Multimedia
Blogs
Forum
Affordable HD
Whitepapers
Newsletters
Events
Advertisers
DCP Directory
MillimeterTest Drive: HP Compaq 8710p, Part 2
Nov 26, 2007 12:00 PM, By Jan Ozer
Editing with Adobe Encore.
Click here for a larger image
Editing
On the editing front, let’s start with Adobe Premiere Pro and Encore CS3 (Table 2). Here I produced two projects, the first my standard 3.5-minute HDV test product with color correction, backlight correction, speed changes, chroma key, still-image pan and zoom, logo overlay, and audio mixing. I rendered this to a Blu-ray-compatible MPEG-2 file on all three computers using the same Premiere Pro output preset. The second project involved producing a 2-minute Blu-ray disc image from HDV source material in Encore, and then recording the results to disc, highlighting the 8710p’s unique Blu-ray recording capabilities.
Subjectively, other than the obvious differences with the screen size (have I mentioned that I love my 30in. HP LP3065 monitor?), in a relatively simple project such as this, I noticed very little difference between the three computers. In other words, in cropping, trimming, setting effects and the like—which comprises most project creation time—I noticed no real difference in performance.
Table 2. Rendering performance for Adobe Premiere Pro and Encore CS3. All times in min:sec.
Click here for a larger image
In fact, as I mentioned in the last segment, the only time I noticed any real difference in subjective performance was when I was selecting camera angles in a four camera multicam project. Even there, the 8710p produced enough frames per second in all streams to help me make my camera angle selections.
When it came to rendering, however, the differences became quite evident. In producing the Premiere Pro test project to Blu-ray-compatible MPEG-2, the xw4600 was 2.68X faster than the 8710p—very close to the theoretical differential—and the xw8400 2.78X faster. When rendering to Blu-ray, the disparity dropped to 2.23X and 2.10X faster, respectively.
Not surprisingly, the record time on the 1X recorder in the 8710p was quite a bit slower than the 2X LaCie Blu-ray recorder I used on the other two desktops, but the difference wasn’t 2X. This was because the project was relatively short, and writing lead in and lead out sections, as opposed to pure transfer from computer to recorder, comprised a significant bulk of overall time. On a 90-minute project, I would expect the disparity to come very close to 2X.
Interestingly, the eight-core system barely outpaced the quad-core system in Premiere Pro, and was actually slower in Encore. If I’m buying a desktop system for the Adobe Production Studio, I’m probably not going beyond four cores.
Table 3. Sony Vegas rendering numbers. All times in min:sec.
Click here for a larger image
In Sony Vegas, I produced a test similar to the Premiere Pro project, using the same or similar effects. As before, display aside, the notebook working experience was subjectively identical to the other systems. If you have to render from the road, however, you’ll pay the performance penalties noted in Table 3.
Specifically, in Blu-ray rendering, the xw4600 came close to achieving the theoretical performance advantage (2.68X to 2.72X), though MPEG-4 encoding was less efficient. Like Premiere Pro, Vegas seems to maximize multiprocessor efficiency at four cores, rather than eight, actually slowing down on the eight-core system.
