So where was I.. No blog updates, no new videos before this. Well now i’ve finally finished my degree in IT, and at the same time, i had numerous video jobs. And those videos, i don’t post them up on my video gallery. So hopefully soon, i get to make some new experimental work. So let’s cut to the chase.

I love HDV. Well not exactly HDV, but the workflow that HDV involves. Despite the compression, there are a wide range of softwares that fully support MPEG2 editing without a glitch. And best of all, i love smart rendering. Smart rendering involves encoding without any quality loss (to the same exact format as the source file). So trimming footage becomes a favourable this way.

So a step of the HDV workflow that doesn’t exist in my steps anymore are tapes. No more capturing and changing tapes. Just offloading files from the SDHC cards with a card reader. But, comes an additional step. I convert all the AVCHD footage to MPEG2 format before i start trimming. And i had to find out the ideal parameters for minimal quality loss.

I noticed blocky-ness in the colour brown after converting to MPEG2. Below is an illustration of the formats experimented with:

1080p test – 1:05minute footage
1. Original AVCHD footage (168MB)
2. MPEG2 VBR 25-30-35mbps (282MB)
3. MPEG2 CBR 30mbps (242MB)
4. MPEG2 CBR 35mbps (282MB)
5. Canopus (633MB)

And the screen caps:
1. Original AVCHD

2. MPEG2 VBR 25-30-35mbps

3. MPEG2 CBR 30mbps

4. MPEG2 CBR 35mbps

5. Canopus

It looks like the winner is no 4. MPEG2 CBR 35mbps but the verdict isn’t very clear. Canopus tends to add a shade of grey which changes the contrast, so it’s pretty hard to judge. But judging from the filesize, Canopus should be the clearer winner since it’s less compressed and wins hands down in the editing room for real-time editing, faster drop ins into Premiere Pro and After Effects and faster rendering of effects.

Now i won’t do a comparison of the same frame between MPEG2 and DVCPRO resized to 720p simply because of the end result:

1. DVCPRO looked blur – but a good format to edit with
2. MPEG2 looked alot sharper with more detail – but loses to DVCPRO in terms of the ideal codec to be editing with

Again, it’s either MPEG2 or Canopus depending on your reserved hard drive space for editing. Now let’s take a look at some screen caps taken with the 720p mode. I’ve excluded Canopus since it would be the clear winner but again, that would depend on the editor’s hard drive space. Note: Pictures below are clickable for full size for better comparison. It’s hard to tell since the frames compared aren’t exactly the same.

720p test – 27 seconds footage
1. Original AVCHD footage (66MB)
2. MPEG2 CBR 30mbps (100MB)

1. Original AVCHD footage

2. MPEG2 CBR 30mbps

This time round, both look almost indentical. Not much difference compared to the 1080p test in terms of blockiness, even in the brown zone. But since this was shot in a brighter lit area, i did a 1080p test too.

1080p test – 19 seconds footage
1. Original AVCHD footage (48.4MB)
2. MPEG2 CBR 30mbps (81.4MB)

1. Original AVCHD footage

2. MPEG2 CBR 30mbps

Blocky-ness is more present compared to the 720p test, but that’s pretty hard to say since 720p is smaller so maybe it wasn’t that visible.

How would you decide on what mode to shoot in? I’m torn between 720p and 1080p. i quote Barry Green of DVXuser:

“As for the chips, Panasonic’s whitepaper on the chips shows them delivering about 1440 x 810 worth of resolved detail. 720p mode captures 1280×720. That means there is a tad more image detail available for 1080 mode to capture, but not enough to fully saturate the codec with detail. The captured resolved detail between 1080 and 720 will be a little different, maybe 10% different. However, 1080 mode allocates a lot more color information (they’re both 4:2:0, but that means 960×540 chroma in 1080 mode, vs. 640×360 chroma in 720 mode). So you’ll have more accurate color rendition in 1080 mode. But more compressed. BUT, not as much more compressed as you might think, because the codec, while having to handle twice as many pixels, isn’t being asked to handle twice as much detail.”

So how would you decide which format to work with? I’d choose Canopus and a larger hard drive to work with. But for a current project, i’m working with 720p/25 footage converted to MPEG2 CBR 28mbps since all my hard drives are filled. But i’m torn until someone answers me this:

Is smart rendering available for the Canopus format? If yes, what programs are needed for the trimming of Canopus footage with the smart rendering feature for the Canopus HQ? A program as light as Womble MPEG video wizard, or if smart rendering is available for Premiere Pro or After Effects, that would be a delight.

So there’s a new contender in point8cam. It’s the Panasonic AG-HMC152 (HMC150 for NTSC version). Pretty unattractive box is comes with, it’s as if I just bought Panny Microwave or something. Knowing that there’s a camera in a huge box like that, there’s no reason not to be excited.

With an upgrade of so many features and technology, few things I was psyched with was the viewfinder and the 3.5inch LCD screen. So much bigger than the Canon HV20. Other things would be built in XLR mics (phantom support), the iris/focus ring, gain control, auto/manual zoom, auto + limit over manual settings, 720p50 are the other little things out of a whole lot more.

Small tripod this is, but i didn’t have my Manfrotto 501HDV + 055XPROB then

Another bit of the party piece is the widest zoom of this camera. The HMC152 has a wider shot than the HV20+Raynox 0.66x Wide Angle Lens. Also it has a max aperture of F1.6 compared to the HV20’s F1.8. F1.8 side by side, the HMC150 makes a difference too.

The bittersweet experience with this camera is tapeless recording wrapped with the annoying AVCHD codec (Adobe After Effects CS4 has problems rendering this, even with the latest updates, so i have no choice but to convert the shot footage into a “friendlier” format since i rely heavily on AE). Let’s talk abit about the shooting modes of this camera. At the highest quality, there are a few recording modes:

HMC152 Shooting modes (at highest quality)
1. 1080i/50i
2. 1080p/25p
3. 720p/25p
4. 720p/50p

At a glance, usually we’d think no. 4 would have the biggest bitrate, followed by no.2 then either one of 1 and 3. Instead, because “they” claim AVCHD is smart in compression, they’re all almost the same in terms of the file size of the end output. Here comes the wonder: “Would no.4 and 2 have compression artifacts?”. Well I’m not too sure, but I notice artifacts in mode no.4, other HMC users please leave a comment if otherwise. Artifacts as in, 1 or 2 blocks of pixels look damaged, or rather, just a different colour then it should be. I’m not too sure if it’s the SDHC card consistency, but I only tested this mode on a Panasonic 4GB Class 4 card. I wonder if a faster card would actually fix this problem.

Anyhows, I ran a test of shooting in mode 2 and 3, to check out how well AVCHD compression is, and how well the different modes are handled when converted to another format. That will be shown in the next update, with screen captures, showing areas cropped at 100% before and after conversion. The point is to find how much loss of detail there is due to conversion of these different modes/compression ratios. These tests will include the following formats:

1080p test
1. Original AVCHD footage
2. MPEG-2 CBR 35mbps
3. MPEG-2 CBR 30mbps
4. MPEG-2 VBR 25-30-35mbps
5. Canopus

720p test
1. Original AVCHD footage
1. MPEG-2 CBR 30mbps

Also, the file size for each format will be illustrated accordingly. So camera operators/editors, stay tuned for the next update.
p/s: I have all the screen caps done