In a September Widescreen Review report on the Planar 8150 1080p DLP projector, Greg Rogers indicates it uses the new Gennum GF9450 video processor. He wrote this processor's motion adaptive deinterlacing for 1080/60i provides a noticeable decrease in resolution pumping while viewing late-night talk shows, plus "the best job I have seen eliminating jaggies and line twitter on the 1080i broadcast video."

FP reviews last year (http://www.avsforum.com/avs-vb/showpost.php?p=11843791&postcount=195) also indicated an earlier Gennum processor used in JVC's RS1 outperformed other processors in minimizing resolution pumping (his follow-on post (http://www.avsforum.com/avs-vb/showpost.php?p=11848094&postcount=196)). "Images were sharp with excellent detail and extremely fine textures, and only minor resolution pumping with movement," his RS1 review noted.

But the current Planar review notes that the newer processor delivers slightly softer motion-adaptive deinterlaced images compared to an earlier Gennum (GF9351) and the Realta. "Strands of hair on static images were also not as distinct as they have been with previous projectors," he wrote.

AIUI, viewing fine details, such as the weave in clothing with static shots, reveals resolution pumping when a slight movement causes sudden excessive blurring. Perhaps somewhere along the line someone will quantify "slightly softer" using test patterns, even though the effective resolution of programs can't match higher test pattern resolutions. -- John

Can you give an example of what quantifying "slightly softer" would consist of?

In motion adaptive de-interlacing, moving subject matter is bobbed while stationary subject matter is weaved (woven). If nearby let alone all stationary subject matter goes blurry when something starts to move, then the de-interlacing algorithm is poor.

It is entirely possible for stationary material to be bobbed slightly all the time whereby the finest detail still shows through and also the transition to full bob for the moving parts is less obtrusive when something just starts moving. It is entirely possible for moving material to have some weave blended in all the time so the resulting softening is less but some combing (motion artifacts) will show up faintly.

Video hints: http://members.aol.com/ajaynejr/viddoubl.htm

^^^Don't watch deinterlaced HD/SD myself. Still clinging to a year-2000 9"-gun 64" RPTV. But if VPs in high-end front projectors perform as Greg notes, it's worrisome for any progressive display acquired, it seems. My link in the OP summarizes his resolution pumping observations for 4 FP reviews last year, and the OP summarizes this month's Planar review.

He seems to use late-night talk shows, consistently for 1080/60i video comparison purposes, and recall one review where he praised the crispness from a static pattern on a guest's clothing, then noted how a slight shoulder shrug kicked in resolution pumping, blurring the clothing pattern. The linked post mentions a seemingly worse instance of pumping. So, while the current Planar review suggests less resolution pumping over the already top-rated earlier Gennum VP, it notes a reduction in overall resolution of images. 'Slightly softer' (the current WSR review) is indeed ambiguous, so it might be useful to have some numerical comparisons. The ATSC test committee, in its 1995 report (http://www.atsc.org/news_information/papers/1995_acats/tsreport.pdf), for example, listed measured results for a 5-rpm test pattern (table 2.3), shown in this simplified table (http://archive.avsforum.com/avs-vb/showthread.php?postid=326565#post326565), contrasting the readings with static resolution. Other rpms were used but not listed. -- John

EDIT: Here's a related comment in Greg's >$3k-forum Planar-review thead:
http://www.avsforum.com/avs-vb/showpost.php?p=14417502&postcount=23