What's missing? Well, if you step up to the HD-XA2, you get everything on the HD-A20 plus multichannel analog outputs, a coaxial digital audio output, and a RS-232 port. Those connections can certainly be useful, but they're not huge omissions if you have an HDMI-capable receiver and don't need the RS-232 functionality. Tech-savvy buyers will note that the HDMI output is only version 1.2, versus the newer 1.3 and 1.3a specs. While this is true, buyers shouldn't worry as there isn't any enhanced functionality on the HD-XA2 because of the HDMI 1.3 port.
The HD-A20 supports 1080p output via its HDMI output, although it doesn't support 1080p output at 24 frames per second. While some have claimed this feature can reduce judder with compatible displays, we haven't seen an increase in performance with the 1080p/24 Blu-ray players we've tested. However, we are continuing to test 1080p at 24 frames per second with different displays, so we cannot say definitively that it won't make a difference in some setups. As the vast majority of buyers don't have displays that are capable of taking advantage of this feature, it's not a big omission as far as we're concerned. Toshiba supposedly is also planning to release a firmware upgrade that would allow the HD-A20 to output at 24 frames per second.
HD DVD Performance
Because the HD-A20's major upgrade over the HD-A2 is 1080p output, the performance of the HD-A20 in 1080p mode is critical to whether the HD-A20 is worth the extra money. To start, we looked at some of the tests from Silicon Optix's HQV test suite on HD DVD in 1080p mode on our Pioneer Pro-FHD1. The results were disappointing. The HD-A20 struggled with the Film Resolution Loss tests, with moirÃ© present in the image of stands at Raymond James Stadium, and a strobe effect occurring on a shifting resolution pattern. The HD-A20 also did a poor job with the Video Resolution Loss test, with the same strobe behavior occuring on certain areas of the test pattern. We even saw significant jaggies (stair-step patterns along the edges of lines that should look smooth) on the Jaggies test with three shifting lines--which is notable, as most players and displays ace this test. We looked at these same patterns with the HD-XA2, and it passed every single test. Additional test discs also confirmed the HD-A20's limitations in 1080p mode--most importantly, the player failed to resolve the finest detail on test patterns from the HD DVD version of Digital Video Essentials.
We also looked at actual content to see how often these failures affected program material. Unfortunately, we found several instances where the HD-A20's subpar 1080p performance had a significant impact on the viewing experience. When we loaded up MI:III, we noticed some flicker on the bottom of the main menu right off the bat, along with some significant jaggies on the computer-generated graphics in the background. Jumping into the movie, one of the worst examples was at the beginning of Chapter 8 of MI:III, where there's very noticeable and distracting moirÃ© on the stairs in the background. At the beginning of Chapter 16, we could see obvious jaggies on the limo as it rolled up to Tom Cruise. These are only the some of the easiest places to spot the jaggies--we saw them popping up constantly throughout MI:III.
While MI:III looked pretty mediocre for HD DVD, other movies exhibited significantly less jaggies. We didn't see nearly many visible jaggies in Batman Begins, and Aeon Flux was also mostly jaggy-free, even on difficult scenes such as Chapter 9, where we've noticed issues before.
In all, whether the HD-A20's 1080p HD DVD performance is worth it over the step-down HD-A2 really comes down to how well your HDTV handles 1080i deinterlacing. If your HDTV has average or better built-in video processing (and most do), the HD-A20 definitely isn't worth it. If your HDTV has noticeably poor video processing, then perhaps the HD-A20's 1080p mode will provide a slightly better viewing experience. Whether that's worth the extra money is up to you, but it's worth noting that very few of the HDTVs CNET has tested over the last year exhibited worse video processing than did the HD-A20 while in 1080p mode.
Load times were definitely an issue on the first-generation HD-A1, but luckily the HD-A20 is a bit faster. We tried four discs (MI:III, Blood Diamond, King Kong, and Aeon Flux) and they all loaded in about 30 seconds from the time we hit the close-tray button until the picture showed up on the screen. When starting from the off position, it took us about 52 seconds to get MI:III playing.
Since most buyers of the HD-A20 will have larger DVD collections than HD DVD collections, the HD-A20's DVD performance is important as well. We kicked off our tests with Silicon Optix's HQV test on DVD in 1080p mode and were a little disappointed. For example, two of the jaggies tests were clear failures, with both a rotating line and three pivoting lines appearing full of jaggies. These jaggies were also present in footage of a waving flag. We were able to minimize some of the jaggies by switching to 1080i mode, which might provide better video quality, depending on your TV. On the other hand, the HD-A20 did a good job with a 2:3 pulldown processing test, kicking into film mode in under a second. The HD-A20 also had no problem displaying all the detail of DVDs, as demonstrated by its solid performance on a resolution test pattern--although again, the image was more stable in 1080i mode versus 1080p on our Pioneer Pro-FHD1.
The HD-A20 faired better with actual content material. It showed off its 2:3 pulldown processing during the introduction of Star Trek: Insurrection, with boats and curved railings rendering smoothly. It also did an excellent job with the difficult introduction to Seabiscuit, looking just about as good as we've seen this sequence. To be fair, the HD-A20 handled all of the film-based discs we threw at it pretty well, so while it may have failed some of the difficult video-based tests from HQV, it should do a good job with standard film-based movies.
DVD load times with the HD-A20 clocked in at 17 seconds from the time we hit the close-tray button until the time a picture came up on the screen.