Testing methods... Your thoughts

[fmctm1sw]

I plan on some in depth testing of my avi to mpeg2 conversions. I have been converting VHS tapes to 352x480 DVD complaint files. I found that my image is softened at an undetermined stage of cap/convert/etc when compared to the original VHS tape on my TV. I have experimented with sharpening (and a lot of it) to see what effect it would have. Still stuff looks great but motion is horrible (yeah, I know... "News flash..."). I noticed that IVTC material with pulldown flag doesn't look as bad as hard telecined encoded stuff. My testing plan is this: make up a bunch of CVD stuff to compare on my TV. I plan on using the "step" feature of my standalone DVD player to step through the frames to compare the picture. Does anyone see anything horribly wrong with this line of thought?

[sacajaweeda]

Read this thread....

https://www.videohelp.com/forum/viewtopic.php?t=199669

[fmctm1sw]

I know that one... I started it! I've migrated to trying to figure out what will look best on my feeble 22" TV...

[sacajaweeda]

I know that one... I started it!

Yep! 8)

[Nelson37]

One suggestion. Make up a key for the disks, noting the methods used and labeling each disk "A, B, C, "etc. When viewing, note "A is better than B, C better on motion, and so forth. This will prevent any bias regarding method from affecting judgement on fine differences.

This originally hit me when judging multiple clips. I used long filenames to describe each clip and my Apex only displayed the first 8 characters, which were often the same. On several occassions the observed quality contradicted what I expected from the method used.

Also, while different disks should not be a problem, I did discover that certain types of clips would alter the playback of subsequent clips, throwing off my results. This was discovered when a clip listed in the middle or end would look significantly worse than the same clip played first. It was play order, not order on the disk that mattered. I do not recall if cycling the power was necessary to remove the effect, but it wouldn't hurt. Lost a lot of sleep determining this problem, I am sorry but I don't remember exactly what characteristics caused this. Very high or very low bitrate, hi-res, interlacing, IVTC, framerate, it was something along those lines.

Take breaks, make lots of coffee, and use other people to confirm results.