Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Underwater Bubble Blowers User Group Thread NEW

Status
Not open for further replies.
Now we have an underwater optical test bench allowing for precise alignment of charts and repeatable testing of optical performance using SFR and other charts underwater. Using industry standard Imatest, we can measure MTF, geometric distortions, chromatic aberrations, in any point of the image frame.

Housings of just about any brand and make can be tested, compared and evaluated in a non-subjective, measurable, repeatable and comparable manner.
 
Tom Hamilton said:
Ross, you owe me a new keyboard, just spit my coffee all over it, that was very funny :-).
Click to expand...

Tom....c'mon---must have been a beer no? I have to get out the windex for the beer spray....btw, in mexico city shooting (of all things) a music video and have to say i miss the water!
 
Paul Russell said:
Hey guys, I'm shooting an ad off the coast of Sabah next weekend, with a Deep Epic and a bunch of Arri Ultrprimes. We're not going to be filming any deeper than 10m and I'll have 4,000w of lights on the housing. Any advice on filtering? My thoughts are to shoot RAW and fix any coloring in post.
Click to expand...

Water is a very effective colour filter of about 0.12 red. This means that you lose 1 stop of red for every 2.5 m. Even if powerful artificial lighting is used, still colour andjustment is necessary. If the lighting is mounted on the camera you lose 1 full stop of red with only 1.25 m from the subject. This is because light travels from camera to the subject and back.

From this you can work out the strength and type of the colour filter required.

full article on the basics of underwater optics (that I wrote a decade ago) is here: http://achtel.com/Optics/
 
When shooting in air, the focus plane is a flat surface. Filming a perpendicular wall and focusing at the centre would make the whole wall appear in focus. However, when shooting through a dome underwater, the surface of focus becomes convex when looking from the lens. Therefore, the same flat wall would be out of focus at the corners when shooting underwater. At small distances the focusing surface approximates to a sphere. Consequently, the subjects that are in focus in the corners would be closer to the camera than if they were in air. To reduce this phenomenon it is important to increase depth of field by stopping down. By increasing depth of field the effect becomes less apparent. To increase depth of field it is best to set the aperture to f/16. This requres good lighting conditions. Shooting through a dome port with large apertures (f/2 or f/4) will cause difficulties in focusing close subjects on the entire area of the image. Pawel Achtel
Click to expand...

No filters, you need all the light you can get going into the lens, left-justify the histogram with just enough iris to keep out of the noise floor. The rest happens in post. Just my opinion.

It actually was beer Johnny, ya got me there.

Music vid huh? How is over there on the dark side? :-)
 
Tom Hamilton said:
No filters, you need all the light you can get going into the lens, left-justify the histogram with just enough iris to keep out of the noise floor. The rest happens in post. Just my opinion.

It actually was beer Johnny, ya got me there.

Music vid huh? How is over there on the dark side? :-)
Click to expand...

As usual, you are incorrect. You have clearly misunderstood the concept of colour balance versus exposure. You also did not seem to comprehend the question in which he clearly stated he would use 4kW of light at less than 10m depth.

I also ask you and the rest of the butt massaging comeettee to stop trolling or I will ask Jim again to take the thread down. Respect is a two way street.
 
We are all friends here Pawel.

It would be interesting to hear from Paul Herrin, just having finished coloring Johnny's upcoming documentary, on the post tools available to the professional colorist for correcting RAW underwater footage.

Paul?
 
It's academic, really, as we're flying off tomorrow at 6 am and it's past office hours here at the moment. I've already advised the client to go no-filter, for the simple reason that we'll be shooting between 5-10m depth and so the differences will be quite variable from shot to shot (not knowing how good the talent's bouyancy control is), so adding in a colour filter is only complicating the problem. I would think that 2-3 stops of red can easily be put back in post in R3D, with much more subtlety than a filter in front of the lens.

Client also wants a 'narrow DOF' Canon 5 MKII look, which is going to give me more problems than colour balance. Especially shooting in shallow water where we're all likely to be moving around a lot in surge/swell/tide/current etc. Some of these shots are tight close-ups, too, so I'll be using lenses up to 50mm, and still trying to focus through a mask on a 5.0 LCD screen. Fun fun fun. :-)
 
For what is worth,

I never understood the aversion, though I fully respect it, of many professional shooters not to use filters. I, for one, am advocate number one of putting as little glass as possible between the image itself and image plane inside the camera, and it kills me when I see topside cinematographers using IMAX and 5K cameras and super expensive Master Primes to get the best image possible and then put all kinds of filters on that matte box, each of which brings with it the risk of light diffraction, glares. AND DUST. And if you never paid your dues by removing dust and sh..t from digital scans at a studio, you never know how much fun you've missed. But, then, it is quite a different thing underwater.

I've been using some custom made filters for the last few months and I'm very happy with them. These are very expensive, experimental units with not only a color hue but also several coatings, but you could get indistinguishably similar results from the old standing filters like UR Pro. Even in the first 10 meters/ 33 ft layer, with lots of sun, I can ASSURE you that you will loose a lot of red light. And fixing in post will only take you so far. RAW is RAW in as much as to the light that hits the sensor, but if a certain spectrum or wavelength is not present in that light by the time it hits the sensor, there isn't much RAW can do to help. A good filter will help correct a lot of this problem, and ironically, it will work best in shallow water, for even the best filters will be rendered useless once you pass 60 ft. I know you loose a stop of light, but then, with a camera that is natively rated at 800 ISO, and can be pushed much farther without much noise, who cares really. I mean, we used to shoot stills on film on Velvia, Kodachrome or even Ektar, 50, 64 and 25 ASA respectively, and for the life of me, I never had much trouble finding enough light for a proper exposure anywhere shallower than 40 ft. I know a lot of this reluctance comes from the fact that, especially with a RED camera, you would have to use the filter in front of the dome port, far from ideal, or use a screw-on on the lens itself that cannot be removed for surface shots. When, or if, RED ever comes up with a filter wheel design iteration, where you could have a filter that flips in front of the sensor, then this would be a lovely solution. But if you have a lens that can accommodate back filters, and there are a few great underwater lenses that can do that, mostly in Nikon or Canon mount, not PL though, I encourage you to do some tests and see how much better color rendition and flexibility in post you get from a filtered image as opposed to the unfiltered one. The differences in noise floor on all channels, not just the red one, are undeniable. And, your blues will be much richer too with that extra magenta...Then again, that's me.
 
BTW,

Being a colorist myself, I'm in the process of recording a series of color correction tutorials on underwater footage for a well known software maker. If anybody here has any RED footage they'd like to contribute, I'll be glad to use it and give them credit for it. The release of those tutorials is a few months away, but I would be glad to post some workflow recommendations right here on this thread if anybody sends some footage my way. I cannot use our own footage for anything outside of these official tutorials, but can work on other people's footage as long as it is not for sale or distribution. So, if you have anything interesting, and which actually offers some possibilities for correction, not something shot in 300 ft of water in the middle of the night, then PM' me and let's talk.
 
Tom Hamilton said:
We are all friends here Pawel.

It would be interesting to hear from Paul Herrin, just having finished coloring Johnny's upcoming documentary, on the post tools available to the professional colorist for correcting RAW underwater footage.

Paul?
Click to expand...

I hate to say it, but this may be one of those situations where the 'fix it in post' mantra comes in due to physical limitations. The details are really a Graeme question, but I'll try to get us started and maybe he can step in. Having just gone through all that footage, I think Johnny and I can both testify that some elbow grease and a little magic fairy dust can really bring some of the muddiest images to life. In that way we're not only able to make reparations, but actually transcend the physical limitations of our own eyes to peer into the deep. Developing & Post-Processing images continues to be important whether you're working with chemistry or mathematics.

I'd rather see properly exposed blue and green channels than have all 3 underexposed and noisy. As Rudi brought up, probably only going to be of any benefit at all to filter close to the surface or in a very controlled environment like a tank or pool. With or without filters and with or without RAW, unless you actually have red wavelengths hitting the sensor, that channel is usually garbage. Basic order of operations... if you put a big blue (water) filter on, don't expect to boost the reds back up by adding another filter in the chain - the light's already gone. All you're doing is subtracting the other wavelengths to 'match' and thus creating a very unnatural scenario that likely doesn't help the debayering process and may also cause you to underexpose. It's very important that the algorithm uses the most correct color-temperature information - and in my experience, whether underwater or on land, excessive color filtering can mess that up sometimes. I'm assuming that Graeme has thought about underwater conditions in his debayering algorithm, but that'll be a question for him. If not, it seems to be doing pretty well interpolating that information anyways. You just have to get the white balance well-set when you process it. Depending on conditions/depth/etc, we were typically seeing kelvin temperatures between 5000 and 9400+ with tint values going +-20. Averages were found between 7,500 - 9,400 and tints usually +-8.

All that being said, if there is ample light to properly expose the red channel with a slight filter - it could potentially help the debayering process and make up for any inaccuracies such a shift might cause in the color science. And I certainly wouldn't complain about a nice, clean red channel to start with. If all of that is in order, and you happen to have the luxury and flexibility to pull it if necessary - I would encourage you to try it out and see what happens. In the real world, I only see any potential value in shallow water and controlled tanks. I'd be happy if that theory could be proven false.

If in doubt I'd say fly without filters. The raw files offer a lot more flexibility for post than other systems, but without naturally red wavelengths hitting the sensor, filtering may not be of much help. If anyone is up to doing some testing I'll be happy to assist.
 
Filters are only ever going to reduce light, never increase it. Underwater you're in a light limited situation (I think, correct me if I'm wrong as I don't dive) so I'd recommend going without filters and balancing in post. As we don't set up the colorimetry for "underwater", you may want to experiment with how you achieve white balance - either using the kelvin and tint as normal, or leaving them alone at a neutral 5600k setting and adjusting RGB gain.

Graeme
 
thanks graeme. any chance there could be some underwater magic in future releases, or are there way too many variables? as i mentioned, your interpolations are doing a pretty decent job already...
 
I've got the feeling that anything we do on improving image quality is going to help in all the adverse environments that the camera is used in.

Graeme
 
Status
Not open for further replies.
Back
Top