Press Release: HMC150, "the new DVX"

BobDiaz said:
There were a lot of comments and I won't be able to answer everyone in detail, but let me provide a general answer...

Scarlet is a big unknown, is it under $10,000 or under $5,000 or ... ???? Plus, how much for a fully working system??? Odds are that Scarlet is to a whole different market that the HMC150 is going after.


The HMC150 is likely to be under $4,000 and its target market is the event videographers, student film makers, VERY small operations (with very little money), and prosumers. This slice of the market is within the $3,000 to $4,000 range. OK, AVCHD is not rocket science, so all of the major camera companies should be able to put something together in the $3k to $4K range.

The problem I see is that the HVX-200 was introduced 3 years ago, but what has JVC or Canon come up with that offers solid state recording for the professional market in the price range of the HVX-200? At least Sony has came up with the EX-1 and the overpriced Z7, but Canon and JVC seem to be doing nothing. (I hope that come NAB they prove me wrong, we need more cameras with solid state recording.)


Flash memory is far more flexible than tape and MPEG-4 AVC H.264 at 24 Mbps is better able to deal with complex motion than MPEG-2 (HDV) at 25 Mbps. It doesn't take a PhD in computer science to realize that an AVCHD camera is the next logical step and HDV is yesterday's technology.

Now if we look at AVCHD's 4:2:0 and 24 Mbps maximum data rate, it looks lame compared to the higher priced cameras, but compared to HDV it is going to be a clear cut improvement. When I look into this market, Panasonic is positioned to answer, but as for the other camera companies.... (That's why I call them 2 steps behind Panasonic.)


Bob Diaz
Click to expand...

Sony may very well be over pursuing MPEG2 for HD, but it does still offer some small advantages over AVC. AVC can be handled by today's desktop computers at HD resolutions, but it is still somewhat taxing on CPUs to support it well in NLEs. MPEG2, while not particularly NLE friendly, is slightly more NLE friendly than AVC, by nature (and still far better supported for at least the moment). The small gaps there will continue to close though.

Canon always lays in the weeds, quite happy to let other companies boldly introduce a new technology to the marketplace. That strategy has served them well, and probably will again here. What would have been surprising would have been if they had gone first with AVCHD in the pro market. What will not be surprising is a solid product when Canon does unveil their own offering.

JVC is the maverick in the bunch. I think the most likely next tapeless offering in the pro market from them will wind up being a hard drive built right into a ProHD camera. JVC's first tapeless offering (of a sort) was an external HDD (just like Sony's first alternative to tape - aside from blue laser disks in the high end cameras). Not really solid state, but not tape either (and in addition to tape, while not abandoning tape). JVC has chosen to pursue 720p exclusively (in the pro market), which is unlike the strategies of any of the others. It's called HDV, but it's not really the same or even very compatible with Sony and Canon's offerings.

Panasonic is a maverick too (perhaps not as much as JVC though). They jumped straight to solid state (at least in the sub $10k range for an HD camera offering in the pro market), skipping right on over tape and HDD recording all together, and way ahead of anyone else. That didn't exactly go off without a hitch though, and didn't really inspire the others to jump right in too, especially with the market successes with HDV.

Where I see Panasonic potentially perhaps having a significant advantage with AVCHD, isn't from being first to market with AVCHD in a pro camcorder offering, so much as that they've embraced AVC Intra as their next major high end codec. That will allow them to leverage AVC R&D efforts from both the pro low and high ends now, and could pay off very nicely.
 
Last edited:
Robert M Wright said:
. . . skipping right on over tape and HDD recording all together, and way ahead of anyone else. That didn't exactly go off without a hitch though, and didn't really inspire the others to jump right in too, especially with the market successes with HDV. . .
Click to expand...

For us the P2 is refreshing and HDD was a disaster. We tried the FS100 Firestore and learned a lot of issues and fortunately it was a rental and we weren't stuck with it. First there is another battery to deal with charging, monitoring, and inventory'ing. Second, another cord to asset and have dangling from the camera. Then there is the whole monitoring issue where one has to keep looking at the thing and figure out where it's at since it doesn't report back to the camera what's going on. Then, it freezes up in the cold cold; and adds more weight to the camera making the camera even less portable/mobile. And finally, if it has moving parts, rest assured that the repair cycle will be faster AND if that occurs after a shoot, you're in big trouble. We had issues with this latter and we know two people who also had issues AND like us they moved away from HDD.

3 P2 cards = continuous record (definition - start recording and don't stop.)
 
My Sony HVR-A1U is good enough for what I need it to do, but I hate tapes and want to go tapeless. I don't trust HDD. The HMC-150 sounds just like what I need. Although Scarlet is a real tease but I'm sure 2x$. So I'm off to NAB for the first time to see what's happening.

I'm just amazed the prices Sony is asking. My HVR-A1U was less expensive two years ago new from B&H and my rebate was bigger. The Z1 is also right up there close to newer and better Sony models. If HMC-150 comes in around $3,500 or a little more, Panny will have another DVX style home run. I'll be one of the first in line.
 
drdimento said:
it freezes up in the cold cold; and adds more weight to the camera making the camera even less portable/mobile. And finally, if it has moving parts, rest assured that the repair cycle will be faster AND if that occurs after a shoot, you're in big trouble.
Click to expand...

Have had the FS100 for almost a year and have had no issues or hiccups at all. I even used it outdoors for a shoot when it was -50 C and didn't even get a dropped frame. The weight isn't too bad if you mount it on the camera and the cord is a non issue if you keep it tied up.

I do agree that having another battery to charge is a pita, though. Perhaps a solar powered FS200 will become a reality.
 
Going from uncompressed video --> MPEG-4 takes a lot of processing power.

Going from uncompressed video --> MPEG-2 takes some processing power, but not as much as MPEG-4 requires.

Going from compressed video --> uncompressed video does not take as much processing power as going from uncompressed to compressed.

So as far as playback of AVCHD, the current MACs & PCs should be able to do it. Those who chose to edit in native format will need lots of processing power or be willing to put up with all the delays. Those who transcode will add a generation to their video, but they will not need as much processing power as native mode editing.

Moore's Law is still in progress, so every 1.5 to 2 years, memories and processing power doubles. In the upcoming years, processing power for editing AVCHD will become less and less of an issue.

As far as improvements in MPEG-2, some small improvements are still possible, but this technology is almost maxed out for improvements. MPEG-4 is only at the beginning stages for improvements. Years from now we are likely to see better performance from MPEG-4.


IF you ever get the chance to listen to David Newman talking about wavelet, take the time to hear what he has to say. Unlike MPEG which works on DCT in little boxes of pixels, wavelet works across the whole image. Thus, over compression in MPEG results in a box like appearance, but over compression in wavelet results in a softer image. Wavelet holds up to multi-generation layers very well.

Bob Diaz
 
CABAC is computation intensive to decode (as well as encode) so it does significantly boost the CPU power required to edit AVC natively. Indeed, as CPU power continues to increase at exponential rates, that becomes less and less troublesome. I'm looking forward to octicores (and software being designed more and more with taking advantage of multiple processors in mind). I'm not sure, but I think AVCHD also uses longer GOPs than HDV.

While MPEG2 does still have some advantages, I agree it is well on the way to obsolescence (and there isn't much room left to improve implementations).

I'm not sure why the camera makers aren't moving towards wavelet compression in-camera. Cineform has shown the viability (and downright overall superiority) of wavelet compression (as opposed to DCT), at the very least, for intraframe only compression. It is much less taxing on CPUs (and should be very well suited for implementing in a camera - less power consumption and heat generated).
 
Barry_Green said:
Red uses wavelet-based compression in-camera.
Click to expand...

The performance is very impressive too. If I recall correctly, they get intraframe only compression down to 28MB/sec on full raster 4K images (10 bit RAW). That would translate to roughly 60mbps for full raster 1080p24 (10 bit RAW).
 
I think Infinity uses JPEG2000; isn't that also wavelet-based?

AVC is much more resilient towards blocky artifacts too. Basically any of the new generation of compressors (AVC, wavelet, JPEG2000, whatever) are all going to avoid the blocks that can happen in DCT compressors like MPEG-2 or DV/DVCPRO.
 
JPEG2000 is a wavelet compression codec (I'm sure Redcode and Cineform are better implementations though). AVC is DCT based, but at bitrates that start getting blocky with MPEG2, AVC is still real clean.
 
CABAC is computation intensive to decode (as well as encode) so it does significantly boost the CPU power required to edit AVC natively.
Click to expand...


First, for those who are interested:
http://en.wikipedia.org/wiki/H.264/MPEG-4_AVC

High Profile (HiP): The primary profile for broadcast and disc storage applications, particularly for high-definition television applications (this is the profile adopted into HD DVD and Blu-ray Disc, for example).
Click to expand...


... and yes, HiP does use CABAC and the less intense CAVLC.

Context-adaptive binary arithmetic coding (CABAC) is a form of entropy coding used in H.264/MPEG-4 AVC video encoding. As such it is an inherently lossless compression technique. It is notable for providing considerably better compression than most other encoding algorithms used in video encoding and is considered one of the primary advantages of the H.264/AVC encoding scheme. CABAC is only supported in Main and higher profiles and requires a considerable amount of processing to decode compared to other similar algorithms. As a result, Context-adaptive variable-length coding (CAVLC), a lower efficiency entropy encoding scheme, is sometimes used instead to increase performance on slower playback devices.
Click to expand...
http://en.wikipedia.org/wiki/Context-adaptive_binary_arithmetic_coding


Come NAB, I hope it will be possible to get a copy of a sample recording from a working HMC150 or if we can't do that, a copy from a working HMC70. That way everyone can see if the playback causes any problems with your system.

So everyone stopping by the Panasonic booth at NAB, please bring an SD memory card and ask Jan very nicely to allow you to record a sample file....


Bob Diaz
 
AVC is DCT based, but at bitrates that start getting blocky with MPEG2, AVC is still real clean.
Click to expand...
AVC doesn't use DCT though. It's similar, but it's not the same (just like DV is similar to MJPG but quite advanced). The block sizes can be MUCH smaller (4x4 instead of 8x8, so 1/4 the size) and it's got an in-loop deblocking filter. Basically, you shouldn't be seeing any blocks in AVC footage unless you're WAY past the bandwidth's ability to resolve.
 
BobDiaz said:
Come NAB, I hope it will be possible to get a copy of a sample recording from a working HMC150 or if we can't do that, a copy from a working HMC70. That way everyone can see if the playback causes any problems with your system.
Click to expand...
You don't have to wait; you could get samples of HSC1U or consumer camera AVC-HD footage right now. Those cameras are already on the market and in the hands of owners.
 
Barry_Green said:
AVC doesn't use DCT though. It's similar, but it's not the same (just like DV is similar to MJPG but quite advanced). The block sizes can be MUCH smaller (4x4 instead of 8x8, so 1/4 the size) and it's got an in-loop deblocking filter. Basically, you shouldn't be seeing any blocks in AVC footage unless you're WAY past the bandwidth's ability to resolve.
Click to expand...

Technically AVC doesn't use DCT, but the basis for the transform is essentially the same. Wavelet transform is a whole different approach (which will never get blocky), and that's where I make a distinction. I don't know of any real disadvantage to using the wavelet approach, but there are considerable advantages.
 
ProjX v2.0 said:
Have had the FS100 for almost a year and have had no issues or hiccups at all. I even used it outdoors for a shoot when it was -50 C and didn't even get a dropped frame. . .
Click to expand...

It's ok, we felt and had experienced just like you until that day when it froze up. Then as Konrad pointed out "I don't trust HDD." When it happens and your goose is cooked, you will more than likely change your mind.

Our burp occurred when we taped a national music star and were under the gun to produce the music video one shoot from a concert and two that were outdoors. The outdoor shoot froze up on us and we didn't know it because the freeze up was intermittent. We'd shot dozens of outdoor stuff without a hitch but then it happened and we couldn't do a reshoot would have cost gads. Then when my friends FS100 HDD failed and he had two tv shows on it he went from a three camera shoot to two and had to put up a screen that said "Technical Difficulties" with a background picture when the had to switch to that camera because the other two were making movies. It's not a pretty picture. When this happens, and trust me it will, the FS100 will be on your hate list as with ours. And, the word "sorry" has never been so shallow a word as it was during those ackward and unpleasurable moments.

Personally, we don't like P2 for that reason but the RAID helps us feel marginally safe for the moment AND we are able to instantly "check" the thumbnails to see that the scenes we shout are actually on the external HDD that we then back up to another HDD while doing the next scene shoots (just because of our FS100 experiences and those of our friends)

Always remember, if it has moving parts, failures occur far more often than for non moving parts. It's a price to pay most certainly.
 
BobDiaz said:
. . . So everyone stopping by the Panasonic booth at NAB, please bring an SD memory card and ask Jan very nicely to allow you to record a sample file....Bob Diaz
Click to expand...

Excellent Idea Bob! Please for those going to NAB (which might be much fewer from recent years as we hear on the streets) DO GET SOME SAMPLE STUFF and post it.
 
Waved all through this discussion of AVCHD seems to be an underlying comparison to both P2 .mxf and HDV. I guess something I'd really appreciate hearing from those who really know (which I admittedly don't) is just how good HDV is? Is it a solid format? Is it comparative to P2 .mxf? And if so, could MPEG2 be as solid as it sounds? We are hearing and seeing a lot of good stuff shot on HDV.
 
A Dream Wish...

One thing that could get me to sell a camera and jump on owning an HMC150 (assuming the price is right), would be built in 802.11g/n WiFi. That would be really sweet. I don't know why that hasn't been done with HDV cameras yet (could work real nice with either HDV or AVCHD). Imagine being able to record to a laptop or desktop computer 50 feet away (without being tethered).
 
You must log in or register to reply here.
Back
Top