The touch screen looks to be the same as on the XR500/520 which is really nice. Hope it also has the other touch screen features like "spot focus" etc that are on all the other Sony consumer cams. The "spot focus " feature is something I use all the time on the XR500 and SR11. It is a pity that they didn't put the "R" sensors on it though as I saw a real difference between the SR11 and the XR500. The Active stabilizer is also something that is on the XR500 and also works well compared to the stabilizer on the SR11. Both are much better than the FX1 stabilizer. It is really close to what I was looking for as a 3 chip version of the XR500 plus. My simple dream had been the XR500 bits in a Z5 body!!! Real close.
They have the opportunity to leave off the SD/HDi interface as well as the flash memory interface to make a FX version as well as upgrade to the EX codec for an upscaled pro version. Then an upgrade them all to the "R" sensor!!!!!!

Ron Evans

Where are you getting the fast motion information? Have you seen Barry Green's latest AVC vs. XDCAM article? It would suggest the exact opposite of what you state here.

The idea that AVCHD can't handle fast motion is indeed a myth.

here is a link to the very thoughtful barry green article about avchd:

XDCAM-EX vs. AVCCAM

the nxcam release date appears to be early 2010.

any guesses on pricing?

over or under $4500?

i'm really looking forward to more specific tech specs.

chris-do u now open a new nxcam section/thread or attach the conversations under ex1/3 excam?

be well

rob

But the new camera reportedly offers true 10-bit 4:2:2 output to an external recorder, which I don't think the Z5U has. That's one of the more interesting features for high-end applications: let's say the NXCAM is ~$4500 and you add the top KiPro unit for $1400 - now you have an excellent recording setup for about the price of an EX1 (just with a smaller sensor).

Well it's not XDCAM, and it will need its own forum (eventually its own category as other models arrive), but perhaps I should re-name our Sony XDCAM category to something like "professional tapeless" and put it in there.

Edit: on second thought, I'm reluctant to use the term "tapeless." It's fast becoming an antiquated word, like "wireless." I'll probably just re-name the category to Sony XDCAM / NXCAM Camera Systems, like we did for Panasonic P2HD / AVCCAM.

Haha, doesn't surprise me at all that sony went with Pro Duo cards.

...And I don't know why I keep expecting Sony to freak people out with 1/2 inch chips in something cheaper than the EX1...doesn't look like that will happen any time soon, but I guess we shall see with the next models coming into the NX line.

Once again: for long-lens cameras, bigger chips = bigger glass, = more expensive.

Therefore "1/2 inch chips in something cheaper than the EX1" is an unrealistic expectation.

From the released information, there are two features that I like. One is the 1920 x 1080 recording. Even when rendering a DVD, this gives more pan and zoom range when editing compared to 1440 x 1080 HDV without the image going soft.

The other is the ability to record 4:2:2 via the HD SDI port to an external recorder. 4:2:2 allows better color editing depth. It is also more green screen friendly.

Other that that, it looks like a Z5 reimagined.

First look at the new NXCAM via Sony virtual trade show
 

Sony USA has asked me to pass along to y'all that senior product manager Juan Martinez will be presenting an eight-minute overview of the new NXCAM tomorrow (Thursday 19 Nov.) during their online virtual trade show event. There's still time to register, attendance is free, and you don't have to leave your house or office. I did this last year and thought it was great; definitely well worth the time.

Sony's "The New Economics of HD" Online Expo

Featuring: Virtual Convention Center with Staffed Product Booths, Technology Demos, Networking Lounge and Group Chats, Product Tutorials, Resources - Downloadable Articles, White Papers, Presentations and More

Thursday, November 19, 2009
10:00 AM EST - 7:00 PM EST

Link: Sony's "The New Economics of HD" Online Expo | Home

The real missed trick is not just missing out 35Mbs XDCAM, but not putting BOTH that and the AVC-HD codecs into the same camera. For some, AVC-HD may be just what's needed, for others it would be good to have the same codec as the EX.

And many people do seem to be overlooking that image quality is normally far more down to the front end than the codec. Sony seem to have adopted Panasonics marketing in the sense they are making a big deal of "1920x1080 RECORDING". The implication is 1920x1080 resolution, and the only way to get that on a 3 chip camera is with 2 megapixel chips.

If the chips are 1 megapixel, that is what limits resolution. You could upscale it to 4k, and it won't be anny sharper.

But that's a marketing criticism. By and large I think this camera will sell well, and I also think that 1 megapixel is a sensible resolution for 1/3" chips. Go higher and lose sensitivity, go lower and it's a bit too soft. And if you really do want 2 megapixel res - get an EX.

The obvious comparison is going to be with Panasonics HMC150, and from what we're hearing this NXCAM seems to soundly trump it in almost every way - the big question will obviously be price.

Are you saying the sensor block is a native 960 x 1080? I don't like it when they market that as 1920 x 1080 recording either, but it is probably a concession to preserve low light gathering.

It's more complicated than that, hence my deliberate use of "1 megapixel", not dimensions.

It can be thought of as 960x1080 - but just as accurately as 1920x540! Either way, it's 1 megapixel.

Confused? It took me a long time to work it out, but the best way to visualise it is as a tiled floor, half white tiles, half black tiles, with the corner to corners lying horizontally and vertically. And a white-black-white-black pattern.

If you wanted to count tiles, the easiest way is to count the number of black tiles, then the number of white and add together. (In fact, there should be the same number of each colour, so count one and double it.)

So, let's assume that along the length of this bathroom there are 960 black tiles in a row, corner to corner, and 540 black tiles corner to corner along the width. Hence 0.5 million black tiles. So then 0.5 million white tiles also, 1 million tiles altogether.

But how would you describe them in an a x b manner? Especially when the pixels on the chip are not black and white, but all the same?

The big advantage of this arrangement (for cameras, not bathrooms!) is that it is one megapixel, but symmetrical horizontally/vertically, and gives the same resolution along both axes, which any arrangement such as a conventional 960x540 wouldn't. It is also very easy to process in a 1920x1080 processor - but the actual output resolution won't be 1920x1080, more like 1440x710.

I'm not being critical of this camera or it's sensor - I think it's absolutely the right approach for a 1/3" chipped camera, a sensible compromise between sensitivity and resolution. But don't think that by recording the HD-SDI output to an external device you'll get a 1920x1080 picture to rival the EX.

Just because a camera records a 1920x1080 raster doesn't mean the front end can produce that resolution, and that's even more true of the HMC151 with only 0.5 megapixel chips.

"Are you saying the sensor block is a native 960 x 1080? I don't like it when they market that as 1920 x 1080 recording either, but it is probably a concession to preserve low light gathering."

If you mean actual sensors sites on the die then its more like 960x1080 diamonds set on a diagonal pattern( 1,036,800). These sensor sites are twice the size of the effective pixels. The DSP uses all the sensor information to create the 1920x1080 effective pixels from each of the 3 chips. I can't find the info now but if you do a search its around somewhere. DSP's are used to create the effective pixels however many "real" sensor sites there are on the chip(s).

Ron Evans

Heath's explanation is good. I also like to think of the arrangement as 1920 columns and 1080 rows in the same way by reversing the process a little. Take 1920x1080 square pixels array now overlay , on a diagonal, the tiles, twice the size of the square pixels the 1920x1080 would give. So each of the big tiles will have a small square pixel in the center. For every group of 4 adjacent titles, there will be the equivalent of one new square pixel between them. Made from the surplus area left over from the small square pixel in the center of each of these titles. Effectively the DSP uses this information to create these extra pixels and level off the outputs from ALL pixels. None of the rows or columns has a full sensor in all of the 1920x1080 positions. They are all created.

Ron Evans