Opinions on the best video port

[mxh]

I watch TV video on my PC in two ways. The commonality is that both methods
use a DCT6412 digital/HD dual tuner DVR converter provided by my cable
company:

1) Y Pb Pr component cables from the converter to the component input on my
Dell 2405fpw.
2) S-video from the converter to the s-video input on an ATI AIW X800XT AGP
card.

I also have the option of running Video out from the converter to to the
video in input on the X800XT.

My question is: Which of these methods provides the best picture quality?
I'm pretty sure I have to use the Y Pb Pr for High def, but that leaves me
with the choice of video out or S-video for the x800xt, unless s-video
carrys an HD signal as well. Which method is recommended here?

Thanks,
mxh

 

[William]

Y Pb Pr is the best, for it is non-color encoded (sorta - not modulated at
3.58 mhz and mixed with Y yet ), offering the highest bandwidth for color
fidelity. With the right encoder, you could get full bandwidth RGB from this
format. It is used in frame stores, special effects generators, and
character generators of a few years old. (Before Digital Video) (BTW - YPB
is a variant from England, where America uses YIQ for NTSC encoding. YBP
became popular from international demand for standards) (Care to know the
difference between NTSC, PAL, and C-CAM? -NO?)

Next best is S-Video. Also known as Y/C. Luminance and Chrominance. The
color is encoded to 3.58 mhz vistisual side-band modulation, but not encoded
to NTSC video, which limits the luminance channel to around 2.8 meg for
low-pass filters, or 4.2 meg for 2 line comb-filters. S-Video can give you
luminance bandwidth in excess of 5 meg, and give multi-generation copying of
acceptable viewer performance. If you are going Hi-Fi this would be a
minimum acceptable format for you.

Lowest is Video. Try to use decoders that use 2-line comb filters verses
low-pass filters if you can. Your TV will have specks that spell this out.
It will affect the sharpness of the video on the screen. Especially when
dealing with analogue Hi-def TV. Makes a difference on recorders, video to
video movement. And ALWAYS keep the encoding/decoding to a minimum.

William

 

[stratus46]

Y Pb Pr is the best, for it is non-color encoded (sorta - not modulated at
3.58 mhz and mixed with Y yet ), offering the highest bandwidth for color
fidelity. With the right encoder, you could get full bandwidth RGB from this
format. It is used in frame stores, special effects generators, and
character generators of a few years old. (Before Digital Video) (BTW - YPB
is a variant from England, where America uses YIQ for NTSC encoding. YBP
became popular from international demand for standards) (Care to know the
difference between NTSC, PAL, and C-CAM? -NO?)

Next best is S-Video. Also known as Y/C. Luminance and Chrominance. The
color is encoded to 3.58 mhz vistisual side-band modulation, but not encoded
to NTSC video, which limits the luminance channel to around 2.8 meg for
low-pass filters, or 4.2 meg for 2 line comb-filters. S-Video can give you
luminance bandwidth in excess of 5 meg, and give multi-generation copying of
acceptable viewer performance. If you are going Hi-Fi this would be a
minimum acceptable format for you.

Lowest is Video. Try to use decoders that use 2-line comb filters verses
low-pass filters if you can. Your TV will have specks that spell this out.
It will affect the sharpness of the video on the screen. Especially when
dealing with analogue Hi-def TV. Makes a difference on recorders, video to
video movement. And ALWAYS keep the encoding/decoding to a minimum.

William

<snip>

You're mostly right. Y, Pb, Pr ('component') is simply RGB after going
through matrix to convert to Hue, Saturation and Luminance. The
constants are a little different betweeen std def and HD. In std def
you add 59%G +29.5%R+11.5%B to get Y you then take Y and subtract 100%R
to get Pr and do the same for blue to get Pb. Note that so far there is
no bandwidth limiting and no loss. You can reverse the algebra to get
the original RGB. So, why bother with this? It is very easy to alter
color saturation, hue, white and black levels -- and if you get
carried away you can control the Hue, Saturation and Luma intensity of
individual R,G,B, Yellow, Cyan and Magenta (and even more). When the Pb
and Pr components are used to quadrature amplitude modulate the
subcarrier, the bandwidth is limited so you are no longer lossless. The
signal is 'Y/C' until the subcarrier is added onto the Y signal to
become the composite video. This is NOT vestigial sideband. That
function is performed on the visual carrier after it has been AM
modulated with the composite video.

Separating the Luma from the Chroma is the most difficult task from a
performance point and your comments about comb (1 line or as you say
better 2 line) filters vs trap are right on.

The digital world, both std def and HD, transcode the RGB to component
but then band limit the color difference components to 1/2 bandwidth,
though there are special full bandwidth units used in special effects.

Is C-CAM SECAM?

GG

 

[William]

<snip>

You're mostly right.

You have not contradicted anything I have said, you have only further
explained what I have stated. Did your explanation help the original poster
as to his query? Or did you only stroke your ego?

Do you feel a need to prove yourself superior to other people? I've been
around the block too many times to play this game with you.

Is C-CAM SECAM?

Yes. Perhaps I sould have got my notes out. Their a little dusty.

William

SMPTE / NAB

 

[mxh]

Thanks to both of you for the indepth responses. I am trying to record
content from my STB to computer. The two options I have are two machines,
one of which has the ATI x800xt (from which I haven't tried capturing from
external video sources as of yet), and the other, a Navis Pro capture card
(which captures some very high quality video). Both cards have s-video
input.

Although the X800XT has plenty of input options, it appears that it has no
component input, but instead offers a component to s-video adapter, which, I
suppose, is s-video quality. So, if I can impose further:

I would like to be able to switch between s-video on the x800xt and s-video
on the Navis pro (the x800 for when I am just watching and the Navis Pro for
recording). Is there some kind of s-video A-B switch that won't cause a
reduction in quality?

Also, some of the captured content will be from a high def source. Will I
experience any problems capturing that via HD to DVD compliant mpeg2?

Thanks again for any information,
mxh

 

[mxh]

Y Pb Pr is the best, for it is non-color encoded (sorta - not modulated at
3.58 mhz and mixed with Y yet ), offering the highest bandwidth for color
fidelity. With the right encoder, you could get full bandwidth RGB from
this format. It is used in frame stores, special effects generators, and
character generators of a few years old. (Before Digital Video) (BTW -
YPB is a variant from England, where America uses YIQ for NTSC encoding.
YBP became popular from international demand for standards) (Care to know
the difference between NTSC, PAL, and C-CAM? -NO?)

Next best is S-Video. Also known as Y/C. Luminance and Chrominance. The
color is encoded to 3.58 mhz vistisual side-band modulation, but not
encoded to NTSC video, which limits the luminance channel to around 2.8
meg for low-pass filters, or 4.2 meg for 2 line comb-filters. S-Video can
give you luminance bandwidth in excess of 5 meg, and give multi-generation
copying of acceptable viewer performance. If you are going Hi-Fi this
would be a minimum acceptable format for you.

Lowest is Video. Try to use decoders that use 2-line comb filters verses
low-pass filters if you can. Your TV will have specks that spell this
out. It will affect the sharpness of the video on the screen. Especially
when dealing with analogue Hi-def TV. Makes a difference on recorders,
video to video movement. And ALWAYS keep the encoding/decoding to a
minimum.

Hi William,
Another thought: May I assume that, even though 'video out' is the worst of
the 3, it still surpasses the quality of just the 75 ohm coaxial from the
STB to the cable input (TV or graphics card)?

Thanks,
mxh

 

[William]

Hi William,
Another thought: May I assume that, even though 'video out' is the worst
of the 3, it still surpasses the quality of just the 75 ohm coaxial from
the STB to the cable input (TV or graphics card)?

Thanks,
mxh

STB = Set Top Box? Then yes, to offer Ch 3/4 TV frequency out, an additional
step of modulation / demodulation is required. ALSO, unless you have a very
expensive STB, you do not have STEREO sound! Dolby, (sp?) the inventor of
stereo TV, has patents on stereo TV and charges extra for use of the
technology, so most Set Top Box - TV modulators skip stereo out. This
includes cable boxes, DVD players, and VCR's - check the specks for stereo
on the TV frequency out - it won't be their most of the time.

A rule of thumb: The less you molest your analogue signal, the better it
will look. This kinda goes to the heart of Digital Rights Management. If
digital video degraded as much as analogue video does, their wouldn't be
DRM. You can only encode / decode video so many times and it starts to look
bad, does the jiggle, tarrying, granny dance for you. I digress.

Keep it short, keep it sweat, and beware of lip-sink. (Where the 'moving
lips' video corresponds to the sound track.) A common problem with moving
video/audio between equipment. BTW this is quite a problem in modern desk
top computers. I run into this problem quite often on my own setup.
Especially computer recorded TV played back later.

William

 

[William]

Thanks to both of you for the indepth responses. I am trying to record
content from my STB to computer. The two options I have are two machines,
one of which has the ATI x800xt (from which I haven't tried capturing from
external video sources as of yet), and the other, a Navis Pro capture card
(which captures some very high quality video). Both cards have s-video
input.

Although the X800XT has plenty of input options, it appears that it has no
component input, but instead offers a component to s-video adapter, which,
I suppose, is s-video quality. So, if I can impose further:

I would like to be able to switch between s-video on the x800xt and
s-video on the Navis pro (the x800 for when I am just watching and the
Navis Pro for recording). Is there some kind of s-video A-B switch that
won't cause a reduction in quality?

Also, some of the captured content will be from a high def source. Will I
experience any problems capturing that via HD to DVD compliant mpeg2?

Thanks again for any information,
mxh

Their are too many variables involved to give you a definitive answer. For
one, one encoder is different than another. I would assume (and I use that
word with fear) that the Navis Pro has a better decoder built in? Or why did
you pay so much money for it. On the other hand, a few years of technology
marching forward will make a two dollar chip run circles around what a 200
dollar board used to do. So without seeing the specifications on the two
products, including the component to s-video encoder, that call is
impossible to give you. (Bandwidth, depth of modulation, cross-modulation,
Signal-to-Noise level, etc.)

My suggestion: It's time for the grand experiment. Try every way
conceivable, take notes. See which works best for you. Then use that
method until something changes.

You have other problems to be concerned about. S-video needs to be kept
short, less than 3 or 4 feet in length, or you start to loose quality.
(Their are distribution amps for this purpose.) A/B switches need to be
hi-frequency shielded, or you may get cross-modulation, (bleeding ghosting)
in the video.

Some video from STB's have anti-copy protection methods installed and cause
great trouble when you go to capture the video. I think it's called
Macrovision. A varying pulse on line 19h that wreaks havoc on decoders that
can't scrub this stuff off. So beware, if everything starts to go dark, and
pops up bright, and then goes dark, over and over, suspect Macrovision.
(They make scrubbers for this infliction.)

William

 

[mxh]

Hello again, William. Thanks for the response, and please forgive the
delayed response.

STB = Set Top Box? Then yes, to offer Ch 3/4 TV frequency out, an
additional step of modulation / demodulation is required.

Unfortunately, yes. I'd really prefer to stream the data via the firewire
port, but it seems that Cox STBs don't have the firewire ports enabled,
according to them, and all efforts to get XP to recognize it via drivers
fail.

ALSO, unless you have a very expensive STB, you do not have STEREO sound!
Dolby, (sp?) the inventor of stereo TV, has patents on stereo TV and
charges extra for use of the technology, so most Set Top Box - TV
modulators skip stereo out. This includes cable boxes, DVD players, and
VCR's - check the specks for stereo on the TV frequency out - it won't be
their most of the time.

I'm no expert by any stretch of the imagination, but the STB looks to be of
a fairly decent quality. It's a DCT 6416 III HD/Dig DVR with dual tuners
(the DVR is nice, but archiving is what I'm after). I haven't had too much
time to play with it, but I did some test records (recording from live TV,
not the DVR) via s-video to the X800XT.

According to TMPGEnc DVD Author, the audio is:

mpeg-1 Audio Layer-2 48000 HZ Stereo 224 KBPS.

I haven't had the chance to try the s-video to the Navis Pro I mentioned
earlier, but it encodes to mpeg-2 and the audio bitrate is 384kbps.

The sample file I recorded with the ATI card doesn't sound like mono.

A rule of thumb: The less you molest your analogue signal, the better it
will look. This kinda goes to the heart of Digital Rights Management. If
digital video degraded as much as analogue video does, their wouldn't be
DRM. You can only encode / decode video so many times and it starts to
look bad, does the jiggle, tarrying, granny dance for you. I digress.


Keep it short, keep it sweat, and beware of lip-sink. (Where the 'moving
lips' video corresponds to the sound track.) A common problem with moving
video/audio between equipment. BTW this is quite a problem in modern desk
top computers. I run into this problem quite often on my own setup.
Especially computer recorded TV played back later.

I've heard of those who have that issue, although I never have myself. I've
done a lot of analogue recording (using the Navis Pro and also an AIW 9000
Pro using a straight Moto digital STB converter as the tuner with a 75 ohm
cable connection from box to card), but that may change now that I'm
fooling with s-video.

Thanks,
mxh

 

[mxh]

Their are too many variables involved to give you a definitive answer.

Yes, I have entirely too many options .

For one, one encoder is different than another. I would assume (and I use
that word with fear) that the Navis Pro has a better decoder built in? Or
why did you pay so much money for it.

Yes. It is a unique card, although I considered the price to be quite
reasonable, given its quality output. It was only around $400 5 years ago.
Oddly, Pentamedia really didn't do much with this card (this is the Navis
Pro with the NTSC tuner). There was only one version of the software
(although they did produce beta version 1.1...) and it is quite quirky, but
most who have had experience with this card will agree that the video
quality can be compared to cards costing much more. Although Pentamedia
abandoned this card not long after it came out, it can still be found at
various sources on the internet. However, there are no software updates
since beta 1.1 that I'm aware of. A shame. Great hardware, pathetic
software.

On the other hand, a few years of technology marching forward will make a
two dollar chip run circles around what a 200 dollar board used to do. So
without seeing the specifications on the two products, including the
component to s-video encoder, that call is impossible to give you.
(Bandwidth, depth of modulation, cross-modulation, Signal-to-Noise level,
etc.)

Here is as much info as I could find:

http://www.cestuff.com/hardware_encoders/navispro_specs.html

My suggestion: It's time for the grand experiment. Try every way
conceivable, take notes. See which works best for you. Then use that
method until something changes.

Yes, I agree. I also will try (once again) to find someone at Cox who is
knowledgable about the STB I have. As I've said, it has firewire ports
(disabled), but there are other ports as well (ethernet, USB, SATA, etc.). I
would much rather stream the digital data, even if via USB2.

You have other problems to be concerned about. S-video needs to be kept
short, less than 3 or 4 feet in length, or you start to loose quality.
(Their are distribution amps for this purpose.) A/B switches need to be
hi-frequency shielded, or you may get cross-modulation, (bleeding
ghosting) in the video.

I wan't aware of the 3 ft. s-video philosophy, but it makes perfect sense.

Some video from STB's have anti-copy protection methods installed and
cause great trouble when you go to capture the video. I think it's called
Macrovision. A varying pulse on line 19h that wreaks havoc on decoders
that can't scrub this stuff off. So beware, if everything starts to go
dark, and pops up bright, and then goes dark, over and over, suspect
Macrovision. (They make scrubbers for this infliction.)

Thanks for all of your input, William. I haven't had as much time as I'd
like to dive into this, but am spending time as it becomes available. If you
have any other ideas on how to obtain the best quality given my set up, I'd
appreciate it.

Thanks,
mxh