Bolex D16: RAW, 16/Super 16 sensor size, costs less than MkIII

It's over $240,000 dollars. I just heard about it and I missed the boat. I was just hours away from a 2k raw camera for $2,500. I grade a lot so this camera would be the camera where I will say, eh don't worry I'll fix it in post. Anyway, despite it's limitations, $2,500 is a bargain. Barry, let me borrow it when you're done playing with it. Thanks. :)
 
17 pages... wow guys, you're outdoing yourself here! :grin:

I don't think having a "640 x 480 B&W" video output in your specs gives some of us comfort, to say the least. Plus, no additional details on the HD-SDI module. How much will they charge for that? "Ok guys, camera is $3300... HD-SDI output will cost you $4000..." Just kidding...

I think Cineform will be happy to hear about the output, even though it's not theirs. Many of us will need to dump the huge files into Cineform before we run out of space.

And the "makers" do exist:

http://www.imdb.com/name/nm1421636/ - doesn't look like his getting jobs that get good ratings...
http://www.imdb.com/name/nm1741056/ - and her height is Height: 5' 4½" (1.64 m) :Drogar-Love(DBG):...
 
many people are skeptic about Dbolex ...
its nothing new about technology...
there are already machine vision camera with kodak sensor..
http://www.illunis.com/products/vmv/vmv.html
http://www.alliedvisiontec.com/emea/products/cameras/gigabit-ethernet/prosilica-gx/gx2300.html
http://baumer.amirada.net/pfinder_vision/scri pts/product.php?pid=SXC40c&cat=Industriekameras&psg=Sensor|1426848949&language=fr&r=1&ptk=&sq=
only difference is that dbolex use some kind of fpga to record data from sensor to dng and ssd buffer...
and how about other companies like KINERAW and open source elphel ...
tech is already here ...
 
Barry_Green said:
Well, to be fair, this digital bolex will also be nothing like 12 bit raw 4:4:4 either. It's going to be 4:2:0 at best. That's the nature of a single sensor with a Bayer pattern.
Click to expand...

Could you explain this? The specs say 12-bit, 4:4:4 color. How do you figure 4:2:0?
 
Barry_Green said:
Well, to be fair, this digital bolex will also be nothing like 12 bit raw 4:4:4 either. It's going to be 4:2:0 at best. That's the nature of a single sensor with a Bayer pattern.
Click to expand...
So an Alexa is at best a 4:2:0 camera? It has a Bayer sensor. C300 has a Bayer sensor, FS100 and F3 have Bayer sensors, not to mention virtually every digital still camera out there capturing 4:4:4 JPEGs. The color resolution of a Bayer sensor is as close to what the human eye can see as any sensor technology out there. It is the processing behind it that determines format color depth and encoding for RGB images not the sensor type.
 
Last edited:
i m afraid that this guys are far off from producing an actual prototype its vaporware for now...
 
Last edited:
vladszp said:
i m afraid that this guys are far off from producing an actual prototype its vaporware for now...
Click to expand...

They have a fully functional prototype that they have made a short film with. The only thing lacking at this point is a body design prototype for manufacturing.
 
if they have camera why they use other camera to shoot promo?
i dont see prototype on that BTS ...
what i see is machine vision camera gx2300 ...
see the pictures i post....
and i m not only one who notice ....
 
vanvideo said:
Could you explain this? The specs say 12-bit, 4:4:4 color. How do you figure 4:2:0?
Click to expand...
Well, Red's specs say 12-bit, 4:4:4, and that's not true either. Maybe they got their specs from the Red page? I don't know.

But it is frankly impossible for a single CCD sensor (or single CMOS sensor) to deliver 4:4:4 color at the stated resolution. Cannot be done, no matter what. The only way you can get 4:4:4 color, at the stated resolution, is to use three chips.

Understand this: an image sensor is a MONOCHROME device. If you see something advertising a single chip at 2048x1152, such as this Digital Bolex, then you know that that is the resolution only in black and white. Period. If you want color out of a single sensor, you have to put on some sort of Color Filter Array (CFA) such as (typically) a Bayer pattern. And when you put on a Bayer pattern, you get a colored filter over every pixel. And 1/2 of those pixels will be covered with a green filter, and 1/4 will be red, and 1/4 will be blue.

And that, right there, is why it is impossible to get 4:4:4 color at the stated resolution -- because only 1/4 of the pixels are blue, and only 1/4 of them are red.

You will not, under any circumstances, get 1920x1080x4:4:4 or 2048x1152x4:4:4 out of a single-chip camera like this. The very most you could get would be a nominal 2048x1152 x 4:2:0, but even then, you won't truly have 2048x1152 because you'll have to de-Bayer that image, and when you do so, resolution will be lost. The typical stated figure is that you'll get about 70% of the nominal resolution, so you'll end up with something around a 1440 x 810 image, with slightly better than 4:2:0 color.

Two things to note about this:
1) If anyone tells you otherwise, they're wrong, so don't even fight it. It is just what it is.
2) I knew all this, and I pulled the trigger and bought one anyway. It's fine. This is just the way things are when dealing with a single-sensor camera.

The Red One was a "4k" chip, and they are extremely proud of their de-bayer algorithm which retains a stated 80% of res, which results in a Red One delivering about 3.2k of real resolution out of its 4K chip. If you could apply the exact same demosaic/de-bayer algorithm to the Digital Bolex, you'd end up with a 1.6k image out of its 2k chip.

As for 4:4:4 -- again, not possible, no matter what you smoke. The Red One isn't 4:4:4 either by any stretch of the imagination, and neither is a Scarlet and neither is an Epic. And neither is a Sony F3 or FS100, regardless of whether you put it in 4:4:4 "mode" or not. Because it's not possible. Now, Red could argue that their RedCode codec *records* 4:4:4, sure, but the fact is that the IMAGE does not have 4:4:4. So I guess DBolex could claim that they're not color-subsampling into a codec so they aren't turning it into 4:2:0, but -- that'd be silly semantics.

Here's the bottom line: 4:2:0 means that you have an image that has twice as many luminance samples as it has color samples, on both the horizontal and vertical directions. And a Bayer-pattern sensor has twice as many green pixels as it has red or blue pixels, on both the horizontal and vertical directions. Therefore, it could be argued (and is argued by Panavision) that a Bayer sensor is a 4:2:0 sensor. The terminology can be argued the other way, that the sensor itself isn't any manner of x:x:x notation, and that argument would be technically correct while disingenuously misleading, because it's getting to the point where we'd be arguing over what the meaning of "is" is.

Graeme Nattress wrote up an excellent lengthy reply to this basic issue years ago about the Red, and he said that it's basically a little better than 4:2:0, because while there are only 2K's worth of color samples in a Red's 4k image, the loss of luminance resolution from the de-Bayer ends up making the color proportion slightly higher. So he argued that a Red's 4K image would be more like a "3.2:1.6:1.6" if you wanted to call it that, but that would be fairly meaningless in common terminology.

So let's just call it what it is: the images coming off a Bayer sensor are basically 4:2:0, for all reasonable intents and purposes.
 
they dont "look alike"...
that IS gx2300 ...
i got e mail from others who notice a same thing ...
maybe they use gx2300 to test concept or the capturing software ...
but they should be honest not to say that they shoot promo on prototype ...
many people are ready to give money for something that exist only on paper....
 
Last edited:
Razz16mm said:
So an Alexa is at best a 4:2:0 camera? It has a Bayer sensor. C300 has a Bayer sensor, FS100 and F3 have Bayer sensors, not to mention virtually every digital still camera out there capturing 4:4:4 JPEGs.
Click to expand...
And any of them, at their stated resolution, will do no better than 4:2:0's worth of color sampling as per the stated resolution.

Your inclusion of the C300 is incorrect though. It does it the "right" way. They oversample and then downsample into a 4:4:4 matrix. The C300 uses a 4K chip to create its 1920x1080. It legitimately samples 1920x1080 worth of red, 1920x1080 worth of green, and 1920x1080 worth of blue. It is a true 4:4:4 chip (but, again, only in context that it doesn't try to make a 4K image! If they tried to use that chip to make a 4K image, it would deliver -- wait for it -- yep, 4:2:0).

The color resolution of a Bayer sensor is as close to what the human eye can see as any sensor technology out there.
Click to expand...
Whether that's true or not, is irrelevant. Marketing-speak. We're not talking about what it simulates, we're talking about what it samples. x:x:x notation is a representation of the ratios of actual color samples in an image.

The color resolution of a Bayer sensor is limited to 1/2 the horizontal and 1/2 the vertical, by very definition. That's what Bayer is.

It is the processing behind it that determines format color depth and encoding for RGB images not the sensor type.
Click to expand...
It is the filtering used over the sensor's pixels that determines the actual color sampling resolution. If the sensor is a Bayer sensor, it can never sample more than 1/2 the horizontal and 1/2 the vertical (which is, by the way, the very definition of what a 4:2:0-encoded image is). That's why Panavision went with the RGB Stripe system -- it uses 5760 pixels wide, delivering one red, one green, and one blue pixel for every one of its ultimate 1920 pixels wide. So an RGB Stripe sensor is a way to get a 4:4:4 image, but only when put in context that it's not 4:4:4 at its stated resolution! It's 4:4:4 at its final ultimate processed size, which is 1920x1080.

You can get 4:4:4 out of a Bayer sensor, but NOT at its stated resolution. If you used a 4K chip, you could create a 4:4:4 2K image out of it.

Now, if you want to get all vague and cutesy with your image processing you could synthesize (well, basically, up-res) your 4:2:0 color into a synthesized 4:4:4 and just call it 4:4:4. But the actual sampled image data would still be coming from a 4:2:0 scan of the sensor.
 
whole thread about this on reduser and if I remeber correctly even on this site where he has disagreed with you on that...

http://www.reduser.net/forum/showthread.php?71931-Is-R3D-4-4-4-or-4-2-2

"Neither - it's RAW. 4:2:2 etc. refers to the situation where chroma bandwidth is reduced in a Y'CbCr encoded signal to achieve a smaller bandwidth.

Graeme"

"...please don't use imprecise terminology. 4:2:2 etc. don't apply to a measure of chroma resolution in any sense, but a process by which resolution is deliberately reduced by some circuitry or process for bandwidth reduction reasons.

For instance, you could record a VHS tape onto Digibeta at 4:2:2. At which point you definitely have a 4:2:2 signal, but that actually tells you nothing about the measured chroma resolution of the signal. Similarly a BetaSP dub to the same Digibeta would be 4:2:2 but the actual measured chroma resolution would be vastly superior. A direct camera recording to that Digibeta would also be 4:2:2 and depending on the camera could have a vastly superior measured chroma resolution.

The reason, of course, why such analogue era compression as 4:2:2 works so well is that we don't see much detail in chroma compared to luma.

The issue of how colour filter array sensors see colour resolution compared to luma resolution is a different issue. In such a system we're dealing with broad colour filters where any detail is visible in all three channels in most circumstances.

Graeme"

edit: I'm fairly sure I could have found some better quotes by him than these but just try and read the theread and you will see it's a rather complex issue.
 
Last edited:
vladszp said:
they dont "look alike"...
that IS gx2300 ...
i got e mail from others who notice a same thing ...
maybe they use gx2300 to test concept or the capturing software ...
but they should be honest not to say that they shoot promo on prototype ...
Click to expand...
Possible that that's what they're doing. But even so, if so, the images it captures should be identical to the images the Digital Bolex captures. If it's truly the same sensor, then there's really no difference, because the DBolex (as I understand it) has no processing at all in it. It just dumps the sensor data directly to the recording media.

All processing, all de-noising and de-mosaic'ing and scaling and edge enhancement and highlight rolloff and gamma adjustment and white balancing and color correction and all that, would happen *outside* the camera. Which basically means, again -- if the gx2300 is using the exact same sensor and dumping the exact same raw data stream that the DBolex will use, then the end result will be identical.

If that's the case, then there's certainly nothing misleading about the posted images or the claims of image quality, but yes for clarification they should have stated that they're using an existing camera to test the concept.
 
Barry_Green said:
Well, to be fair, this digital bolex will also be nothing like 12 bit raw 4:4:4 either. It's going to be 4:2:0 at best. That's the nature of a single sensor with a Bayer pattern.
Click to expand...

Barry_Green said:
Possible that that's what they're doing. But even so, if so, the images it captures should be identical to the images the Digital Bolex captures. If it's truly the same sensor, then there's really no difference, because the DBolex (as I understand it) has no processing at all in it. It just dumps the sensor data directly to the recording media.

All processing, all de-noising and de-mosaic'ing and scaling and edge enhancement and highlight rolloff and gamma adjustment and white balancing and color correction and all that, would happen *outside* the camera. Which basically means, again -- if the gx2300 is using the exact same sensor and dumping the exact same raw data stream that the DBolex will use, then the end result will be identical.

If that's the case, then there's certainly nothing misleading about the posted images or the claims of image quality, but yes for clarification they should have stated that they're using an existing camera to test the concept.
Click to expand...

So...why not just buy the GX2300?
 
Makes me wonder if allied is the OEM for it. It makes sense that they would also be manufacturing the d-bolex they have a OEM service.
 
You must log in or register to reply here.
Back
Top