Los Angeles (CA) – So today is the day we tell you which camcorders shot the clips we featured in yesterday’s HD shootout. We admit that this was a rough test and definitely not scientific, but it did give us a sense of the cams’ differences in image quality and handling. You can read yesterday’s article here.
To give you a bit more background on our tests, we had just purchased the JVC HD-110, Canon HV20 and Canon HG10 camcorders. Yes our credit cards are really hurting at this point (Capital One loves us). We’ve owned the Panasonic DVX-100B for several years and it has definitely proven its worth in the hundreds of videos we’ve shot at TG Daily and at our previous jobs at Tom’s Hardware Guide.
But isn’t the 21st century all about HD? Well the problem here is that we shoot exclusively for the web and all clips are downconverted to 420 by 315 pixels. Even worse is the bitrate is lowered to 650 kbps so that most (not all unfortunately) broadband customers can stream the footage. So the massive increase in initial resolution might not matter that much when transferred down to a lower resolution.
You can view the four clips we shot, all downrezed to 420×315 and at 650 kbps bitrate, on yesterday’s article. We’ve also popped up 1280 by 720 pixel HD clips from the JVC HD-110 and Canons HV20 and HG10 on Vimeo.
Video #1 – Sony DCR-HC28
Ok video #1 was probably the easiest one to figure out as the standard definition Sony DCR-HC28. Most people recognized the standard definiton aspect ratio along with the shakiness of the video. We actually handheld all the cameras to test their image stabilization effectiveness. The JVC is a shoulder camera and we used it as such. The Sony cam does have ‘Steadishot’ image stabilization, but that doesn’t do much when you are at the maximum 20X optical zoom put out by the Carl Zeiss lens.
The Sony was primarily used as a deck for our Panasonic DVX-100B, but we did shoot some videos at Demofall 2007 with the camcorder. However, given our move the HD we’ll be retiring this cam to the electronic pasture.
Videos #2 and #4 – Canon HG10 and Canon HV20
Distinguishing between the HG10 and the HV20 in clips #2 and #4 was very difficult for most people probably because the two are essentially the same camcorder, except for the recording medium and codec. The HG10 records video that is compressed with the AVCHD codec in various bitrates – our camcorder was set to the highest 15 Mbps bitrate. This camcorder also records to a 40 GB hard drive which should make transferring clips easy in the field, but we ran into several quirks.
The HG10 connects to any PC through a USB cable, BUT you need to connect the AC power adapter while transferring clips to the computer. You need the adapter even if you have a fully charged battery. How crappy is that?
But after connecting everything, the camcorder reads like a external hard drive and you just drag the clips over to the computer. This is much quicker than rewinding a tape and twiddling your thumbs for an hour while a MiniDV camcorder captures.
The AVCHD codec is fairly clear, but we were able to see some artifacting in the rendered HD footage. If you click on the Vimeo clips, you’ll see blockiness in the skies and sometimes on flat areas of the ground. AVCHD-encoded clips also require quite a bit more processing power when editing and rendering.
The HV20 records HDV-encoded video to regular MiniDV tapes. It accomplishes this trickery compressing more than four times the information into the same 25 Mbps per second DV bitrate. The cam also uses some fancy (we think aggravating) magic to encode 24p footage onto the DV 29.97 NTSC stream. We captured the clips by connecting the HV20 to our PC via Firewire and then used the built-in capture program in Sony Vegas.
Since the 24p clips are embedded inside of a 29.97 NTSC stream, you have to do some insane steps to rip out the signal. We used TMPGenc to re-encode the clips and eliminate the ‘pulldown’ frames. You can read this thread on the HV20 for the exact steps. This of course adds extra time to creating a finished video and that’s unacceptable in online journalism.
The HV20 does shoot in 60i and I think we’ll just use that mode in the future for time-sensitive events. We don’t need the extra low-light sensitivity that 24p provides because we just bought a Lite Panels Micro which gives us plenty of fill light.
Both the HG10 and HV20 had image stabilization turned on and while it did a decent job, nothing really can take out the shakes while handholding at 10X optical zoom. From the Vimeo clips, it almost seems like the camcorder buffers some shake and then it tries to catch up to the position. This produces a delayed jerking that is evident in the footage.
Video #3 – JVC HD-110
The third video was obviously the JVC shoulder cam. This bad boy records splendid 1280 by 720 pixel video at 24p, 30p or 60i. For our test we used the 30p setting which made for easy editing. We just captured through the camcorder’s Firewire into Sony Vegas, edited and rendered. It also turns out the exposure could have been much better because we forgot the camcorder was on manual iris… DOH!
Of course we aren’t done yet and plan on doing a few more tests thanks to some suggestions by our readers. ‘Eijah’ asked why we didn’t include the Panasonic DVX-100B and there’s a good reason for it. We already had four camcorders on hand (and the JVC is a beast to carry around) and adding the Panasonic would have been a big hassle. But for our next outing we’ll be sure to take the DVX because we still think it is the best standard definition camcorder available. Max Wilson, our cameraman, actually still gets job offers requesting the DVX.
We also plan on using a tripod for the next round to cancel out all the shaking. For the LAX test, we handheld/shoulder mounted all the shots because it was easier and quicker for us. There’s actually a method to our madness in not using a tripod because all of the handicams were purchased with hand-holding in mind. Plus the JVC was purchased mainly because it is a shoulder mount camera and reduces fatigue on the cameraman. You really don’t want a tired cameraman… it’s quite bad!
‘Guy Barwood’ suggested that we do a low light test and we’ve bought a nice package of tea candles from IKEA to help us do that test. Ohh candles, how romantic.
Product pages for the HD camcorders