Blaise Douros
Forum Replies Created
-
In Cine EI mode, the ISO does not change when you change the setting. It is always set to 2000. However, you can “rate” it the way you would a film stock by pushing or pulling the image in post based on the Exposure Index (EI, get it?) you’ve set in the camera.
If you have an MLUT applied, and you change the ISO/Gain setting, the MLUT will show you what the image would look like if you had set that ISO, but the camera will still record in 2000. What this does is allow you to see a clean, correctly exposed image in the monitor, while still shooting to protect either highlights or shadows (depending on the situation). Let’s say you want to protect highlights; you raise the Exposure Index by a stop, which means you will end up adjusting the exposure down a stop to compensate. In post, the image will be underexposed by a stop, but the highlights will be protected.
-
Definitely make sure, too, that if they’re in a situation where they need shadow detail, that they overexpose at least one stop so you can crush the blacks later. Same for highlights–underexpose for extra detail there. Of course, they should notate this if possible.
On other piece of advice I can offer you when working with log footage: if you need to make global adjustments to exposure or white balance, make sure that adjustment comes BEFORE the LUT in the effect order. The LUT applies a logarithmic curve to the flat footage, so the highlights and shadows get crushed faster if you’re adjusting brightness after the LUT in the effect order.
Lumetri Color applies the LUT before any other effect in its panel. This means that if you need to globally adjust the footage, you need to use two Lumetri effects: one to apply the exposure adjustment or white balance correction, and then another one to actually apply the LUT and tweak refinements. I know this sounds overly complex, but you get very different results depending on where you apply exposure or WB adjustments in the effect order.
-
Okay, sounds like you guys are mostly getting it right. A couple of things to check–Make sure you’re on the latest firmware, version 4.0. Custom mode on the FS7 has noise reduction applied automatically. In versions 3.1 and up, NR is available in CineEI, but you have to turn it on; I can’t remember the menu path, but you can set it to Low, Medium, or High. That helps right off the bat with some of the shadow noise–it’s a very good NR algorithm.
You need to verify that the LUT the cam op is using is really the technical LUT from Sony, not one that was generated via LUTCalc. It’s easy to make critical mistakes when creating those from scratch. There are also a bunch of LUTs floating around from Alister Chapman that are designed to make the camera op reactively over-and-underexpose the image to preserve highlights or shadows depending on the situation, and if one of those is being loaded, it may be causing them to underexpose unintentionally.
Also make sure that the LUT is the correct one for the Slog3 flavor they’re using; there’s a difference between Slog3.Sgamut and Slog3.SgamutCine, and the MLUT that they use on set should reflect that.
Finally, to answer your question, yes, the FS7 has a very robust profile system that allows you to save a bunch of different profiles to the SD card–that’s why it has the slot in the first place. You can save everything from Paint and color setups to full camera setups.
-
Yeah, I haven’t really shot the C-series much, so I’m in the opposite (or same?) boat. I’m sure you’ve got a good shooting setup dialed.
It’s such a common problem with new FS7 shooters, Sony should supplement the manual with an Slog3 exposure guide. Luckily there are guys like Alister Chapman out there to do it for them, thought I don’t completely agree with all of his shooting advice.
-
Hey DJ, I think I can help.
First, I need you to tell me what color space, resolution, and codec you’re shooting in. Then, I need to know what mode you’re shooting–Cine EI, or Custom? If Custom, how is it set up? I also need to know how you’re exposing the image–are you exposing just by looking at the image in the monitor, or using scopes, an 18% grey card on set, etc…? Using MLUTs in the viewfinder?
The bottom line is that I think whoever is your DP doesn’t really understand how to expose for log footage. Here’s what happens: you get the log footage on the monitor, and when correctly exposed, it looks super washed out. So in order to make it look good on the monitor, the DP drops exposure, which makes the image look contrasty and better on the monitor. Only it’s not–it’s underexposed now, so you end up with ultra-dark shadows with grainy noise when you try to brighten it. So even when using a LUT, you have to bring the footage way up in brightness for it to look good…and consequently, you get grainy shadows.
You MUST understand how log color spaces relate to Rec709 before you shoot in them. Skin tones, for example–we’re trained to keep them at 75-80 IRE on the hottest spots. But in Slog3, the hottest skin tones should sit at 60-62 IRE. That’s a HUGE difference. Blacks, which sit at 6 IRE on Rec709, end up around 18-20 IRE on Slog3. So if you place the blacks down at 6, where we’re trained to put them, we grossly underexpose the rest of the image.
Does this sound like a plausible explanation for what might be going on in your footage? I have seen this issue probably 15-20 times with shooters new to shooting in log color spaces, and that’s exactly what it sounds like to me.
To be clear, I’m not saying anyone is negligent here–I learned this lesson the hard way myself. It’s an incredibly easy mistake to make, and I made it on several shoots myself before I figured it out.
To address some of the prior points:
Yes, the native ISO of the FS7 is 2000. It has very minimal noise at that ISO when exposed correctly. It shoots at that ISO to preserve the native 14 stops of dynamic range that the FS7 can produce when correctly exposed at that native ISO shooting Slog3. You can certainly drop to lower ISOs, but you sacrifice a bit of dynamic range (not a significant amount, but some).
And to address your comments, Todd, yes ISO 2000 does seem high until you know that the FS7 has three built-in physical ND filters: 1/4, 1/16, and 1/64, so exposing at ISO 2000 in full daylight is no problem whatsoever.
Of course, since ISO is totally unstandardized, ISO 2000 on the FS7 seems to be closer to 1200 on some of the other cameras I shoot with it. That’s a standard that REALLY needs to be resolved!
-
Your shutter speed is mismatching with the flicker rate of the fluorescent lights. There are a hundred articles out there explaining why this happens, so I won’t try to explain it here–go do a bit of Googling about fluorescent lights, framerates, and shutter speeds.
Once you understand it, bookmark this tool and use it when planning shoots: https://www.red.com/tools/flicker-free-video
-
I would say it’s better to conform both the footage and timecode so you can work seamlessly within the timeline–you just have to record it in your production notes. Anyone trying to re-create a sequence will have to conform the footage to 23.976 anyway, so make the timecode conform a required step in the process, too. Otherwise, if you leave the timecode in the other timebase, you risk having to parse out weirdo fractional frames from the “stretched” timecode. Ain’t nobody got time for that. If you were transcoding/conforming, it would do this anyway by default.
I just much prefer this method to transcoding because it saves time, money, storage space, and only requires good record-keeping, which is cheap (and you should probably be doing it anyway!). I think you just make a space in your production binder for post-production process notes, and call it good.
-
Import your footage to Premiere. Select all your 25P clips, and color-code them to a certain color–let’s say Orange. Then, Modify > Interpret Footage to 23.976. Now you have footage that is native framerate, with slowed-down audio, color-coded orange for easy recognition.
Edit your doc. Keep everything that is your 25P orange colored clips on one audio track. In the meantime, test out the pitch-shifting algorithms of as many different software packages as possible.
Once you’ve got Picture Lock, crunch out an audio-only track of your 25P orange clips. Run it through your favorite pitch conversion software. Add another audio track in Premiere, mute the old orange ones, and import your pitch-shifted track to replace them.
Now you’re done!
This workflow minimizes your transcoding, and keeps the audio round-trip to one single pass.
-
Sounds like what you really need to do is roll your laptop back to the previous version of Premiere, if you can. That would tell you whether the update is the culprit without borking your iMac.
Alternately, do a Carbon Copy Clone of your iMac, update to the new version, and then restore your system if it breaks.
-
God, there is so much confusion. I’m so confused at what you think you’re going to accomplish. I’m confused at whether you understand the words I’m typing.
If the lens has +1 barrel distortion on one camera, it will have EXACTLY the same +1 barrel distortion on another camera. Why? Because the lens will always refract light the exact same way.
If the lens has a certain number of lines of resolving power on one camera, it will have EXACTLY the same resolving power on another camera. Why? Because the lens will always refract light the exact same way.
Do I need to go on?
Do you really, honestly believe that when you put it on another camera, the results are going to be different? Because IT WILL NOT CHANGE WHEN YOU PUT IT ON ANOTHER CAMERA. All you’re doing is putting a different sensor behind the exact same glass.
Sheesh. Maybe you’re just trolling. If so, you got me, good one.