Industrial CMOS camera as an entry-level HD cinema camera (1080p)

[Stuart McCammon]

I have to wonder how the color reproduction off a single large CCD will compare with a 3-CCD camera - I was going to buy a really neat JVC video camera about a year ago (that also had a single, large sensor), until I saw that the picture quality was not all that - anyway, worth checking in to)

[Cesar Rubio]

I have to wonder how the color reproduction off a single large CCD will compare with a 3-CCD camera - I was going to buy a really neat JVC video camera about a year ago (that also had a single, large sensor), until I saw that the picture quality was not all that - anyway, worth checking in to)

 

Stuart:

 

Actually a Single CCD MVC camera has it's advantages over it's 3-CCD's counterparts.

 

For starters, the use of inexpensive C-Mount prime lenses, which gives a better color rendition and resolution.

Also using a single CCD eliminates the optical aberrations caused by a 3-chip prism assembly.

 

I own a Nikon D70s camera with a single Sony CCD sensor and it gives extraordinary results!

 

If you check all High End Digital Cinema Cameras, Dalsa, Red Camera, Arri, etc, they ALL have a single CCD or CMOS sensor!

 

Thanks,

Cesar Rubio.

[Cesar Rubio]

This is good information extracted from the Dalsa Image Sensor PDF.

 

http://www.dalsa.com/dc/documents/Image_Se...18-00_03-70.pdf

 

In designing the sensor and electronics for our Origin® digital

cinematography camera, DALSA drew upon its 25 years of

experience in CCD and CMOS imager design. Given the demands

and limitations of the situation, we determined that the best image

sensor design for our purposes was (and still is) a frame-transfer

CCD with large photogate pixels and a mosaic color filter array. It

is not the only design that could have succeeded, but it is the only

design that has succeeded. No other design has demonstrated a

similar level of imaging performance across the range of criteria

we identified above. This is not to say that no other design will

reach those performance levels; to bet against technology

advancement would be short-sighted. On the other hand, the

performance Origin can demonstrate today is several generations

ahead of the best we?ve seen from other technologies and

architectures, and Origin?s design team is forging ahead to

improve it even more.

 

 

Thanks,

Cesar Rubio.

[Stuart McCammon]

So, given that 1.2" is as big as a CCD sensor is probably ever going to get due to lens coverage of C-mount lenses, why not get the Pike with the big sensor? Just wondering)

[Cesar Rubio]

So, given that 1.2" is as big as a CCD sensor is probably ever going to get due to lens coverage of C-mount lenses, why not get the Pike with the big sensor? Just wondering)

 

The Pike F-421c has a 1 1/2" sensor. You can use that camera with almost any regular lenses with adapters.

For example, the Nikon mount adapter only cost like $70 dlls. And you can use pretty much all Nikon lenses to shoot video with extraordinary quality!

 

Our "friends" who make and sell the 10K and up lenses, don't want you to know the truth...and that is, a Nikon prime lens has more than you need resolution for "High Definition"....Actually in my "High Def" Nikon D70, I have to use a soft filter to take pictures of some people!

 

thanks,

Cesar Rubio.

 

The psychology behind the "High End" lenses prices is this....Hey if the paid 80K or more for the camera...sure they will pay 10K or more for the lens...lets just "convince" them that with this only lens, they can get their cameras full potential resolution....yeah right.

 

Cesar Rubio.

[Stuart McCammon]

Yah, I got that one - C-mount lenses, thousands of happy still photogs - right with ya man)

 

So, tell me if I get the picture here: One of these cams, mounted on a base plate, with C mount primes (or a zoom with no follow focus), hardwired to a mini-computer with video feeding out of the computer to a reference monitor as the viewfinder?

 

Sounds, um, somewhat challenging in any other shot than absolute locked-down solid. Pan, tilt, dolly, put the thing on a crane or arm and suddenly you are wielding a heavy beast with cables that constantly get caught and a viewfinder that needs repositioning. And that doesn't even take into consideration the Steadicam version - which kind of seems like it might end up looking like this guy:

 

http://www.bpmlegal.com/w5901666.html

 

All I am saying is, this really sounds like quite the engineering task guys - best of luck)

 

PS - you might want to get started training a monkey to be the focus puller now, he has his work cut out for him -

[Troy Warr]

Troy:

 

I apologize because I don't have clips taken with the Pike F-145 yet, there's some problems with my distribution channel that need to be resolved. As soon as I get my camera, I will post some footage on my forum.

 

I think that 720 is enough resolution for most circumstances, that is 4 times as much information of SD with letterbox (16:9 aspect ratio 0r 360 x640 pxs.)!

 

If we consider that we are recording uncompressed RAW video, that we could convert to RGB 4:4:4 color space later in post...that is a LOT of info, and then if we shoot at 14 bit, oh well needless to say!

 

I think that the folks that make the Viper, CineAlta, etc won't be very happy with us!

 

Even Sony is the one who produces the CCD for the Pike F-145c...

 

Thanks,

Cesar Rubio.

Thanks, Cesar - no problem at all! I appreciate your help and there's no rush. I'm eager to see the footage but I still have a lot of research and testing ahead of me before I'm able to make a purchasing decision.

 

I'm also starting to think that 720p may be good enough for my purposes, at least for now. I wish there was a more affordable way to get 1080p, but I would assume that these cameras will gradually come down in price as tech products do. 1080p would give me the most options in terms of project output/distribution, but 720p still leaves a lot of options open, too. 720p allows for some niceties like higher frame rates, lower cost, less data to work with (as you mentioned), quicker editing, etc. It may be best to get my feet wet with a 720p setup for now, and then upgrade to 1080p once I've gained some experience with this type of setup and can afford to build a higher-end system.

 

Thanks again!

[Cesar Rubio]

So, tell me if I get the picture here: One of these cams, mounted on a base plate, with C mount primes (or a zoom with no follow focus), hardwired to a mini-computer with video feeding out of the computer to a reference monitor as the viewfinder?

 

Sounds, um, somewhat challenging in any other shot than absolute locked-down solid. Pan, tilt, dolly, put the thing on a crane or arm and suddenly you are wielding a heavy beast with cables that constantly get caught and a viewfinder that needs repositioning. And that doesn't even take into consideration the Steadicam version - which kind of seems like it might end up looking like this guy:

 

PS - you might want to get started training a monkey to be the focus puller now, he has his work cut out for him -

 

Stuart:

 

With Gig-E MVC, you can get hundreds of meters away from your camera with a simple cable+power, and still have control like if it was there with you. For focusing, there is some c-mount auto-focus lenses. A LCD monitor as big as 24" or more, with 1920x1080 resolution can be used for focusing....do you know a better "viewfinder" than a live preview of that size?

 

I don't know how they focus film cameras in a dolly...do they use a monkey? Whatever they do, it may work with MVC. Film cameras & MVC have "similar" specs, like they don't record sound, use of MF lenses, RAW film or digital workflow etc.

 

Thanks,

Cesar Rubio.

[Troy Warr]

Yah, I got that one - C-mount lenses, thousands of happy still photogs - right with ya man)

 

So, tell me if I get the picture here: One of these cams, mounted on a base plate, with C mount primes (or a zoom with no follow focus), hardwired to a mini-computer with video feeding out of the computer to a reference monitor as the viewfinder?

 

Sounds, um, somewhat challenging in any other shot than absolute locked-down solid. Pan, tilt, dolly, put the thing on a crane or arm and suddenly you are wielding a heavy beast with cables that constantly get caught and a viewfinder that needs repositioning. And that doesn't even take into consideration the Steadicam version - which kind of seems like it might end up looking like this guy:

 

http://www.bpmlegal.com/w5901666.html

 

All I am saying is, this really sounds like quite the engineering task guys - best of luck)

 

PS - you might want to get started training a monkey to be the focus puller now, he has his work cut out for him -

Hi Stuart,

 

It's actually not as complicated as you might think. I would go so far as to say that this system is in some ways *easier* to work with than most digital cinema cameras, because the components are modularized and the camera head is so small and lightweight.

 

You can use C-mount lenses on the camera, but as Cesar suggested, 35mm SLR lenses may be the better choice. You can either use a straight adapter (such as C-mount to Nikon F), or, as I'm hoping to do, use a 35mm DOF adapter like the Cinevate Brevis35 or SGPro for even more "cinematic" shallow DOF. There are a number of affordable follow focus solutions out there for 35mm SLR lenses. Obviously those aren't quite as practical as a cinema lens follow focus, mainly due to the shorter focus pull of modern 35mm lenses, but they can still be quite usable.

 

As for the recording PC, that does need to be "hardwired," but that doesn't necessarily imply reduced mobility. Depending on the camera's interface, that cable can be as light and flexible as standard Ethernet cable, and in some cases up to a few hundred feet long. All that a camera operator would need to worry about would be the camera head (as small or smaller than most consumer miniDV palm-corders) and lens assembly. The cable would need some protection from getting snagged, but I don't see how that would be any more detrimental than lighting cables that are usually scattered around a movie set. As for monitoring solutions, you can use everything from a huge HDTV on a table next to you (in a studio setting - in fact, I think they did that in some scenes with Zodiac, shot on the Viper) to a 7" monitor hung with an arm off your camera or rails. The latter is used every day in film shoots and can be quite versatile and even preferable to a viewfinder.

 

There is a good blog that details the use of an early version of the Silicon Imaging SI-2K, and they have several examples of how versatile and portable the camera can be, even handheld or with a Steadicam. Obviously it's not ideal for all solutions - documentary work, for example, could become a nightmare - but for studio work or narrative shooting on a set, it really doesn't seem that limiting to me. It will take some study, hard work and a little ingenuity to set up a system like this from scratch, but that's why I'm doing my research to see if it's feasible, or if my money would be better spent on a ready-to-run camcorder.

 

BTW, how the hell did you come across that link? :P

[Stuart McCammon]

OK, I have to admit, this idea is fascinating to me. As far as the link, I Googled it up, having seen a TV short some time ago about some nutjob who actually made a wearable Habitrail - but that wierd patent site was a wonderful thing to stumble in to)

 

So, all that is mounted on a dolly, say, is the head on a baseplate, with a C-mount zoom (if a follow focus exists, I would think it can be easily modified or extended), with a matte box on rods and a small viewfinder, and handles if you decide to shoulder mount it - sounds very workable!

 

I guess the only remaining question would be a cheap, quiet, stable power source that wouldn't fry the whole thing without having to rent a huge genny. Also, how many amps would that thing suck up? I would think the comp alone would need lots of power to run that fast. Are Honda generators quieter than they used to be? I will have to check that out.

 

I would think another concern is potential focus problems caused by the alignment of the lens and how far away from the sensor it is, and would still want to make sure that there is no problem with vignetting.

 

BTW Cesar, I was only kidding about the monkey. You would have to look at a picture of a film camera on a crane or dolly, and try to imagine how to pull focus without a follow focus to understand why a monkey might be neccessary) Also, autofocus lenses would not be practical for a cinema camera.

 

All the best,

[Cesar Rubio]

Stuart:

 

I have tested regular 12V batteries for cars, using 2 MVC (for 3-D work) and a SFF PC with a 17" LCD monitor and it runs for hours. Check out this topic please:

 

http://www.davidrubio3d.com/view_topic.php...amp;forum_id=10

 

Thanks,

Cesar Rubio.

[Stuart McCammon]

BTW you should consider a comp running Vista to get past the XP RAM cap issue)

[Cesar Rubio]

BTW you should consider a comp running Vista to get past the XP RAM cap issue)

 

Stuart:

 

Actually RAM use is minimal while recording to disk. HD Speed is the determining factor for recording great amount of Raw uncompressed data.

 

I don't know about making the Vista jump just yet...with all the negative reviews and lack of drivers and software support, I think I'll wait a little longer. Actually, if Windows 2000 pro was still available for sell, I will use it in all my machines!

 

Thanks,

Cesar Rubio.

[Cesar Rubio]

Troy:

 

With the Pike F-145c you can achieve:

 

At 1388x1038= 25 fps.

 

And to calculate other resolutions this is the formula:

 

1388x1038= 1,440,744 pxs.

 

1280x720 (720p)= 921,600 pxs.

 

Now divide 1,440,744/ 921,600= 1.56

 

Multiply 1.56 for the full frame resolution at 25 fps....1.56x25= 39 fps.

For the Pike F-210c is this.

 

1928x1084= 2,089,952 pxs. at 30 fps.

 

1280x720= 921,600

 

2,089,952/ 921,600= 2.26x30=67.8 fps!

 

That formula is the same for any MVC and any resolutions.

 

Thanks,

 

Cesar Rubio.

Cambridge Wisconsin, USA.

http://www.davidrubio3d.com/

 

Troy:

 

I've been told by an engineer of another MVC manufacturer, that only the vertical resolution has an impact on the increase of frame rates. On Tuesday when the Engineer from AVT will be back, I will double check with him this info to see if applies to all MVC or just that manufacturer in particular.

 

If that is true, then the formula to get the frame rates would be as follows:

 

Example, Pike F-145c at 720:

 

Divide 1038 (max. vertical resolution) into 25fps (max. frame rate)= 41.52.

 

1038-720 (V resolution) =318/41.52=7.65 fps. (additional frames)

 

25+7.65=32.65 fps.

 

 

For the Pike F-210 at 720p:

 

1084/30=36.13

 

1084-720=364

 

364/36.13=10.07

 

30+10.07=40.07 fps.

 

Please don't take either rule as correct, until I double or triple check them.

 

Thanks,

Cesar Rubio.

[Cesar Rubio]

Troy:

 

I've been told by an engineer of another MVC manufacturer, that only the vertical resolution has an impact on the increase of frame rates. On Tuesday when the Engineer from AVT will be back, I will double check with him this info to see if applies to all MVC or just that manufacturer in particular.

 

If that is true, then the formula to get the frame rates would be as follows:

 

Example, Pike F-145c at 720:

 

Divide 1038 (max. vertical resolution) into 25fps (max. frame rate)= 41.52.

 

1038-720 (V resolution) =318/41.52=7.65 fps. (additional frames)

 

25+7.65=32.65 fps.

For the Pike F-210 at 720p:

 

1084/30=36.13

 

1084-720=364

 

364/36.13=10.07

 

30+10.07=40.07 fps.

 

Please don't take either rule as correct, until I double or triple check them.

 

Thanks,

Cesar Rubio.

 

I think these are the correct calculations. This portion is extracted from the AVT Pike Manual:

 

Area of interest (AOI)

The camera?s image sensor has a defined resolution. This indicates the maximum

number of lines and pixels per line that the recorded image may have.

However, often only a certain section of the entire image is of interest. The

amount of data to be transferred can be decreased by limiting the image

a section when reading it out from the camera. At a lower vertical resolution

the sensor can be read out faster and thus the frame rate is increased.

 

 

I apologize for the first wrong information.

 

If you want to read more about this you can download the Pike Manual here:

 

http://www.alliedvisiontec.com/files/pdf/p...n_V3.0.1_en.pdf

 

Thanks,

Cesar Rubio.

[Troy Warr]

Hi Cesar,

 

Thanks so much for relaying this info - I appreciate it. I haven't had a chance to read through the Pike Technical Manual, but I'm very interested in doing so and I've bookmarked it to read as soon as I can.

 

So, just to clarify on the calculations for ROI/AOI increased frame rates, I'm not sure that I follow your calculations exactly. Using the Pike F-145c as an example, I would think that you could basically perform the calculation the same way as you showed me before (dividing the maximum sensor resolution by the AOI resolution, then multiplying that by the full-sensor frame rate) but now simply use the full horizontal resolution instead.

 

 

In other words, where the old way we were doing it would be like this (based on F-145c):

 

Maximum Vertical Resolution (1040) x Maximum Horizontal Resolution (1392) = 1,447,680

AOI Vertical Resolution (720) x AOI Horizontal Resolution (1280) = 921,600

Full Sensor Resolution (1,447,680) / AOI Resolution (921,600) = 1.571

Maximum Full Sensor Frame Rate (25) x 1.571 = 39.271 (OLD)

 

 

The new way would be like this:

 

Maximum Vertical Resolution (1040) x Maximum Horizontal Resolution (1392) = 1,447,680

AOI Vertical Resolution (720) x Maximum Horizontal Resolution (1392) = 1,002,240

Full Sensor Resolution (1,447,680) / AOI Resolution (1,002,240) = 1.444

Maximum Full Sensor Frame Rate (25) x 1.444 = 36.111 (NEW)

 

 

Does that seem to make sense? Based on the latter equation, I would think that you could extract more frames per second than your most recent AOI calculations determined.

 

By the way, this website lists the max frame rate for the Pike F-145c as 30fps - do you think that's accurate, or is it actually limited to 25fps as you mentioned?

 

Also, does this new revelation for AOI calculations only apply to CCD sensors? Since CMOS pixels can be addressed individually, I'm wondering if you could still calculation maximum frame rate based on the old method - i.e. the vertical resolution wouldn't be the limiting factor.

 

Thanks again, Cesar! Keep up the great work ;-).

[Cesar Rubio]

Hi Cesar,

 

Thanks so much for relaying this info - I appreciate it. I haven't had a chance to read through the Pike Technical Manual, but I'm very interested in doing so and I've bookmarked it to read as soon as I can.

 

So, just to clarify on the calculations for ROI/AOI increased frame rates, I'm not sure that I follow your calculations exactly. Using the Pike F-145c as an example, I would think that you could basically perform the calculation the same way as you showed me before (dividing the maximum sensor resolution by the AOI resolution, then multiplying that by the full-sensor frame rate) but now simply use the full horizontal resolution instead.

In other words, where the old way we were doing it would be like this (based on F-145c):

 

Maximum Vertical Resolution (1040) x Maximum Horizontal Resolution (1392) = 1,447,680

AOI Vertical Resolution (720) x AOI Horizontal Resolution (1280) = 921,600

Full Sensor Resolution (1,447,680) / AOI Resolution (921,600) = 1.571

Maximum Full Sensor Frame Rate (25) x 1.571 = 39.271 (OLD)

The new way would be like this:

 

Maximum Vertical Resolution (1040) x Maximum Horizontal Resolution (1392) = 1,447,680

AOI Vertical Resolution (720) x Maximum Horizontal Resolution (1392) = 1,002,240

Full Sensor Resolution (1,447,680) / AOI Resolution (1,002,240) = 1.444

Maximum Full Sensor Frame Rate (25) x 1.444 = 36.111 (NEW)

Does that seem to make sense? Based on the latter equation, I would think that you could extract more frames per second than your most recent AOI calculations determined.

 

By the way, this website lists the max frame rate for the Pike F-145c as 30fps - do you think that's accurate, or is it actually limited to 25fps as you mentioned?

 

Also, does this new revelation for AOI calculations only apply to CCD sensors? Since CMOS pixels can be addressed individually, I'm wondering if you could still calculation maximum frame rate based on the old method - i.e. the vertical resolution wouldn't be the limiting factor.

 

Thanks again, Cesar! Keep up the great work ;-).

 

Troy:

 

Scott Smith the Enginner from AVT is back, and actually he send me an e-mail a few minutes ago with this issue. You are welcome to register in my forum and probably he would help to resolve this issue for the benefit of everybody! He is a registered member there, but there hasn't been any questions directed to him, if you go there and ask more questions he would be very happy to help you out.

 

With this I am not saying that the discussion here is closed, we can participate in both forums of course!

 

I am waiting for my Pike F-145c that it will be delivered by the end of this month, and then I will do a lot of tests my-self. I will keep you informed here and on my forum of course!

 

Last thing that Scott is telling me, is that he can achieve 39-40 fps on 720 V resolution! so probably my first assumption was right!

 

The Enginner from another company made me doubt, but I think I was right!

 

I am going to have a last and final result when I test the camera my-self...but I trust Scott Smith 100% anyways!

 

Keep in mind that we are talking about RAW-8 bit acquisition, with Raw-16 bit he told me that he is getting 33 fps, which is good for me also! At RGB-24 bit (8 bit each color) he's getting in his camera 22 fps...but who cares? Raw is the way to go man!

 

Cesar Rubio.

[Cesar Rubio]

Troy:

Keep in mind that we are talking about RAW-8 bit acquisition, with Raw-16 bit he told me that he is getting 33 fps, which is good for me also! At RGB-24 bit (8 bit each color) he's getting in his camera 22 fps...but who cares? Raw is the way to go man!

 

Cesar Rubio.

 

My last comment was incomplete. You can shoot in Raw mode to make adjustments in post, and then convert (or debayering) your video to RGB-24 or RGB-48 bit mode. (depending if you shot in Raw-8 or Raw-16)

 

Thanks,

Cesar Rubio.

[Troy Warr]

Thanks, Cesar! I registered in your forum and will post some of my AVT-specific questions there shortly.

 

RAW is definitely the way to go, in my opinion as well. One of the major reasons that I wanted to investigate this path myself is that I couldn't stand the idea of HDV or AVCHD compression (at least in its current stages) of HD video. Unless one's workflow dictates it - which is a perfectly valid reason, but doesn't tend to apply to my projects - I can't see any reason to invest thousands of dollars in high-quality lenses and cameras, only to have the image highly compressed in its final stages. In my view it's sort of like shooting with a high-end digital SLR kit set to medium-quality JPEG mode.

[Cesar Rubio]

Thanks, Cesar! I registered in your forum and will post some of my AVT-specific questions there shortly.

 

RAW is definitely the way to go, in my opinion as well. One of the major reasons that I wanted to investigate this path myself is that I couldn't stand the idea of HDV or AVCHD compression (at least in its current stages) of HD video. Unless one's workflow dictates it - which is a perfectly valid reason, but doesn't tend to apply to my projects - I can't see any reason to invest thousands of dollars in high-quality lenses and cameras, only to have the image highly compressed in its final stages. In my view it's sort of like shooting with a high-end digital SLR kit set to medium-quality JPEG mode.

 

I agree wit you Troy!

 

Actually this assumption also applies to Raw compressed files. But if you want to shoot uncompressed Raw, be prepared to pay the price in HD storage space requirements!

 

Thanks,

Cesar Rubio.

[Troy Warr]

Actually this assumption also applies to Raw compressed files. But if you want to shoot uncompressed Raw, be prepared to pay the price in HD storage space requirements!

Definitely. Months ago, I briefly corresponded with a salesperson from Blackmagic Design about using their Intensity card to capture uncompressed 1080i60 HD from HDV camcorders, and I quickly began to realize what kind of storage and data bandwidth that was required! I'm all for moderately compressed RAW - I think it's a good balance between visual quality/integrity and data volume.

 

Incidentally, have you (or anyone else still following this thread) come across the early announcement of this "MicroHDTV" camera? I'm mostly intrigued because it should be capable of handling 720p60. I can't find any information on price, image quality, or other features (I don't think that's been released), but apparently some details should be coming out in a week or two.

[Jason Rodriguez]

I've seen it for 18.500 Euros (so $24,290.65), and it's using the AltaSens sensor (same as the SI-2K and SI-2K Mini from Silicon Imaging).

[Troy Warr]

I've seen it for 18.500 Euros (so $24,290.65), and it's using the AltaSens sensor (same as the SI-2K and SI-2K Mini from Silicon Imaging).

Well, then! Scratch that idea. B)

 

Thanks, Jason - I had feared that it would be expensive, but the previews I'd read didn't even hint at price. Obviously considering that price, it would make a lot more sense for my purposes to buy an SI-2K or SI-2K Mini (assuming I had the budget).

 

By the way, I've tried contacting AltaSens and related companies for more information about their sensors, including ProCamHD, but haven't heard back. Based on your experience, and given my budget in the $5-6K ballpark, do you think that there's any chance that I could find an AltaSens-based camera for that amount? I looked into JAI Pulnix, Imperx, etc. per your earlier recommendation, but so far the most promising cameras (pending footage to review) appear to be the AVT models that Cesar is testing.

[Cesar Rubio]

Troy:

 

Tomorrow I will be receiving my long waited Pike F-145c (after this, we expect to have plenty at hand for future sales)....Please check for results on my forum tomorrow night! (at least some stills, and later on some videos)

 

Thanks,

Cesar Rubio.

[Troy Warr]

Hi Cesar,

 

Awesome - that sounds great! Thanks for letting me know. I'm very eager to see some stills and footage. I'll keep an eye on your forum tomorrow night.

 

It may be a little early to ask - and you probably already know what I want to ask, anyway - but at some point (no rush at all) do you think you can make a short clip taken at the maximum frame rate for 1280x720px (I'm guessing it would be 8 bit?) at a "cinematic" shutter speed like 1/48? I would love to see that, to see how it handles motion, what the maximum slow-motion effect looks like, etc. I think that we're probably both interested in the same kind of applications/uses for the camera, and I'm eager to see maximum-quality footage to get a feel for the limits that the camera is capable of. If it's a large file and you're not able to host it, let me know and I can probably host it.

 

Thanks so much, Cesar!