The future of HVX200

[Barry_Green]

Barry,
Does the single-sensor guess (systematically) the needed colors of the image?

Yes, but it does it intelligently. It's really rather good, but you're right, it is ultimately a guess. An intelligent, systematic guess, but a guess.

If you want absolutely accurate color information for every pixel, that demands three chips -- or a Foveon-style sensor.

I wonder what the cost difference would be between this sensor and CMOS technology?

Not sure what you are asking. A color filter array can be used on a CMOS chip or a CCD chip. The Dalsa Origin is a 4K digital cine camera that uses a CCD with a color filter array (CFA) and the Red is a 4K digi-cine camera that uses a CMOS with a CFA.

Do you think the difference in costs of the sensor is why RED (and now Sony) chose CMOS?

I believe the surge in CMOS is due to two factors: lower cost/higher yield (which is, basically, the same thing) and faster scan rates. CMOS imagers can produce lovely images; the only gripe I have about them is that stupid rolling shutter design. I'd consider spec'ing the dream camera with a CMOS if we could get a global shutter to avoid any rolling shutter issues.

 

[der4]

I'd consider spec'ing the dream camera with a CMOS if we could get a global shutter to avoid any rolling shutter issues.

Why don't CMOS based cameras use global shutters? CMOS technology (rolling shutter) would be a nightmare for some event videography work. Especially with fluorescent lighting, and not to forget the flashes from all the still cameras.

You are right on the money. A single chip with a global shutter would be the perfect compromise. We get interchangeable lenses and it could be used for event videography without the inherit dangers of the rolling shutter. The camera could be sold with a standard zoom lens for the average joe. The company could also sell additional lenses if they want. And we could use Nikons/Cooke's etc.

I think this camera would be a BIG hit. Panasonic are you still listening?

 

[Barry_Green]

Why don't CMOS based cameras use global shutters?

A few reasons, mainly more complex design = more cost.

Interesting little article from Micron about their TrueSNAP global shutter CMOS. They tested by using a can on a rotating surface, to show the difference between a conventional global shutter (CCD), a rolling shutter, and their TrueSNAP. TrueSNAP looks like it's just a super-short shutter speed, but the interesting part to me is the difference between the CCD shutter (which looks like what you'd get from film as well) and the rolling shutter (which looks unlike anything you'd ever get in film).
http://download.micron.com/pdf/articles/truesnaparticle.pdf

CMOS technology (rolling shutter) would be a nightmare for some event videography work. Especially with fluorescent lighting, and not to forget the flashes from all the still cameras.

Agreed. As more rolling-shutter CMOS cameras hit the market you'll hear more and more about this.

The camera could be sold with a standard zoom lens for the average joe. The company could also sell additional lenses if they want. And we could use Nikons/Cooke's etc.

I'd recommend no lens. Make one available, but don't include it. I think that was the mistake JVC made with the HD100 -- putting in a lens that was a piece of doggy doo-doo. If they'd offered it with no lens (HPX500-style) and had people select what lens they want, people would rightly have attributed the shortcomings to the lens and not to the camera itself. So, no, I say no lens in the box, let people pick their own.

Another way to go, much less tasty but an option, would be to use an existing chip and incorporate a Letus Extreme type of GG adapter in the design inherently. That way we could use multiple chips for 4:4:4 color possibility, no Beyer image degradation, etc. and still use film lenses. It would probably be much cheaper than using a large S35-size sensor. Drawback would be the aperture restrictions of a GG adapter, as well as not getting the great sensitivity gains that larger chips would allow.

 

[Barry_Green]

Another excellent discussion of why global is better than rolling:

http://www.photonfocus.com/html/eng/cmos/linlog.php

 

[mikkowilson]

Nice link Barry.


I think this picture says it all:

A sationarry and a moving fan, shot with global shutter.


- Mikko

 

[Tzedekh]

Yeah, I'm not thrilled with the design compromises involved, but frankly I think it's the best choice for anyone wanting to use film lenses. Which is us. We just need to make sure that there's enough oversized resolution so that the de-mosaic'd image has enough resulting resolution. 3 megapixels ought to do about right.

I was thinking that an interesting "compromise" would be a dSLR-sized (about 24 x 16 mm) chip with a dSLR-style lens mount. Assuming that the spatial resolution were comparable to that of a typical dSLR, you would get 6-10 Mp. Now if only the chip manufacturers could get the chips to chug along at up to 60 fps (with all the useful rates between 1 and 60 fps).

Why don't CMOS based cameras use global shutters?

A few reasons, mainly more complex design = more cost.

There's also an equally troubling compromise: at any CMOS chip size, the additional circuits needed for a global shutter reduce the space available for pixels, meaning less dynamic range and light sensitivity. I suspect, though, that the various CMOS imager manufacturers are diligently working to reduce or eliminate the problem.

 

[electricpig]

I'd love to be the fly on the wall of a client edit, where he suddenly asks why his fans look as though they're about to blow off screen, and the producer just starts digging himself a hole with his non-sensicle answers.

hilarious.

 

[TedRR]

That picture gave me more concern about rolling shutter then I thought was called for. Good example.

 

[Luis Caffesse]

I think this picture says it all:

Global-shutter on the left, rolling-shutter on the right.

Actually, from looking at the site I believe both those shots were taken with a CMOS camera - in the first picture the fan is off, in the next picture the fan is on.

It clearly shows the problems with a rolling shutter - so the point is still valid.
But I don't think its a 'side by side' example of global shutter vs. rolling shutter.
It was an example of how a rolling shutter deals with stationary objects versus fast moving objects.

From the site:
Figure 2a Image of the fan at 0 rpm with a traditional logarithmic response sensor.
Figure 2b Image of the fan at ~ 1600 rpm with logarithmic response. The fan is strongly distorted due to the implicit rolling shutter mechanism.

 

[mikkowilson]

Thanks for the corretion Luis .. makes sence now that I read it again.. Duh.

- Mikko

 

[alpi69]

Interesting thread. I can see you guys are already "working on a new camera" while the EX1 is just emerging as the HVX only contender.
I for my part made a decision this week. It had to be made for taxreasons. In Austria I can have 10% of my profits taxfree when either used to buy gear that stays in the company for 4 years or buy certain papers / bonds that stay for 4 years (and make 5-7% profits every year).
I bought the HVX200 in early 2007; so that means I "have to keep it for 3 more years roughly.
My thinking was this: will Panasonic bring a new cam to rival the HVX and what will it do? Will the EX1 help me make better profits merely by its 1/2`` chip and manual lens? And mostly: will the HVX and its P2 workflow be obsolete within 3 years or can I "milk it" so long?
I maybe should add that I rent cameras (mostly DigiBetas and HDCam) on a regular basis anyway for certain jobs.

The future of the HVX in my eyes: If Panasonic brings another HVX with a better LCD the upgrade will probably cost me what a monitor like the marshall costs (around 1500). "Upgrading" to the EX1 will open a new can of worms, because I need to pay at least 3500up, then add new solid state, get a even stronger PC to edit long-GOP, lose 4:2:2, get new batteries and accessoires that I bought for the HVX (raincoat, .6 adapter, 1.6 adapter) etc. It will cost me another HVX in the end. And even if these cams pay themselves within a couple of months for me, it is silly money spend when you think as a businessman instead of the tech-nerd that we all are.
Finally the HVX does all I need in Europe already: 720p50 will be the broadcast standard according to the EBU. The HVX is an awesome 720p camera as we all know. Tests by the EBU have shown that viewers prefer compressed 720p to uncompressed 1080i. So for all network-work the HVX will do for at least 3 years, because most networks are not even shooting more than SD. When a client wants "real full-HD" I will anyway have to rent a RED or HDCAM 2/3´´ camera as even the EX1 won´t be sufficient.
To avoid that as often as possible I will get myself a 35mm adapter with the money I save on that "tax-trick". Something I would have to do with the EX1 also.
And finally: the HVX makes me happy now. It will be state of the art for at least 18 more months and in the worst case I can get the "new HVX" then and keep the current one as B-cam.
At a recent event I was there were many stations with brandbew HVX. I could see the operators have still been discussing workflows and such. P2 seems to finally gain momentum in Europe (which is funny as here on dvxuser it sounds like it already lost to XDCam) and the P2 prices keep getting lower (relative per GB).

Finally a clip shot mainly with HVXs. It was/will be broadcast worldwide (from Brasil to UK to to Germany to MTV Russia to Foxsports USA). It is produced in 50i, because there was also a DSR450 involved, but it shows nice images from our HVX s clearly ;-)

http://www.youtube.com/watch?v=9vdgsSCY7co

 

[SPZ]

http://www.youtube.com/watch?v=9vdgsSCY7co[/url]

Nice video!

 

[Stevet]

alpi69,
You're in a great position to stay with the HVX200. The HVX200 is a nice camera and in a few years when it comes time to upgrade, you're going to have a lot of nice stuff to choice from!

 

[Sumfun]

Will there be a prototype of the Green One camera showing at NAB?

 

[Barry_Green]

Er... no.

 

[soarprod]

Barry - I'm building the balsa wood mock up aka vaporware

 

[Barry_Green]

Well, paint it green! :thumbsup:

 

[TimurCivan]

im gonna make a green camera mock up too... mine is gonna have chopper handles....

 

[soarprod]

and twice dice

 

[SPZ]

I'm posting this following the Red Scarlett announcement from Jim Jannard. What will this mean for the future of the HVX?

One thing is for sure: stated as a professional pocket camera, this camera will obviously compete with the HVX and other handheld prosumer cameras, independent of the size. Its performance, even if its pocketable, should be superior to any of the sub 10k cameras. So what does this mean for the future of the HVX?

In my view, there is still room for Prosumer 10k cameras besides the Scarlett. From the description, it is a pocketable camera, and this, even with fantastic 4k or 2k resolution, is not good ergonomically speaking for much professional Broadcast work. Pocket means helmet cam, or POV cam to be able to be called "Professional". Either that or they have a heck of a OIS that can prevent handheld shake from managing such small gear.

The bar for the next gen HVX has just been raised.

However, the P2 workflow, Panasonic's cinegamma and fantastic 25p/24p motion rendering and excelent track record can still prove to be a very viable choice.

With pristine 1080p with variable frame rates, avc-intra, HVX ergonomics and better LCD, as well as bigger chips (2/3 now seems more viable versus a possible misterium powered pocket cam) and a good fixed lens, this would still be my camera of choice vs the RED.

Notice, once again, that I'm talking about FIXED LENS- this will not hurt into the rest of Panasonic's product line of cameras. A fixed camera would never look as good as an interchangeable camera with top glass.

Panasonic, don't disappoint us loyal costumers!