Just rented an Ursa 12K and have noise/exposure questions...

Hey all.

Long time Blackmagic User (had a Cinema and Production cam) - who was one of the dreaded EARLY 4.6K owners with all kinds of issues (don't want to bring up a certain shade of red for starters).

I had a Helium for about 3 years - and grabbed a 6K pocket. I was so impressed with the results for quick set ups - that I have a major corporate gig Sunday - and we decided to go with the 12K due to the overcrank abilities (job has lots of firearms and is at a gun range). The last corporate 6K job I had to deal with horrible mixed lighting temps...shooting in hangers with overhead 2700 fluorescents while the hanger doors were open - sending shards of bright light in - and the camera performed like a champ considering.

Here's our issue.

I'm planning on using our Schneider Xenon Primes - and during EVERY camera test - exposure is WAAY off judging by false color and our waveform displays. Monitoring with the Extended Rec 709 on camera false color sent to external monitor.

Noise is off the charts - and the footage - when the same lut is applied in post - has a crazy crush to it.

Is there something we are missing in camera for these results?

We shot mainly in 8K full readout - as that's where we plan to be - lowest compression settings.

Everything we bring in is to the right for exposure - but can't be salvaged. We tried shooting in what usually is a dark room with horrid fluorescent overheads - and outside in and out of shade - with hard light banging the background to stress test (which I've used many times as it's in my office) - and also shot dusk the day before. We can't gauge anything.

This job will be mainly run and gun - almost all steadicam - daytime exteriors - and we're nervous. That's why we were trying to shoot stress tests; but - we can't seem to gauge it for normal shooting without tremendous noise.

The camera is a rental and we shoot early Sunday morning.

Things are clipped that were well below 90 percent on the zebras (according the camera's monitoring) .

Indoors we tried to expose to the left - only to find it totally covered in crazy digital noise (same as the outdoor stuff). If we go to right - the shirt has crazy noise and the skin clips fast.

There's so much crush with the extended video lut applied (which looks nothing like it from the camera out) - that I thought there was a hidden gain function or something crazy.

I created a Google Drive folder with a RAW clip - and 2 stills in log - with comparisons to the 6K run and gun shoot we did (which was in extended rec 709) - and the same lens on the RED (graded - which I shot with the same AC we did our setup tests on). All were monitored in false color and I had tremendously different results with the Ursa 12K.

What is super concerning to me - is how the low light shot - when in extended video - the wall in the back clearly reads 50 on the waveform - but is covered in noise as if we pushed it stops and stops.

I'm sure it's just something we are missing; but I'll be honest - totally nervous as we're just trying to find a safe exposure range to ride during the day - with decent dynamic range within a stop or 2.

Any help would be appreciated!

https://drive.google.com/drive/folders/ ... sp=sharing

I'd like to add - for the daytime shot - I graded the footage from LOG - to where it looked on our monitor and the camera on set - and uploaded that still - VS - the insanely contrasty "extended video" that looks crazy. The regular REC 709 looks almost like a sepia music video from the 90's. But heavy noise remains on the shirt.

- Mike

Sounds like you have "Apply LUT in file" enabled?

I do not - no.

Unchecked everything.

I wonder if the previous renter could have baked something into the settings?

This is driving us crazy!

I agree with newbie. What are you using for the remote monitor and is the remote monitor showing false colour just to judge the exposure or all the time?

Interesting you got decent results shooting with the BMPCC6K but the UNP12K is bringing you to your knees.

What ISO are you using? Don’t exceed ISO 800. Sounds like you are using 3200 and exposing for 3200. What shutter angle are you using?

What is your project frame rate and what will be your maximum off speed frame rate? Do you have the artificial light to manage the highest frame rate you want?

What recording media are you using and have you tested it at the highest off speed frame rate you need?

On the off chance the camera is at fault, did you verify with the rental
house that previously they’ve had no complaints about about the behaviour you’re seeing?

We all wish we were there to help you get this straightened out quickly.

Rick;

Using an external monitor - sending the camera's false color out - but using the monitor's waveform. We are using my 1st AC's TV Logic via wireless and splitting to a Feelworld for my viewing monitor (I'm directing). I guess I can also bring my Flanders in - but that's big -- and I trust it implicitly - but I usually reserve it for clients.

Shot everything ISO 800. Using lowest compression.

Project frame rate is 23.98 - highest frame rates we will shoot is 120 fps full frame lowest compression possible and possibly 240 in super 16 crop mode.

Did you check out the noise samples I uploaded? Also - if you just look at the stills - you'll see how insanely crushed the Extended Video appears VS having to adjust each shot normally - which looks closer to what we were monitoring - but is still very very noisy (especially the reddish shirt).

I also included the RAW shots in that folder.

We are super stumped. Our goal was to just load the camera onto the steadicam with like a 1.2 ND - and ride around 5.0 on our primes to keep them tack sharp. We aren't looking for "perfect" exposure for this shoot - just close enough and wanted to see how things looked in a "run and gun" environment. Right now - not good. We can also see quite a bit of IR noise using internal ND's - which isnt' surprising - so we just planned on using our Firecrest IRND's on the shoot - but were using the internals on tests.

I did not verify anything with the rental house. It was from an online company as no one in Florida even had the camera.

Thanks!

1.2.1_1.2.1.jpg (895.5 KiB) Viewed 8254 times

1.1.1_1.1.1.jpg (837.23 KiB) Viewed 8254 times

BTW: You have used a 324° shutter on the outside shot.

I didn't use the BMD Gen 5 extended video LUT from Resolve, but my own (made for the Pockets and color science Gen 4 - but works with Gen 5 too) and a bit modified BRAW settings:

Outside clip:
ISO 800 / 5173 K + Tint 0 / Sat 0.85 / Contrast 1.2
Play with the High/Shad Rolloffs to taste.

Added a denoise for chroma - with the UMP12k you can set it pretty strong without destroying anything:
Spatial denoise luma 0 / chroma 25 / Mode Faster / Radius large

The rest is done by my free LUT LBK-Neutral-v10 - get it here:
viewtopic.php?f=2&t=97623

Inside clip:
ISO 1600 / Exposure +1 / WB default / Sat 0.85 / Contrast 1.2

Plus denoising.

Interesting on the shutter -

We flipped into High Frame Rate mode for the shot before - then toggled it off - but the shutter angle stayed?

Hmm.

I will try your settings out to see what I get.

Thanks!

- Mike

Michael, just got up but need to leave; I’ll take a look at your files later. Based on what we’ve seen, from Robert, I suspect they’re useful.

I am concerned you’ll be able to get a good exposure at 240 fps as that’s more than 3-stops less light. But if those scenes are outdoors in daylight, then you could shoot that footage without ND 1.2 and a wider shutter angle.

You need to keep an eye on the camera HUD to check your shutter angle and overall exposure at the beginning of each shot even when on a Steadicam.

Hi Michael,

8k is likely a good choice if you anticipate lots of gunfire as it has a faster integration time (rolling shutter) and you're less likely to have issue with muzzle flash. Though it won't entirely go away.

I took a quick look. As I do most always I applied extended video in the RAW tab.

Your shutter is at 360 there for both shots.

For me to get anywhere looking like a "straight" exposure on the corridor shot I had to lift the exposure more than 2 stops. Not sure what your lens was at there exposure wise. Also the White point had to go even lower for me, which says a lot about the lighting in that corridor. Between the two stop lift and the 360 shutter, that's 3 stops of "miss-exposure" - for me....

The exterior shot is containing a lot of contrast and if I'm hearing you, you're finding that it's crushing ? As in too contrasty ?

I find in shots like this I will usually LOWER the exposure in the raw tab and then also lower the contrast as well. Then when I go to the primaries I find I've usually got the skin tones in a better place and I've elongated the roll off in highlights and the blacks won't feel so crushed.

The noise is the noise. Even on normally exposed shots there's a texture that's always there. Even a mild amount of NR can remove it, I usually go for an extremely low amount and it still cleans up well.

But I also haven't been using it as much as I used to. I used to find the sensor was "noisy" but I don't think it's as noticeable once you grade and render out and I think with such high resolution it's better to have a little bit of texture.

JB

Michael, I’m using HDR waveforms with Dolby Vision enabled;
here are my project settings:
Colour Management tab
Colour Science DaVinci YRGB
Timeline Colour Space Rec.709
Timeline Gamma Rec.2100 ST.2084
Mastering display
1000-nit P3 D65 ST.2084, Full
No LUT

Project Settings: Camera RAW tab
Colour Science Gen 5
White Balance Custom
Colour Soece Rec.709
Gamma Rec.2100 ST.2084 (PQ)

CLIP C001
Colour Page
Camera Raw tab
Decode Quality 8192x4320
Decode Using Clip
Colour Science Gen 5
White Balance Daylight
Colour Space Rec.709
Gamma Rec.2100 ST2084 (PQ)
ISO 800
Highlight Recovery On
Gamut Compression On
Colour Temperature 5500-10 or 6500-16
according to your desired look

Colour Wheels
adjust according to taste as John discussed.

CLIPC005 as above except
Colour Temperature 2850-0
Exposure +3

To deliver 4K, change the timeline to 4K DCI and the decode quality to 4096x2160. I used Deliver
h.264 multipass
Advanced Settings:
Force Sizing to Highest Quality On
Force Debayer to Highest Quality On
Tone Mapping Dolby Vision
100-nit BT.709, BT.1886, Full

Very quick adjustment to taste on the Colour Page, for the HDR Colour Wheels:
set Shadow Expisure to 1.00 or 1.50.

This will get you in the ballpark where you want to be.

As for noise, you can reduce as Robert and John indicated, but this is your ‘film grain’ so go easy so you don’t produce a plastic looking image.

As Howard is fond of saying “Good luck.”

Graded in Resolve 17.0b2
Apple Pro Display XDR set to HDR Video (P3-ST 2084) but works well as Rec.709.

Thanks guys.

Appreciate it.

John - yes - by "crushed" - I mean - the blacks are super contrasty and whites appear blown - totally unlike the same lut applied in camera when we were monitoring via zebras and false color.

Shoot is in the morning.

Obviously - we are going to go for better exposure - but it will be run and gun - and was just trying to figure out safety zones as to ride shots out - as it will be in and out of sun all day - with rain inbetween - so "good enough" will have to be just that.

Am playing with the other takes some more - and trying to get a handle on things.

From the looks of it - I should upgrade to Resolve 17.

Best!

- Mike

The noise is there and it looks too strong. It is not extremely low light situation and it looks like something is really wrong. Maybe it is something related to BRAW compression? Try do a Black Shading if this option exists on 12K model, maybe it helps...

Guys;

Finished our shoot - 1st off - THANKS SO MUCH for the advice. We did shoot 8K and for the most part - got everything we needed (always a challenge and weather has been iffy here in Florida).

However - our team did their best - but something is wrong here - with the camera's own internal monitoring. Unless we are all this off (and this was a commercial DP that shoots all day on the Gemini doing informercials/infomercials - and they also have a pocket).

Our setup was:

Camera 1: The 12K on his Steadicam with a5 axis gimbal attached. Firecrest 1.2 ND in the matte box and we used my 35mm Schneider Xenon.

Camera 2: Our Studio's Pocket 6K on sticks with a Firecrest .9 ND - matching Xenon 75mm.

We sent both images out to our external monitor with the overlays via the same lut (Extended Video) and went off the cameras' internal Histogram.

HERE IS WHERE IT GETS FUNKY.

Both cameras were set the same. 400iso (we set them for base on the 6K and were willing to give up some headroom on the 12K) - and just adjusted the iris brighter or darker on each camera (which usually wasn't too far from one another) to match the cameras' histograms to one another - flipping between the images.

NOW - in RESOLVE - when I bring in both cameras - they are VERY CLOSE color wise (with a bit more contrast on the pocket) - but the 12K is overexposed - and has to be developed at about 200 iso darker to match the 6K (which we were usually pretty close on - but have to go to 320 sometimes - in those shots - the 12K had to go to 160 or so VS 200).

AND HERE IS WHERE IT IS WEIRDER - the Loop monitor has it's own histogram - and on set - the Cameras' internal one matched the monitor's on the 6K - BUT - on the 12k did NOT. The monitor's built in waveform would show the highlights as over 100 - while the 12k's waveform that was being sent were always under 80. I thought something was happening during the loop - but, in hindsight, the 6K's matched the monitor's.

My 1st AC had a waveform up on his TV Logic (he was looping to my Feelworld with the Histogram) - and had his waveforms matching as well (according to the monitor) - with nothing really exceeding 80 (aside from the odd sky or small areas of ground hot spots). That did NOT match the footage from the 12K upon import - but did the 6K just fine.

I have created a folder here - showing the side by side development (same luts and color - so one is greener and a smidge punchier) - between the 2 and the different ISO's I had to use to get them to match (copy and paste grade with just an ISO change and slight contrast adjustment). DO NOT TRUST THE CAMERA'S FALSE COLOR or HISTOGRAM. It will burn you. My monitor was correct. Wish I had snapped a photo of the monitor - but wasn't thinking at the time. We had SO much to shoot - and were at the mercy of the sun with just bounces and silks. My brain just made a small note of it between the shoot chaos and said "Hmmm - that is wierd". We didn't have time to do dumps once the shoot started the way things were.

I also threw some fun Steadicam grabs in there shooting into flares at end of day that look nice.

It's all usable - just not much highlight retention at that ISO and a few things clipped (mainly on skin highlights on arms or shoulders) the camera said were there.

VERY STRANGE. And yes. The 12K is def nosier in the flare stuff - no question - but the Noise Reduction helps.

Pics here.

https://drive.google.com/drive/folders/ ... sp=sharing

OHHH - And we used a Juggler on set - but had mixed results. It would just turn off (we didn't push compression that much) - but the camera's record light would stay on. You had to look for a little exclamation point. So - we missed some things. We didn't shoot 12K - or push the overcranks that hard - and we did lower compression and faster speeds in 8K on our camera tests, By the end; the Juggler would freeze and we loaded a card. And that TOO did the same thing. Not an overcrank. Just inserts of the product (the headset).

This is a VERY weird issue - and the DP wasn't very happy when he left set upon seeing the clips in the raw player/in resolve. LOL. But he is now seeing the images correctly - considering our time; budget and tiny crew.

I took a look at the .jpgs. Good results in spite of the first impressions. I was looking forward to a few short original clips to grade! I actually made a video with the quick settings I described in my previous posts and was hoping I could just add a few samples from your actual shoot.

Some takes are long - due to the steadicam - and we let it roll - but I will see if I could post something we aren't using shortly.

This is a corporate gig - and I don't want to upload anything important - etc - before client gets an edit - but I'm sure there is something like a start stop - I could upload.

Best!

- Mike

Some takes are long - due to the steadicam - and we let it roll - but I will see if I could post something we aren't using shortly.

This is a corporate gig - and I don't want to upload anything important - etc - before client gets an edit - but I'm sure there is something like a start stop - I could upload.

Best!

- Mike

You can export a single braw frame from Resolve.

Are you aware that the UMP12k is using a newer color science (gen 5) than the PCC6K (gen 4)?
If you are on Resolve 16 than the extended video LUT is still on gen 4 - afaik. And not matching gen 5.

That's right - I forgot.

I just need to clear it's okay to share like that - as the product is in the shot - but I'm sure it's fine.

I am not completely sure, but I think only with Resolve 17 beta came the extended video gen 5

So - both cameras are in Gen 5 VIA resolve 16.3 Beta.

It is automatically applied by Resolve when I copy the initial grade from the 12K.

That doesn't explain the strange (correct) exposure discrepancy between my monitor and the camera's Histograms on set.

I've made a few RAW grabs - awaiting sign off tomorrow from my company to upload them.

Best!

- Mike

I downloaded few 12K clips shared Eugene Belsky here https://drive.google.com/drive/folders/ ... ffJCESixzM and i can see same rainbow noise everywhere in the shadows. Noise is almost gone when debayer quality set to 2K. In 4K and higher it became visible at 1:1 zoom ("Center Crop with no resize" project setting) as well as in downscaled to HD project.
At HD resolution image is perfect and mind blowing. Color science gen5 is great, colors and dynamic range looks nice in one click, image looks crisp and there is no any visible noise.
At 4K image is a little bit sharper and noise is became visible.






On Michael Hoffman's examples a lot of noise visible at the dark purple-magenta cloth. That color patch on ColorChecker usually the noisest with any other BM camera as well.



So i think this is more like perfect HD/4K camera, but not a real 8K in terms of noise. It is also not a camera that allow you to bring back 3-5 stops of dynamic range hidden in shadows, probably because the sensor dynamic range is already "overclocked" to maximum limits. I guess same image quality with dramatically less file sizes and processing power could be produced with HD/4K dual gain sensors with large pixel size. But participation in megapixel race is probably some marketing strategy...

By the way, same (and even worse) situation goes with RED Komodo example footage. It looks OK when downscaled to HD but at higher resolutions it is very very noisy in shadows.

I can definitely say the 12K image is noisier than our pocket 6K - without a doubt. And it probably is how the RAW data is allocating the data it collects (it is super punchy and saturated in Extended Video).

I've owned a Red Helium and Dragon - and while the dragon was noisy at base ISO - it was pretty clean at lower ISOs. The same for the Helium. When it originally shipped; the base ISO was 1280 for dynamic range - but the tradeoff was a "noisier" image. HOWEVER - the noise wasn't this saturation noise -- it was more like a desaturated version of film grain. The one thing about Helium was - in 8K - the image was NEVER sharp in resolve - and almost noisy/blocky - UNTIL EXPORT - despite highest debayer settings - and the recommendation from RED was actually to debayer at the 2nd highest settings - not top - for least noise if working on a 4K timeline. Go figure.

I will say though - that film gets noisy in digital codecs FAST. I'm cutting a Super 16/35mm feature - both are obviously "grainy" - but Super 16 gets crazy amounts of noise and grain in the blacks/blues/purple when scanned at 4K - especially anywhere there is a push (and you can't push 16 well at ALL - you actually should overexpose it almost 2 stops) - as the blacks are thin -- so all these cameras really are a blessing to work with VS film (remember - when you scan film - you are also inheriting the traits of the scanner's sensor as well as the texture of the stock and other stuff) - unless on newer Super 35 stocks (those are pretty clean).

The last feature I directed that went to cable (Showtime/The Movie Channel/Syfy) was almost a decade ago - but then we chose the Varicam 3700 over RED (as back then - the Red One would overheat) - and it was nice - but low light - and also noisy in our darker scenes. Trade off was shooting an ENTIRE feature at around 100 ISO. Base was 320 - and you had to use an adapter for PL lenses in the day - which had rotating mirrors to simulate film grain. So all our night shoots required an 18K - especially when our action scenes had a 90 degree shutter (that camera was $60K and didn't overcrank - so we needed the 720P 2700 HPX to do 60 frames on slow mo days). So we have come VERY FAR - and all these cameras offer better images than a $100K+ plus rig a decade ago. I would have rather shot that film on the new Pocket Cam - easily. So much less light. Such a nicer image. LOL.

HOWEVER - I will say that my Helium - after IPP2 and the ISO Calibration - where old 1280 was the new 640 -- looked great -- and was VERY clean at 400 ISO/320. Downside there was; loud fans during long interviews and set-up time. We had to black shade and save each Shutter angle and any major temp changes. Or that camera got lots of hot pixels. It was definitely a "sharper" image than all the URSA cameras.

Here - despite our monitoring issue and shooting waaay more to the right than we were told (still have no idea what is going on there) -- this camera definitely has noise and needs some minor post sharpening.

There is no perfect camera for sure - I've had noisy Alexa images land accross my desk -- but I would say that the saturation noise here is what's high. In Reds/Blues/Greens. Even at such a low ISO for development (200/160) - in our flare tests - the purple flares have chunk noise for sure in them as they cross his black shirt in the haze. I'm used to that being clean.

That said - the image is very nice on the camera; if anyone could figure out how to monitor the image (so far 5 out of 5 I've hired are stumped - and literally scared of the camera). Hoping it was just that rental. Not sure. I haven't called LensProToGo yet to see if there were any other complaints.

The good news is - I'm used to noise - and I still think just job looks great.

The downside is - the Schneiders I own do fringe on super highlights under 5.0 (especially the 35mm) - and that purple color is where the noise is super high on this camera. We tried to keep it at 5.6 just because we were on steadi for the sharpest images with some depth of field - knowing about the fringe (i generally shoot around it - KNOW your gear - LOL) - but we clearly were off.

I'm not digging the Komodo footage for some reason - and I have a $1K discount if I want it (I got burned on the hydrogen phone). I'd rather keep the pocket as a B cam despite hating it's ergonomics and weird weight distribution - the ports are a PAIN - in a terrible place and it's odd with bumps the sides of my gimbal.

Michael, perhaps you need to avoid using that Extended Video gamma in post. I thought I got better results with the settings I suggested for you earlier. When delivering, use tone mapping with the 100 nit ST.2084 setting which is SDR friendly.

I tested it with R16.3 on my workstation now and Extended Video gamma definitely does not fit the new camera - either use R17 or stay on BMD film gamma.

What I did now is to change the color science back to gen 4 and just apply my LBK neutral v10 LUT.
Keep gamut compression on. That works instantly well with your samples.

Along with gen 4, the other thing to try is RCM or ACEs. I believe Mr. Hook said they're both configured to handle the 12K footage correctly as of the 16.3 betas. Am not equipped here to try it, but it would interesting to see how the 6K and the 12K compared in that workflow.

Guys;

When grading I had to do Manually adjust the sliders VS using a present like Extended Video (it is just so crazy).

I created a folder on that drive called RAW Screengrabs and posted my corrections in a subfolder there. Used a go-to cinema lut on top - and made simple adjustments. These are just for proxies - but will note maybe there is a better way to expose.

Have fun!

Still would like to know how to accurately monitor(!) - unless I upgrade to Resolve 17 and it suddenly matches the camera's Histograms on set in Extended Video.

Oh - and it was a pretty rough day out. Crazy overcast - then sun would beam through - so white balances were always changing and funky. We did our best to silk and bounce for interviews - but there was a 4 man crew - and we couldn't fight the sun (or lack there of). LOL.

Robert - BTW - Thanks for all the help.

Do you have a version of that lut for the 12K yet - or am I using the Pocket 4K one I found on the forums?

Because - that already helped a great deal - just made things a bit off in color. But I saw sky come back I was having a hard time recovering.

Best!

- Mike

Robert - BTW - Thanks for all the help.

Do you have a version of that lut for the 12K yet - or am I using the Pocket 4K one I found on the forums?

Because - that already helped a great deal - just made things a bit off in color. But I saw sky come back I was having a hard time recovering.

My LUT was designed for the PCC4k - but seems to work with most BMD cameras on color science gen 4 it seems. I did some corrections specific to the PCC4k like moving very saturated colors like tail lights a bit down and towards orange (because camera makes it a bit too magenta). And a correction for greens in foliage to make it less brownish.

I used almost a hundred different real world clips for the PCC4k version and my experience in shooting with that camera for over 18 months.

It seems to work fine in general with color science gen 5 if the BRAW contrast gets increased a bit.
But for creating a version for the 12k I would need many more samples to test it on.

I think if you provide me privately with some samples of your shots I can use them to tweak the current LUT for a 12k version. It will need a stronger contrast curve and taking back some of the corrections I made for the PCC4k.

OK, here you go - a first shot:
LBK neutral for Gen 5 - Version 1

LBK_neutral_High-desat-Gen5_v1_33.zip

(351.21 KiB) Downloaded 179 times

The raw exposure should sit about in the middle of the scopes for this LUT.
It does not alter white balance (hence the name neutral), but amplifies an color cast if present.

Thanks!

Did you find the raw clips I had put up?

They were in the original folder - just down in the chain -https://drive.google.com/drive/folders/1p64pqFbYFDFhyU7p_HyfdczCa-JoZ0Oj

[CaptainHook]

Hi Michael,

I had a look at your frames, the following is just the new Gen 5 Video and Extended Video in Resolve 17 (with some minor raw adjustments here and there for wb/tint/exposure). The only other thing is a slight bit of chroma NR applied. UHD timeline with a 1:1 crop for the first image.

Gen 5 Video:


Gen 5 Ext Video:


1:1 Crop in UHD timeline, same processing as the Video version above but with some slight (luma only) sharpening (source file is 8K capture):


Gen 5 Video:


Gen 5 Ext Video:


Gen 5 Video:


Gen 5 Ext Video:


Gen 5 Video:


Gen 5 Ext Video:


Gen 5 Video:


Gen 5 Ext Video:


Gen 5 Video:


Gen 5 Ext Video:


Gen 5 Video:


Gen 5 Ext Video:


You can right click the images and view in new tab or download to view at full size.

I'd recommend using False colour and/or Zebras for exposure. The new Gen 5 log curve is much flatter so if you are looking at a waveform the clipping point will be much lower than a Gen 4 camera. Our camera histograms are typically scaled to show you clipping for the log curve, so comparing histograms across Gen4/5 cameras may confuse you, but using False colour and/or Zebras should not.

Thanks for sharing the frames Michael.

The G4 OFX CST gives lots of options for gamma/gamut compression, more than G5 in the raw tab. When everything is nice, the G5 to BM Video is practically one click, with clipped skies, I find more handles are better.

Good Luck

Grady.png (844.44 KiB) Viewed 7452 times

I've heard from multiple Ursa 12k users that it would help to have more in-camera metering options and controls. Like a full screen histogram. I'm in that group too. It's crucial to nail your exposure.

I hate having to bring extra gear on long and exhausting shoots. I'll often carry a light meter and do what I can with false color. Speaking of false color, I'd really like to be able to adjust it to my own code.

Captain Hook - Thanks for the info.

Those obviously look HEADS and TAILS better than what I originally was seeing in resolve 16.3 VIA extended video.

In regards to the curve - that's interesting.

We did try and use false color - But it wasn't the kind of shoot where we could leave it on - and after set-up we were at the mercy of the sun. The camera was mainly on steadi - so no one could look at that screen during operation and my AC won't pull to false color. I wouldn't either.

I find that works best when lighting talent - then toggling it off for the shoot.

We were definitely to the right of where we wanted to be - and yes - matched waveforms - which was probably a mistake - but had no idea that in extended video out - the camera would read it's data so much lower down the scale.

Appreciate it.

Appreciate the input, Howard.

And wanted to say thanks to Robert. The lut you passed did help tremendously as a starting point in 16.3 (not sure if I want to upgrade yet - almost done with a feature) and I got more out of the footage no doubt - especially saw SOME blue back in the skies - but that shoot was a mix of a rough day (almost all clouds - bright - in and out of storms) -- and us being about a stop or 2 from where I wanted.

Client is still VERY happy with the initial stills.

Best!

- Mike

[CaptainHook]

The G4 OFX CST gives lots of options for gamma/gamut compression, more than G5 in the raw tab. When everything is nice, the G5 to BM Video is practically one click, with clipped skies, I find more handles are better.

There is no support for Gen 4 Film gamma from the 12K camera in the CST plugin (or RCM) currently (nor any support for Gen 5 yet although that information has been sent to the Resolve team), so while you may be happy with the results just be aware its not "correct". You might be better off setting the RAW tab to linear gamma and using linear gamma in the CST plugin if you want it to be "correct".

Or with the Video mode you could also try the highlights control in the primaries tab which is the same algorithm as the DNG raw panel (the shadows and highlights slider in the DNG tab were never "raw" adjustments so they are actually a double up of the primaries tab controls). Video mode allows 9 stops over middle grey before clipping (3 are mostly just for specular highlights) so its still very 'grade-able' as well.

I would also suggest trying Ext Vid and playing with the custom gamma controls in the raw tab to try control the highlights, and if you prefer "accurate" colour for 709 monitoring then set the colour space to 709 (which is what Gen 5 Video in R17 is) and adjust the saturation in the gamma controls to taste.

I feel like I may be the oddball and never touch the BMD>Video/ExtVideo(etc) LUT option.

I adjust my images with very little manipulation of the Raw tab and really utilize the full color grading features of Davinci

I agree Adam.

Though many don’t want to work that way, and it took me a while to covert, but my workflow as introduced by Hook above.

It’s not the most intuitive but the RAW tab gets you 90% there.

I start with extended video but also modify it as required in the raw tab, grading to the waveform.

It’s a difference in workflow and to get away from the node / LUT mentality.

I use the RAW tab now to get me 90% there and use LGG and curves for the final seasoning.

I wish the RAW tab controls were more intuitive but they are more powerful than just trying to grade though a LUT.

JB

I used to use LUTs with Colour Science 4 in previous releases, but with the recent refinement of Colour Science 5 and the new grading tools of Resolve 17b2, I haven’t seen the need. You’ll recall John Brawley’s 12K footage of the angry actor carrying the chain who was surrounded by brightly saturated lights. I relooked at those clips in Resolve 17 and the glowing spillage of light is now under control without use of a LUT.

You’ll recall John Brawley’s 12K footage of the angry actor carrying the chain who was surrounded by brightly saturated lights.

Those were actually shot by Hook

JB

Thanks for the correction, John!