Postmaster
Dark Side of the Camera
+1 I´really wonder why they don´t doing that.
Bolex D16: RAW, 16/Super 16 sensor size, costs less than MkIII
- Thread starter underpressure
- Start date
+1 I´really wonder why they don´t doing that.
S. Matthews
Well-known member
Thanks for the reply! yeah it is once again another oversimplified and not entirely accurate statement to say it best replicates human vision, and just adds to the confusion. With regards to Bayer sensor's you are correct, I just think some people don't understand the importance of the color science that interprets the data that comes off a Bayer sensor. (then again I haven't finished my homework Barry gave me so... I got some reading to do!)
Either way it's a time with a wide array of options, to suit many needs, and more than ever we have no excuses for not just getting out there and producing great content on any of the available tools.
Stay passionate, and keep on creating!
Peace
Either way it's a time with a wide array of options, to suit many needs, and more than ever we have no excuses for not just getting out there and producing great content on any of the available tools.
Stay passionate, and keep on creating!
Peace
Last edited:
Barry_Green
Moderator
It is an understandable presentation of a Bayer color filter array. I wouldn't call it "overly simplistic" any more than I'd call your assessment "overly complex". A
"much" higher? No. Maybe a little bleedover, which (when countered by the resolution loss) results in a relatively higher proportion of color to actual resolved detail, which is why I already said it would be a little better than 4:2:0. But not a lot. And I doubt it'd have more color information than the NOMINAL resolution would dictate. It may have synthesized up-rezzed intermediate pixels, but not higher actual sampled information. How can it?
What it will have, when it's all said and done, is that your 2K image from a bayer sensor will have 1K of red and 1K of blue actual resolution. If there's uprezzed sampling being added in, well, that may look good, or it may not, but it isn't actual resolution, any more than uprezzing a 2K image to 4K makes it a "4K image."
You keep saying this, we all understand it, but what I want to know is -- why keep saying this? What does this have to do with the notion of whether or not the DBolex (or the Red One, for that matter) actually deliver 4:4:4?
Maybe we need to term it more in the form of a formal debate to clarify where our positions differ, so -- here's the question. "Can a 2048x1152 Bayer-pattern sensor deliver 2,048x1152 resolved detail, with 2048x1152 resolved green, 2048x1152 resolved red, and 2048x1152 resolved blue?" I say that it cannot. Do you say that it can? Because that is the only acceptable definition of "2K 4:4:4" that I can think of.
Samuel H
Major Contributor
my take: this is a hacked version of that industrial camera (GX2300C), what they added is: a film-making friendly casing, an audio recorder, an interface to easily control the industrial camera, a simple way to record the RAW signal that the industrial camera outputs through its ethernet port, and a small screen that displays the analog image-out from the industrial camera *** GENIUS!! *** (as long as that GX2300C can be bought for $1K or so, which I'm not sure is the case...) *** http://www.1stvision.com/cameras/AVT/Prosilica-GX2300-GX2300C.html http://www.1stvision.com/cameras/AVT/dataman/70-0067_GX_TechMan.pdf
Barry_Green
Moderator
Graeme Nattress has written as to why. It was really rather funny how he blatantly pointed it out. The Foveon is basically three chips sandwiched together, as opposed to a conventional three-chip camera where there's an optical prism that splits the light and sends each wavelength to the appropriate chip. So what splits the light in the Foveon? The silicon itself! And Silicon is a fairly lousy light filter, at least as compared to an optically pure prism. There are reasons that Foveon has had very limited acceptance in photography in general. It's one of those ideas that seems perfect, but in reality, well, there's always tradeoffs and compromises. There's no one simple magic bullet that solves all problems.
http://www.graftek.com/pages/GX2300.htm
$5660
repackage and lose 3000$
and one more thing ...
how they going to be ready for production when they even have problems to record image from finished industrial camera ...
http://www.digitalbolex.com/color-grading/
$5660
repackage and lose 3000$
and one more thing ...
how they going to be ready for production when they even have problems to record image from finished industrial camera ...
http://www.digitalbolex.com/color-grading/
Last edited:
Zephyrnoid
Veteran
Looks like you are right. Some shortcut hey ? repackage and let's hope for their sake that they got it right.
Bassman2003
Veteran
I am not interested in this camera but I always learn something when Barry is engaged in a topic. Great read!
and also ive read somewhere foveon sensors are much expensive to make
wouldnt an S35 with 2k photosite foveon sensor be awesome? those pixels would be huuge. may compensate for foveon's sensitivity.
Barry_Green
Moderator
Well, yes, that would be kind of fantastic actually...
I don't really understand why they're going for a B/W output. I mean, is it solely for nostalgic purposes? Does it really take that much processing and effort to show colour (color) video. I mean, they speak about how little effort it is to get XLR audio added to a camera (in Phillip Bloom's interview) but then they can't provide a decent monitoring solution.
To be honest, it's kind of a deal-breaker to have a camera without decent real-time video monitoring. I mean, even my 550D provides better monitoring then this spec-based DBolex, and that's not even a "cinema camera"...
Film is the hardest way to get an image. I don't see our current generation trying to use this thing like a 16mm film camera... along with B/W video tap...
If I go by what you're saying, Barry, then I'd have to assume that there will be many disappointed people coming from the world of hacked and feature-rich DSLRs and Video Cams. Let me just go along with this then... I've seen (actually spoken to - come to think of it) people who think that the REDs produce horrible flat images and that their DSLRS and AF100s give you a better picture. And they're the people, the exact kind of people, that would jump onto this and throw money at it. I wouldn't be able to begin to imagine the disappointment that these guys are going to experience when they watch B/W video at 640 x 480 on a field LCD and then play it back on their PCs (packed to the brink with HDDs, btw) in 2k with almost no contrast at face-value (and not to mention the effects of using their 50mm EF lenses). I think the guys making the camera are trying to be honest, but hype and the mere word "RAW" attracts a certain crowd who knows not what RAW is.
Also, since I'm in to this, what's up with the whole old Bolex look actually. I mean, does anyone really want an old school looking camera - or has he just done it to get some sort of brand behind him. Because I can't imagine many modern-day filmmakers really careing about this nolstalegic value he keeps going on about. I may be wrong, but that crank - on a 2K digital camera, just looks weird to me. I'd say that most guys just want a decent looking black box...
Last edited:
A couple things:
First, they came out and said they are partnering with an electronics company. They are using off the shelf parts. Is it that big of a leap in logic that the electronic company could be Graftek? (Or who ever makes the original camera?)
Second, this will be one of the best Film Student cameras available (if true.) As Barry points out, you have to do it yourself! When you do, you will learn. Which I think is awesome. I know my knowledge gap is lacking in some areas because the camera "does it for me" so to speak. This looks like a truly "fun" tool.
First, they came out and said they are partnering with an electronics company. They are using off the shelf parts. Is it that big of a leap in logic that the electronic company could be Graftek? (Or who ever makes the original camera?)
Second, this will be one of the best Film Student cameras available (if true.) As Barry points out, you have to do it yourself! When you do, you will learn. Which I think is awesome. I know my knowledge gap is lacking in some areas because the camera "does it for me" so to speak. This looks like a truly "fun" tool.
Last edited:
I don't think the look really matters that much to most people. (Well if it looks like a Star Trek hair dryer it may matter to some.) The Retro thing will go over good with the hipsters though.
Exactly my point.
Also, that prosilica looks awfully similar. Too much so to be a coincidence. In fact, I think I feels a conspiracy coming along...
rogar-Dum(DBG):... but seriously, they may need to answer to this.
S. Matthews
Well-known member
The hand crank is the most intriguing part for me, and the design too (I always wanted to own a Bolex). But back to the hand crank ever since "man on fire" I have wanted a digital hand crank camera, heck even if it just controls fps that would be amazing, but to be able to set it to control other aspects of the camera or effects would be very interesting. I see it as a potentially wonderful creative tool for the low end indi and student market as well as people looking for that nostalgic feel in a fun little camera. However as Barry points out unfortunately a large amount of people will most likely end up buying this camera only to find out that it is not quite the camera they were looking for or expecting.
... oh and if it had a three lens turret with removable mounts for each hole(pl,cmount,b4... cannon?) that would just be my dream mini run and gun camera.
... oh and if it had a three lens turret with removable mounts for each hole(pl,cmount,b4... cannon?) that would just be my dream mini run and gun camera.
Barry_Green
Moderator
Yes, it does. Now, look, I'm going out on a limb here because I don't know what the engineers are doing, but if I was doing it (and I'm not a hardware engineer, I'm an ex software engineer) I would think it would be relatively easy to scan the image to just pull out the green pixels and display that as a rough black & white. But doing an entire Bayer de-mosaic? That's not a trivial task. That will take some processing horsepower to do, which means work, and chips, and then it'll start to spiral out of control. And from their design philosophy of "keep it as simple as possible" -- I don't think they'll do that.
Here, let me give you an example. The Red One has an SDI output, and an HDMI output. The video quality is fairly lousy. Why? Because they never meant for you to use it! It's a quick and dirty de-bayer, it's a scratch color grade, it's something you can slap up on a monitor to at least see AN image, but it is in no way representative of what the real recorded image will be. The engineers understood that. I understood it. But the public in general? They'd look at the monitor output and say "that looks like @$$, this Red's a piece of junk." That was a grossly unfair characterization, but I've seen it happen over and over. I've even seen people try to use an SDI recorder to record the Red's output! I try to tell them it's pointless, but they do it anyway. So here we have a case of Red spending more than they probably shoud, to create something, which bit them in the butt. I bet they almost wish they'd gone with a black and white output, because it would have kept expectations in check. People would have realized "oh, this is just a video tap" instead of "this is what your image looks like? That's awful!"
So, without putting words in the DBolex guys' mouths, I suspect that they're keeping it as simple as possible, and that probably means a quick grab of the sensor, and scratching something together to show basic framing. About all the processing you're likely to get will be a focus assist, which I would guess would be a black and white extraction of the center of the sensor, so even then not really any processing at all.
Well, right, but -- think of the can of worms they avoid by not trying to provide a "decent" monitoring solution. I am frankly quite surprised that they are including XLRs, they must have found an inexpensive all-in-one solution that just dumps digital data out that they can then simply embed in their stream. I'm glad it's there, but I'm surprised it's there!
Well, and that goes to the core of people not understanding what a cinema camera is. If someone thinks that a "cinema camera" must inherently be more powerful or higher on the scale than a "video camera", that's that person's error. A cinema camera is different from a video camera, which is different from a DSLR. A regular video camera (like the HPX3700) does things a Red One can't dream about doing, but the Red One is a "cinema camer" and the HPX3700 is a "video camera". I said earlier that a DSLR doesn't do a tenth of the things a regular video camera does, and I also said that the DBolex won't do a tenth of the things that a DSLR does. And I stand by that. The DBolex is going to be the most stripped-down thing you can imagine. All the features are going to be done not by the camera, but by you, in post.
Which is why I said earlier, this unit is not going to replace an AF100 or F3 as your "A" camera, even though this shoots raw. This is a special-purpose tool, not an all-purpose tool. And yes, I fully believe that 95% of the folks on this site who evaluate this unit, at its MSRP of $3200, against an AF100 at its current street price of $3500, will be better off and happier with the AF100 type of product.
The DBolex is for old-guard film shooters, or for those who must have raw at any cost (and it's a surprisingly low cost, in terms of dollars, but it's a much bigger cost in terms of post).
That is exactly what I anticipate happening, yes.
I predict this is exactly what will happen. Which is why I say, the key to the DBolex's success will be for them to manage expectations. They have to let people understand what it is they're really getting. And that's hard to do nowadays. There are people out there who think that a DSLR is "better" than a video camera, and now there are those who will think that the DBolex is "better" than a video camera too. And they're in for a real surprise.
Think of it this way -- think of the massive hype that the Hyperdeck Shuttle got on day one. People could not BELIEVE how amazing it was to get an uncompressed recorder for $345, and they were ready to kiss the Nano, Ninja, Samurai, all of them "goodbye." Then after a few days people started to think about it, and the reality set in, of how cumbersome and expensive it was going to be to have to buy the drives and work with the massive, massive file sizes. And lots of folks started thinking that the Ninja, with its $995 price tag, and its monitor, and its edit-friendly ProRes, started looking awfully good, right? Well, the DBolex (as I understand it) is basically a Hyperdeck Shuttle with a lens on it. Great for some, but hardly ideal for all. There's a place and a market for all these products, I just hope people think it through and realize what they're really getting. I think what they're doing is cool, and I plunked down my money already (which, I acknowledge, is crazy) but I fully understand what I'm getting into. I'd hate to see people thinking that this will be more than it really is, and then being disappointed. I hope the DBolex guys put out some more information so people see exactly what they're getting. What they're getting is, in reality, quite cool, but it may not be what they're expecting. Reading the comments and questions on the kickstarter page, I think the developers have a long road ahead of them to address those expectations.
I'm surprised the guys are as young as they are, if they were old coots who used to shoot on the Bolex then I guess I'd understand it better. The Bolex was a revolution when it came out, specifically because it was such a simple thing. But man, it was NOT an "A" camera by any stretch of the imagination! It sounded like a coffee grinder, you couldn't shoot sync sound on it, the prism in it robbed you of about 1/3 of your total light, I wasn't ever really a big fan of the Bolex. This unit seems closer to what I want, but it's definitely in the spirit of the old Bolex.
Barry_Green
Moderator
That's exactly how I see it, although I also acknowledge that it does have that 12-bit raw, and that's a nice bonus.
Make that suggestion on the Kickstarter page! A three-lens c-mount turret would be great, or, better yet, a three-hole universal-mount that we could add our own mount to. But a new-old-stock Angenieux 17-70 f/2.2 zoom is going for $799, and that zoom is tiny, so that one zoom is probably enough that I'm not gonna be stressed if they don't offer the turret. But a turret would be fun.
S. Matthews
Well-known member
new prototype pic...
this is their current prototype, they say the GX was used to test the sensor, basically proof of concept and they liked it so much they shot the short on it. the source is the comment section of the kickstarter page.
P.s. could someone post the turret idea for me, not on kickstarter yet, does it take pay pal? not fond of amazon as I've had a few issues dealing with them being from Canada.
this is their current prototype, they say the GX was used to test the sensor, basically proof of concept and they liked it so much they shot the short on it. the source is the comment section of the kickstarter page.
P.s. could someone post the turret idea for me, not on kickstarter yet, does it take pay pal? not fond of amazon as I've had a few issues dealing with them being from Canada.
Last edited:
Thanks for the reply. It actually gave me a better idea of where this camera is/will be standing.
It's going to be tough for us non-film guys to get decent exposure and colour (color) without any on-board waveforms and vectorscopes. Many more light-meters being sold...
It's going to be tough for us non-film guys to get decent exposure and colour (color) without any on-board waveforms and vectorscopes. Many more light-meters being sold...