Source problem or bad workflow: DVD quality vs VHS quality
 

Hello,

Just did some observation on my 40" 720p LCDTV playing DVD.

First I played a commercial movie DVD; Meet my parents (2000).

Passing signal via RCA (AV) give what we should expect from a NTSC movie in term of filling the screen. Playing with zooming option (to fill the 720p screen) altered the IQ for sure, but nothing too scary.

Then passing the signal via YPbPr (component), for sure image look more "noisy" " grainy", but not a disaster if you don't have choice.


==============

Second I played a DVD that have been made from a VHS tape and the tape himself I been made from a VCR camera capture from the 80' (I don't have the model).

Was okay at native resolution via AV. But zooming was less forgiving. For the YPbPr it was just very bad.

==============

I am wondering if this test just show the limitation my original source, basically we comparing 80' technology vs 00'?

Or that should not be that bad, meaning that the workflow was not optimal, and normally with the right tool and knowledge you should be able to get commercial DVD IQ?

Or maybe I just constating the effect of i vs p?

http://hometheater.about.com/od/befo...essivescan.htm

Thanks

http://hometheater.about.com/od/home...vidupscale.htm

Try 1970s technology. ;)

What's rather sad is that picture recording quality got worse as time went on, and not better. There was this idiotic demand for "detail enhancement" that boosted high frequency noise, giving a false appearance of detail. This was then recorded into the video, resulting in a grainy image that compresses poorly with modern MPEG-2 and H.264 compression.

Only audio improved over the life of the VHS format.

Commercial source is

• clean, from either a film scan or broadcast master tape format (1" tape like D1)
• color-accurate and calibrated at every step
• high in resolution, at least double/triple of VHS, if not closer to 2K or 4K resolution (film)
• encoded with high-end hardware, using custom matrices and GOP
• 4:4:4 colorspace

VHS source is:

• noisy, full of frequency noise
• weak and inaccurate color, due to color-under method
• low in resolution, about one-third to one-half of maximum DVD resolution (~200-360/x480 vs 720x480)
• generally encoded with software or consumer-grade hardware, which is ideal for this source
• 4:2:2 colorspace

Garbage in, garbage out.

Not that what you see is "garbage", but simply that you can't really improve it when the limiting factor is the inherent flaws. Signal restoration removes as much as possible, sometimes even removing inherent flaws (grain, chroma noise, etc), but some factors are too damaging to repair as you cannot create data where none exists. At most, you can remove flaws, you can't add what is missing.

No, it's not that -- not an issue of interlaced vs progressive.
What you're observing is simply an issue of resolution, color depth of the source, and purity of the noise.

Keep in mind, a lot of HDTVs have notoriously poor video up-scaling and sometimes poor de-interlacing. My VHS and Betamax transfers look quite lousy when I have the Panasonic plasma here upscale them, yet when the BluRay player up-scales them they look noticeably "nicer". High quality video scalers exist, but few TV manufacturers are going to incorporate them into mainstream price competitive models. Standalone scalers are also sold and many home theater receivers incorporate the same chips. Handy because so many TVs are dropping support for S-Video and composite inputs.

Image quality on my DVDs is closer to the original source when viewed on a plain old CRT TV (which I use for monitoring capture and final DVD testing). These low resolution sources were never intended to be viewed on large progressive scanning HDTVs.

Here's an interesting related post, regarding quality of VHS vs DVD: http://www.digitalfaq.com/forum/vide...ercial-vs.html
From a few years ago.

The inverse is true, too! I have a rather nice Sony HDTV, and it makes the scaler in a Philips DVD recorder look like amateur hour. Some of the better scalers are found in Sony HDTVs, which is why I've long promoted and suggested them to others. Some high end HDTVs use scalers made by Faroudja, a well-respect image processing company (similar to Snell & Willcox's reputation).

So very true. This is why I detested (for years) the baloney marketing that "HDTV will make your viewing experience better." Well, no it won't, not if you watching anything non-HD. All a high resolution display does it show you video errors in their full glory -- errors previously hidden by the flaws of CRT SD technology.

Who says that truth will make us happier :D;)