IBC 2006 Pre-show discussion

mikkowilson

mikkowilson

Steadi-Guru
We are just 47 hours out from IBC 2006...
As we speak, thousands of people are in Amsterdam or on the way in preperation for a huge show this weekend. Many more are preparing to travel or standing by online for the flood of new info


In 2 days time we will know what RED's first footage looks like, and the web will be flooded with news releases of new products. Exciting times once more.


Some of us are lucky enough to attend IBC in person.
I will be attending from Saturday onwards. I'll be posting news from the show from Sat PM onwards.



Who else is going and when? Who can we count on for news from the show?

Are there any locals in Amsterdam who could recomend a good place for a possible DVXuser party?

What are people looking forwards to? What news releases and new products are poeple looking forwards to?


It's almost showtime!

- Mikko
 
I'm going to IBC at Friday only, it's the first time I'll be going to a big video event.
And there is only one reason I'm going... ... ... ... ... RED :)
I can't wait!
 
tnle2 said:
that records in true progressive CMOS 1920 X 1080 25P!
Click to expand...
It does not "record" 1920 x 1080 x 25P. It uses 960x1080 chips, up-rezzes them internally to 1920x1080 (note, not using a spatial offset technique, but using digital uprezzing because CMOS is a fundamentally different technology that includes discrete pixel information), so it then handles 1920x1080 internally, then scales that down to 1440x1080 for recording.

What you get recorded is going to be 1440x1080 4:2:0 HDV. What's new is that Sony is for the first time offering a progressive-scan camcorder at a filmlike speed. They've offered 15p before on the VX2000, but never 25p. It will be interesting to see what appears on the American side, whether they'll make it 30P or if they'll work some magic to include a genuine 24p.
 
And..........beng!! low power 3xCMOS, low noise, 20x lens, HDMI out and real 25P. Sony again. Europe.
 
Barry_Green said:
It does not "record" 1920 x 1080 x 25P. It uses 960x1080 chips, up-rezzes them internally to 1920x1080 (note, not using a spatial offset technique, but using digital uprezzing because CMOS is a fundamentally different technology that includes discrete pixel information), so it then handles 1920x1080 internally, then scales that down to 1440x1080 for recording.
Click to expand...

Whoops, sorry, forgot the HDV specs are limited to 1440x1080. Also, I was incorrect about the sensor resolution. AFAIK there is no explicit mention of the actual sensor resolution, only that the signals are "internally processed at 1920x1080p 4:2:2".
 
Last edited:
You know that Fujifilm is capable of making an ultracompact digital camera with their SuperCCD technology that shoots noise free images at 800 ISO? Its a different placement orientation of the pixels, the ClearVid CMOS is doing for the CMOS technology 'the same' thing and uses lower power consumption. Result is that you can shoot noise free in a medium- low light situation. Just what you need most of the time.

:thumbsup:
 
CLEARVID_1_S.JPG
 
Dont get it. Increased cell size makes sense, but what would a rotated grid do?

(Wouldn't an increased cell size also give shallower DOF?)
 
wow, what a dreadful example

in the left example, there are 24 Green, 16 Blue, 9 Red .. sum = 49 samples

in the "clearview", there are ~17.5 Green, 3 Blue, 3.75 Red!! sum = 24.25 samples

... no wonder the (sub)pixels are bigger.

they should've added a third example-- an SD cam .. "huge pixels, outstanding low-light capability, the way of the future!"
 
Something else-

Why would you have this kind of green-biased pixel-gridding on a 3-CMOS cam .. obviously you wouldnt, so this doesn't seem like it's relevant to the FX7/V1 cams.

I don't see how twisting the grid really fundamentally changes anything to do with pixel size. It changes the balance between horizontal and vertical resolution, and it also creates a lot of complicated problems for the DSP to solve (rastering these weirdly-arranged sub-pixels into an artefact-free 4:2:0/4:1:1 image)... Who knows, we can only presume (at this stage) that sony has done a decent job of the processing, but at-best I can only see potential improvements in anti-aliasing. My guess is that it would introduce more problems than it would solve.

Does anyone remember Sony's "emerald" pixels in their digicam CMOS sensors? Does anyone know if that's still around, or whether it died as a fad? Basically, IIRC, they were touting that RGB-E (red/green/blue/emerald) pixel grids gave truer representation of natural greens. Sounded to me like a complete fad- surely the need for an extra (slightly-different) 'green' sensor pixel is just highlighting the fact that the current r/g/b pixels aren't very good at isolating their specific wavelengths, and that you're using a complicated and expensive 'hack' to try to separate them better...
 
Oh, right...That makes sense. But then why do people say 1/4, or 1/3 CCDs provide more DOF than 2/3 or larger CCDs?
 
Does anyone know if you're allowed to bring your photo- or videocamera to IBC? I'm visiting tomorrow and sunday and would be more than happy to make a short impression movie of the RED exhibition.

But I won't be so happy anymore when a big security bouncer takes my cam away.
 
Fugitive said:
Oh, right...That makes sense. But then why do people say 1/4, or 1/3 CCDs provide more DOF than 2/3 or larger CCDs?
Click to expand...
Because of the focal length of the lens used on smaller-chip cameras.

The smaller the chip, the smaller the field of view. So you need a correspondingly wider-angle lens in order to get a usable field of view. It is the wider-angle lens that causes the deeper DOF.

Look at the focal range on a 35mm movie camera, typically 25mm-250mm. Then look at a 16mm camera, usually around 12mm-120mm. Now look at a 1/3" camera like the DVX, 4.5-45mm. Then look at a 1/6" camera, something like 2.8mm-28mm. All have a 10:1 zoom range, all deliver roughly equivalent field of view. So which is going to give you shallower DOF -- the 250mm on the 35mm movie camera, or the 28mm on the 1/6" CCD camera? Obviously it's the 250mm that's doing it.
 
Right right. Thanks. That makes alot of sense.

So, in a round about manner, I would be right in saying that a larger cell size would cause shalloower DOF because a larger cellsize would mean a larger chip, that would require a larger FOV lense, and hence provide a shallower DOF?
 
Well, yes. The generalization is "smaller chip = deeper DOF, larger chip = shallower DOF". But that's not what causes it. It's like saying "coats kill plants, you can't put a coat near a plant or the plant will die". Obviously that makes no sense, so you have to ask further, and then the person explains "well, I put my plants in the closet with my coats for a week, and when I came back the plants were all dead."

Ain't the coats that did it, it's the dark.

And thus it is with chip size and DOF: it's not the chip size that does it, it's the lens. To get a usable field of view you need wider-angle lenses.

Could you get identical depth of field on a 1/6" CCD and a super35 film frame? Of course -- just use the same lens at the same distance. But you won't get the same field of view, and therein lies the rub. To get equivalent field of view, we have to use wider lenses on the smaller sensors, which gives deeper DOF.
 
You must log in or register to reply here.
Back
Top