GH2 In the end, which camera has better light equality?

[RPLace]

A big question here:

Basically, with my style of shooting, I find myself always needing the most natural lighting equality, in-camera, that I can get - documentary style shooting, quick interviews, or short films where I don't have the buget or time to film on extensive sets. I've seen beautiful cinematography in my day to know that an eye-catching story (or just an interview or documentary shoot) can be told, to the point needed, by simply using tricks in cinematography or a little know-how in editing (for example, if you can, throw any blaring light sources or windows out of focus, or dial down the white level in post.)

However, sometimes certain things just can't be done with contrast levels that are way off the mark of any sort of reality - for example, I rate Canon's Standard profile to be unrealistic even in the most lit scenarios (daytime), it is so contrasty that it makes the most vibrant, sunny field of flowers look harsh and way too punchy - almost robotic. This is a downside of having too much contrast, no matter which situation you're under. Anyway, the GH2 profiles have the same problem, so I'm not trying to pinpoint cameras.

But my problem is this: like I said, I like having a realistic light equality, I like being able to get a truly professional image if I'm on the go, but do take my time to set up some very eye-catching cinematography - you can be a cinematographer who's expertise lies in documentary shooting or Dogme shooting or just whatever, but if your camera has so much contrast that it cuts out information where flatter settings would without any doubt pick it up, you'll have an image that really is out of balance no matter which way you slice it, unless maybe you're shooting RAW.

Now, with my Canon DSLR, I've fell in love with the light equality of CineStyle. Holding off all other opinions about this picture style, the lighting equality of it is simply amazing, it picks up information in the whites and blacks that Standard (or say, Cinema, for Panasonic) just do not. I've seen the tests, contrasty picture styles do indeed create seas of blacks and whites where CineStyle and other ultra-flat picture styles do add stops of light. (See: http://www.similaar.com/foto/flaat-picture-styles/index.html ) This creates an image that is a few stops closer to film, and is also possibly half-way toward what more advanced digital cameras like Red, Alexi, and others churn out.

Here's where my other opinions come into play though - for CineStyle to look natural in most situations, you do have to add contrast to the image, without a doubt. Different films have different levels of contrast, but there is a certain point where, even just judging by skin tones, you can tell the image is suffering from a lack of contrast, thus lacking any sort of realism or professionalism. The Tree Of Lifeis probably for example the flattest theater film ever made, it truly bordered on beautiful and too-flat most of the time, and it looks this way because hardly any artificial lighting was used on the film - the filmmakers had to go flat as they could and teeter on the edge of contrast and flatness. What this means is that no matter how much you want to preserve your lighting equality, you had to add contrast to CineStyle, and it will slowly make it's way back toward Standard, Cinema, or Neutral, whatever factory-made profiles there are.

But here's my main point - I'm positive that even if you do have to add contrast to CineStyle, that you don't have to add as much as to make it look like Standard, or even Neutral. There probably is a point well below the contrast of those picture styles where almost anything from CineStyle looks realistic. In the end, because of complex editing techniques, I'm even sure that you can merely add contrast to the mid-tones, leaving the whites and blacks alone. I would bet money that even in large budgets and expensive shoots, that somewhere down the line, this is what the professionals want to do as well. For example with Black Swan, with the subway train scene shot on Canon, a production photo reveals that the scene was shot all-natural with the exception of a little light panal shining on Portman's face. I'm pretty sure they didn't want to use Standard for that shoot.

Now this brings me to the Panasonic Gh2. We all know that the Gh2 has higher levels of contrast than any of the flat styles from Canon, I've seen all the tests to prove it. However, what I'm not sure is if this contrast level (I'll pick Nolstalgic, as it seems to be Gh2's flattest profile), is over or under the point where CineStyle requires contrast. I guess a rephrase of this question would be - if you took a flat CineStyle product, and added contrast to the mids while preserving the blacks and whites, would this equal a more information-packed, more cinematic image than if you shot on Nolstalgic for Gh2 and flattened it a little in post (or even just left it alone) or would the Gh2's product come out on top with the advantage of less work (already applied contrast?) In the end of the uber-professional day, with the professional editing techniques applied and finding the realistic point for CineStyle or the Gh2's limited selection of Picture Styles, which camera comes out on top in terms of lighting equality?

P.S. Please keep this based on lighting equality as I know there's plenty of benefits to both cameras, like how beautifully and film-like the Gh2 handles highlights, has clearer video resolution and less moire, and has sound monitoring, while the Canon DSLR's have more professional LCD/Viewfinders viewings and have hundreds of picture style options, and possibly are able to pack more visible information into a picture style.

 

[jeree]

P.S. Please keep this based on lighting equality as I know there's plenty of benefits to both cameras, like how beautifully and film-like the Gh2 handles highlights, has clearer video resolution and less moire, and has sound monitoring, while the Canon DSLR's have more professional LCD/Viewfinders viewings and have hundreds of picture style options, and possibly are able to pack more visible information into a picture style.

yeah, i woould like more info also. i have a canon 7d

 

[Kholi]

I don't even get what you're asking.

What does lighting equality mean?

And natural light doesn't mean NO LIGHTING, you still use gear to shape...

 

[RPLace]

Basically, I'm asking which camera, after a hired uber-professional editor goes in and edits all the right tones of light and color (we're talking a cinema-grade editor here, not a high school kid using free contrast/saturation software), bringing out and yet preserving all the right informations in an image, which camera would prove more cinematic when shooting mostly natural environments where not much manipulation other than good cinematography tricks and correct weather can be added on set and location: Canon, where a few picture styles give you the option of adding another stop of information in the blacks and whites at the sacrifice of realistic contrast and color in the non-edited product, or the GH2, which seems to record around two stops less of information than Canon's CineStyle in the blacks and whites because it, plain out, is about as contrasty as Canon's Standard profile.

Wether anyone likes it or not, higher end digital cinema cameras record, what, 8+ more stops in the blacks and whites than a DSLR? I believe I even seen a test once that proved it was 12. I can't vouch for what the big folks in the industry buy those cameras for (beyond 4K), but I'm also sure that the ungraded footage that comes out of these cameras aren't merely edited with iMovie with a few contrast-based "Film Look Modes" applied. I believe that the footage is taken and the contrast is edited at every light interval between white, grey, black, with the colors closely monitored because, after all, some colors merely by being certain colors can register with more luminence on a black-and-white scale, depending on the white balance and in-camera saturation, and such. That's how so much information in a flat image is "brought out", while you cannot "bring it back" if say, you shot with a HV40.

But since we're stuck comparing Canon to Panasonic, the question pretty much is - is it better to shoot with Canon, where picture styles can give you a couple more stops of information that *may* be able to be preserved through the levels by an experienced editor in post, or does the possibilities of Cinestyle only create, with professional post editing, what you can pretty much accomplish with the GH2 right out of the box? In terms of achieving a real-eye look, thus opening yourself up to a few more styles of shooting, which camera's product equals a better visual in the final, edited product? Does the Gh2 "handle everything best" in-camera (since contrast is already applied), or does a better product come out of taking the time to shoot flat on Canon, with added information, and detailedly grading in post?

 

[Kholi]

Your posts are confusing.

If you're asking if, all things created great and equal, which camera will produce the better image it's going to be a hacked GH2.

Cinestyle isn't giving you anything but a headache in post. What you think you gain in shadows is just noise and an extra pain in the butt to correct. And you will never get any of the Canon DSLRs to match the GH2's resolution, especially the hacked camera. It will always have that edge.

I'll put paypal cash on it that you will get a TECHNICALLY better image out of a hacked GH2 despite the film modes, but you make like the soft, mushy, "dreamy" look of Canons. That's subjective.

 

[simonpwood]

The GH2 definitely produces a sharper image. I had a play with one this week and its quite shocking to see the difference on a big screen in terms of the resolution. It absolutely blasts the Canon DSLR's out of the water in the detail.

However the flip side to the sharper image seems to be sharper noise and sharper banding.

The noise can be quite distracting, especially on a bigger screen (its not so noticeable on smaller computer screens, so vimeo and youtube are not the best formats to base the image on). Different ISO's and careful exposures still seem to produce noise in the dark areas; but noise reduction software should deal with it when it can't be minimized in camera.

The banding is really evident in the sky under certain conditions, and also surprisingly evident on flat surfaces (like a wall that has a light source from one side). Harder to work around that in post so you have to shoot with it in mind.

So everything is still a compromise and there is still no perfect DSLR solution; pick which image works best for you under the circumstances.

As for me I'm going to try to use both a GH2 and a T3i for an upcoming shoot, and using the 2 different images under different conditions to use their strengths (and avoid their weakness).

Has anyone tried working out some color profiles that would match them closely in camera?

 

[Kholi]

Hence, subjectivity: hack noise is awesome to me. Looks great big-as hell and on a small screen (seen both). Even when it gets bad, a very light noise reduction is sweet. Thing is with the Canons, the noise is a bunch of mush so it looks cleaner. That was my final discovery with the 5D. Yeah, it gathers more light, but it's noisey still, just uber soft so it looks cleaner.

Really isn't. GH2 Hack footage you can denoise and still end up with a better image.

I will agree on the banding, but again if any of the Canons actually had better compression you'd see a lot of the banding there as well. It's just being covered up by sub par product output to cards.

Just don't shoot below 320 and banding's not an issue.

Because we're selling the 5D off (will just be shooting Epic and GH2 now) I forced myself to see how much of a loss I'm taking on the larger sensor... and realized I wasn't going to miss it that much at all. Nothing that fast glass and backing up can't take care of... and I don't have to deal with the soft image, aliasing, moire, overheating, so on and so forth.

 

[RPLace]

I'm not asking about the resolution (which I know is better on the Gh2, though the tests I'm running so far don't exactly seem to show any major differences - and all the tests I've seen online sometimes have me questioning if lenses, slight focus adjustments, and such come into play...), but I'm asking about which camera simply stores more information in the picture.

Basically, digital cameras pick up light in different levels of white, grey, and black. The most expensive cameras out there, can shoot an image so dense with detailed information that literally it has a "pixel" for every 10 levels of light (think on a scale from 1 - 100). Now, in terms of how sensitive an electronic sensor is to both sunlight and electronic light, they usually are created pretty equal, same sensitivity as in a HV40 as in a Red, as long as contrasty picture styles aren't making things brighter or darker than what the sensor actually can pick up - Canon's Standard profile is proven to be problematic this way. CineStyle, with 10 points of contrast applied in post out of 100 (or whatever), is a very decent representation of what an electronic sensor captures in terms of being sensitive to electronic light, sunlight, and shadows. If anyone likes the light equality (less harsh shadows, less harsh highlights, more of a "real-eye" look in terms of sensitivity to shadows and lights) of CineStyle, they like what electronic sensors are indeed capable of.

Where, say, an HV40 is different from a RedOne in terms of information and light equality is this: for one, cameras like the Red and Alexi can produce an image that has the color and realism of Canon's Standard profile, but with the light equality (less harsh highlights, more sensitive shadows) of CineStyle. It can just do that. And if you don't want the camera to do it for you, it also can give you an extraordinarily flat image that, even without shooting Raw, has so many information-packed stops of light that you can pretty much change it into whatever you want - you literally have pixel representation of every 10 levels of light or so, from a 1 - 100 histogram. Now, what makes a DSLR or Camcorder different from this, is that even in the flattest of picture styles, the sensors and such can't process such information, and the information is cut down to where there are only like 5 different levels of light that a pixel is capable of before, in terms of changing the brightness or contrast, the pixels "darken or lighten" and blend with the lighting levels of the surrounding pixels (that's why post work is limited in the black and white ends of the spectrum).

All in all, long story short, with cheap cameras, you usually have to choose between having an image that looks more-or-less realistic, with either the choice of having seas of whites and blacks OR artificially lighting your scene really well (or shooting during the day) or you have the choice to shoot flat as crap, have a much more realistic light distribution over your image (the distribution that your sensor is trying to output anyway), but well, your image will probably look flatter than professional looks really call for unless you have yourself an editor or editing program that can edit the image's lighting levels individually (think of curves). Your DSLR sensor, wether Panasonic or Canon, tries (long story short) to give you the same image that a Red can, but, because its lack of power, you either have to settle with A. Quite a bit of flatness to achieve the native sensor image., B. Seas of Whites and Blacks at the ends of the lighting spectrum of the image (1 - 100), or C. Accept that the trick to DSLR filmmaking is that really, using much more artificial lighting in combination of a contrastier picture style (say, Panasonic Smooth) than you would for a Alexa or Red, gives you the best product.

However, from what I've seen, in the right circumstances especially, that picture styles like Cinestyle do give more information in a final product than say, Canon Standard or Panasonic Smooth. It's just what I've seen. The cost is that CineStyle is flat in color. It has an amazingly powerful and life-like contrast level (meaning this as saying out of any sub-$1,000 camera, this picture style represents best what our eyes can see, which is what people expect to see in a cinema), however, the price is horribly flat color and that the contrast is aesthetically flat and almost inverted-looking. But there is more information there, tests prove it. My big question is again: in natural situations where contrasts can get a bit crazy (forests, cities, bridges, parks, whatever) and you just can't artificially light from most angles, can the few extra stops of information make a huge difference? Or does the GH2 pretty much cover the "final product" zone in-camera for what any DSLR product is really capable of in the end? What I mean is that if you are editing the curves and levels of an image, does having a few stops more of information lead to a more realistic image in the end?

And most of all, since the Gh2's limited picture styles do have a few stops less of information than CineStyle, is it a proper trade-off for slightly better resolution, less moire, horribly-confusing camera controls, finicky exposure (what, every picture style of the Pana jacks the exposure up and down a stop? wtf?), finicky colors, a horrible LCD screen, and camera files that don't work with anything on my computer? The downside to Canon seems to be moire and a bit less resolution, so I'm still unsure which camera suits me best. In the end, I want the most options in post (am VERY critical of an image), but I also want things to be easy and I don't want to go around shooting CineStyle all day if Panasonic's Smooth generates something extraordinarily cinematic right of the box. So far, my tests show that switching from Canon CineStyle to Panasonic Nolstagic is like switching from Canon CineStyle to Canon Standard (not good!) but if the Panasonic picture styles at least have a little unseeable information in there that can be brought out with de-contrasting, then...who knows.

 

[Kholi]

I'm not asking about the resolution (which I know is better on the Gh2, though the tests I'm running so far don't exactly seem to show any major differences - and all the tests I've seen online sometimes have me questioning if lenses, slight focus adjustments, and such come into play...), but I'm asking about which camera simply stores more information in the picture.

No, they don't come into play. Not a single Canon DSLR out right now (or any other DSLR for that matter) can resolve anything near what the GH2 is resolving. There are charts etc floating around and, if you don't like those, there are real world tests. It is what it is, you won't find anything different than anyone else has already found.

I own a 5D and a GH2, same glass on both cameras (newest, fastest canon primes) and the 5D cannot hang with the GH2 shot for shot IQ wise. The advantage is larger sensor real estate, that's it.

Basically, digital cameras pick up light in different levels of white, grey, and black. The most expensive cameras out there, can shoot an image so dense with detailed information that literally it has a "pixel" for every 10 levels of light (think on a scale from 1 - 100). Now, in terms of how sensitive an electronic sensor is to both sunlight and electronic light, they usually are created pretty equal, same sensitivity as in a HV40 as in a Red, as long as contrasty picture styles aren't making things brighter or darker than what the sensor actually can pick up - Canon's Standard profile is proven to be problematic this way. CineStyle, with 10 points of contrast applied in post out of 100 (or whatever), is a very decent representation of what an electronic sensor captures in terms of being sensitive to electronic light, sunlight, and shadows. If anyone likes the light equality (less harsh shadows, less harsh highlights, more of a "real-eye" look in terms of sensitivity to shadows and lights) of CineStyle, they like what electronic sensors are indeed capable of.

My primary acquisition system is RED. That means I reach for an Epic before I shoot anything else. I understand RAW, and Cinestyle doesn't come anywhere NEAR it. It's a farce, and probably the worst picture profile you can shoot for any serious grade. The only thing it does is flatten the image out to simulate raw, but introduces a good two dB of noise to mids and lows, further decreasing the IQ of the image.

Applying NR to this footage kills your detail.

However, from what I've seen, in the right circumstances especially, that picture styles like Cinestyle do give more information in a final product than say, Canon Standard or Panasonic Smooth. It's just what I've seen. The cost is that CineStyle is flat in color. It has an amazingly powerful and life-like contrast level (meaning this as saying out of any sub-$1,000 camera, this picture style represents best what our eyes can see, which is what people expect to see in a cinema), however, the price is horribly flat color and that the contrast is aesthetically flat and almost inverted-looking. But there is more information there, tests prove it. My big question is again: in natural situations where contrasts can get a bit crazy (forests, cities, bridges, parks, whatever) and you just can't artificially light from most angles, can the few extra stops of information make a huge difference? Or does the GH2 pretty much cover the "final product" zone in-camera for what any DSLR product is really capable of in the end? What I mean is that if you are editing the curves and levels of an image, does having a few stops more of information lead to a more realistic image in the end?

Read above. There's no difference shooting GH2 or Canon, you've got a minimal leeway with both systems in the BEST finishing environment. Any high-end post house (since that's what you're talkin' about) will tell you that you should shoot as close to your final look as possible in camera when dealing with Canon footage. That's just fact, you aren't fooling any DI process by shooting Cinestyle, you're just fooling yourself into thinking you're getting a ton of DR and you really aren't because of the SNR.

And most of all, since the Gh2's limited picture styles do have a few stops less of information than CineStyle, is it a proper trade-off for slightly better resolution, less moire, horribly-confusing camera controls, finicky exposure (what, every picture style of the Pana jacks the exposure up and down a stop? wtf?), finicky colors, a horrible LCD screen, and camera files that don't work with anything on my computer?

My LCD shows me near to what I'm getting as far as color goes, and colors are never finicky. I nail skin every time I go out to shoot now. No Finicky exposure, I shoot Smooth and everything's fine. LCD articulates, works great for me and so does the EVF when I go to use it.

If your computer is having trouble seeing camera files I suggest you look into updating your NLE. I don't see anyone else having trouble with it.

The downside to Canon seems to be moire and a bit less resolution, so I'm still unsure which camera suits me best. In the end, I want the most options in post (am VERY critical of an image), but I also want things to be easy and I don't want to go around shooting CineStyle all day if Panasonic's Smooth generates something extraordinarily cinematic right of the box. So far, my tests show that switching from Canon CineStyle to Panasonic Nolstagic is like switching from Canon CineStyle to Canon Standard (not good!) but if the Panasonic picture styles at least have a little unseeable information in there that can be brought out with de-contrasting, then...who knows.

You were already given information in another thread, I gave it to you. Others told you what to do. This seems like a rehash of that thread.

You shooth Smooth or Nostalgic, balance to a grey card and either use 5DtoRGB to convert your files with a REC709 setting OR use an NLE like Premiere and drag the low end of your image up. It's really that simple.