"Standardized" clip for color/clarity comparisons?

[InMaster]

I hope some of the heavyweights (among others) on www.videohelp.com read this forum . . . this isn't a Newbie question, but a question that didn't fit in a more specific Forum . . .

Everybody's got opinions on what hardware/software/techniques yield the best color and clarity in output.

Is there a still/clip (downloadable?) that could be used for comparing, say, Capture Card A with Capture Card B, or Encoder A with Encoder B?

I'm envisioning some kind of circle of colors and/or some kind of grid which would make color changes, 'ghosting', and other changes more obvious and less subjective. And maybe some kind of fast action to tax a system/method.

If there's not anything close to a 'standard' testing clip, what sort of torture test clips (from a movie, for example) have you found to make weaknesses show up in your capturing/encoding/converting experience?

[flaninacupboard]

if you google you should be able to turn up plenty of test patterns. there's such a huge number of variables with this that a site wide comparison covering people on different continents won't end up very useful. person A in a 50hz zone usng expensive cables and a good dvdplayer will get totally different results from person B in a 60hz area using crappy cables and a supermarket dvd player, even using the same capture card and software.

encoders however can be tested ok. the problem is that you need a clip with high detail, fast motion, dark scenes, long enough for VBR to be effective, some high noise areas and some low noise areas, saturated colours, realistic fleshtones and natural movement. it also can't be heavily compressed already. really you want a Huffyuv compressed file that's full D1 and about ten minutes long. i don't fancy hosting that

it's very difficult to pinpoint what ill be troublesome, it just comes with experience. as an example, encoding the monty python movies i had to totally adjust my encoder settings. a lot of scenes were poorly lit and they shot on cheap film stock. this involved a lot of visible grain which made my encoder crap out. a few teaks later and everything was ok.

[BJ_M]

also downloading a clip will not help much on comparing capturing cards ...

you would have to have all capture the same thing - with the same cables and connection

[InMaster]

Yep, good point -- for comparing captures, one would have to start with something analog, which won't be downloadable . There'll never be a "standard", although here on this website as well as at other sites I've seen references to certain scenes in some movies, such as The Matrix, as showing strengths/weaknesses in conversion/compression/etc. processes.

There is a site http://www.epanorama.net/links/videotest.html which contains some good links to test patterns (hey, including the old Indian Head that used to be the signal for TV stations after hours!) and video setup tutorials. The best I found, anyway (outside of this website, of course) . . . no torture tests, that I saw; one would have to do the movie route for that.

Thank you for the input!

[BJ_M]

ive posted this here before - but i will repost it as it has been awhile ///

huge collections of test video (and audio) @

ftp.tek.com

look in the /tv/test/streams/ as a start


this is the :


T E K T R O N I X F T P A R C H I V E