|
|
|
HDV:What You NEED to Know 2nd Edition
|
This month we've begun focusing on the future of consumer digital video; AVCHD. The Advanced Video Codec High Definition format is rapidly replacing the DV format in the consumer space, and will soon work its way up into the pro-sumer with forthcoming products.
In this article, we're looking at the Sony HDR CX7 and
its various applications in the pro-sumer workflow.
In short, there is a dearth of information about the format and an abundance of ignorance. Opening the CX7 box, it was a surprise to see the
camera uses a docking station similar to previous Handycam-styled models
from Sony; the recent HDV camcorders and other AVCHD camcorders had not
included this feature. It would seem that Sony was intent on being in
the lead of the "who can make the smallest camcorder" race rather than
providing ease of use for the user of the product. The box includes RGB,
Composite, and USB cables, plus the docking port. Additionally, the
package provides a power supply, software for browsing, a "Let's Enjoy
Video" DVD, battery, remote control, and a very thin owners guide.
There is no memory card in the box, which means purchasers of the
camcorder will need to plan on spending at least 100.00 for a 4GB card,
and around 175.00 for an 8GB card. We decided to test the CX7 in a few situations. The first encounter was in daylight, shooting green foliage being blown by heavy winds. The camcorder held up very well. You can see in the reduced image below, that colors are bright, detail is fine, and motion blur due to the encoder is minimal. Next, I shot into the sun using manual exposure options in the camcorder. The manual exposure mode is easy to work with, if you don't mind a touchscreen menu. This could prove to be problematic on a cheap tripod, but overall, I don't find touchscreens to be bothersome. An appreciated aspect are the number of manual options. With manual focus, manual exposure, and manual shutterspeed, users can fairly easily control the image and get great pictures. However, many manual adjustments such as shutter speed, audio levels, and the camera dial found on comparative Sony models is missing. This may be an issue for those familiar with previous camcorders. At higher bitrates (15Mbps, which is all we ended up using), the image is nearly the quality of HDV, at roughly double the compression. Lower bitrates were definitely challenging to the image quality with the 7 Mbps and 5Mbps bitrates being so poor that even the small playback screen on the camcorder displayed macroblocking. I did not test the Standard Definition modes in any of the various bitrates such as 9, 6, or 3 Mbps. Following the first opening and a few test shots in manual modes, the next day brought the test we were most interested in performing; finding out how well the camera performed in high speed sports such as skydiving. The small size of this camera make it optimal for rock climbing, BASE jumps, skydiving, kayaking, and other sports requiring lightweight and good quality, and where hard drive or DVD-format camcorders cannot be used. (Hard drives fail at altitudes greater than 10,000 ASL/Above Sea Level.) We strapped a Royal Lens .5 to the front of the camera, which has a 37mm thread. The Royal lens is not a high def lens, but is one of the better single element lenses we've found. Ground footage with the cam came off very nicely, although significantly softer and distorted due to the width of the added lens. This lens is very common in the skydiving world as the its low profile reduces snag hazards during deployment.
Placing the camera on a Bonehead Flattop Pro camera helmet, I jumped with Mike Semanoff, a newly crowned tandem instructor and his student. The camera performed well, although the lack of a LANC is a serious shortcoming for most sports uses, as it doesn't provide a means for the camera operator to know whether the camera is functioning or not. The Optical Image Stabilizer was turned off for this jump, but it appears the gyro system isn't locked down when this feature is disabled, as evidenced by the softer image, which should appear sharper rather than less sharp. The OIS system appears to be the same as the OIS found on the HC7 camcorder. This alone makes for a challenge with this camera in some situations such as skydiving, where EIS is generally preferred, or no stabilization at all is used. Once we were on the ground, it was time to test the workflow. I shot seven minutes of footage for Mackenzie and Mike's jump, this included prejump interview, in-aircraft interview, skydive, and post-jump interviews (the final video can be seen here). Media was captured at 15Mbps on a DuoPro 8GB card.
Strangely enough, the Sony VAIO laptop's internal slot used to transfer the media would not recognize the 8GB card. Using a no-name brand card reader, the laptop immediately recognized the card, and transfer began. The seven minutes of footage took just under 2.5 minutes to transfer, thus shaving off quite a bit of real-time transfer that tape would have required. The next step was to open these files up in Sony Vegas 7e software. Vegas immediately recognized the files, generated audio peak files, and was ready to edit. With anticipation, I rubbed my hands together and set to editing the skydiving video. Generally, a skydiving video takes approximately 12-15 minutes to capture, edit, and render to SD MPEG for DVD output, but obviously I was way ahead of the game due to the transfer vs capture of the media files. How wrong I was... Playback of the AVCHD files on my 1.83GHz VAIO was around 8fps. Incidentally, the files couldn't play from the card, they MUST be transferred to a hard drive prior to editing. Editing of the files ran no different than editing of DV or HDV media, excepting that due to playback restrictions, 10key editing wasn't as easy. In Sony Vegas, the project properties were set to the 1080/60i template, and that was all that needed to be done. Out of curiosity, I set the project settings to NTSC Widescreen, but the highly compressed AVCHD files still had to be decompressed of course, and playback speed didn't improve at all, so I went back to the 1080/60i template. Completing the edit took about four minutes vs my normal three minutes, due to the lack of ability to speed through the playback, but trimming was identical to trimming HDV or DV, so this wasn't too much of a consideration. It's more difficult to park the cursor during shuttle playback when the playback/framerate is low, so this required a small amount of pixel by pixel shuttling, which accounted for the extra minute or so during editing. Then the true pain began. The speed I'd gained in the transfer process was lost in the render process. A DV widescreen file typically takes around four minutes to render to an SD Widescreen MPEG file. The seven minute AVCHD timeline took over 20 minutes to render, due to the extremely high decompression/re-compression ratio. As BD (Blu-ray Disc) becomes more available, this will speed significantly, particularly with improvements on the editing/NLE technology being developed and announced. A render to MP4 for uploading to YouTube took nearly 30 minutes. For the best AVCHD experience, you'll definitely want a fast computer. iMovie 08 fared a little better than the PC side of things, due to the convert-on-import of the format, but still wasn't nearly as fast overall as I could do with the native AVCHD format in Vegas. However, output/render time was nearly half that of the render to MPEG in the Windows world. Granted, part of the render time involved some special FX up front, but I wanted to see in both Apple and PC applications how the codec held up to blurs and other spatial shifts. It did fine. By comparison to it's HDV counterpart, the codec, the camera did well, particularly when cost factors are taken into consideration. The HDR CX7 is a fabulous small camcorder, if you don't mind a few limitations and some gimmicks. The video quality is very good, even though it's extremely compressed. The buttons/touch screen are easy to navigate which is a good thing, as the owners manual is extremely light on details and pertinent information. There were quite a few questions about the camcorder that I still had/have, but repeated emails to Sony Electronics public relations team went unanswered for nearly a month, so, the questions will have to suss themselves out over time/use with the camera. Offering bitrates of 15Mbps XP, 9Mbps HQ, 7Mbps SP and 5 Mbps LP, it is extremely flexible, although I can't recommend using any bitrates other than 9Mbps or 15Mbps. Theoretically, 15Mbps should be better than HDV, although when comparing images from an HC5 (HDV/25Mbps) with the image from the CX7 (AVCHD/15Mbps), the HC5 still provides a more robust image, and is of course, significantly easier to edit.
Testing in low-light, the nod still has to go to HDV. The DSP and lesser compression of the HDV camcorders (HC3, HC5, and HC7 models) offer a far superior image when compared to the CX7 in low light. Greater compression means a greater reveal when it comes to noise, and the CX7 definitely shows noise in darker areas. In reasonable or good lighting, the CX7 looks very impressive. Speaking of lighting conditions, the camera offers a physical button for backlit situations, which is very thoughtful of Sony to provide. Nothing is more frustrating than fumbling through touchscreen menus when the lighting isn't right and the scene is moving away. Audio quality on this camera is lacking. There is no other way to describe it. Due to the faux 5.1 surround recording system, the slightest wind from any direction kills audio from any other direction. With no microphone input, this could be a significant challenge for anyone using this as a primary camera. The menu does offer a gain reduction option, but this is a fixed option. We're experimenting with the camera now to see if we can find a way to improve the audio. Intial tests indicate we can significantly improve the audio quality via technique and some small tricks. Another option of course, is to purchase the Sony wireless system made for the AIS input/shoe on the camera, or purchase the VMC-k100 adaptor for around $20.00 and will allow users to plug in any mic they wish.
Battery life is impressive, particularly for the skydiving world; I was able to shoot for 10 jumps without replacing/recharging the battery. Of course, the LCD panel isn't used for most of this sort of work, so the battery drain is significantly less. Expect around 90 minutes at best with the LCD panel open. I did quickly test the still image
capabilities of the camera; images look good for most purposes such as
web use, small prints, and so forth, but don't plan on replacing your
DSLR with this smaller camera. It is 6.1 megapixels, but the lens and
processing aren't up to side by side comparison with a dedicated still
camera. Mentioned previously, I did not test the Standard Definition
modes of the camcorder, as SD isn't something we ever find ourselves
wanting to deal regardless of the type of work we're doing.
The only AVCHD competitor in its class is the new (but already discontinued) Panasonic SD1, and there is no comparison. In my opinion, the Sony CX7 left the Panasonic SD1 seriously wanting for more in terms of image quality, ease of use, and overall build quality. Additionally, the ProDuo card used by Sony did not require any specific driver software or hardware as the Panasonic SD1 requires with the SDHC card. Overall, this camera is a great buy for
the low cost, and very, very convenient. It's clearly aimed at the point
n' shoot crowd, but can be used for other purposes. A couple of recommendations;
use a tripod whenever possible, especially in highly dynamic scenes such
as shooting over sunlit water. Purchase an extra memory stick; storage
fills fast at 15Mbps. Avoid the lower bitrates for the most pleasing
results. If you're shooting frames for YouTube or other UGC sites,
you'll likely be very pleased at the 9Mbps quality level, which provides
for longer/more media on the disc. *Skydive footage is shot with a non-HD wide angle
lens adaptor, thus reducing optical resolution. |
||||||||||||
|
BOOKS from VASST
|
|||||||||||||
|
|
|
|||||||||||