Best Video Input to Capture SVHS Tapes

[bingmel]

I'm currently using an ATI 7200 to capture AVIs from my SVHS tape collection, convert to MPEG2, then eventually make SVCDs from them. The Card currently has a Composite Input, but I'm thinking of upgrading to a card that has a SVHS input. That's when I found out that Component Inputs, apparrently the best way to view videos, existed although I haven't seen any Capture Cards with Component Inputs. Taking into account picture quality, Composite input is good, SVHS input is better, and Component input is best.

Does it really matter what type input I use to capture my SVHS videos when I make SVCDs? Although I'd like to have the best picture quality when capturing my videos, has anyone really made tests to determine whether the same video captured via different inputs noted above yields the expected video quality? Just wondering before I spend the money to upgrade my card.

Thanks in advance for your thoughts!

[easydvd]

Canopus DVRaptor

http://vcdhelp.com/capturecards.php?CaptureCardRead=Canopus%20DV%20Raptor&Search=Search

I think this one has an "S" video input on it

[snowmoon]

Componient is the native format of DVD's, but composite is the native format of VHS. SVHS just has better bandwidth that VHS, but it's still composite as it's native format.

You CAN do better, but it would require investing money into a GOOD SVHS deck with a TBC, 3d comb filter, and S-Video out.

[Nelson37]

Description of cable type quality comparison is accurate SFAIK, BUT

Unless I am very mistaken, there is no such thing as "component format" video (stream or file as opposed to signal), nor is any video formatted for any cable type. Better quality hardware will better show the differences but there is no reason you couldn't have coax out on a DVD deck or component out on a VHS deck.

The first would be like putting low-grade gas in your Maserati and the second like running Nitro on an old VW.

S-video is a cable/connector type, SVHS is a type of tape deck. The deck puts out a higher resolution signal, which will look better on S-Video cable, but which will all come through on composite OR coax, just with some slight signal loss. The same signal would look even better thru component connectors, although as someone recently pointed out some decks have decent-quality S-Video connectors but really poor quality Component connectors and in such cases the S-video is superior to the Component in terms of visual quality.

This false connection between cable type and source type is really becoming a pet peeve of mine, I have done a fair amount of research on this and determined the only real difference is signal loss and/or blending, NOT resolution or specific equipment type.

Most ATI cards have S-video input, as do many inexpensive ones. Component input is a rarer bird.

[snowmoon]

neilson...

Their are 3 signals that are used for most color video systems luma, chroma1 ,and chroma2.

In RF all of those signals are matrixed together and then placed on a FM carrier wave.

In COMPOSITE they are matrixed together and send over RCA. Color's bandwidth is a third as mich as the luminence.

In Svideo the colors are matrixed, but the luma is seperate. Colors are only about half as strong as the original, but they are no longer affected by the luminence signal.

In Componient all three parts of the video signal get their own wire.

TV's have to convert the signal back to componient and then to RGB for the raster of your set. Matrix and unmatrixing operations are both lossy, you can never guarntee the same signal out.

If your VCR TAPE is composite, then the best carrier you can use is composite. You should only use S-video is the color filter in the source is better then the destination since that is one of more important operations when it comes to color and picture fidelity from composite sources. Filters go from simple comb filters, two and three line filters, and then 3d comb filters are the best.

Mistmatching your cables can lead to degridation of the signal from multiple unessary conversions.

[Jester700]

But VCR tape is NOT composite! It has separate luminance and chrominance signals. Therefore, following your "matching" theory, all VHS should use S-video cables. The only difference with S-VHS is the luminance signal has a greater bandwidth.

[snowmoon]

Another no-brainer occurs when interconnecting any two consumer video tape devices together or to a display device. You should always use the Y/C inputs and outputs that are provided with Hi-8 and S-VHS camcorders and VCRs. All consumer analog VCRs and camcorders store the Y and C video signals separately on the tape using what is called a color-under system. So if you connect any two of these devices together you can avoid Y/C separation artifacts by using Y/C interconnects. The same holds true for connecting any of these tape devices to your monitor. Use the Y/C outputs and you will avoid the Y/C separation process, and the resulting artifacts, in the monitor.

Interesting.... all this time I had though they were just laying down a composite signal onto the tape.

[crahak]

Nope. They're separate. S-VHS VCR's will generally give better picuture too (sharper). But I think it comes down to, you get what you pay for... A 200$ VCR usually looks better than a 50$ one :/

[solarfox]

Oooo-kayyy, I think it's time for me to put on my electronics engineer hat, drag out my copy of the "Television Engineering Handbook" (an incredibly heavy tome, featuring over 1500 pages of everything you could ever want to know about the inner workings of this stuff in the most excruciating, mind-numbing detail ), and correct a few things...

INACCURACY #1: In the color-under recording system used by VHS/S-VHS (and also by 8mm/Hi-8 ), the signals are processed separately -- but they are not recorded separately. The chrominance signal (C) and the luminance signal (Y) are separated from the incoming video (if you're recording from the TV tuner or from a composite-video source), processed through their associated circuits, then added together again and recorded to tape as a combined signal through a single recording head. During playback, this single signal from the tape is again split into its Y and C elements for processing, then combined once more on its way out to the RF modulator and composite-video outputs.

The S-Video connections on a VHS/S-VHS/8mm/Hi-8 unit simply bypass one of the separation/combination stages. The signal flow looks something like this (warning: ASCII art ahead!)

COMPOSITE      ---------           ------------       -------
VIDEO IN ( 0--|   Y/C   |**(Y)****|Y PROCESSING|-----|COLOR- \      --
        _|_   |SEPARATOR|       *  ------------      |UNDER   \____[  \RECORD
         =    |         |..(C).... ------------      |COMBINER/    [  /HEAD(S)
               ---------      . * |C PROCESSING|-----|       /      --
                              . *  ------------       -------
                              . *
S-VIDEO _o  o.................. *
       |                        *    S-VIDEO OUT
IN     |_o  o********************    _o  o.......          
      _|_                           |           .       
       =                            |_o  o***** . 
                                   _|_        * . 
                                    =         * . 
          __     ---------      ------------  * .   ---------- COMPOSITE VIDEO
PLAYBACK /  ]___|COLOR-   |----|Y PROCESSING|******|  Y / C   |-------O ) OUT
HEAD(S)  \__]   |UNDER    |     ------------    .  |QUADRATURE|        _|_
                |SEPARATOR|____ ------------    .  |MODULATION|         =
                 ---------     |C PROCESSING|......|(COMBINER)|
                                ------------        ----------

That being said, the color-under encoding is somewhat different than the Y/C-to-composite encoding, so they are somewhat easier to separate cleanly during playback -- but it's not accurate to say that they're stored separately; Y and C still get combined into a single signal, and thus can interfere with each other during the multiplexing and demultiplexing processes.


INACCURACY #2: DVD's "native" format is not composite, S-video, or component. The correct answer is, "none of the above, because all of those are analog signal formats and the video on a DVD is, of course, stored digitally as MPEG-2 data."

Once the RGB information for each pixel has been decoded, then it can be made composite, S-video, or component video as required, using the same methods used to encode broadcast-TV or analog-videotape signals, but none of these is any more "native" to DVD than the other.

[snowmoon]

Once the RGB information for each pixel has been decoded, then it can be made composite, S-video, or component video as required, using the same methods used to encode broadcast-TV or analog-videotape signals, but none of these is any more "native" to DVD than the other.

DVD's don't decode to RGB, the decode to YUV ( YCbCr to be more specific ) and that is a digital representation of componient signaling. DVD's would have to be colorspace converted to get to RGB.