Underwater Bubble Blowers User Group Thread NEW

[Johnny Friday]

Well, takes me about 5 minutes to change lenses on my Gates R1 or Epic housing....If i'm sleepy...but if on the move, probably closer to 2mins....But then i'm usually shooting a wide 10-17 fisheye...or might be on a 16-85 for all round work....so i may not need to change lenses as you suggest Pawel....from your 15 to ? ......So, your suggestion of taking an hour to change lenses is kind of a stretched TO SAY THE LEAST...maybe it could take an hour to change lenses clean the mirror and re-calibrate a 3d underwater rig....if that's what you might be referring to...but even then, hasn't taken me that long either....unless say working in snow or rain and water or other elements get on the mirror--then i'm inclined to agree with you--ok, that point is shot down...so maybe keep that one in your pocket a bit because it just doesn't fly....

I'll bet MOST of us are shooting R1 & Epic and delivering for television....95% of what i do I'll admit is just that or stock. AND shooting 2:1 I'll center crop almost EVERY image to get some very crisp images.....rarely if ever do i deliver a raw full size image...it's cropped and delivered as 1080p. So there's a LOT of room to work images. That's just a plain fact. And this is also an area that will set many shooters apart. Those that will tweak images into the morning hours....and those that will do a simple transcode to the desired deliverable codec.

I was going to continue on here. but i know the outcome....But just wanted to point out a few fair replies to your comments pawel.

 

[George Evatt]

Hi,

The IMATEST charts that Pawel posted here yesterday comparing sharpness etc of a 15mm Nikonus and a Zeiss 14mm prime behind a dome were not really fair as the Zeiss' nodel point was not aligned with the domes centre point. We are gong to do these test again next week with the correct alignment and see how it goes.

Regards

George Evatt
Sydney, Australia
www.coralseatv.com

 

[Pawel Achtel]

... Wider is better, without a shadow of a doubt....

Sharper is better, without a doubt

Well, takes me about 5 minutes to change lenses on my Gates R1 or Epic housing....If i'm sleepy...but if on the move, probably closer to 2mins....

Thanks Jonny for pointing this out, Yes I exaggerated in respect to some housings but still it takes some 50 times longer than with Nikonos system. I thougt the adjustment of position of the entrance pupil going from one lens to another would be time consuming. I guess this varies from housing to housing and, in case of a beam-splitter housing I imagine the time taken to do a lens change would be an hour if not more.

 

[Michael Hastings]

Diffraction limit aside, next test I'm going to do will be at f/4.0 as, in my experience, this is more realistic f-stop when shooting in deep water, particularily at higher frame rates.

Maybe we need to send you some brochures on the dozens of spectacular dive areas in the world that have clear water and interesting things to see in less than 100 feet (where the vast majority of serious underwater filming is done, and you couldn't get to F4 on an Epic without major ND)

Well, just as putting a Nikonos lens on an underwater housing is nothing new or revolutionary, neither would a corrected optical system be.

I smile when I read the incessant rant Pawel is crusading about, that a dome and lens combo don't resolve past SD quality, when in essence that's what the 15 mm Nikonos is, an old 20 mm lens housed under a dome port that properly calculates the lens nodal point and adds correction/magnification in the areas where is needed. Didn't you ever crack a 15 mm against a coral? I did, and afterwards, my secret dream of taking the lens apart and inspect its components came true, and what I found was nothing short of simple. The dome is not regular glass, it is magnified all over and it thickens slightly on the edges, increasing the magnification to fight off the corner distortion. Beautiful in its simplicity. Something as logical, but somewhat more complex is what lies behind the Fathom optics, which are wonders of design and optical quality. Unfortunately, they are designed for zoom lenses and 3 chip setups, which makes them large and heavy, but something similar to a Fathom could be made for a zoom like the 14-24, or better yet, a prime like a 14 mm or 15 mm. Not having to take into account the 3 chip design and instead making the image plane converge onto a flat area such as for a single sensor camera like a RED would make such an optic relatively easy to build and not any bigger than existing HD Fathom designs. I have actually contemplated the possibility of sinking some money into having some such prototypes built for me in China, where optical glass fabrication is (surprise) very cheap, but have held back on it. I have info on fabricators and pricing which are relatively doable, what becomes a little expensive is the design of the optics in question and the testing that must be performed. But again, nothing exorbitantly prohibitive and well within the financial scope and technical acumen of MANY companies and individuals out there.

How nobody at this stage and age has undertaken this relatively costly and difficult, but totally reachable, task and come out with the first truly revolutionary underwater optical system for cinema cameras in many decades is really beyond me. I hope Mike is not blowing smoke up our rear ends...

Well, some of it is out of my control so not 100%, but I've never been a smoker, and I think it's about 98.8%

BTW I never broke a Nikonos 15 but I suspected the dome had some correction in it - it would have to to get the kind of results Pawel has shown.

A simple diopter probably provides some correction since the distortions would be in offsetting directions.

I also suspect that you could make a lens optic that would go on the camera lens - similar physically to a thick diopter but with graduated curvature (might need to be aspherical?) - that would do the same as the inner surface of the Nikonos dome and could take it a lot further toward correcting the edges and could be used with multiple lenses - just never got around to pursuing it because there is a limit to what this market will support in terms of volume and price versus the time/effort to get it done. It's probably something for an optics specialty guy that likes diving to pursue - and right now for us the time is better spent on the full "Fathom-type" optic.

Go wide underwater or shoot macro. Wider is better, without a shadow of a doubt. Nikonos 15mm is not wide enough.

Sharper is better, without a doubt

Sharper AND wider is better without a doubt but so far it is a trade off. 4K resolution of a subject that has been diffused and desaturated from shooting through an additional foot or two of water generally isn't as good as the same with better clarity, contrast, color saturation, and some loss of pure resolution. As I have said before, I salute the idea of the Nikonos on a 36x24mm full frame sensor at 94 degrees FOV (these days we are used to significantly wider but 94 is close enough) but 75 degrees maximum on Epic 5K or 6K dragon with Nikonos lenses is just not enough. If it were, John and I would probably both be selling a Nikonos front plate option - it is pretty easy to do, although you still have the serious issue of any leak in the lens or around the mount of those lenses goes right into the sensor of the Epic. And those lenses and mounts have been known to fail - not such a big deal with a $300 nikonos camera body, a lot more so on a $30K Epic .

Not to mention getting it approved by RED so it doesn't void the camera warranty since it is effectively a lens mount.

Has Jarred approved your setup? - I think that does require some full disclosure if you are offering something here on Reduser that is going to void the warranty.

 

[Pawel Achtel]

[

...I smile when I read the incessant rant Pawel is crusading about, that a dome and lens combo don't resolve past SD quality, when in essence that's what the 15 mm Nikonos is, an old 20 mm lens housed under a dome port that properly calculates the lens nodal point and adds correction/magnification in the areas where is needed. Didn't you ever crack a 15 mm against a coral?...

Sorry, as a scientist I try to assess the optical performance of lenses by measuring their MTF, not by cracking them open against a coral.

Your statement is entirely incorrect, Rudi.

The 15mm Nikonos lens (among other characteristics) is specifically designed to eliminate image plane curvature and other defects caused by domes. There is no way for any land lens placed under such tiny dome (or even a large dome) would produce sharp pictures in the corners.

You can actually make the Nikonos 15mm lens to focus in air. Just increase the back focal distance by about 1.20mm and terrestial objects come into focus. The problem is, that it has massive inverse plane curvature, distortions and chromatic aberrations in air that I haven't seen in any 20mm lens. The Nikonos 15mm was specifically designed to produce sharp and undistorted images in water and, unlike land lenses placed behind a dome port, it produces perfectly flat image with no soft edges.

It is not so with terrestial lenses placed behind dome ports.

To illustrate image plane curvature caused by a dome port, I have placed a 5k Epic frame grab of SFR chart taken with Master Prime 14mm through 8" Aquatica dome port at f/2.8. The lens has been perfectly aligned with the dome.

Indeed, as pointed previously, the corner sharpness does not even reach standard definition. As you stop down it becomes better before going worse due to diffraction, but it never reaches even half of the Nyquist frequency limit of the sensor. The Nikonos 15mm does.

centre and corner 1:1 crops are below.

Of course, flat ports perform even worse.

 

[Pawel Achtel]

Maybe we need to send you some brochures on the dozens of spectacular dive areas in the world that have clear water and interesting things to see in less than 100 feet (where the vast majority of serious underwater filming is done, and you couldn't get to F4 on an Epic without major ND)

No, thanks. I like my underwater housing to take sharp pictures at any f - stop.

Particularily at apertures allowing the maximum sharpness offered by the camera senor and not only those that actually limit the sharpness due to diffraction.

The whole excitement about filming in 4k or 5k is to actually be able to use the optics that can take advantage from such incredible sensor resolution, not cripple it with optics that cause significant degradation of the image.

 

[Michael Hastings]

No, thanks. I like my underwater housing to take sharp pictures at any f - stop.

Particularily at apertures allowing the maximum sharpness offered by the camera senor and not only those that actually limit the sharpness due to diffraction.

The whole excitement about filming in 4k or 5k is to actually be able to use the optics that can take advantage from such incredible sensor resolution, not cripple it with optics that cause significant degradation of the image.

I agree with all of that. The part you keep ignoring is that in most circumstances your solution of using a lens that gives only a 75 degree FOV creates more problems than the corrected corner sharpness solves.

 

[Frazier Nivens]

I have a Nikonos 15mm for sale if you want it. The newer style.

 

[Frazier Nivens]

BTW I'm at NAB right now. The show is all 4K. The next technology wave. If you've not seen a 4K panel showing quality 4K images, it is...well....stunning.

Wanted to be at NAB this year but had a job filming. I've yet to see 4k imaging shown on a 4k panel. I think Clark Dunbar from Mammoth HD might have submitted some uw imaging on his reel. Next time. Thankful to be working, diving and filming. Enjoy the show.

 

[John J Ellerbrock]

Wanted to be at NAB this year but had a job filming. I've yet to see 4k imaging shown on a 4k panel. I think Clark Dunbar from Mammoth HD might have submitted some uw imaging on his reel. Next time. Thankful to be working, diving and filming. Enjoy the show.

Yeah, next time Frazier. Looking forward to a visit.

Good health,

J-

 

[Michael Hastings]

Here are some clips that Richard Hall shot a couple weeks ago of two olympic swimmers in an olympic type pool. The lane lines and cross hatch lane markers provided probably the most extreme "resolution chart-like" subject you will ever see in a real world situation. We shot this on epic 5K with an old 90's version of the Sigma 14mm F3.5 with Optitek Nikon mount, (since my Canon mount hadn't arrived yet) through an AquaVideo 8" glass dome. Nothing special was done in terms of nodal point positioning, lens optimization, etc. as this was put together pretty quickly - I just picked a set of holes on our mount plate I thought would work OK, put in two controls to match iris and focus and we went and shot. (Also This was just a quick cut/grade Richard did to show the swimmers. - I will try to get some .r3D frames to post later.)

It is only vimeo but while Pawel would have you believe that anything shot on Epic underwater through a dome will be horrible - it seems that this indicates it isn't so bad.

And again, we are working on even better solutions, but the current setups are quite good in the real world and one should question the validity, motivation, and possible bias of chart tests that could be designed to make the competition look bad - since virtually every lens/optical system has distances, f-stops, etc. where they don't perform as well. Not trying to make any accusations - just saying that not everyone else that makes housings or shoots pictures underwater is FOS.

https://vimeo.com/40817727

password: blue

Thanks, to Richard and Gary Hall from the http://www.theraceclub.net/ for letting me post this.

 

[Pawel Achtel]

I care about picture quality and "isn't so bad on Vimeo" is not the kind of benchmark I like to compare to. I like my pictures to hold up on a large screen.
It is my understanding that no housing other than DeepX can resolve even standard definitionin on the entire frame. DeepX resolves 5k corner to corner (possibly more). It's just that simple.
I think that having a choice of hundreds of lenses and none of them actually producing sharp images, it is like having no choice at all. Standard definition is not sharp enough for me.

But, it is not just the sharpness. Domes introduce field curvature that changes with the distance. I like flat, undistorted and even look of rectilinear lenses.

 

[Michael Hastings]

But, it is not just the sharpness. Domes introduce field curvature that changes with the distance. I like flat, undistorted and even look of rectilinear lenses.

We understand that which is why we are working on true optical solutions, with the appropriate super wide FOV necessary to get good clarity and color saturation - which is what your use of the Nikonos 15 on Epic is lacking with only a 75 degree FOV which is NOT what the people at Nikon designed it for - it was designed to have a 94 degree FOV on a 36mmx24mm frame. And even 94 degrees is considered a bit narrow by the standards of the past 20 years.

 

[Pawel Achtel]

Michael, you are incorrect.

Nikonos 15mm, while not specifically designed for S35mm sensor, is the only wide angle underwater optics able to resolve more than standard definition across the entire Epic frame. It actually out-resolves the Epic sensor corner-to-corner. Using the Dragon sensor it will cover over 82 degrees underwater and I expect will out-resolve the Dragon sensor at 6k too. This is more than ten times the sharpness available from any other underwater optics currently available.

On the other hand, terrestial lenses were not designed for underwater. They pose great challanges to achieve sharp and undistorted pictures and I haven't seen any evidence suggesting that any corrective optics would change this. This poor performance means that, while pictures may look acceptable when played on a computer, they will not be satisfactory on a large screen.

I do have and published the evidence that even a Master Prime lens is unable to resolve standard definition on the entire frame when placed behind an 8" dome port. And, this defect gets worse as the size of the frame increases. The image curvature caused by dome ports is also impossible to correct. This is because the image plane curvature depends on the distance of the subject from the dome.

I realise that there are different distribution formats, including the Internet, and I accept some people aiming for such formats. There is nothing wrong with that.

We believe that Red Epic deserves 5k underwater optics because it is a 5k cinema camera. We currently have a great solution producing spectacular wide angle images even in worst water conditions imaginable. It took us over a year of intensive research and development, considerable financial outlay and a great deal of testing. We are not making housings out of sewage pipes like some manufacturers do, we set the standards in precision machining, materials and superior engineering. Like RED we are focused on delivering products for professional cinematographers and accepting that not everyone cares about their picture quality.

 

[Johnny Friday]

Pawel, well stated...very well, but then i got to you last paragraph and you just had to be cynical. You begin pouring a nice cup of wine and then it turns to sour grapes....unbelieveable....you always seem to do this. Hell, i'm no manufacturer--merely a consumer, but so here you talking about "other" manufacturers and sewer pipes and then"not everyone cares" you just trash yourself. You want me or others to take you seriously as a manufacturer? good luck sir. I can imagine someone having an issue with your housing one day and calling back for service and they then get your attitude. Wow! how you persist. unbelievable. that will be what gets in the way of your success in the manufacturing field.....good product or not...

 

[Michael Hastings]

Michael, you are incorrect.

Nikonos 15mm, while not specifically designed for S35mm sensor, is the only wide angle underwater optics able to resolve more than standard definition across the entire Epic frame. It actually out-resolves the Epic sensor corner-to-corner. Using the Dragon sensor it will cover over 82 degrees underwater and I expect will out-resolve the Dragon sensor at 6k too. This is more than ten times the sharpness available from any other underwater optics currently available.

On the other hand, terrestial lenses were not designed for underwater. They pose great challanges to achieve sharp and undistorted pictures and I haven't seen any evidence suggesting that any corrective optics would change this. This poor performance means that, while pictures may look acceptable when played on a computer, they will not be satisfactory on a large screen.

I do have and published the evidence that even a Master Prime lens is unable to resolve standard definition on the entire frame when placed behind an 8" dome port. And, this defect gets worse as the size of the frame increases. The image curvature caused by dome ports is also impossible to correct. This is because the image plane curvature depends on the distance of the subject from the dome.

I realise that there are different distribution formats, including the Internet, and I accept some people aiming for such formats. There is nothing wrong with that.

We believe that Red Epic deserves 5k underwater optics because it is a 5k cinema camera. We currently have a great solution producing spectacular wide angle images even in worst water conditions imaginable. It took us over a year of intensive research and development, considerable financial outlay and a great deal of testing. We are not making housings out of sewage pipes like some manufacturers do, we set the standards in precision machining, materials and superior engineering. Like RED we are focused on delivering products for professional cinematographers and accepting that not everyone cares about their picture quality.

Ah, we are back to the sewage pipe again - full circle. (BTW it is fresh water pipe - the sewer pipe is green )

I/we are not suggesting that we accept lesser picture quality. My point was that you have published charts of tests YOU made; with the cheaper domes that YOU chose - the specific one which is not used on any Epic housing I know of; at the distances YOU chose - which can affect performance since different lenses perform differerntly at different distances. Your point about above air lenses not being correctable is ridiculous - the Nikonos is a 30+ year old above air design with a combo dome/corrective optic (you say the dome curved field is impossible to correct yet it is the water contact surface of the dome that creates the curvature and of course the Nikonos water contact surface IS JUST A DOME! - the correction is done behind it with other elements so obviously it CAN be corrected) . Modern wide angles with current glass formulations, CAD and CNC technologies that allow much more flexibility in aspherical elements, nano coatings, etc. are FAR superior to older designs. The current 14-24 Nikon zoom is better than the ultra primes and is pushing the Master Prime (AND covers full 36x24 sensors) and to suggest it can't be well corrected for underwater use flies in the face of common sense given what has already been developed for other lenses over the past 20 years.

I simply published a real world sample, which while only on Vimeo (SO FAR) looks better/sharper than any real footage I have ever seen from you. Further even the Redone footage that I have seen on large displays shot by Peppermill, myself, or Frazier Nivens (both Frazier and I also have Epic footage) looks way, way better than anything I have ever seen from you.

I have said before that there are some situations where you will be able to get some spectacular, extremely sharp footage with the Nikonos 15 on Epic at 75 degree FOV. But I will say it again - 75 degree FOV (or 82 on Dragon) is so limiting relative to the normal 95-110 degree FOV (that most u/w motion pictures use) that it will create more image degradations than the supposed resolution increase will benefit, in a great many (probably most) normal underwater shooting situations.

 

[Frazier Nivens]

The footage from the swim shoot looked totally awesome when I viewed and graded some of it in Red CineX Pro. I seriously doubt that a Nikonos 15mm lens and the 14mm Lens we shot on my Epic during this shoot could compare image wise. The 14mm lens gave us a beautiful image that the 15mm Nikonos lens couldn't have matched. Too bad we couldn't shoot the two systems side by side. The water better be crystal clear to get good clear imaging with the Nik 15.

Wider is better underwater, less distance between you and your subject allows for a clearer image. Less water between your subject and camera.

Michael makes a great housing.

 

[C.H.Haskell]

Agreed that shoot looks good in the pool, nice work...now where is my housing Mike!

 

[Tom Hamilton]

I recently came back from a Baja shoot. Inside my sewer pipe was an Epic kitted out with all necessary accessories, safely and protectively secured in a package that not only weighed within airline cabin bag limits, but easily fit into the overhead bins as well. No ramp rat damage, no extra bag fees, no luggage mistakenly diverted to Tokyo, $50K worth of kit within arm's reach at all points in the journey. Can't begin to tell you the de-stress factor in that, the absolute guarantee that you'll arrive at the job with everything you need.

On site, just a couple of standard scuba weights in specially designed holders on the housing, and it's perfectly neutral and balanced.

The rig even took great vids underwater, in fact spectacular ones. The client is happy.

That, to me, is an impressive accomplishment, something I've always wanted in my frequent travels.

Mike delivers. Count on it.

 

[Ross Isaacs]

Michael those swimming pool shots look fabulous the sharpness is stunning as your eye follows the lane seperators. You obviously know how to marry up lens with domes to make wonderful images.