Revenge of the Great Camera Shootout - blind test

[Mattbatt]

Originally Posted by Tom Wilkinson Please help me understand why B is bad. I thought it lookd the best, you could see every face, highlights seem good, shadows all had some detail. Why is the "video" look bad for film. I guess I shoot "video" and like the look. But I need to understand the "film" look so I can learn to appreciate it.

Personally I hated the way "B" looked, it looks to much like video. Real sharp and everything fake, film in my opinion looks more organic/natural which is why I like the the other scenes better. Personally I like A the most with E and F coming in and a tie for second. B by far looks clean and razor sharp but I truly do not like that image at all, truly hate it.

The reason why B upon any second glance is a weaker shot over H, F, C and A is that B has so much light, the scene becomes a staged set rather than a scene. The scene for B is easily depicted as a lit set rather than a person's house, and it took me out of any type of visual story (I know it is a test but still, the set was designed to look like a house and B looked less like a house than any of the others). The best language I can use for it is, "soap opera style" - too much artificial light.

Now, nothing is wrong with that per se, unless you were shooting a "24" style scene or something naturally lit for a soothing drama from a movie - now instead of a room scene you have a stage set and it doesn't not look real.

Overlighting can be just as bad as underlighting - and we all know this. I find most people who are not film majors call it, "green screened" or "too much CGI." What I finally figured out from friends, family and strangers when they use this lingo (because there is no CGI or bad keying going on) is that the light looked 'fake' to them.

As a professional graphic designer, in school we learned how the eye thinks and moves on its own, far more than we consciously think about. Think of it this way: the eye knows where to go for facial recognition, you do not have to think about it; the eye knows natural light and shadow extremely well - even if you don't know the terms - and can generally pick real vs fake very well from things as well, which is why CGI for realistic characters is so tough!

So when a scene is overlit, the eye begins to communicate to the brain that something is 'artificial' or plastic or CGI. Most people think greenscreen in their basic terminology, and I suppose that is because many greenscreen scenarios have been lit wrong and it looks different from the previous scene or natural light. Most people never see keying, edging or anything along that sort, nor know what that means.

So how does this fit with the camera? Well, the GH2 is strong in resolution, but weak in DR and its codec, therefore more lighting and layers in post had to be added.

If you go off of each face, the background, the exposure, "B" is very polished. (HINT: there is aliasing slightly on the window grid, nothing like the 7D in G though.) But if you compare to the more experienced DP's gunning for more of a cinema style, B stands out like a soap while H, F, C, A look more like a movie.

Yes, DP skill is 90%, but the tool has to be able to deal with natural, harsh light as well. The original scene was high contrast simulating daylight. B handled it far less well and therefore more artificial lighting had to be used, thus creating a staged look.

The F65 by contrast had NO EXTRA LIGHTING, and if it is H by my guess, is the sharpest, has the highest range of DR, has the best flesh tones and is the most unique in look out of them all.

If H ends up being the GH2, I will buy one (I highly doubt it is).

 

[Mattbatt]

Now I want all of you to re-watch the test twice using this technique: watch the front lamp on the left. Before the crane pans down, you can see inside a bit where the lightbulb is. Look at the highlight, then the background window (take a peak at the top edge ring too, that seemed hard on the cameras). Then look at the lamp's shade: F and H really have a nice, pink, flesh tone emanating from the center outward. B is grey by comparison. G which is the 7D based off moire artifacts, keeps the pink a bit too which makes sense for two reasons: 1, Canon has good skin tone, 2, the footage has saturation cranked.

What does this lamp shade tell us? It shows dynamic range (when looking at the lightbulb and background window) and shows how well a camera can "see" light emanating through substance (the lamp's shade). D and B were dark grey, what does that tell you? Well, skin tone is going to have to be lit extra well because that nice fleshy-orange/pink-glow that film gets down so well is not gonna be as good on those cameras.

You say, "Matt, the faces all seemed decent in the tests." Well, wait, in post, am I going to bother putting a mask and power grade on a lamp or a face? This test shows us what people did to make their scene look best, so of course they are going for the faces. I'm looking at that lamp because people missed it and it shows me some hidden facts about the camera.

Why do I say this? Well, I'm a lowly indie filmmaker hobbyist with a 5dmkII and BMCC on the way (yaya!) and I don't have huge sets, lights or awesome skills in color grading. So I look at how well a cheap camera can handle harsh conditions. Canon handles flesh better than GH2 but is softer and moires (still softer with MKIII). I'm hoping my BMCC does great in this area.

It is all about skin tone IMHO and for me, using available light as best as possible.

 

[TheDingo]

Canon handles flesh better than GH2

For video straight out of the camera without modification, I would agree with you, though the difference is not as big a people make it out to be.

Once you add post production grading, then it all comes down to the skill level of the colorist.

I'm not a fan of the "Canon" color, but I can understand why some people are.

 

[maranfilms]

Now I want all of you to re-watch the test twice using this technique: watch the front lamp on the left. Before the crane pans down, you can see inside a bit where the lightbulb is. Look at the highlight, then the background window (take a peak at the top edge ring too, that seemed hard on the cameras). Then look at the lamp's shade: F and H really have a nice, pink, flesh tone emanating from the center outward. B is grey by comparison. G which is the 7D based off moire artifacts, keeps the pink a bit too which makes sense for two reasons: 1, Canon has good skin tone, 2, the footage has saturation cranked.

What does this lamp shade tell us? It shows dynamic range (when looking at the lightbulb and background window) and shows how well a camera can "see" light emanating through substance (the lamp's shade). D and B were dark grey, what does that tell you? Well, skin tone is going to have to be lit extra well because that nice fleshy-orange/pink-glow that film gets down so well is not gonna be as good on those cameras.

You say, "Matt, the faces all seemed decent in the tests." Well, wait, in post, am I going to bother putting a mask and power grade on a lamp or a face? This test shows us what people did to make their scene look best, so of course they are going for the faces. I'm looking at that lamp because people missed it and it shows me some hidden facts about the camera.

Why do I say this? Well, I'm a lowly indie filmmaker hobbyist with a 5dmkII and BMCC on the way (yaya!) and I don't have huge sets, lights or awesome skills in color grading. So I look at how well a cheap camera can handle harsh conditions. Canon handles flesh better than GH2 but is softer and moires (still softer with MKIII). I'm hoping my BMCC does great in this area.

It is all about skin tone IMHO and for me, using available light as best as possible.

Much of what you say is subjective. Especially considering we just watched the test, So i'm not sure how you can say Canon has better skin tone than gh2. Based on this test, I thought the GH2 DESTRYOED the canon. In pretty much every catagory. And I shoot both, Canon and gh2. Brands don't mean squat to me. Second, If you think they didn't take the light bulb into consideration when grading, and only focused on "skin tone" shadows and exteriors. Well.... You would be sadly mistaken. That bulb and lamp shade was looked at very very closely in the grading process. I can assure you there wasn't any part of that image that wasn't gone over with a fine tooth comb. I won't disagree that skin tones are important, But to say "It's all about skin tones would be somewhat silly. What good are great looking skin tones if the remainder of the image looks like crap, and vice versa. I would say, it's all about the overall image, as a whole.

 

[evilthought]

Much of what you say is subjective. Especially considering we just watched the test, So i'm not sure how you can say Canon has better skin tone than gh2. Based on this test, I thought the GH2 DESTRYOED the canon. In pretty much every catagory. And I shoot both, Canon and gh2. Brands don't mean squat to me. Second, If you think they didn't take the light bulb into consideration when grading, and only focused on "skin tone" shadows and exteriors. Well.... You would be sadly mistaken. That bulb and lamp shade was looked at very very closely in the grading process. I can assure you there wasn't any part of that image that wasn't gone over with a fine tooth comb. I won't disagree that skin tones are important, But to say "It's all about skin tones would be somewhat silly. What good are great looking skin tones if the remainder of the image looks like crap, and vice versa. I would say, it's all about the overall image, as a whole.

According to someone who watched the screening, in the empirical test, where no one was allowed to change the lighting, GH2 had worse DR than 7D but better resolution.

 

[Gillvane]

There is something about Epic footage that draws me to it as well. But I will say that I didn't see that "something" in any of the footage from that test so who knows.

I would love to see what Jim Jannard chose as his favorite. It would be interesting if he chose the Epic and it would be even more interesting if he didn't. I guess we'll never know because he's not much of a team player unfortunately.

That does have me slightly confused. Normally I can say Aha! Epic! Looks great! Not so much in this test. If we blow it up on a 4k screen it's still not going to look pretty.

The scene just looks terrible in general. No matter which camera it is I don't like the way this looks. I would never purposefully light anything like this.

I'd like to see a similar test where the scene looked gorgeous and this just doesn't.

The test seems to be, which camera and DP can make this scene look the least crappy?

 

[evilthought]

That does have me slightly confused. Normally I can say Aha! Epic! Looks great! Not so much in this test. If we blow it up on a 4k screen it's still not going to look pretty.

The scene just looks terrible in general. No matter which camera it is I don't like the way this looks. I would never purposefully light anything like this.

I'd like to see a similar test where the scene looked gorgeous and this just doesn't.

The test seems to be, which camera and DP can make this scene look the least crappy?

I disagree. If lights are were good, it will be even harder to tell the difference. It's only in challenging situations where you can see the difference between a better sensor vs worse sensor. Lets wait for empirical part where no one was allowed to change the lighting. That will clearly show the differences between DR and sharpness.

 

[TheDingo]

The test seems to be, which camera and DP can make this scene look the least crappy?

The set was not built to be pretty looking, but rather to be challenging for the cameras to record.

In August the 3rd part of the shootout will be released and you will get to see how the different cameras handled the SAME identical lighting. ( I liked the GH2 camera a lot more when the lighting was the same )

 

[Gillvane]

The set was not built to be pretty looking, but rather to be challenging for the cameras to record.

In August the 3rd part of the shootout will be released and you will get to see how the different cameras handled the SAME identical lighting. ( I liked the GH2 camera a lot more when the lighting was the same )

Well it better look gorgeous because then there's no excuse to make it look crappy to just to make it challenging.

 

[agcohn]

I can see why they wanted each crew to light the set differently for each camera in the first test, as different cameras might have different lighting requirements. However, after recording each set-up with the intended camera, I wish they had gone on to record the same set-up with the other cameras as well, for a point of comparison.

 

[TheDingo]

Well it better look gorgeous because then there's no excuse to make it look crappy to just to make it challenging.

My guess is that the GH2 crew were not experienced at lighting a large set, and went crazy with way too much directional light ( the over-lit guy in the green shirt ) and with too much fill light. ( the sun-lit shadows on the floor by the woman standing by herself are so filled in that they are almost gone ) These lighting changes look extremely "fake" to my eye. No subtlety at all.

 

[Mattbatt]

Much of what you say is subjective. Especially considering we just watched the test, So i'm not sure how you can say Canon has better skin tone than gh2. Based on this test, I thought the GH2 DESTRYOED the canon. In pretty much every catagory. And I shoot both, Canon and gh2. Brands don't mean squat to me. Second, If you think they didn't take the light bulb into consideration when grading, and only focused on "skin tone" shadows and exteriors. Well.... You would be sadly mistaken. That bulb and lamp shade was looked at very very closely in the grading process. I can assure you there wasn't any part of that image that wasn't gone over with a fine tooth comb. I won't disagree that skin tones are important, But to say "It's all about skin tones would be somewhat silly. What good are great looking skin tones if the remainder of the image looks like crap, and vice versa. I would say, it's all about the overall image, as a whole.

Did you even read my entire two synopses' before judging me with your biased opinion that I am a brand loyalist or somehow bashing the GH2? If you read my synopses, I am all for H and F, certainly not G. If you read what I wrote, you would of noticed that one area we can look at is that lamp - there is a similar color temperature and overal exposure in all of the shots but that lamp differs greatly, from cold, dark grey to fleshly pink life eminating forth from the center.

Now, I am sure people went through the image, but not as fine as you think. You can see the time spent on each, an hour here and there is not enough time to go through it with a fine toothed comb, but certainly, one could of added something to the lamp.

But I don't see it. I do not see someone artificially coloring the lamp - I see someone pushing the camera and balancing the color and some cameras held the light, others did not.

Photography and cinema is all about light, so any camera that crushes, looses or polutes light and color is a looser in my book.

Here my profession is to craft beauty from shadow and tell a story but yet my tool forces me to fix in post? That is not very efficient.

The GH2 is a colder, harsher image out of the camera than the Canon from many tests I have seen and done, but sharper. And yes you can color grade and push it. So what? The less post for me the better.

Back to the test, again, only a few cameras hit that dynamic range, sharpness, light, curve, flesh and contrast I'm talking about. It is clear F and H do such things very well. If the GH2 is F or H, I will eat my words and gladly buy one.

 

[squig]

Is this some kind of Sesame Street horror spoof? People fighting over letters. C is for cookie that's good enough for me.

 

[Steve R]

Do people here WANT me to reveal which camera is which? Or would it be more fun if we all waited for Zacuto's reveal next month?

(FWIW - I was at a trade show this week where Zacuto had a booth - and they ran all of the clips in a loop with the actual Camera names on them, vs. the letters. It doesn't seem to me that this was a mistake on their part; they MUST have known the information would get out there...)

 

[squig]

Noooo! Watching people fight over letters is so entertaining.

 

[TheDingo]

Do people here WANT me to reveal which camera is which?

Hey Steve,

I was at the same trade show and I think it was a "goof" on Zacuto's part. They were showing an early edit of the shootout footage that identifies which camera is which, and showed footage from all the cameras under the same lighting.

Out of respect to Zacuto, I have kept this info to myself. Zacuto spent a ton of money to make this shootout happen, and are still paying a small crew of editors to complete the full versions of Part 2 and Part 3.

If we spoil the surprise, this will definitely be the last shootout Zacuto does, which would be a shame.

 

[evilthought]

Do people here WANT me to reveal which camera is which? Or would it be more fun if we all waited for Zacuto's reveal next month?

(FWIW - I was at a trade show this week where Zacuto had a booth - and they ran all of the clips in a loop with the actual Camera names on them, vs. the letters. It doesn't seem to me that this was a mistake on their part; they MUST have known the information would get out there...)

No, but can you answer one question? In creative test, which one did you like better? C300 or FS100?

 

[indiawilds]

Who ever knows the actual please keep it to yourself. Let's respect Zacuto's decision. Where as someone may think why they are releasing the three parts over a period of three months and why not earlier, lets give them the benefit of the doubt. They must have thought about it. It is not ethical to spoil the show by revealing names.

Also, it is interesting reading the reasonings of people trying to dissect the clips etc. Let the interest levels remain.

 

[evilthought]

I still think I is the 7D, it has tons of aliasing, and the kind of noise that I hated CineStyle for

I don't think so .... Here is lighting for Canon 7D

Now which one matches that setup?

 

[Samuel H]

Well, G matches that diagram better, that's for sure. We'll see... (also: as others, I'd rather wait for Zacuto to unveil the mistery; it's their shootout, and they've asked those in the know to please remain silent)