (compressed or uncompressed) video stream + audio, and digitise that into 352 x 288 1.5 MBit/s MPEG stream in high quality without dropping frames or sailing off into artifact-land.

as for MPEG-2 in realtime in software?? no way, unless your definition of either 'realtime' or 'software' are very differenet from mine.

Re:realtime MPEG encoding

[Ryokurin]

That isn't possible with Mpeg-1. The max that it can take is 352x288. To do 720x576, you need at least Mpeg-2 or Mpeg-4.

Details on what's already do-able and available

[Vito]

This question has been asked no fewer than two times before, and one time, I even answered in +3 detail on exactly what would be needed to make a PC-based TiVo.

But that's okay, let's rehash.

Since we're going entirely software-based, e.g. you're sitting a normal, icky PC in your stereo rack, or you're just using your PC as normal, you probably don't have a hardware MPEG encoder. The best you've probably got is a Matrox card with onboard MJPEG compression, and I don't think the Linux drivers support that.

Now, assuming you already know how capture a video stream and pipe it to an MPEG encoder (and trust that your system is fast enough to not drop too many frames; think P3/500 or better), the only thing you really need to do is add in TV listings, and integrate them into channel changing and record functionality.

Copy and pasted from my previous post, channel guides are easy. Just have a Perl script rip and reformat any of the listings from the online providers, including Excite TV, Ultimate TV, GIST TV (which also provides the Yahoo TV listings), Ask TV (in the UK), Click TV (what TiVo uses), TV Quest, TV Grid or TV Guide Online.

As for integration, a lot of this work has already been done, at least for satellite TV streams. Klaus Schmidinger produced his Video Disk Recorder which performs channel guides and VCR functionality on his Linux PC, for his satellite TV using a PCI card. All GPL'd, so feel free to port it over to plain old TV cards, too.

--Vito

Re:Details on what's already do-able and available

[Xylantiel]

Several notes. Last time I checked, I think they had gotten MJPEG to work for the matrox cards, though it was still a little flaky (not all resolutions and such) but it should be improving quickly. And there are other video capture boards that do work for MJPEG (the MPEG ones are the hard ones to get specs for I think.)

Now on to your second point. You assume real time compression. Why? I typically watch things 1day+ after they are broadcast (videotape). That combined with the fact that I typically record less that 10 hours a week mean that as long as I had a big enough buffer, I could spend several hours compressing a show (overnight). Also MPEG compression is very parallelizable, why don't I just distribute it over all the (4 or so) computers I have. Makes a lot more sense than attempting to do it real time, and you can do a LOT better quality MPEG compression too. As soon as you go asynchronous the options really open up.

The script thing -- I've already written a script to parse search output for clicktv.com. Works great. Certainly makes it easier to check for missed reruns. The utility of something like this is a lot smaller for a vcr though, since the real big deal is that you have to program it and feed it tapes.

The real issue for me at this point is where do I put all the stuff I record. DAT tapes are the cheapest, but the reason is the drives are really expensive. Right now I just can't compete with videotapes. (I typically record half a season of all the shows I watch and then watch them all at once.) That and I haven't had time to set it up, which is the real kicker, but maybe sometime in the future.

Maybe sometime... I'm trying to help out with LiViD right now.

Re:Details on what's already do-able and available

[clackamas]

Microsoft has a patent on parallel encoding of MPEG. With the wrath of MS you won't be able to make it fly. http://patents.uspto.gov/cgi-bin/ifetch4?ENG+PATBI B-ALL+0+972171+0+8+48226+OF+8+98+1+"vide o+encoding"

Used a TiVo?

[NetJunkie]

If you're asking this, I don't think you've used a TiVo. For the 10th time, it isn't just a digital VCR. First, they cost $199 for the 20 hour model right now. $199!!! You aren't going to build anything close for that. Take that 20 hour and add a big HD to it sometime. Second, there is an enormous amount of effort put in to the software and interface. It does more than just record a show at a certain time. Check one out, and you'll be hooked.

Re:Used a TiVo?

[djweis]

Where do they have them for this price now? I've not been able to find them.

Re:Used a TiVo?

[NetJunkie]

Best Buy has the 20 hour for $299, and you get $100 back from TiVo.

This might be practical

[X]

This might actually be practical if you do it with a digital satellite receiver card, much like the DirecTiVo systems that are expected any week now.

Re:Cost Effectiveness -- is not the issue

[Xylantiel]

I don't know about the guy that submitted the question, but cost isn't really the issue for me. I have other issues. I want a digital vcr but I want features that are not being sold. I'll give some examples with explanation.

Fully programmable. Mr. Malda's chief complaint about his tivo was the lack of programmability. I have figured out how to look at the TV guides and know what is and is not a rerun and what is new and what is syndication. I want to be able to tell my box this too, not wait for somebody to decide to provide some variant of this for who knows what price. I want control!

No monitoring. I don't want to pay to have somebody record my viewing habits for who knows what purpose, no matter how good their intentions or what I get in return. I want to be able to anonomously download TV schedules and use them as I see fit. Without targeted advertising.

An uncrippled box. In order to archive tivo you have to use a VCR! This is asinine. I want the raw digital. I don't care if they have 'issues' with it. This is like them selling me a car, but putting in a limiter or something that prevents me from exceeding 45mph so that they won't be sued if I exeed the speed limit. Sounds pretty ridiculous huh?

My display. Face it. TV's suck. 60 hz interlaced. 3:2 pulldown on movies. Also, compared to computer monitors, TV's (except for the really expensive ones) are really behind the times. I'd like to be able to play with refresh rates and have digital level color quality and undo the 3:2 pulldown if possible.

Control over compression quality. I'm not stuck in the dichotomy of SP vs EP. I'd like a scale in between. I'd like to be able to adapt based on content. Also I'd like be able to update to better compression. MPEG-4 inhancements? better encoding algoritms? even the basics, like are they using an IEEE compliant DCT? Stuff like the last are probably irrelevant on a TV, but when you start displaying on computer monitors the fine points count.

Automated commercial removal. Not fastforward, I mean no commercials. I think it could be done. But I don't expect some large company to do it. Basically for the same reason they don't provide a digital out.

So there's what I want. I don't think these are going to appear in a box any time soon. There are just too many legal 'issues'. That and the fact that the more freedom you have the less the box makers are going to make off of you. They don't want a customer to sell a box to, the want a leash with a 'consumer' attached to the other end.

Sorry this post has been pretty bitter. It's nothing personal to you, I just turned into a rant :)

Later.

Re:It exists today and it uses Windows Media

[clackamas]

MSMPEG4 is really H.263 with a wrapper. This is the same videoteleconferencing codec that has been around for years. What I want to know is why everyone is hung up on the encoding? The killer solution is an ethernet port on your TiVo hooked up to your DSL/cable modem and just download the already encoded flick. A 1 to 1.25 Mbit encode produces a VHS quality encode, and would download in just a couple of hours for a feature length film. Say for instance your download speed was 500Kb (an example) and the movie was 90 minutes at 1Mbit you could start to watch in 45 minutes. Would not be hard to do if the TiVo had a PCI slot. Napster for movies anyone? You could call it vidster if the domain was still avail (not). What is needed is someone in the braindead media world that realizes that a low quality encode is not a threat. I could do much better decrypting DVD's. In fact this could be the way to get the content. Decrypt the VOD files and do a software encode to the desired bitrates.

Electronic TV Guides? Not so big an issue.

[JCCyC]

What I'd really like is to be able to simply record at predetermined times, watch the result and then... edit the output. Yeah, that means cutting out commercials at the exact frame, translating the close-captions to Gaelic if I were Irish and burning a beautiful video CD to store on my shelf.

The way it is now, to have my collection of, say, X-Files episodes, I sit in front of the TV and press the pause button at the right times. I have only one chance to do it right (I ruined some episodes figuring out my new VCR's pause delays) and it's VHS, which will degrade over time. Season 1 and 2 episodes have muffled sound already.

I pay for cable TV. As long as I don't go selling those CDs, why shouldn't I be able to do as described in the first paragraph? Yeah, yeah, I know, there's still some money left in my pocket and Jack Valenti doesn't like that.

We have the technology for this now. It's not in the TiVO because they're scared stiff that the MPAA will sue them into oblivion. Wait and see the shit fly when someone figures out the storage format they use.