Cinematography Mailing List - CML
    advanced

Cine SpeedCam

16th November 2004

So what's the deal with the Cine SpeedCam? Has anyone used it?
The specs look enticing, but I'm wondering what the image looks like, what the dynamic range is, does it put out an image that gets corrected later or can it be painted on set, can it match Varicam/Cine Alta footage, etc.

I'm also curious what the price advantage is over shooting film.
Anyone have any idea who rents it and for how much?

Art Adams, DP [film|hdtv|sdtv]
San Francisco Bay Area - "Silicon Valley"
http://www.artadams.net/


Art

Funny you should ask. In our ongoing efforts to provide viable working and real solutions to Electronic Cinematography community we are introducing the Cine Speed Cam worldwide. Band Pro Burbank is responsible for the Americas and Band Pro Munich handles the rest of the world, for sales.

There are several companies that are about to jump in to rent the camera. The cameras will be arriving this weekend.

The camera captures raw data at up to 1536 X 1024 at 1000fps or smaller images at higher speeds up to 10,000fps.

The data then needs to be transferred to the processing computer and can be rendered then or later for use as an AVI or Tiff sequence in post.

You can build a "look" from the data when rendering to intercut with HD or Film footage, but this matching is done as a post process.

It is ASA 64 and daylight rated.

You do not have camera controls other than start and stop the control is done in the computer.

The basic camera function and image has been available for a while, what Cine SpeedCam is, is a re-functioned re-featured version of a machine vision system that has been in use very successfully in the industrial machine vision industry (car crashes etc).

All of the tests I have seen and been involved in have produced very nice images that were filmed out or intercut with HD. We will have demo material available soon for viewing.

Some of the advantages over high speed film cameras is image resolution, near realtime playback for immediate feedback/review, no film or film processing needed (saving many thousands of whatever currency you trade in)and it is silent in operation.

More to come

Hope this helps

Michael Bravin
Chief Technology Officer
Band Pro Film & Digital

US  www.bandpro.com/products/hdtools/cameras/cinespeed/
EUROPE  www.bandpro.de/news/de/553


Michael,

do you know if it's going to be available to rent in Spain?

Pol Turrents (ACTV)
DoP Spain, Barcelona


Michael Bravin wrote :

>The camera captures raw data at up to 1536 X 1024 at 1000fps or >smaller images at higher speeds up to 10,000fps.

And the images at NAB were quite beautiful.

Can you get more record time running at slower speeds, say 128 or 256 fps?

Jeff "hey, it's single-sensor, isn't it? What about all my
DigiPrimes?" Kreines


Pol

Contact Gerhard Baier for info on who will rent it to you in Spain. I believe Ludwig Kameraverleih Gmbh in Germany will/does have one to rent.

Michael Bravin


Michael,

If, at some point, an NTSC demo reel becomes available I'd love to see it, as would many others here. I don't think I'll be able to make it down there for the demo.

64 ASA... wow. Why does it need so much light?

Art Adams, DP [film|hdtv|sdtv]
San Francisco Bay Area - "Silicon Valley"


Art wrote :

>64 ASA... wow. Why does it need so much light?

Art

Compared to what?

Michael Bravin


>Compared to what?

How about, "Compared to most modern things that record images."

I'm not sure if you've kept abreast of modern technology but there are film stocks that require half, or a quarter, or even an eighth as much light for an adequate exposure. Even old HiDef cameras only require about a sixth of that much light to create a decent picture.

So this time when I ask "Why does it need so much light?" I'd prefer a technical explanation instead of "Compared to what?" which seems a bit of a snipe. I think it's reasonable to ask why something so apparently technologically advanced needs more light than any other electronic camera on the market today.

Additionally, I'm curious as to why it's daylight rated. (I assume that means its native color temp is daylight.) I assume that belies its industrial roots, but there may be another answer.

Art Adams, DP [film|hdtv|sdtv]
San Francisco Bay Area - "Silicon Valley"


Art Adams wrote:

>So this time when I ask "Why does it need so much light?" I'd prefer a >technical explanation instead of "Compared to what?" which seems a >bit of a snipe. I think it's reasonable to ask...

So do I but...

Is it 64 ASA at 1,000 fps?? If so, then that's not too bad at all IMHO.

After all, 500 ASA film is 16 ASA at only 800 fps.

Jim Sofranko
NY/DP


>After all, 500 ASA film is 16 ASA at only 800 fps.

As I read the post 64 ASA is given as a general sensitivity gauge and is not related to a capture speed.

Art Adams, DP [film|hdtv|sdtv]
San Francisco Bay Area - "Silicon Valley"


Art Adams wrote :

>As I read the post 64 ASA is given as a general sensitivity gauge and is >not related to a capture speed.

Are you saying that it's 64 ASA at any frame rate?? or at 30fps as a base?

That's what I'm questioning.

Jim Sofranko
NY/DP


Michael Bravin writes :

>up to 10,000fps.

Awesome. Will a hand crank be available?

Dan Drasin
Producer/DP
Marin County, CA


Art Adams wrote :

>So this time when I ask "Why does it need so much light?"

Was that ASA 64 at 1,000, or ASA 64 at 10,000?

Steven Gladstone
New York Based D.P.
www.gladstonefilms.com
East Coast CML List administrator


Art Adams writes :

>"Why does it need so much light?"

Hey Art,

You shot any film at 1000fps. Did you need much light. Was it hot under those lights?

I suspect it's daylight balanced to be used with strobes.

Brian Heller
IA 600 DP


>You shot any film at 1000fps. Did you need much light. Was it hot under >those lights?

Yeah...but a chip doesn't change in sensitivity when you run the recording medium faster, does it? Or do the time slices decrease significantly when you run the recording medium faster? How much control does one have over those slices? The Varicam always records at 60 fps and the chip doesn't get any less sensitive at higher speeds. (They do adjust the shutter for you but you can work around that if you want.)

Michael didn't say anything about 64 ASA at 1000 fps; he implied it had an overall sensitivity of 64 ASA in the same way that we would say that a film stock had a sensitivity of 64 ASA. When I shoot 5218 at 96fps the EI label on the can doesn't change to 125, so I'm assuming he's describing an ASA (or EI) that describes how sensitive the chip is, period.

Defining an ASA for a frame rate seems a little odd, and I don't think that's what he's doing, but if he is it should be noted that there's a huge difference between ASA 64 at 24 fps and ASA 64 at 10,000 fps. We don't know which one he's talking about, and when I asked I felt like I got a snappy comeback instead of a real answer. An EI of 64 is the exception and not the rule these days, especially when it comes to electronic mediums. How many other HD cameras are there with an EI of 64? And how popular would they be if they weren't able to capture extreme frame rates?

>I suspect it's daylight balanced to be used with strobes.

Or 18k's. It makes a certain amount of sense, but I'd still like to know the reasoning behind that decision.

Art Adams, DP [film|hdtv|sdtv]
San Francisco Bay Area - "Silicon Valley"


> Was that ASA 64 at 1,000, or ASA 64 at 10,000?

Having had the chance to shoot it and it seemed to me to be 64ASA @ 30FPS

So that comes out to 2 ASA @1000 FPS I did have 6 1K's at 3 ft, 4 were Full Spot. I was shooting White Milk with Creamer splashing into the bowl.

If it was 64ASA @ 1000 frames that would make it 2000 ASA @ 30 FPS, Less Likely

But there is some partitioning going on where the ASA and time may not be working as we conventionally know it with a Film Camera.

The better question here is what is the Pixel Count at 1000 FPS and again at 10,000.

Best to Test it for yourself.

B. Sean Fairburn SOC
Director of Photography
Castaic Ca


Since everyone is guessing and making stuff up and guessing what I inferred or implied or really meant let's clearly state the FACTS...

The ASA is 64 at ALL capture rates from 1-10,000, it just is. This is a feature of how this system operates. It is daylight rated because it was designed to be used in a daylight application or with strobes capturing car crashes for industrial use, we repurposed it and accessorized it. (PL mount for instance)

It isn't film doesn't work like film and isn't constrained like film to require more exposure when the film moves faster. It is designed to replace high speed film cameras and the millions of feet of film run up in Photosonics cameras.

There is no recorder so those issues are nonexistent it captures to RAM and dumps to a hard drive. The RAM and the hard drive do not care or involve themselves in the decision as to what the capture rate is. They can handle 1 frame and 10,000 frames only the pixel count goes down above 1000fps.

Perhaps some of you have a little trouble embracing and accepting new technology without "fitting it into" what you already know, try something different here. We will expose you to what this camera is designed to do you will take it and create with it and express with it and create something new and exciting.

The introduction of the Cine Speed Cam is an exciting new technology and no amount of talk or conjecture will improve it or make great images. You need to see it work and test it and then discuss it's merits and its weaknesses, please, just this time here on CML humour me and wait till you have used it or at the very least seen it before CML'ing it to death.

Thanks for your interest

Michael Bravin


>up to 10,000fps.

Dan Drasin wrote:

>Awesome. Will a hand crank be available?

Dan

Actually at 10,000fps it requires 10-12 people all cranking as fast as they can to keep up with the pelicle mirror and the magnesium flash powder

One other thing while I am at it...

I would think that when we introduced the Zeiss DigiPrimes and told you that no lens makes a better picture, and then many of you tested and shot with them and found out it was absolutely true, I would think I gained some credibility here on CML.

Trust me on this one too. D'ya think maybe?

Michael "isn't really snippy in person" Bravin


There are several hurdles with film based high speed cinematography:

1/. Creating enough light for the exposure. This is obvious higher speeds = more light. However, many high speed cameras, rotating prism for instance, have a limited prism aperture of f:5.6, thus considerably reducing the effective ASA. (Just to put this in
perspective, 500ASA @ f: 5.6 @ 1/20,000 sec requires approx 20,000 footcandles.)

2/. Co-ordinating the event you're trying to capture with the camera, or the camera with the event, and knowing that you "got it".

3/. Closely related: co-ordinating the event within the capacity of a roll of film. 400' 16mm = 16,000 frames 10,000 fps = 1.6 seconds Allowing .6 sec for the camera to get up to speed, leaves approximately 1 second. (It is not practical/possible to accelerate longer {heavier} rolls of film to anywhere near these speeds.)

What's the point? A man comes up to me (on CML) and says he has a camera that basically solves all of these problems, but it's only ASA 64.

To me this is like a guy coming up to me and saying he has a car that runs on water, but only gets 15 miles per gallon. I don't really see that as much of a problem.

Let the discussions begin. Marquis of Queensbury rules. No biting, no gouging.

Brian "So's your old man." Heller
IA 600 DP


Brian Heller wrote :

>To me this is like a guy coming up to me and saying he has a car that >runs on water, but only gets 15 miles per gallon. I don't really see that >as much of a problem

Unless you are trying to cross a desert.

Steven Gladstone
New York Based D.P.
www.gladstonefilms.com
East Coast CML List administrator


Brian Heller writes :

>To me this is like a guy coming up to me and saying he has a car that >runs on water, but only gets 15 miles per gallon. I don't really see that >as much of a problem

Steve Gladstone :

> Unless you are trying to cross a desert.

How so? When you can drink the fuel.

Brian Heller
IA 600 DP


Brian Heller writes :

>How so? When you can drink the fuel

...and if the car filters the "recycled" water after you drink it you could keep on driving... lol

John Babl


John Babl writes:

>And if the car filters the "recycled" water after you drink it you could keep >on driving... lol

Another excellent point. Of course some might prefer to drink fresh fuel at that stage.

Brian Heller
IA 600 DP


>require more exposure when the film moves faster.

From  www.cinespeedcam.com

1536 x 1024 pixels at up to 1000 fps = 1.572 Megapixels
1024 x 768 pixels at up to 2000 fps = 0.786 Megapixels
768 x 512 pixels at up to 4000 fps = 0.393 Megapixels

Dynamic range :
10 bit monochrome / 24 bit color depth

The laws of physics are maintained after all, twice the speed equals half the pixels.

I'm relieved for one.

Jim Houston
Pacific Title Imaging


Jeff Kreines writes:

>John "picture in In Camera this month" Babl wrote:

>And if the car filters the "recycled" water after you drink it you could keep >on driving...lol

With big thanks to David Heuring and Mike Brown(from Kodak)!

To keep on topic, is there a pic of the Cine Speed cam we could see? I'm a big fan of Photosonics (the movement/12 pulldown arms/4registration pins on the 4ER is incredible).

I'd love to see comparisons on high speed film/numérique cameras(pardon my French lol)

John"dealing w/record labels...what a nightmare"Babl


Pics... http://www.cinematography.com/index.asp?newsID=158

Nick Hoffman ICGDP NY


Dammit !

>numérique cameras(pardon my French lol)

Now *there's* the name for future CML - 2K+ 444 list: CML-Numérique

Classy...

Sam Wells


I did a music video a while back that used the camera.

It is an amazing piece of gear. Great image quality, easy to use.

Only problem is that it does need a lot of light :

http://www.ironwolf-legends.com/drivethru/dayofthedead/P6130093.JPG

http://www.ironwolf-legends.com/drivethru/dayofthedead/P6130097.JPG

http://www.ironwolf-legends.com/drivethru/dayofthedead/P6130100.JPG

The above is what happens when a high speed stunt scheduled to happen during the day, happens at night.

But the resulting images looked great.

Kevin >"I was the gaffer on that show, so don’t yell at me about the schedule" Zanit


Nicholas Hoffman writes :

>Pics... http://www.cinematography.com/index.asp?newsID=158

Wow, I thought it would be a big bulky camera! Nice

John Babl


John Babl writes :

> Wow, I thought it would be a big bulky camera! Nice

Yes, but the crank is huge.:):) :):) :):) :):) :):) :):) :):) :):) :):) :):)

Michael Bravin wrote:

>Actually at 10,000fps it requires 10-12 people all cranking as fast as >they can to keep up with the pelicle mirror and the magnesium flash >powder.:):) :):) :):) :):) :):) :):) :):) :):) :):) :):)

:):) :):) :):) :):) :):) :):) :):) :):) :):) :):) :):) :):) :):) :):) are

courtesy of Steve ":):) :):) :):) :):) :):) :):) :):) :):) :):) :):) :):) :):) :):" Gladstone

Brian ":):) :):) :):) :):) :):) :):) :):) :):) :):) :):) :):) :):) :):) :):)" Heller
IA 600 DP


Jim wrote :

>The laws of physics are maintained after all, twice the speed equals >half the pixels.

Some laws not all the pixel count is 1536 X 1024 from 1 to 1000

BTW here is a better picture

www.bandpro.com/products/hdtools/cameras/cinespeed/cinespeed.jpg

Michael Bravin
Chief Technology Officer
Band Pro Film & Digital
www.digprimes.com


Michael Bravin wrote:

> BTW here is a better picture

www.bandpro.com/products/hdtools/cameras/cinespeed/cinespeed.jpg

The camera looks fine, but to me, that big box just screams DALSA!

Jeff "likes Dalsa, likes BandPro, likes 'em all" Kreines


>The camera looks fine, but to me, that big box just screams DALSA!

I would bet there is some corollary of Moore's law that applies to
boxes.

Dave Stump ASC
VFX Supervisor/DP
LA, Calif


Is there a chance to see a frame obtained from the camera at 1536 X 1024 1000fps?

Pol Turrents (ACTV)
DoP Spain, Barcelona


David Stump wrote:

>I would bet there is some corollary of Moore's law that applies to boxes.

In this case, it's perhaps Michael Moore's law...

Jeff "it takes a big guy to make a big guy joke" Kreines


Jeff wrote :

>The camera looks fine, but to me, that big box just screams DALSA

The "big box" is the rendering station, a rack mount PC a hardware capture and hardware/software rendering platform and a 20" LCD computer monitor. It is MUCH smaller than ANY other 1536 X 1024 1000fps camera I am aware of. DALSA is Canadian The Cine Speed CAM is German and is much more closely related to the ARRI D20 as a matter of actual fact.

I get the sinking feeling we need a CML-what it's not list now

Michael
"likes new stuff, doesn't complain, asks lottsa questions"
Bravin
Chief Technology Officer
Band Pro Film & Digital


Its too good to be truth.

Manny questions need to ask.

How come just from no were, single cmos chip can replace the job of film stock.

Shooting day is : lighting time and cost (64ASA daylight).
Shooting time to capture the action (very limited ram maybe 8sec in 1000fps).

Downloading time from Ram to the computer (can take minutes).
Very poor playback quality you need to wait few minutes to se it.
To render the shooting day can take hours.

How match, can you save on shooting day?

Quality of the image: Can it hold a nice Dynamic range. Many questions only test will answer that.

I hope to do test in a month.

Benny Mali (DOP, DIRECTOR)
High Speed Production


Michael Bravin wrote:

>I get the sinking feeling we need a CML-what it's not list now

Aw, Michael, it was just a joke about big boxes. don't take it personally!

I did say some very nice things about your single sensor, PL mount, hard-disk recording camera!

Jeff "handtruck optional" Kreines


Brian Heller writes:

>at 10,000fps it requires 10-12 people all cranking as fast as they can to >keep up with the pelicle mirror ...

But pellicle mirrors are fixed. Does that mean the whole camera is moving at high speed, with 10-12 people pedaling after it as fast as they can?

Dan Drasin
Producer/DP
Marin County, CA


Benny Mali wrote:

>How come just from no were, single cmos chip can replace the job of >film stock

Benny

This system is not from nowhere and not to replace film.

This is a high speed camera that offers a compelling option in terms of resolution, costs, speed, and on set playback, to high speed film cameras. It also will provide high speed electronic images for HD productions. It is for film and HD productions.

The camera is made by a company in business since 1946 (imagine how cutting edge this camera was back then)
http://www.cine-speedcam.com/

Hope this helps. Thanks for your interest.

Michael Bravin
Chief Technology Officer
Band Pro Film & Digital


Benny Mali writes :

>How come just from no were, single cmos chip can replace the job of >film stock.

I've been involved with high speed work since 1972. Mostly film based.

R&D on high speed video/motion analysis was well underway at that time. There were several high speed video systems available then. Many more between 1972 and now. Many more are available now. They are of extremely high quality and capable of extremely high frame rates 100,000 + Incidentally, Kodak was a pioneer in this field. I'm sure Mr. Pytlak can furnish the appropriate websites.

The difference between the other video systems and the CineSpeedCam is that this system has been modified and adapted to allow it to work more easily with current commercial/feature production techniques as opposed to scientific or industrial photo instrumentation.

> Shooting day is: lighting time and cost (64ASA daylight)

This compares very favourably to ASA 500 film in high speed cameras with limited apertures.

>Shooting time to capture the action (very limited ram maybe 8sec in >1000fps)

Maximum shooting time with 16mm at 1000fps = 16seconds, then reload. Maximum shooting time with 35mm at 1000fps = 7 seconds

Do you know of a 35mm camera capable of 1000fps with a similar run time?

>Downloading time from Ram to the computer (can take minutes). As >opposed to reloading a film camera.

>Very poor playback quality you need to wait few minutes to se it. How >long do you have to wait for your film to come back from the lab?

>To render the shooting day can take hours.

Again what's the turnaround time for film.

> How match, can you save on shooting day?

Doesn't that very much depend on what you're looking for. I have been involved in many filming situations where we did not get what we were looking for after an entire day of filming, and didn't know it for a couple of days.

If the peak motion event you are trying to capture, say for instance a golf club striking and compressing a golf ball, occurs in an extremely brief interval --1/10,000 sec --and you are shooting at 1000fps the odds are you're going to miss it most of the time, if not all of the time.

On the industrial/scientific side, days and weeks can be spent preparing an event for high speed analysis.

>Quality of the image: Can it hold a nice Dynamic range.

I doubt it can do any better than an ordinary HD video system.

>Many questions only test will answer that.

That is absolutely the case.

One final note, R&D in high speed video is going on at a very rapid rate all over the world. I don't know of anyone currently putting any sizable amount of R&D funds into developing new high speed film cameras.

Brian Heller
IA600 DP


The Phantom v9 system & (By reading) the Cine SpeedCam features, its looks that they have some similar functions and features in common.

Benny Mali (DOP, DIRECTOR)
High Speed Productions LTD


>This compares very favourably to ASA 500 film in high speed cameras >with limited apertures.

Brian makes some good points in regards to this system and it's comparison to film.

The Photosonic high speed systems use a tremendous amount of film and are limited by the film stock ASA and frame rate. I have been on many shoots which required several days of high speed shooting simply to get one shot. Often the action is hit or miss on high speed shots because of the limitations of control over gravity or chance. This camera system sounds very promising for many applications.

So I have several questions about the system to help understand it's use and/or limitations.

Is there a fixed shutter angle or capture rate which can help determine the crispness of the image at high speed?

Is there a limitation as to the amount of hard drive space in acquisition which may limit the amount of shots one can store during a shoot day?

Are there any strobing issues similar to film that need to be addressed ?

What are the lens choices for the camera?

At what frame rate does the resolution become unsuitable for film out or for TV out?

How many techs from Band Pro are needed to go out with the system?

I think the 64 ASA is a huge lighting advantage over the film systems in shoots of 1000 fps or more. My experience with high speed shots with over 1000fps has been mostly tabletop. 64 ASA will provide greater depth of field and less heat with lighting. Both are the common problems one encounters in shooting high speed shots.

I, for one, am looking forward to seeing this camera in action. It sounds very promising. Are there any in NYC?

Thanks.

Jim Sofranko
NY/DP


>closely related to the ARRI D20 as a matter of actual fact.

And since that demo was quite impressive...

John Babl


Brian Heller writes :

>at 10,000fps it requires 10-12 people all cranking as fast as they can I >have an adapter for my Mitchell GC, an external gearbox, that lets a >user handcrank at 128 fps -- which would ordinarily be (at 8 frames per >rev) 16 turns per rev -- pretty fast.

Rather than adopt this approach for the Kinetta (which will, with
faster sensors, be capable of greater speeds than our current 60 fps progressive) we gearbox electronically -- so you can just set the number of frames per crank revolution in a menu. Not exactly rocket science, that one...

Jeff "cranky revolution" Kreines

PS : I was very impressed with the Cine Speed Cam.


Jeff Kreines wrote :

>I have an adapter for my Mitchell GC, an external gearbox, that lets a >user handcrank at 128 fps --

Jeff, were those gearboxes fairly common accessories?

And Photosonics also has some new high speed cameras
http://www.photosonics.com/Sales_HTLM/Phantom9.htm

I'd like to see their new demo if it's available -

John F.Babl
Miami


Now look, this has just gone too far, we have facts from Michael Bravin and a reasoned argument from Brian Heller.

It's got to stop!

Lets get back to conjecture and opinion.

Also closed minds and intolerance for another's viewpoint or approach would be good.

Facts! who needs 'em?

Cheers

Geoff Boyle FBKS
Director of Photography
EU Based
www.cinematography.net


Yes, Kodak Spin Physics near LaJolla CA was a leading supplier of high speed video motion imaging cameras for scientific and defence research applications. They also supplied high precision magnetic heads for video and data tape recorders:

http://www.sprayresearch.com/capabilities/ektapro_1000.htm

http://web.umr.edu/~boom/facilities.html

http://www.bmius.com/img/thumbnails/20010725193309.jpg

John Pytlak
Eastman Kodak Company
http://www.kodak.com/go/motion


Sean Fairbairn wrote :

> I did have 6 1K's at 3 ft, 4 were Full Spot.
> I was shooting White Milk with Creamer splashing into the bowl.

This brings us back to the color balance issue. The 6 tungsten 1K's (four of them spotted) started cooking the milk right away. We saw smoke within minutes. Strobes or even HMI's wouldn't give off as much heat.

Alan A. Hereford * Cinematographer
Marin Co., CA USA


>...The 6 tungsten 1K's (four of them spotted) started cooking the milk >right away. We saw smoke within minutes. Strobes or even HMI's >wouldn't give off as much heat.

Yes, but HMI's can't be dimmed.

I prefer to work with tungsten for high speed shots because they can be instantly brought up for the shot and dimmed between takes.

Just my 2 cents...

Jim Sofranko
NY/DP


>I prefer to work with tungsten for high speed shots because they can be >instantly brought up for the shot and dimmed between takes.

Hey Jim, that's a great idea.

We were using whatever was at hand for our impromptu test. Six 1K's instead of one 6K. That is how we ended up dropping an individually wrapped creamer into a bowl of milk.

Who knew the flash point would be so low?

Alan A. Hereford * Cinematographer
Marin Co., CA USA



Copyright © CML. All rights reserved.