Standard vs Superflat

Beeblebrox

Well-known member
Someone, I think it was Barry, was asking for a sample of a dramatic difference between superflat and standard image modes.

Here's a sample from a shoot from today. We shot superflat, but I grabbed a shot in standard mode just for this comparison. Everything else in the shot is identical, including all other settings on the camera and the lighting.
 

Attachments

  • test_flatvsstandard.jpg
    test_flatvsstandard.jpg
    51.2 KB · Views: 0
That is seriously an ENORMOUS difference. Almost to believable, but it seems to be clearly the same settings.

Bravo +1 for SuperFlat preset.
 
I like the first image over the second. I would rather add some light or a faster lens to the standard setting than get a washed out image from the superflat. I prefer images with more contrast and saturation. This will simply boil down to personal tastes I think, some people seem to like the faded look of flat picture styles some people like a more punchy and colorful picture.
 
You're basically stuck with the first image- the second leaves your options wide open in post. Kinda the idea of stretching your gamma out by not shooting with the standard in-camera settings...

Noah
 
I think the benefit of superflat though isn't to just have that "washed out" look...its more so that when they put it through color grading, it holds up better due to more dynamic range being retained.

Kegan

EDIT: Noah beat me to it...heh.
 
the only problem I have when I shoot super flat is that the footage looks really grainy and the colors are all grey shifted... plus I am new to Color 1.5 someone should put together a Color 1.5 correction so that we can take that amazing superflat footage and bring back that beautifuly colored footage through color correction while keeping the apparent range
 
I like the first image over the second. I would rather add some light or a faster lens to the standard setting than get a washed out image from the superflat. I prefer images with more contrast and saturation. This will simply boil down to personal tastes I think, some people seem to like the faded look of flat picture styles some people like a more punchy and colorful picture.

Hi mtan,

The superflat setting is not intended as the final output, the standard setting is a look intended when you don't plan intensive grading. The superflat is similar to the reason for using black stretch in traditional video cameras to brighten the shadow areas and lowering highlight intensity so you can control how much to add or remove later in post. This is suppose to emulate the characterestics of film stock which gives you a flat milky like look upon developing to give the colorist enough info to work with.

Of course you are right when you say the image using standard setting looks better because its suppose to give a finished look right after shooting and not intended to be further processed but if you do put that image in grading its already crushed blacks can't provide enough detail in post.

BUT speaking of this camera, the 7D, in particular I'm worried in my own tests due to the fact that creating superflat curves in the eos utility somehow gives me inaccurate skin tone rendering.

Also, i believe this has something to do with the codec, stretching the shadow areas using custom curves apparently produces more noise, than the more robust codecs. But subtle curves and or simply lowering contrasts on the 7D NEUTRAL presets seems to hold the noise in check.

Ted



p.s. whoah! when i was typing only mtan comment was there! :-)
 
Last edited:
Yeah I also noticed the falsities in color that is what I am wanting to know how do I bring back these colors in post I am having a very hard time doing even simple color corrections unless I just place in MB Looks pre-sets... I shot a coke can next to a black heel and it was great I could see all of the detail (ruffles) in the heel, but the color on the bright red coke can was off. I tried my hardest to bring it back but unless I physically did a mask in geometry then I would just lose the detail in the shoe??? I dont know what im doing wrong.
 
Perry,

I believe this is the inadequacy of the codec's color space. I'm also experiencing a caked like, pastel like skin tones which is next to impossible to correct. This was shot in bright daylight and I was hoping , using the flat curve used in superflat can give me detail in strong shadow areas.

Apparently the trade off is not worth it. I think the codec/bitrate dumps so much info during encoding to the memory card that our best bet would be to get the nearest "normal" look
in-camera during the shoot.

We simply cannot treat this beast the way we do with traditional HD cameras with more robust codecs.

Even with the sample above, look at the wall on the superflat setting, you can already begin to see greenish blotches and the tell tale signs of "noise" beginning to build up.

Ted
 
My experience with the picture style profile tool is that it screws too much with the curves, leaving crossed colors when things get pulled back up in post. I think it is better to simply use a preset and dial everything down. Working with the neutral in this way tends to yield a good base image for post. The codec is deifinitely awful though for certain things. Green fields with a lot of darker tones is a nightmare. Better to boost the saturation to separate the values in this instance. I think it is often more important to bring down contrast and sharpening than almost anything else for the majority of circumstances.
 
I would rather add some light or a faster lens to the standard setting than get a washed out image from the superflat.

This makes more sense to me, get the image as close to your final image as possible with proper exposure. Surely that would yield maximum image quality?

...unless you know you want to do some heavy post work to get a look you can't get in camera - presuming superflat does stand up to heavier post work that is.
 
I think the benefit of superflat though isn't to just have that "washed out" look...its more so that when they put it through color grading, it holds up better due to more dynamic range being retained.

I think it actually has less dynamic range. Instead of using the full range of the h264 codec it just uses a small part in the middle as all the highs and lows are pushed in.

So I'd much rather use a contrast setting of -2 or -3 which has worked extremely well for me than this extremely pushed superflat.
 
You know what I'd honestly like to see, is Barry or Kholi, or the like take their time, and analyze this confounded Superflat vs. Standard picture style setting.

I mean we've seen people do some tests but those have either been biased, or extremely unreliable, and thus non-credible.

So an open letter request to the Green Machine, if you have some extra time, I'd love to see your take, or better yet your observation regarding this topic, so we can FINALLY have an answer if its more beneficial to have it one way or the other.

Who's with me?
 
Before I say too much, keep in mind that I haven't even figured out how to use the custom picture profiles yet, so I'm not going to speak to how they work with the 7D.

Strange way to start a post...

However this doesn't mean I don't know anything about color grading. A lot has been said about how the "super-flat" picture profile increases or decreases dynamic range, it does neither.

What its doing is a process called "Black and roll off," its setting the white point value a bit lower and the black point value a bit higher then rolling off the highs and linearizing the blacks. When you set the exposure your providing the camera with the information it needs to fill in the middle.

So if your scene is predominantly dark and you expose for it correctly, the math in this profile (the curves) are such that the shadows will show more detail but it will appear that the whites are blown out. But when you bring that clip into color correction, [OK a quick aside here, I'm not talking about the three way color corrector in FCP or the equivalent in PPRO, I'm talking about Apple Color, Synthetic Aperture (AE), or DaVince], you should see that the histogram is more evenly distributed and when the luminance is corrected (first step in primary color correction) you get a nice full histogram and there is more details in the whites than first appeared.

If you shot the same scene with the standard settings, what you'd see when you brought that clip into grading is the histogram would be pushed heavily towards the blacks and when you adjust the luminance, bring up the mid tones you run the risk of clipping the whites and introducing noise. How much noise and clipping depends on your exposure.

Which kind of brings me to my point. Color grading is all about tradeoffs, it can't magically add something that isn't there. If you use the super-flat picture profile exposure is critical, because although the histogram is more evenly distributed, it is very flat (hence the name) and if your exposure wasn't correct it will show up in the histogram by either the whites or the blacks being more to the center. And if I have to add luminance to correct for it I'm potentially adding noise and at that point I might as well have used the standard profile. Also, what you'll notice is that if white balance is off, even a little bit, the image will push toward a particular color. The result is that although the grade looks good the reds might be a little more vibrant for example. This can be quite easy to correct for but think about how this might effect skin tones.

So, if your "really" good at getting the right exposure and white balance, and you have a good color grading application then shooting with super-flat can be really rewarding. However if your not then I'd suggest staying with the more standard profiles until you are. Also if you don't have a good color grading application [or don't know how to use it] you can use the 3-way color corrector in FCP, however you won't have nearly the options for correcting any mistakes in exposure or white balance and the results can leave you a little flat :D
 
BUT speaking of this camera, the 7D, in particular I'm worried in my own tests due to the fact that creating superflat curves in the eos utility somehow gives me inaccurate skin tone rendering.

+1

I have found the same thing. I don't like the look of superflat. Even after grading I cannot get the skin tones to where I want them. I shoot with a dialed down neutral now.
 
Have you guys tried the Marvel Cine picture style it is great the colors are correct and the range is far better than the standard not as good as the superflat though... but I really dont like the super flat, just light it up people... I think it is funny that in the original video "How to increase dynamic range" the only way it looks good is when he slaps on a MB looks grading filter. But then the range goes right back to the way it was... what the F*!K is the point then, and then the color is of any ways becuase of the MB looks styles.
 
Back
Top