EVA1: Footage seems overly bright when ETTR, even when not blown?

F

filmguy123

Veteran
This may be more of a 14 stop DR range, or the joys of working with my first S35 camera, or a vlog quirk - or maybe just EVA1?

Compared to working on GHx line, when I utilize ETTR techniques and basically expose the image just below clipping on WFM and zebras, maintaining all highlights but bringing the image as bright as possible, the image tends to look unpleasantly bright.

On the GHx line, this generally yields a nice image - maximimizng lattitude, etc.

On EVA1, it tends to create a pretty washed out looking image where the whites are just too bright and almost seem blown out, but can still be recovered.

1. Why is this?
2. Where should I be setting my whites? Should I be using ETTR techniques, or focusing more on keeping things centered in the WFM in the middle between extremes?

In post, I find I can generally grade the image back down to a more pleasing look. When comparing two shots - one exposed to what looks best to my eye on the monitor, and one exposed by "instrument" and that looks overly bright on the monitor, and then grading the overly bright version in post, it nets a similar result to having exposed by eye. Not entirely the same, but similar. Is that the intended workflow? Or should I expose by what looks best, so long as I'm not clipping darks or whites as best as possible?

I'm just going to assume I'm doing something wrong here ;)
 
I would expose just right, over exposing will hurt the quality of your colors in the highlights.
Also it will make differently exposed scenes (some exposed right, others over exposed) harder to match in post.
 
Thanks Cary!

So I've been so long in the lower budget camera market that ETTR tended to always be "right" due to the limited DR it seems...? (or was I misunderstanding that, too?). In either case, it always seemed to turn out great.

But this is a new ballgame. If exposing to the edge of highlight range isn't "right", what is "right"? Other than eyeballing it, how do I know what proper exposure will be via instruments or something easily and quickly repeatable?
 
filmguy123 said:
Compared to working on GHx line, when I utilize ETTR techniques and basically expose the image just below clipping on WFM and zebras, maintaining all highlights but bringing the image as bright as possible, the image tends to look unpleasantly bright.
Click to expand...

The ETTR exposure technique is not about getting the correct exposure in camera, it's about retaining highlight detail while providing as much exposure to the shadows as possible. The idea is to then bring everything back down to their proper exposure values in post, which should give you proper highlight exposure, proper mid-tones, and proper shadows, with as little noise as possible in the shadows. It can be a b*tch in post to figure out what the proper exposure values should be if you didn't shoot a reference card every time you a changed the ETTR exposure.
 
^Listen to that.

Also, different cameras will behave differently with how much you can push them. Some may clip hard and fast...others are more forgiving later on.

If you're using your Inferno with the camera it is a must you load LUTs on it and monitor your filming with them. You'll see how different LUTs react to exposure, which is absolutely critical for them to look correct. You'll then realize how the values are distributed and how your exposure could vary.

If you choose a LUT/look you like, you'll then monitor the exposure for it and never miss exposure again.
 
Interesting. Does it tend to pretty much always produce proper exposure on GHx line just because of the limited DR then?

My experience was that only rarely would this technique not yield proper exposure, and on those occassions just a 1 stop decrease seemed to bring everything in line. Would also do some mild tweaks shot by shot in post but was never an issue.

With EVA1 it's a totally different ballgame. I guess an extra 2 stops is exponentially more latitude, by a massive factor.

When on set, gray card, no problem, careful monitoring, no problem. One thing I liked on the GHx line is when doing live or docu work it was crazy quick to nail exposure by instrument and it was always pretty "on". I don't suppose there's any quicker "cheats" with an EVA1?

For example, keeping most of my brights around no more than... 70% IRE? Skins around 55% IRE on EVA1? (I know the old standard was 70%, and it was an older thread from a while ago, but I remember reading how on newer cameras mid 50s was actually often best?).
 
IMHO, what you are seeing is the difference between latitude between the two cameras. ETTR is for footage that will go through comprehensive CC and grading. Not sure if all of your footage sees that? Basically depending on the types of shooting you do, many of us live by waveform and zebras, not Histograms. I am almost always more concerned with what my talent's skin tones are exposed at more than protecting for highlights but most of the footage I shoot doesn't get a pro level CC and grade either.

I hate to sound vague but these exposure skills come through experience and experimentation, not from having a pedantic setting/workflow that works all of the time for each situation, regardless of what those settings are. The EVA 1 is a forgiving camera with all of that latitude as is the C200 when I shoot CRL. You need to be in the ballpark but you don't have to nail it perfectly each time in regards to exposure, that's the whole idea of having a camera with extended latitude, you have more signal that is able to be recovered. For me personally, I find that some of the expsoure recommendations for shooting RAW to be underexposed, rather than overexposed but I know you're not yet shooting RAW, right? I mostly shoot images of people and for me, if I can get skin tones between 65 and 80 IRE on the waveform, without grossly blowing out the BG, I am usually going to be happy with what I can pull out of that signal.

Keep in mind that the dynamic range of the eye in the shadows is about 20 stops so even at 14 stops ideally with your camera, you still have to compensate for what you see versus what your device can capture as photographers always have. Once you shoot a LOT more with your EVA 1, probably in about a year, you'll be able to compensate with your eye to somewhat understand what your camera will see. Sometimes it will still surprise you, you'll think the camera can record an image that looks beautiful, no problem but when look in the monitor and shoot it, it will look like a mess whereas it will look beautiful to your eye. I mostly shoot documentaries, run & gun and BTS type footage so I am not always as concerned about perfect exposure as I am when we are on a stage shooting a music video, dance number or commercial, there I use a gray card, white card, chip chart but you rarely get a chance to use any of those on seat of your pants documentary shoots where we don't always know where and when we will even shoot. YMMV.
 
filmguy123 said:
Interesting. Does it tend to pretty much always produce proper exposure on GHx line just because of the limited DR then?
Click to expand...

ETTR originally came from digital still photo cameras, when the only exposure tool you had was the histogram.

I think the in-camera footage will vary quite a bit from one camera model to another. I generally don't use the ETTR exposure technique because I rarely shoot high ISOs, and I don't mind a little bit of grain in the shadows. ( if it really bugs me I will use a NeatVideo NR profile to take out the noise in post )

filmguy123 said:
With EVA1 it's a totally different ballgame. I guess an extra 2 stops is exponentially more latitude, by a massive factor.
Click to expand...

I think it's just a matter of getting used to how each camera responds and using a good calibrated monitor to get your exposure right. ( I "eyeball" my exposure from my monitor when I have to work quickly, and I've gotten pretty good at this )
 
Also a fan of exposing correctly. Not sure what tools the EVA1 has but on the GH cameras I use the exposure meter with an 18% gray card. Set the meter to average when you want a general exposure reading and don't have an 18% card. It will never be 100% accurate since a small bright light source will taint the results but it at least gets you close. On the meter the middle or zero is a balanced even exposure. Anything to the left is under exposed and anything to the right is over exposed. When I use my 18% gray card I use spot meter mode and move the point around the screen so it lands on my 18% gray card in the scene. I then adjust the lighting or exposure so that point lands at 0 or the middle of the exposure meter. Good way to keep the exposure consistent across the shots and keep a sort of standard across projects.

The exposure meter is a poor mans light meter.

The optimal solution is false color which is the best way to judge exposure. The false color allows you to judge multiple areas of the shot all at once so you can keep adjusting light until every portion of the shot is exposed as desired. Harder to do this with zebras and the spot meter unless you keep adjusting the zebra values and move the spot meter point around the screen to know what each portion registers as. Yes a ton of work but the end results are worth it if you use light to sculpt the scene and not just as a means to eliminate noise.

In run&gun situations I find the average exposure meter pretty accurate and just keep in mind if I'm point at a light source the results are likely off and I need to manually compensate or switch to spot meter to isolate the section I want exposed in the middle.
 
Get a Sekonic l-758 C...they've been discontinued for a touch screen version...so super affordable used now....the C is the Cine version....may seem like a quaint idea, but all guesswork is completely removed...it's both a spot and incident meter....next best is of course false color, but then you have to set a camera up wherever you scout....with a lightmeter you arrive on any location...plot out your lighting package or what time you have to shoot, if using only available light...and other reasons to numerous to mention...one can wing it and get lucky...until one day with important footage you don't...lastly.your camera's spotmeter can never compete with a dedicated light meter...IMO one of the most important things in your bag
 
I've always been a "get it right in-camera" guy. This comes from back in the day, you had to. And being a "quick turn" broadcast guy. I rarely shoot log and even when I do, I shoot with a 709 lut and aim for nailing the exposure. I also never use waveform while running around shooting run & gun(sit downs are a completely different story). (Very)Occasionally I may turn zebras on briefly. Histogram, I'm not a still photographer. I don't need no stinking histogram.

In all seriousness, though, the best thing to do is learn your camera and VF. If you do, you won't need any exposure tools besides your VF and your eyes.
 
fboonzaier said:
Get a Sekonic l-758 C...
Click to expand...

That's a great idea. And if you want to get your head around using a light meter and have an iPhone, you could check out Adam Wilt's genuinely useful Cine Meter II:

"Cine Meter II turns your iPhone®, iPod touch®, or iPad® into a reflected light meter, an RGB waveform monitor, and a false-color picture monitor. Cine Meter II works on any iDevice with a camera running iOS 6 or higher. It expands the original Cine Meter app..."
https://www.adamwilt.com/cinemeterii/

CMII-colorMetering.jpg

Cine Meter II doesn't replace the need for a very good light meter, but it's not a toy and is a great and (if you already have an iPhone) affordable way to start learning more about light.
 
Run&Gun said:
I also never use waveform while running around shooting run & gun(sit downs are a completely different story). (Very)Occasionally I may turn zebras on briefly.
Click to expand...

I'm just curious (and I don't intend this to sound critical because I often shoot the same types of situations), but if you are not at least using zebras most of the time, what are you usually exposing for? The subject or the scene? Or maybe better asked, what are your zebras set at when you do use them?

Sorry, don't mean to hijack the thread!
 
I don't subscribe to ETTR. Why put your subjects up where there is so little data? Protecting the highlights has its place, but ETTR is not a panacea. The colorist me :) also doesn't like having to unnecessarily grade each and every clip because middle gray is never in the same place. I'm with Cary and Tom, my $0.02

And it's probably not that the images are too bright as they are too flat, the consequence of recording with not enough bits, which are allocated to the meaty part of the curve in the middle.
 
This is helpful!

Tell me if I am understanding this right: Thinking of latitude not in stops but in contrast ratio sheds some light. If I'm used to working with a 4000:1 contrast ratio (at BEST) on a GHx line, that means a 1-stop over exposure (even if nothing is blown) may be 2000 brightness levels "off". Being 1-stop over exposure on an EVA1 w/14 stops, thus a 16,000:1 brightness ratio, may mean being off by 8000 brightness levels. Thus, the differences I am seeing in being a tad off are exponential.

Further, on the GHx line, which is a bit starved on DR and can have less low end shadow fine detail, pushing things brighter has generally been a good strategy. Yes, it's meant bringing a number of shots down a bit in post to get the contrast and levels back in line, but because of the smaller contrast ratio and less desirable low end, its generally looked close and any adjustments on levels have always felt more like making some tweaks rather than "correcting".

What I am gathering is that with the EVA1, the low end fine detail is SO much better and richer, that it's SAFER to utilize the lower end of the curve more than before, and in fact, even desirable. It's also got so much more latitude that pushing things uuppp into the higher regions is, proportionately, a more dramatic gap (ie 8000+ brightness levels) away from "middle grey" VS maybe 2000 levels away on a smaller latitude camera?

Apologies if I've got that all wrong, just trying to wrap my head around it. This is my first time working on a camera with this much DR. My camera progression was HVX - AF100 - GH4 - GH5 - EVA1.
 
Those numbers you're throwing out, 14 bit, 12 bit; I don't know what for the EVA1 but don't kid yourself, it applies to the GHx as well if you are shooting with one of the log gammas which are 4-6 stops more than 709. All you have to understand is this; each f-stop is a doubling of light so it would be intuitive to think that you would be doubling the number of shades with each brighter stop but unless you're recording raw, that's just not how it works, beyond the mid range there are fewer shades allocated with each successive brighter stop, exactly the opposite. ETTR is okay when you're shooting with just the six stops of 709 but with log, 12-14, if you expose your skin tones into the range of stops with diminishing shades, although not clipped they still become flat. It became popular to overexpose by 1 stop with S-Log2/3 because the DSLRs at least were only 8 bit and noise is a problem.
 
So now that noise isn't such a problem like on 8-bit codecs and limited latitude, there's less need to keep the image so bright, basically?

And now, the goal is to keep the majority of the image centered around middle gray? Not because you can't bring it back down in post for proper levels (you can, it's just unneeded work without benefit), but also because it actually harms the footage to push it brighter than needed, since most data on the logarithmic curve is allocated to the lows and center, and since on a higher end camera avoiding lower end noise is not so critical?
 
Seanik said:
I'm just curious (and I don't intend this to sound critical because I often shoot the same types of situations), but if you are not at least using zebras most of the time, what are you usually exposing for? The subject or the scene? Or maybe better asked, what are your zebras set at when you do use them?

Sorry, don't mean to hijack the thread!
Click to expand...

Good question. And the answer is, it depends. If it's an interview with someone, even just 'off the shoulder' sound like post game, etc., you obviously put priority on the subject, BUT sometimes you have to evaluate the entire scene and sometimes find a "compromise exposure" so that you're not blowing out a huge portion of the background(if it's hot). Of course, in these run & gun scenarios, this is like driving a race care and these decisions are made almost subconsciously in a split second and comes with years of experience of what you want and need to do in these less than ideal situations. My answer may not be much help, because it's not a black and white, one or the other answer, but thats how I shoot in those situations. On the local news, I see shots all the time, where guys gave 100% priority to the subject and the background is literally a hot mess and the entire shot then looks bad.

As far as zebras, I have them set to 70% and I usually should not see them on someones face. In sit-down situations, I'm usually exposing skin in the 50%'s(maybe 60%'s).
 
You must log in or register to reply here.
Back
Top