Proper Exposure with Slog2 and the Odessey 7Q

cuervo

Veteran
I'm wondering how other users of the O7Q are managing exposure with the Slog2 profile. I've found the similar situation is true as recording Slog2 internally. That is, it's best to expose to the right, bringing overexposure down in post, rather than underexpose on try to lift the shadows in post. My exposure process really uses the zebra settings on the O7Q set to 108%, combined with turning on the LUT and the WFM. With the LUT on, I set exposure so the highlites are just kissing the 109% limit on the WFM. This method seems to yield the best results.

How is everyone else doing it?
 
I do the same basically nothing below 30 on the O7Q scope. It is even OK to above 110 on some highlights on the O7Q, the scopes are misleading at this point . This will look terrible on the internal FS700 screen and internal recording will be to bright.

There will be a fix in future O7Q FW for this. To make the internal recording match the O7Q RAW 2K. And that will be nice because then you can use the internal for dailies. In the future the Odyssey may even be able to capture file names from the FS700 according to Dan Keaton.

The white point is also a bit misleading in the current FW on the O7Q. It is important to set white point even for raw as Sony does some compression even in RAW.
 
I generally agree with your methodology.

It's always easier to bring down exposure than it is to bring it up. There is less information in the shadows, meaning that literally there are fewer digital bits of data for the lower exposures then there are for the brighter ones. So an excellent way to work is to keep your highlights just short of clipping and let the 14 stops of exposure range capture the rest. In post you can then darken the exposure to where you like it, reducing the overall noise levels as you drop them down along with the rest of the image. Of course consistency from shot to shot is important, so you just need to find the levels with which you are comfortable.
 
ETTR works for short form, but will be a bit of a ball-ache for the colourist on long form stuff - for this it's better to assume a set ISO.

So far the only Raw I shot with the FS700 was 4K with the Sony Recorder. Checking out the rushes I'd say the 2000 ISO Sony intimates as the 'base ISO' is waaay off: very noisy blacks and mids.

However, I noticed that Sony Raw viewer seemed to default to 800 ISO when the footage was added in, and this looked much cleaner - I wouldn't be at all surprised if Sony have changed their mind about the camera's base ISO.

In future I'll probably meter to this. I'm not sure where this will place skin tones on the O7Q, but I'll check when I get one (it sounds like my dealer has finally got some coming in, but I'm not sure if I'm far enough up the list to get know in this batch).
 
Colin

My minimal tinkering with the F5/Raw revealed that Raw CineEI mode was 640 at lowest, using that cam I would therefore meter at 640 and a half aka 800, (the half is to give a little headroom) - just an observation.

I cant see why ETTR is a PITA for long form, or any form, or rather IETTR (IntelligentExposeToTheRight) IETTR you dont wiggle your stop on every take you just have the histo/wfm heavily populated at the top end.

Seems that a colourist can make a single correction and paste it to all clips to get a base look (or rather he/she can with IETTR that is what IETTR is)

--
Mitch - I thought the point of SLog is that there is equal data allocated to stops so ETTR for the reason you suggest is flawed, ETTR to avoid the noise floor is, however, not.

S
 
Last edited:
I read an implication from both Colin and Morgan that ETTR means a constantly changing exposure level. On the contrary, once exposure is set, according to ETTR, at the beginning of a shot. It remains fixed for the duration of the shot. Provided there are no drastic changes in lighting, ETTR stays pretty consistent. Heaven forbid the colorist would need to keyframe sequences to bring them all into line with each other. This does require some foresight into what lighting change is expected as the shot progresses.
 
IETTR tends to keep it stable across not just a shot, but a whole scene, or in fact shoot or indeed life of camera or in fact model of camera.

Colourist hate ETTR where every shot is dingled to be exactly on the clip point.

As for ramping in shot - that would be crazy (except sometimes.. like a specialist long steadicam shot through various locations)

S
 
ETTR, as I understand it, means just below clipping on each shot. It's a PITA for colourists as it means there's no consistency in the skin tones from one shot or one scene to the next - which means every scene or shot will need correcting before grading.

It it also makes it easier on the DP as it means you don't have to worry about the colourist undoing all your work by 'normalising' exposure you've deliberately set off.

Using a fixed exposure base is just about consistency. Any move away from consistency makes more work for the colourist (which could be significant increase on a 90-120min picture) - what you're sacrificing is flexibility in post as a fixed exposure base means you will end up clipping scenes where there is plenty of range available to cover it, just above or below this exposure.
 
Yep that is ETTR, I do IETTR - where a single consistent transform (aka lut, aka look) should get each image to a good start point

S
 
Exposing to the right works good for linear gammas as found with the REDCode codec shooting on RED One, Scarlet, Epic, or RAW Sony R5 module. This is also true because generally you're working with 12 or 16-bit depth codecs and you have exponentially increased data in the highlights end of the image. These are easily brought down in post, as anyone who has shot RAW stills on a Nikon, Canon, etc.. and then processed them in Lightroom would know.

When exposing for a log based gamma curve there really is no reason to expose too far to either end, except in the case you would like to add extra dynamic range to the low or high end in which case you might shoot 1 or 2 stops above true middle grey 50% IRE. However in the case of slog, slog2, and clog what people need to realize is that what is normally 50% middle grey gets mapped down to 32% on c-log and slog2, and 38% on slog. If you light what should be middle grey at 50% in slog/slog2/clog then you have already over exposed your image by approximately 2 stops.

The more you over expose with a log based gamma, the more plastic looking your highlights will be and the more you negate from the benefits that shooting log begins in the first place. Also consider that on the F3, or Canon C100/C300 log is shot in only 8-bits so to begin with there are some limits to how much you can adjust in post. If extremely diligent with exposure you can get away with 8-bit recordings, if you are more under the gun for shots and time, or you REALLY want ample flexibility in post for grading and adjusting exposure then you might want to look at an external 10-bit 422 recorder.

If you want a demo on just how bad you are making log look when you expose to the right, watch this video by Alister Chapman who has side by side shots of REC709 and properly exposed log.

http://pro.sony.com/bbsc/video/related-home_page_featured/video-F3_S_Log_Stage_Presentation/
 
Has AC ever shot a commercial job? and has he used Resolve for more than six months?
He is useful but pinch of salt IMO about what image he reckons can go where..
 
Allister is a pretty smart guy and his images speak for themselves. While I do not agree 100% with his practices I do find that on the whole his is spot on in his observations and explanations.

BTW, for most work I personally rate the FS700 at 800 ISO, perhaps sometimes 1250, and that's in RAW or SLog2.
 
It's interesting, at 0db, CINE4 is ISO640 with about 5.2stops of DR over middle grey. SLOG2 is supposedly ISO2000 also at 0db which implies a rebalance of about an extra 2 stops over, meaning about 7 stops above middle grey. It would definitely leave me questioning how much room is left in the shadows and if the noise is acceptable once graded... Maybe it's fine once you compress the noiser stops a bit? 7 stops over is alexa territory so I'm skeptical...
 
I'm not sure what the difference between what Sam calls IETTR and rating the camera at ISO 800 is. It the both sounds like they do the same thing to me: placing mid grey slightly above where Sony says it 'should' be (over exposing) - but at a consistent point.

I hear where Noah is coming from in terms of the dangers of over exposing a log image, but he's not really taking into account the issues with exposing the FS700 log 'correctly' - namely that the image produced is simply too noisy for many people.

Overexposing by a stop and a half means you lose some of the highlight head room (gaining DR in the shadows) but also pushes the noise from the mids down in to the shadows. At the end of the day it's a bit of a compromise. The trick is to understand the compromises you're making so you can make informed decisions about what is best on the day.

If you're in a controlled lighting situation, it's really not hard to produce a decent image from a camera with little highlight head room (some of us having been doing it for years on HDcam and DVcam): you just control the highlights! However, in these situations not controlling noise can be a big problem - and will seriously effect the perceived quality if the work you produce.

In other situations - where you have less control of the highlights (usually documentary style or low budget interiors with windows in shot) having nasty blown out highlights might look worse than noisy mids - so in those situations you might expose to the left/rate the camera at a higher ISO.

But, as I said, for long form work you want to try and be consistent. This is doubly true with log-in-8-bit as there is so little room to correct.
 
It's interesting, at 0db, CINE4 is ISO640 with about 5.2stops of DR over middle grey. SLOG2 is supposedly ISO2000 also at 0db which implies a rebalance of about an extra 2 stops over, meaning about 7 stops above middle grey. It would definitely leave me questioning how much room is left in the shadows and if the noise is acceptable once graded... Maybe it's fine once you compress the noiser stops a bit? 7 stops over is alexa territory so I'm skeptical...

The DR doesn't change when you change gammas - it just slides mid grey up and down the curve. So you get the same 12/13 stops with both gammas, it's just that Slog2 has more stops above mid grey and Cine4has more below it. Adding DB (going off the base ISO) doesn't increase DR much either - it just spreads fewer stops over the full image range.
 
BTW, for most work I personally rate the FS700 at 800 ISO, perhaps sometimes 1250, and that's in RAW or SLog2.

That's interesting. Since the O7Q has to debayer the 4K raw before recording it, I'm assuming it's making a decision about the camera's ISO when it does it (since that's part of a regular debayer). Is this ISO selectable? Or does it read it from the camera (which sort of sucks as SLog2 only goes down to 2000).

I'm assuming you can change this in the user defined LUTs, in which case how easy is it to alter the standard LUT provided in the O7Q?
 
I'm not sure what the difference between what Sam calls IETTR and rating the camera at ISO 800 is.

Pretty much the same, I just get annoyed when people think that stills togs wildy vary exposure on every frame - of course we dont because we have a need for cost effective workflow too.

Often the value of avoiding the noise floor is confused with the lack of value of twiddling exposure and creating a post 'mare. The good (noise floor avoidance) gets lost in the noise (chatter) of avoiding the bad (widdling exposure).

I would tend to expose in sets/scenes/locations sequences. Yep im sure 800 is a good start point, but some scenes one would be needing to dig the blacks or protect the highs and choose a different ISO suited to the scene.

In terms of post there is no 'good' way of passing the 'exposure choice' to a colourist (slate?) which does not help matters.

(for example if LUT was applied as meta data (is it?) one could pass the colourist files that would not vary wildly but might have one of three or more start points like the 'iso' tag on Red)

As for over exposing log Im sure a stop is good but beyond that needs some experimentation, skins too high and you get wood, not in a good way.

S
 
I'm assuming you can change this in the user defined LUTs, in which case how easy is it to alter the standard LUT provided in the O7Q?

Currently, there is ONLY the C-D installed LUT available. In future firmware upgrades, I'm told, we'll be able to install whatever LUT we choose.
 
When the custom LUTs become available you will be able to alter exposure level.

In rating the FS700 at 800, I don't do anything to it. The RAW signal at ISO 2000 and SLog2 goes to the Odyssey7Q. I still have tons of room in my highlights, as evidenced by the various exposure tools on the Odyssey. But I still use a proper lightmeter to design and preset lighting, as I have done for more than two decades. With that experience, I actually don't need the tool all that much. I know when I place lamp X at a certain distance it will deliver me T-stop Y given ISO Z. I've lit entire features by eye with a consistent stop on the lens, and even when the Gaffer insists on doublechecking I'm rarely off by more than 1/10th of a stop. It's just something you get used to with a trained eye and experience.

All that said, I consistently prefer the noise level and performance of the FS700 when I rate it at 800 ISO while shooting RAW & Slog2 to the Odyssey7Q. That's just my taste, but it works for me. The shadows tighten up and I still have plenty of range in the highlights. Technically it's 14 stops but usably it's more like 11.5-12. I chop off a similar number of stops on the bottom for other cameras as well, so again this is a consistent thing for how I like to expose for RAW and/or Log images.
 
Back
Top