no true - if you put HDMI stream to a DVI receiver chip that does not understand the bit pattern, the the whole stream will be rejected and you get no screen.
and HDMI DRM encryption is different than DVI HDCP.
DVI has 2 links (1 clock and 2 sets of 3 channels to carry R, G, B) where HDMI is 1 link (1 clock and 1 set of 3 channels to carry R, G, B, or YUV in varies format)
although the spec for HDMI is physical layer looks the same link DVI single link, however, since the payload has to reserve for audio stream, it is at a discounted video rate.
now top that with the encoding overhead (8B10B), etc. you are very close to just get 1080P working in most cases but not always.
the problem you see most the time when pushing 1080OP is on audio, not video. the audio either get out of sync on timing and you need to flush buffer and re-sync or the bits are corrupted that you lost a segment (silent).
this is why DVI connector supports dual links.
BTW, 1080p is not end of the story; there is a new digital cinema spec called 2Kp24 already been in work for 3 years.
unfortunately, the transmission is one way; the return path is I2C at very low rate (relatively)
thus there are no true error correction but only error detection.
the spec calls for 10 to -9th order of SER on a transmission link that meets the eye pattern requirement. that means even with the best cable, you may experience some bit error in every few seconds.
so unless the SER creep up to 10-e3 or 10-e4, you won't see too much difference.
DDWG did very extensive study on the error rate as well as the error symptons before move to 1.0 spec release.
unfortunately, not all cable are created equal.
if you read the DVI or HDMI or the raw TMDS spec (if you can find one), then you notice there is no "cable spec".
it is spec at the quality of eye pattern at the receiver end. so good cable gives you longer distance; bad cable gives you weird displays on screen and really bad cable gives you no screen.
this was not surprised as the stream cipher was designed so that old process technology is able to manufacture it.
anything man made can be broken. there is always someone smarter out there. 8-)
Slashdot Mirror
words on the street is 135 Watts; I heard someone said Intel CPU may hit 150 Watts pretty soon.
no true - if you put HDMI stream to a DVI receiver chip that does not understand the bit pattern, the the whole stream will be rejected and you get no screen.
and HDMI DRM encryption is different than DVI HDCP.
spec may says that but in reality, it's not.
DVI has 2 links (1 clock and 2 sets of 3 channels to carry R, G, B) where HDMI is 1 link (1 clock and 1 set of 3 channels to carry R, G, B, or YUV in varies format)
although the spec for HDMI is physical layer looks the same link DVI single link, however, since the payload has to reserve for audio stream, it is at a discounted video rate.
now top that with the encoding overhead (8B10B), etc. you are very close to just get 1080P working in most cases but not always.
the problem you see most the time when pushing 1080OP is on audio, not video. the audio either get out of sync on timing and you need to flush buffer and re-sync or the bits are corrupted that you lost a segment (silent).
this is why DVI connector supports dual links.
BTW, 1080p is not end of the story; there is a new digital cinema spec called 2Kp24 already been in work for 3 years.
unfortunately, the transmission is one way; the return path is I2C at very low rate (relatively)
thus there are no true error correction but only error detection.
the spec calls for 10 to -9th order of SER on a transmission link that meets the eye pattern requirement. that means even with the best cable, you may experience some bit error in every few seconds.
so unless the SER creep up to 10-e3 or 10-e4, you won't see too much difference.
DDWG did very extensive study on the error rate as well as the error symptons before move to 1.0 spec release.
unfortunately, not all cable are created equal. if you read the DVI or HDMI or the raw TMDS spec (if you can find one), then you notice there is no "cable spec". it is spec at the quality of eye pattern at the receiver end. so good cable gives you longer distance; bad cable gives you weird displays on screen and really bad cable gives you no screen.
actually, you can pick up EM waves from a distance. that's why we have tempest and FIPS.
this was not surprised as the stream cipher was designed so that old process technology is able to manufacture it. anything man made can be broken. there is always someone smarter out there. 8-)
this is incorrect at best, mis-informed at worst.
it is NOT line compatible DVI even though it does support DVI coding format input.
and it is NOT 5 gbps - it is basically 4 TMDS links at (slightly) lower rate than single channel DVI.
no - it has major problem for 1080p videp at current state.