Prototype HD Camcorder From Sony

[Jason Dunn]

<div class='os_post_top_link'><a href='http://www.pcworld.com/news/article/0,aid,115266,pg,1,RSS,RSS,00.asp' target='_blank'>http://www.pcworld.com/news/article/0,aid,115266,pg,1,RSS,RSS,00.asp</a><br /><br /></div>"Sony is demonstrating a prototype camcorder based on the HDV high-definition video format at the Cebit exhibition in Hanover, Germany this week. The HDV format is a recording standard that was jointly developed by Canon, Sharp, Sony, and Victor Company of Japan (JVC). HDV uses MPEG-2 compression and is based on DV tape, a commonly used consumer video format. The HDV format can record and play back video at a higher picture quality than DV while using the same cassettes, track speed, and track pitch."<br /><br />I'm all for the coming of HD video cameras, but I'm a little iffy on the whole "let's record in MPEG2" idea. Perhaps once this really happens I'll be less inclined to care if the quality is high enough, but to me MPEG means "compression", and the thought of compressing my source file is distasteful - compression should occur at the end of your project, not at the beginning. Then again, shoot in JPEG rather than RAW mode, so perhaps it's a matter of how <i><b>much</b></i> compression is used. We'll see! What are your thoughts on it? Any MicroMV-based video camera owners out there? Those wee tapes are for MPEG2-based content as well.

[Gary Sheynkman]

It is not ARE they are compressed.... it is HOW they are compressed

Compression is great if you dont loose fidelity

[Jason Dunn]

It is not ARE they are compressed.... it is HOW they are compressed. Compression is great if you dont loose fidelity

Right. MiniDV also has compression when it's moved onto the PC, but it's very light, not MPEG2 compression...

[michael]

I don't know what resolution they are recording in, but if we assume it's 720p, then uncompressed it's about 158MBytes a second (for 24bit colour) - 720 x 1280 x (colour depth, 3 for 24bit colour) x (frames per second, 60)

Even 480p is about 55MBytes a second...

It's just far too much data to deal with on a portable device uncompressed.

[michael]

Am I blind or is this story not on the front page?

[Suhit Gupta]

I see the story there as the top story now. Are you sure you have refreshed your browser? Try exiting and restarting.

Suhit

[Jason Dunn]

It's just far too much data to deal with on a portable device uncompressed.

That's true, I didn't factor in the additional size of the HD signal - the 5 to 1 compression ratio of MiniDV wouldn't be enough. 20 to 1 and higher on MPEG2 gives them the space they need.

Speaking of compression though, and this has always confused me, when a camera is recording a 720 x 480 MiniDV stream to tape, is it compressed there? I didn't think so - I thought that the light 5:1 compression was applied during the Firewire to DV-AVI process. But now I'm not so sure...can anyone confirm? :-)

[Doug Johnson]

Speaking of compression though, and this has always confused me, when a camera is recording a 720 x 480 MiniDV stream to tape, is it compressed there? I didn't think so - I thought that the light 5:1 compression was applied during the Firewire to DV-AVI process. But now I'm not so sure...can anyone confirm? :-)

Yes, it is compressed on the tape. The video data being transferred over FireWire is the same as what's on the tape. No loss in quality during the transfer. The tape is actually the limiting factor that makes compression necessary. FireWire could handle uncompressed standard definition video, but it would be hard to fit ~ 60 GB on a tape the size of a MiniDV tape, so it is compressed 5-to-1.

One thing that is nice about DV compression though is that it uses intra-frame compression. Each frame is completely independent of all others. In addition, it is a constant bit rate (somewhere around 25Mbps), so randomly seeking within the data is very easy to do.

With MPEG, this isn't true. MPEG only stores whole frames a fraction of the time, and other frames are just stored as deltas (changes) over previous frames. Not only that, many MPEG implementations use a variable bit rate, so randomly seeking within a stream is somewhat harder.

The combination of the two makes editing MPEG a lot harder to do and the potential for quality loss much higher than DV. The bit rate with HD though is pretty high, so normally there aren't any visible artifacts with professionally encoded HD sources. (normally at least... there are certainly visible artifacts on some of the HD feeds that I watch here) Makes you wonder about how good the MPEG encoders in consumer-oriented cameras are, though. I guess we'll have to wait and see.

I'm sure that MPEG was selected for consumer cameras because broadcast HD is transmitted using MPEG, and MPEG is more friendly in a bandwidth-limited environment such as tape storage, DVD, or transmission. Probably the best logical choice considering the storage mechnisms we have today. If DV-like compression was used, you'd get about 10 minutes of footage on a MiniDV tape.

[Doug Johnson]

Here is some more info about the camera, including a picture:

http://www.camcorderinfo.com/content/sony-hdv-prototype-camcorder-03_17_04.htm

[oom]

All of Sony's (and I presume others) cameras compress the signal as it goes to tape to some degree.
Even their high end F900 24P camera.
If I remember correctly it is a function of the Tape being able to receive the bandwidth thrown at it.

craig

[bryhawks]

I WANT ONE!!!!!!

I've been wanting this camera ever since I bought my TRV-900 four years ago, saw it had wide screen and progressive-scan, and subsequently discovered that progressive scan was 15 fps (choppy) and widescreen just chopped the top and bottom off the existing image (as opposed to true anamorphic 16x9).

Since in 10 years (or much less) every TV will be HD and bandwidth/storage will be so much greater, I'd LOVE to be taking HD video now - even if it is just of my kids growing up - heck, ESPECIALLY since it's my kids growing up. Like Jason, I have a serious archival streak, and the higher the quality, the better.

As the article mentions could happen, I'd love to see a consumer version of this camera, like the TRV-900 and -950 models.

GO SONY!

[Jason Dunn]

Yes, it is compressed on the tape. The video data being transferred over FireWire is the same as what's on the tape. No loss in quality during the transfer. The tape is actually the limiting factor that makes compression necessary. FireWire could handle uncompressed standard definition video, but it would be hard to fit ~ 60 GB on a tape the size of a MiniDV tape, so it is compressed 5-to-1.

Ah, thanks for the detailed response! I appreciate it - that's something that has always confused me a little. :-)