The Gentleman's Underwater Bubble Blower Thread

[Greg F]

I'll get you 6K footage Rudi as soon as my Epic is upgraded.
Cheers!

Same Same, I am unsure what water you are shooting in Michael? I will be dragonizing the clear waters of Hawaii, and sometimes the Murk Murk of Australia

 

[Michael Maes]

Same Same, I am unsure what water you are shooting in Michael? I will be dragonizing the clear waters of Hawaii, and sometimes the Murk Murk of Australia

I'm based in Grand Cayman Greg so: +120ft visibility and low plankton level... almost like diving/filming in a swimming pool )

 

[Pawel Achtel]

All,

Further to my post, a few more observations regarding this issue:
....

.... I am praying for a miracle with the Dragon, I really am.

Thanks Rudi,

I have some ideas what may be causing this.

Let me do some investigations.

In order to fix any problem we need to understand it first, quantify it, find a solution and measure it again. Only then we can be sure that we fixed it.

We should have pretty good idea what is happening in about 2 week's time.

Stand by...

 

[Michael Maes]

Thanks Rudi,

I have some ideas what may be causing this.

Let me do some investigations.

In order to fix any problem we need to understand it first, quantify it, find a solution and measure it again. Only then we can be sure that we fixed it.

We should have pretty good idea what is happening in about 2 week's time.

Stand by...

You've got my attention Pawel!! )

 

[Greg F]

I'm based in Grand Cayman Greg so: +120ft visibility and low plankton level... almost like diving/filming in a swimming pool )

Well I guess he will have two separate levels of ridiculous visibility with low plankton shot almost on opposing sides of the planet.

 

[Pawel Achtel]

That's your problem. You guys should be shooting in less than 5m (15ft) visibility and water so dense from plankton you could use it to make fish soup

 

[Greg F]

LOL.. Then you cant see the wildlife very well and then backscatter is a evil bastard!

 

[Jeff Whitehurst]

Hope this helps
Jeff W

 

[Rudi Herbert]

Jeff,

I don't know, though there is a difference in colorimetry, that could be attributed to subtle variations between the temp/tint values of both settings. Saturation is also lowered on the Production sample. I did several tests after you suggested to me that setting white balance during shooting in fact would have an impact on these problems but I couldn't really see a difference that I couldn't match later on in post. To assume that WB settings do make an impact during shooting is to assume the camera is not really shooting in RAW, kinda like the Canon RAW where WB is burned in, but I don't think this is the case. Truth is, a color chart can be made to look good, but the noise ad degradation those changes introduce are most of the times unacceptable. I remember those sample R3Ds that were posted on the Epic thread, about the whale shoot for the Nat Geo gig, and where many people took a crack at them and got some nice warm tones on the whales, BUT the water and everything else around had huge amounts of chroma noise, the kind that is either near impossible to fix or requires complex windows/masks around objects like the whales in this case to prevent them from getting too soft with all the denoising manipulations. Pushing colors out is not that difficult, doing it so that the whole scene remains in balance is the trickier part.

 

[Jeff Whitehurst]

No, actually the temp/tint values are identical in the two samples I posted from the test. My point is simply that WB in production and "fix it in post cause itʻs RAWʻ arenʻt the same thing. I would also point out that many RED UW shooters, including the poster with the trouble whale footage and your colleague with the "bad Epic", didnʻt even record a white/grey reference card in the scene, which places them in real trouble in post. Setting WB in post with no reference and attempting to correct with a compromised color palette, now thatʻs tricky.

 

[Rudi Herbert]

No, actually the temp/tint values are identical in the two samples I posted from the test. My point is simply that WB in production and "fix it in post cause itʻs RAWʻ arenʻt the same thing. I would also point out that many RED UW shooters, including the poster with the trouble whale footage and your colleague with the "bad Epic", didnʻt even record a white/grey reference card in the scene, which places them in real trouble in post. Setting WB in post with no reference and attempting to correct with a compromised color palette, now thatʻs tricky.]

Jeff,

Please don't take this to be confrontational or adversarial, far from it, I'm trying to see if I can understand your rationale so I can learn something new. But you've lost me there. White balance is a matter of tint and temperature, combinations of which produce different values at infinitude. Of course you can arrive at the same exact color balance combination by manipulating other things instead of temp and tint, curves, HSL, etc, etc, etc, but then that value still corresponds to a defined temp/tint setting. Thus, if color balance in your footage simply reflects the existing light conditions AND the way the camera interprets them, but said values are not burnt in, then if you change them in post later your footage will look the same as if you set them up before or during shooting. That's not me, that's the mechanics of real RAW.

Now, if for reference's sake you prefer to set up a WB value that represents the actual conditions while you're shooting, or record a reference such as a grey card, that's understandable and you're welcome to do it. BUT this does not guarantee that you will be able to achieve the values that card is telling you. I am not defending the people who shot the whale footage, but my friend with the "bad epic" did things just the way I would have done them, and the way I've done them for many, many years. An d I assure you I've gotten many jobs due to my ability to color correct underwater footage. Which does not mean that I can't be wrong of course. But recording a grey card in anything deeper than 10 ft or extremely clear water in a studio tank without color bias is a waste of time simply because you will never have enough information on your recorded image to balance your channels to match what the grey card is telling you to. You may come close enough, and when you know how to work with underwater images, you will figure out ways to use the information you actually have to come close enough to the color balance you want, but that's about it really. It is simply not feasible to adjust temperatures to values of 100 000 and tint to -70 as most software would ask you to do when you shoot a grey card in 80 ft of water with ambient light with an overcast day, no camera records enough information to allow you to push its footage into that territory, that's a fact. And that's not just me. There are many professional underwater cinematographers that shoot reference cards at the head of each clip, and others who don't, and usually those who don't have a better understanding of color correction besides the shooting aspect of it. And while I will praise your thoroughness for shooting a grey card on your dives, and while I will not bother with it, neither method will change the fact that WB in RAW is simply a mathematical function.

 

[Greg F]

Jeff

I am looking at the coral in the background and i am seriously questioning if thats 70ft... I have never seen those coral species at 70ft.

Also if you are ever on maui would love to do some experimentation so you can see my results?

 

[Tom Hamilton]

My preference is to correct in post, and once I achieve the look and feel I'm after, load that look into the camera for subsequent dives in similar conditions. What I then see on the monitor underwater is close to what I'd like to eventually deliver to the client. Also picks up nuances one might miss otherwise.

All very depth/light/viz/etc - specific of course, but my library of looks is growing, and with my housing, am able to change on the fly, as necessary.

 

[Johnny Friday]

Shooting reference chart:
Here's my take and how i do this or don't do it. If on a job where most of what we shoot is on sticks or with minimal distance to subject and plenty of ambient light and finally an assistant and (but not a must) artificial daylight rated lights....then i would and sometimes to use a reference chart. But it is by no means a deal breaker and often enough i find myself drifting away from any charts other than a burst chart to perform focus which i think is much more important.
As pointed out by Rudi, as depth increases and we loose light, then a chart really ends up doing no good and values jump all over the place like 20,0000 wb and -70 on tint or something crazy---often just not enough information to get a proper WB. But, if at depth and using lights, then a chart can work out fine. However, as pointed out again....as you spend more and more time behind the drivers seat color correcting, you find a chart is so often not necessary. Well....i do. That said, i almost always and exclusively in the past shot with a reference chart...this always helped in in some way to see what values or settings would look like if shooting a chart 3 feet from the camera. Now, the flip side is the chart might look great, but the back-ground does not look natural which i would say is often the case. Again, not enough information to get a proper reading...less light, and all the other factors that play underwater. Therefore at the end of the day you are left with judging what looks good an an all round look.

I also don't do as you do Tom...save looks for different depths, light values etc....or rather i don't save them to use diving and setting the camera. My feeling is there are far too many variables that come into play and look at the water and say i need look #256 for light blue Caribbean water with small particles....What i only use looks for now are when i am coloring a reel on the same day and same dive....then save looks from one clip to the next. The reason is simple. all it takes is for the tide to change to change the water conditions or a cloud to pass over or you name it.....

I typically recommend to new red shooters to shoot default settings always, then change them in post when coloring. I've yet to find any other way to shortcut or make this easier....and as Rudi pointed out....hours and days and weeks and months or years working on images in post will reveal a lot of things.....and jobs will come due to this ability to correct shots on location for producers.

 

[Jeff Whitehurst]

To be clear, the color chart is only a reference for this test to show that the colors are more accurate straight off the camera when I set WB to the other side in the scene. I donʻt normally even shoot that chart, itʻs only a visual reference for this test.

@Rudi, donʻt let the conversation be construed in your mind as anything negative, Iʻm trying to learn as well. I reread your posts several times trying to understand your position regarding color correction with Epic footage, and thereʻs a lot there I completely agree with. However, I noticed you never once brought up WB. Iʻm just baffled how RED shooters Iʻve talked with are claiming they never set WB in the field with RED because itʻs RAW, and double down on this approach by not even recording a reference to use in post. Then, they get real frustrated when the footage "canʻt be fixed".

Perhaps you are misunderstanding what I posted. The image labeled "WB in Production" is exactly how it looks when I open it in RCX. The one labeled "WB in Post" is what I get when I use the RCX eyedropper tool to pick the white value from the recorded image. The resulting tint and temp values are at this point identical, I set WB in post and got the exact same temp/tint values as I would have gotten had I set WB in the field. However, the histogram clearly shows the color channels are not "the same". The third labeled "No WB" is what someone gets when they donʻt set WB at all, which is a bad place to be if they donʻt have a recorded reference, yet thatʻs the situation I keep seeing RED shooters faced with due to a complete reliance on post production to set colorimetry.

@Greg, No trickery here, I assure you I made these images minutes apart at Honaunau "Two Step" at 70 feet, and youʻre right the coral population there is really unique. As for diving, anytime buddy

I typically recommend to new red shooters to shoot default settings always, then change them in post when coloring. I've yet to find any other way to shortcut or make this easier....and as Rudi pointed out....hours and days and weeks and months or years working on images in post will reveal a lot of things.....and jobs will come due to this ability to correct shots on location for producers.

@Johnny, I must be misunderstanding what youʻre saying. I shot the image labeled "WB in post" with daylight default settings, and the image is inferior to the one shot with WB Auto Calc. Are you saying that you would not set WB in the field or post, but rather grade without setting it at all?

 

[Rudi Herbert]

No, of course I'm not advocating shooting "without" WB, a proper setting is a must to help you shoot and get images on the monitor that somehow resemble what you have in front of you. But at the same time, were I to set two Epics at vastly different settings, say Epic 1@ 2000K/+50 and Epic2 @ 15000K/-60, I should be able to achieve the same results in post provided I know what I'm doing. No footage will be worse than the other, or so the theory goes with RAW, which Redcode really is no doubt about it. I may have said this before, but I find that with RED footage, optimal temperature settings for ambient light shooting will be between 7000K and 9000K, and tint between -10 and -30, with anything over 9000K/-30 really degrading the image as a whole regardless of water clarity and optimal conditions. Again, for ambient light shooting, lights is a different animal. BUT, when you have problematic footage in your hands (strong magenta and/or green casts), I find that WB settings even within the parameters I describe can really nuke the images into fugliness easier than any other manipulation. In those cases, I will sometimes work with temps of 4000K to 6000K and tints of -10 max and be able to extract some more colors in other ways, with curves and Hue-Saturation-Lightness and lowered local contrast, i.e: negative alchemy values, etc.

My biggest point of concern over all this is that I've shot film, 8, 16, 35 and 65mm, and any flavor of HD cameras and, provided my WB settings were proper since those werent RAW cameras, I have never once encountered the colorimetry anomalies we are encountering with the Epic so frequently. If this were common then it would be what it is, but it really isn't the norm with any other camera I've used, down to crappy camcorders, thus it is all the more surprising and disheartening that it happens with the mighty RED.

 

[Johnny Friday]

T
@Johnny, I must be misunderstanding what youʻre saying. I shot the image labeled "WB in post" with daylight default settings, and the image is inferior to the one shot with WB Auto Calc. Are you saying that you would not set WB in the field or post, but rather grade without setting it at all?

Hey Jeff: I actually have or had no reference to your photos...but what i had meant or intend to say is i stick with default values: 5600k WB & 0 Tint when doing 99% of shooting. Then when i take those images into the computer to do any color work, THEN i change those values. Point i was trying to make is that those values change each time you point the camera in a different direction or change depth. So really there is no reason---or I find no reason to put time into setting my WB & Tint underwater at all....in fact i think it is mostly a waste of time since i can accomplish my finished look at the computer. I find that setting WB & Tint underwater does not help or assist me in anyway when color correcting later.

....sorry if i was not clear on that. Now, if there is some proof otherwise, i'm all ears. But i do think at times having a reference (certain shots/scenes) comes in handy. And that i find mostly with talent and/close up work. But even then, a skilled colorist (underwater colorist) can whip that out fast. It just takes time to see what settings are usually needed to do this....

My almost normal settings for getting grades to look warm are 10500 WB & -10 to -20 tint.....so a bit different from Rudi, but same idea. All of this does change when adding lights though.

 

[Michael Maes]

Hi Jeff,

The Epic records RAW, which essentially means every 'bit' (0/1) which the sensor produces by capturing light is recorded 'as is'.
Whichever setting you chance in the Epic happens after the image has been recorded.
Even changing your ISO doesn't change a thing for the recorded image: every Epic-X (or M) always records at 800 ISO, the baked-in value.
So, in essense, every change in setting you make, only reflects to the monitor output.

Additionally it is good to know the every change to settings you make is stored in a separate file, a 'meta-data' file called RMD.
This file helps the editor (or you) to see how you intended to shoot a particular frame.
In fact: in RED CINE-X you can reset each parameter to either the recorded value (through the RMD file) or to the camera default value.

So, if you could shoot two identical frames but one with WB X and the other with WB Y, the two RMD files will be different but the R3D image files will be identical.
I realize this sounds weird, especially for the ISO (try changing the ISO on camera, your histogram won't change), but that's the way RED created this: record the clean sensor-data before manipulation
Does that make sense? Well, it took me a while ;-)
Cheers!

 

[Jeff Whitehurst]

Ok, I give up. I posted test samples that show that the results are not the same if you WB in the scene or in post. Pretty simple to see the difference.
The image recorded is not 1 to 1, there is a debayer process as well as compression at play.
I'm not a liar, and not a noob, sorry it didn't help you guys, it helps me greatly. Cheers, Jeff

 

[Michael Maes]

Hey Jeff,

No one says your a liar! Certainly not and your input is greatly appreciated!

It is true that if you WB underwater or not will give a different result in what you see when you load the images in RED CINE-X.
That is an indisputable fact!
The thing however is that there is a big difference in what you 'see' and what's 'recorded' by the Epic.
If you take your WB-ed shot in RCX and reset every setting with the 'D' button, then you will see what the Epic actually recorded.

So, technically, it doesn't (or better "shouldn't") make a difference weather you WB underwater or not.
There is nothing wrong in doing so as it does not affect the actual R3D files.
But it can certainly help you getting a better idea of colors (on your monitor) while shooting.

Personally I'm in the "Johnny camp": I never WB the Epic, nor do I pre-load settings in the Epic for exactly the same reasons Johnny states.
But that is merely my personal workflow and doesn't make me a better (or worse) shooter.
On every other camera I do WB before every shot when shooting ambient.
Cheers,

Michael