Firewire is able to handle 400Mbit/s transfers easily, and has a defined standard for DV transmission which is very usful for people who use DV cameras. Firewires video standard should mean you can get a firewire-equiped DVD player or set top box, connect it to a firewire-equiped digital video recorder or TV, and get a digital connection between them for fantastic quality at HDTV resolution. But this transfer is unprotected. There is a slim chance that a technical genius with a logic analyser costing ten thousand dollars and a cray supercomputer could monitor this bus and capture content on it. More seriously, if an appliance tells a firewire DVD player it is a Sony Digital-Non-Recorder, ordinary firewire provides no authentication system to prove its not really a PC. As usual paranoia takes over and various studio and TV executives respond to the horror of an open digital interconnect so the DTCP standard appears.

Plain firewire connections on appliances are almost unheard-of, because of massive resistance from the movie industry. DTCP is basicly an encryption and authentication system for firewire which can be used for digital connections between appliances. DTCP can also be used to encrypt USB connections, although with the limited bandwidth of USB I cant see much use for it. DTCP-encrypted external sound interfaces perhaps, or DTCP-encrypted MP3 player links?

Actually DTCP does have an understandable use. Its part of the CPSA, so any appliance which uses DTCP connected input or output will need a CPSA license, which will require it respects all the CPSA rules such as only saving it to an encrypted disc, only outputting on protected outputs and running watermark checks.

I brought a firewire card which contains absolutly no protection. But as DTCP is implimented in software for PCs it doesn't really matter. (The spec says "CEs", that is STBs, televisions, digital recorders, etc, can use either software or hardware implimentations). Just in case though, never know what might appear in them next. Sony released a firewire chip with DTCP internal, though its targeted at cable and satalite decoders, not PCs.<.p>

DTCP was devloped by the 5C condortium, that is Sony, Toshiba, Intel, Hitachi, and Matsushita together with the MPAA.

Firewire/DTCP interfaces on appliances have been very rare. This is essentially a support and license problem. The CSS license doesn't allow digital output of any sort, except the modified S/PDIF surround audio, and requires SCMS protection on that. And without other appliances to connect to why would anyone want a digital interface? This problem should be solved in a few years though, as firewire/DTCP slowly appears on home entertainment appliances and creates its own demand. And perhaps through scared studios demanding more encrypted interfaces instead of the macrovision/cgms combination which has failed to protect DVDs. Already some studios are looking with intrest at the new blue laser discs and an excuse to replace the ineffective CSS system

Just to clear this name mess up, The technical name for firewire is IEEE1394. The name firewire is used because its more consumer friendly then a number. The i.link interface manufactured by sony is just firewire by another name, and it is identical in every way. So IEEE1394, firewire and I.link are interchangeable. Sony has recieved heavy criticism for its rebranding of firewire, which appears designed to create consumer confusion. If a non-techie buys a Sony digital camera with a "i.link" interface they will probably start looking for compatable equipment with the familiar "i.link" logo. Its another popular trick, Intel did the same thing rebranding 802.11 interfaces as "centrieno" and there are plenty of other examples.

DTCP is in very active devlopment so I havn't put any specs up on this site. The DTLA website contains some fairly technical specs, though important bits are missing. The website explains they "omit specific sections and sensitive information" which can be obtained "by executing the License Agreement and paying the associated fees as prescribed in the license agreement.". Obviously that license agreement includes a NDA, but if someone finds a way to trick it out of them it could be quite intresting. DTLA is the licenceing administrator in charge of DTCP.

DTCP, being a CPSA technology, comes in two forms. Strong and weak encryption. Strong is a paranoid 512-bits with an equally paranoid authentication system (public key). Weak is weaker, designed to avoid the large processing requirements of strong which could be a problem on some equipment, such as portable players, and uses a different authentication system (common key). copy-never flaged content can only be sent over strong. Copy-once and no-more-copies using either, but strong is prefered when available. In practice, it is unlikely any major copyright holder would choose weak encryption when strong is an option.

High-bandwidth Digital Content Protection is a new standard for protected transmission of very high bandwidth video over short distances. Raw video over a VERY high speed link, with an option to use two interfaces for even bigger displays. In this system monitors, HDTV TVs and anything else able to display video but unable to record it are classified as DVIs, or Digital Video Interfaces. Every two seconds the source checks its still connected to a compilent, certified, non-recording DVI. HDCP also includes the self-destruct System Renewability and is planed for integration into the CPSA.

I have two PDFs about it, a simple one here and a more technical one here. The upstream protocall is in a seperate file here. Im not quite technical enough to fully understand the upstream part, but it appears to me about establishing a secure channel within the computer and graphics card, presumeably to stop anyone pluging a logic analyser into the bus and recording video that way, but could also present various backwards compatability and OS compatability problems. If linux doesn't have a secure driver for a video card, how will the card respond? HDCP is designed for intel DVI ports, a digital replacement for VGA. Anything with a whole lot of bandwidth will do but it is primarily designed for DVI. Both the DVI connector and HDCP are from Intel.

One crypto researcher found a vulnerability which will allow anyone with access to 50 monitors and four computers to break it. The paper was delayed for a long time because of DMCA fears, but is now available. I have it mirrored here. Dont get too excited though. Althrough it will reveal the master key, thus allowing anyone to make a fake-compliant device, actually using the key still needs a logic analyser capable of reading the high-speed HDCP connection and a computer able to process the amazeingly fast raw video stream, so its not practical. Perhaps some of the more skilled hobbyists can find a use for this when making their own hardware, but its not much use to the people who want to back up their DVDs. Its only use is making uncertified but fully functional hardware, and that requires a large-scale operation which would probably suffer legal action from Intel.

HDCP has been criticised because the only signal it protects is the one the user is watching. The aim is to stop people recording DVD or other computer-displayed video by intercepting data in the monitor cable. The main fears here are possible problems for monitor repairs setting off anti-tamper systems and giving intel far too much power over DVI. DVI is of course intel-controled, while the familiar 15-pin monitor connector was an open standard. DVI would give intel even more control of DVI through key control, forceing all monitor and video card manufacturers, as well as many HDTV equipment manufacturers, to buy licenses.

If you have any DVI video cards look after them before the encryption arrives.

HDCP can also be used for connecting appliances and sending raw video. This has the advantage of allowing HDTV onscreen displays without having to go through the quality-reduceing decompress-edit-recompress cycle. But its an expensive system, requireing an enormas bandwidth and a correspondingly high ammount of processing power, so expect firewire (DTCP-protected) in the vast majority of consumer equipment. HDCP will probably be limited to professional equipment needing lossless interconnects and computers.

Video enthusiasts will hate HDCP most though, because it greatly limits wireing systems. For example, I have a cable decoder connected to TVs in three rooms by composite cable. In a HDCP (or DTCP) system I would not be able to do that because only one TV could be authenticated at a time. The cable companys probably wouldn't mind that, because people would have to rent one STB and subecription for every TV in the house, but it would be very annoying.

Very little DVI+HDCP equipment is available currently, though DVI video cards and monitors (no HDCP) have been available for a few years. Its a common problem. Theres no demand for DVI equipment because there is no equipment available to connect to it. JVC manufactures some D-VHS recorders with DVI+HDCP support, but calls the system "D-theater". Sony is not the only company that can cause deliberate nameing confusion.

If an HDCP source cannot authenticate the reciever and establish an encrypted connection it will play at greatly reduced quality. No specification says precisely how reduced. The CPSA likes to define reduced quality vaguely as "analog VCR quality", which actually means 480p, the lowest quality standard available, or normal broadcast TV. Other documents suggest (but only suggest) 480p as a maximum quality for an unauthenticated device, which is about normal TV quality.