Why many says that Alexa has real 11,5 stops and Epic only 5,5 indeed?

[Jonas Nyström]

As you may know I'm a Epic shooter, owner and a Red fan myself. Just curios.

 

[Alvise Tedesco]

wide dynamic range and decent latitude.

I thought dynamic range and latitude were the same thing. Where's the difference?

 

[Mark Britton]

Dynamic range refers to the ratio between the smallest value and largest value that can be recorded. In photography, this is measured in stops.

Exposure latitude refers to how much underexposure or overexposure still produces an acceptable image. It really isn't a measure, it's more a judgement call based on the scene, recording media, and aesthetic tastes.

 

[Peter Moretti]

Keep in mind that I said it was about a stop difference without HDRx, which is not all that significant except in the extreme highlights -- in practical shooting, either camera would be considered to have a wide dynamic range and decent latitude.

David, do you feel the same could be said for the F35/Genesis and maybe D21? Because from my humble perspective, those cameras were capable of producing incredible images as well. And then there's the F900 going even farther back.

I'm honestly starting to feel that the revolution going on digital cinematography has been mosly about equipment availability and price. While image quality has certainly advanced, I feel perhaps so far it's been more of an evolution.

Put another way, looking back at films like "Secretariat" shot on the Genesis, I don't ever recall thinking that the image was lacking. But looking back at the fact that an F35 was about $150K and that paractically all the of the digital cinema cameras were rental only, I think "Oh my!"

 

[Mike Kushner]

So is resolution if you do digital image manipulation in post like `THE HOBBIT´ etc.

Besides that, these 5.5 stops latitude are just misinformation, 5K versus 2K is not.

It's not that pixel count or number of "K" is misinformation - the pixels can be physically counted. The misinformation or confusion stems from the fact that the pixels are different sizes, shapes, have different color reproduction and sampling specs. A 10K camera with bad color reproduction and big chunky pixels might not look as good as a lower pixel count camera with superior photo sites with better bit depth.

Wh

 

[David Mullen ASC]

David, do you feel the same could be said for the F35/Genesis and maybe D21? Because from my humble perspective, those cameras were capable of producing incredible images as well. And then there's the F900 going even farther back.

I'm honestly starting to feel that the revolution going on digital cinematography has been mosly about quipment availability and price. While image quality has certainly advanced, I feel perhaps so far it's been more of an evolution.

Put another way, looking back at films like "Secretariat" shot on the Genesis, I don't ever recall thinking that the image was lacking. But looking back at the fact that an F35 was about $150K and that paractically all the of the digital cinema cameras were rental only, I think "Oh my!"

Sure, you can do cinematic work on an F35, D21, etc. as has been proven already. Those cameras were more in the 12-stop range, not 14-stops, so there is a bit more clippiness in the highlights to address. Plus there are a host of other image characteristics due to the recording formats, compression, bit depth, resolution, color science, etc.

Regarding resolution, a "K" label can also be misleading because generally they aren't using measured line resolution figures after debayering but are just counting pixels on the sensor. But obviously the Epic records more resolution than the Alexa, before or after debayering. By the way, a DR measurement should be fairly scientific -- most people should get results within a 1/2-stop of each other if they shoot a chart correctly. The main difference often comes in whether they count steps visible in the noise floor, and that's where some subjectivity comes in because what's "too noisy" is a bit debatable. But generally that just affects the DR results by a stop.

I think there are definite image improvements with the cameras over the years that make them behave more like film in terms of DR -- to the laymen, some of it may be subtle improvements but if the goal was to replace 35mm film with something equivalent or better, we couldn't settle for HD cameras with 12-stops or less of DR, only the newest generation has really hit the goal of being similar to 35mm, if not better in some areas.

 

[Marc Wielage]

40% of the producer around here who had extreme bad experience with Alexa swear for their life, they wont use Alexa anymore. We did rescue their project in post. But the damage has been done. I have seen and work with Alexa footage a lot and it is good, but EPIC is better.

I have never in my life heard Alexa criticized by producers for technical problems at the LA post houses at which I've worked. The only issues I've ever heard debated were 1) cost and 2) size & weight. Everything else was pretty stellar, and the workflow is very straightforward and pretty easy, even with ArriRaw.

I have defended Red several times against people who speak against it out of ignorance or misinformation. There are some things that are actual issues (fan noise, timecode unreliability, audio issues, lack of documentation), but there are workarounds for everything and the camera can look and function great if you know what you're doing, and the LA office has good support. The last five or six Epic or Scarlet shoots I've been on were flawless -- I'd guess not even 20 minutes of downtime in weeks of shooting, and even then, that was mostly for battery changes and drive changes. Not a big deal.

Producers are very, very superstitious, and once they're burned, they're very reluctant to go back and try again. This was no different than the 1990s -- I can recall a project shot on Fuji film that had shedding and dirt problems, and the producer swore never to use Fuji and again and went with Kodak. Problems were gone. He badmouthed Fuji for years afterwards, while we continued to use it on other projects without a problem. Go figure.

I constantly say, "it's very sad when a non-technical person is in a position of power -- like a producer -- and he or she has the final say on technical decisions that affect the work of the crew and the cinematographer." This is not a trend I like to see, because to me, choice of camera should be the DP's decision.

 

[James T Mather]

Maybe true, I had only shot Fuji on maybe twenty jobs. However, all stock related issues necessitating heavy post fixes or reshoots happened on Fuji - odd exposure fluctuations on one batch, dirty film necessitating post paint-out, peeling swarf in the gate visible during high speed shots, a single roll that looked like it had been exposed back-to-front (as in through the back of the emulsion), weird intermittent neg marks down the centre of the film etc. I personally share your producers reluctance to use Fuji. Through these events, I came to distrust it - although producers were always pushing it because it was less costly than Kodak. That all said, I knew DOP's who swore by it and I quite liked the cooler look it produced - I just didn't trust it's consistency.

 

[Álex Montoya]

I don't really like the Alexa's look. I believe that even in the flicks shooting RAW with it, like 'Avengers', it has some kind of pastel-ish colors that I've started to dislike.

I read somewhere that some colorists stated the same, that it doesn't give too much room for tinkering.

IMO, the Epic is more flexible and more realistic.

 

[James T Mather]

Possibly true - although I sat through In Time and was completely unaware it was shot digitally - I thought it was 35mm film - and I liked the look and it had that softness and "film barrier" I felt was absent in Total Recall and Ghost Rider I mentioned earlier. That got me interested in the Alexa.

Although as I often think, it's the guy behind the camera who makes all the difference and they had some guy called Deakins. He's not bad. If he sticks at this lighting thing he could really go somewhere.

 

[Mark Toia]

Are you guys blaming the camera for what the colorist had created ?....

 

[James T Mather]

No Mark, I'm not - if you read my earlier post you'd understand what I was talking about. If you still don't get it, I understand.

EDIT: Sorry Mark - I assumed you were referring to our recent posts as opposed to the thread starter regarding 5.5 stops of latitude - if so, then as above, if not then I agree with you. Many a slip between the camera, the grade and the print. This 5.5 stop stuff is just nonsense, plain and simple. Although, in one sense, could be a product of Red itself - let me elaborate (before the Red Faction shoot me): Initially, Red cameras were affordable - suddenly many people who wanted to be DOPs who hadn't been to college nor stood on a set but watched Blade Runner a bunch of times decided to cash in their bank balance and chase their dream of becoming DOPs - all the while claiming that the elitist Old Guard were afraid to let younger blood come in etc - in short, there was a flood of really inexperienced cameramen hitting the market holding an ace that was in demand at the time - the camera itself. As a result they got jobs and possibly didn't know how to shoot to contain latitude or with any clear understanding of the limits of the uptake medium, contrast ratios, fill light, basic three point lighting, noise floor etc. Mishandled, any camera will clip or crush and digital cameras offer a lot of buttons that will get you out of a tight spot - for a price (360 degree shutter, gain function etc). The price is can be a number of things : Latitude, Grain, DR etc.

Recently, I saw the most shocking instance on a set which would have tested any format by a DOP (I was merely observing - although the confused horror must have been apparent on my face) which would have led to zero latitude options in rushes due to the extreme nature of the lighting. This is supposition on my part - I'm not dissing anyone's work and have seen some great stuff here on the forum - but I wouldn't rule out pilot error and generally poor work. One of the positive aspects of the so-called elitist system was that people had to earn their stripes in some fashion or another before declaring themselves photography geniuses and getting work. I've been at this twenty years and I'm still learning every day.

None of this is what I was referring to in my earlier post. Paul Cameron is one of my favourite DOPs and the lighting in TR is terrific.

 

[jake blackstone]

I don't really like the Alexa's look. I believe that even in the flicks shooting RAW with it, like 'Avengers', it has some kind of pastel-ish colors that I've started to dislike.

I read somewhere that some colorists stated the same, that it doesn't give too much room for tinkering.

IMO, the Epic is more flexible and more realistic.

That's the definition of hearsay with no basis in fact...
I can't imagine ANY colorist stating such nonsense. I use Alexa images on daily basis and I couldn't be happier with the look and the amount of flexibility Alexa image affords me, even with Prores. BTW, same goes for Red, unless black shading wasn't performed correctly, which, unfortunately, I see quite often...

 

[Vadim Bobkovsky]

I was interested to see and finally went to Total Recall yesterday in the cinema - the mix of Paul Cameron, C-series lenses being interesting - and thought the image looked great. Really great. Almost too great if I'm being honest.

One aspect of it I felt was that digital somewhat removes that barrier that film creates - leading to that "distance" that makes the so-called film look a desirable holy grail for digital sensors. I'm talking about image feel here. A kind of film softness. In terms of resolution there is no question that the camera is up to the challenge but in terms of look there is something almost too clean looking about the image in TR - a lack of texture or something. To me it was almost too perfect. I Would love to see this aspect examined in the Dragon sensor. I never quite believed the movie (and trust me, I'm an easy sell in that regard). The art direction, effects and lighting are incredible (props to Paul Cameron) but recently there is something happening in cinema where I find myself observing digitally acquired films as being "actors and a set" as opposed to "these events are really happening". This aspect varies from one digital system to another - assessing what this ingredient is has been preoccupying me recently - is it DR, colour rendition, digital projection, camera gamma curve or dare I say it, too much resolution? (gasp). I realize the pixel counters are going to murder this one but do understand that I mostly shoot Red these days and own 2 Epics. I'm not evangelical about cameras - only pictures.

Ps: I have noticed that using old glass knocks this effect back a lot as does shooting RedGamma 2 as opposed to 3 and generally shooting further down the exposure curve - which is why this surprised me as C-series are quite soft compared with, say, Hawks or E-Series and Cameron is pretty low on the exposure scale. Maybe it's time to drag out those old black pro-mists, or low-cons and start shooting at higher ASA settings to introduce grain. it could be that simple.

That's interesting, because Anamorphics and film grain appiled in post gave Total Recall (2012) pretty film-like "tactile" dynamic feel, especially considering it's action genre. I never felt it was too sharp or too clean, to be perfectly honest with you. Or maybe I just saw 2K copy on a really big-ass movie screen, so for me it was way softer and grainier than I wanted, hehe. Then again, there was so much CGI backgrounds and elements in almost every shot, I can't really say it was too "artifically digital" or whatever, that'd be strange. If anything, film grain applied at the final stages gave those sick chinatown scenes and Quaid running up and down the roofs more textured realistic feel. Just saying.

 

[Mike P.]

I was always under the impression that you could lower detail/sharpness simply enough, whereas to add detail/texture/sharpness (and have it look natural) is somewhat more difficult.

For example, lowering the "detail" settings in camera (or in RCX) usually gives the r3d images a very pleasant "film-like feel" look. Conversely, with 4k/5k footage, adding unsharp mask and then downscaling to 2k ends up looking incredibly natural even though, objectively, it's excessively sharp... I think you can do a similar thing with ArriRAW/2.8k, just not to the degree of tweak-ability that 4k/5k provide.

Realistically it's the same old story (although kind of counter-intuitive): the more post potential these camera systems provide, the more important proper cinematography and colour correction/grading become (as the formats themselves, in terms of "look/feel", are less forgiving than film despite being infinitely easier to "process"/tweak in post.)

 

[Marc Wielage]

I don't really like the Alexa's look. I believe that even in the flicks shooting RAW with it, like 'Avengers', it has some kind of pastel-ish colors that I've started to dislike. I read somewhere that some colorists stated the same, that it doesn't give too much room for tinkering.

I didn't find Avengers to be "pastel" at all, and I saw it here in LA both at a digital screening and later at a 35mm screening. Both looked fine to me.

I'd be curious where you read "somewhere" that colorists don't like the Alexa. The only colorist discussion forum I'm aware of is the Telecine Internet Group, and I can't recall ever seeing a negative comment about Alexa there. There have been discussions about workflow issues with all cameras, but picture quality is ephemeral.

 

[Nick Morrison]

Game of Thrones is shot on Alexa, and I think that show looks amazing. I think it goes w/out saying that it's possible to film remarkable imagery on both Epic and Alexa. KNOWING your camera is more important than fretting too much which one is "better". Master the tool. That being said, RED's remarkable pricing and stunning specs for Epic and Scarlet makes RED the only choice for me.

 

[Maxi Claudio]

A Xerox has 5.5 stops of latitude

 

[Blair S. Paulsen]

Pushing quickly past the DR nonsense (the troll has been fed) we find ourselves once again addressing how fundamental differences in acquisition methodology send different messages to the viewer. Whatever the technical specifics, content creators are inevitably doing an interactive dance with the audience where stylistic choices function as cues referenced to what has come before. Perhaps James is more aware of inferences that feel like a "mis-match" because like many of us he is charged with making myriad choices in photography to enhance (or counterpoint) particular aspects of the story/message/sub-text so he notices when they seem out of sync. The general audience may have little conscious recognition of these aspects but that doesn't mean that they aren't following the scent.

FWIW I think grain management is a more "narrow" factor than people imagine and while it can provide an analog "cloak" for the digital source its not a panacea.

In any case, my guess is that digital acquisition of live elements facilitates a "keep it clean approach" on VFX heavy shows to simplify (and lower the cost of) acceptable matching of elements. One might even argue that tentpole titles with video game and other tie ins are seeking greater marketing synergies (reusing plates for video game backgrounds, making a cartoon version, etc) by keeping a consistent "look" across various presentations of the "brand". (Let's leave the art vs commerce ruminations for another day ;-)

IMO the operation of the "shutter" (or its virtual equivalent) imparts a subtle yet powerful quality to the perceived motion characteristic that is the biggest "tell" of the digital vs film experience. I also agree with those who contend that the constantly changing random grain pattern in film stock gives the image "life" and depth. At the end of the day, whether digital formats "should" be manipulated to more closely mimic film may be a moot point as the visual language and display technology evolves. For now it seems to be more of a litmus test of aesthetics and generational conditioning.

Cheers - #19

 

[Karim D. Ghantous]

I think it is all what you are used to seeing. If you've been watching film for 40 years, it's "comfortable" and "normal." Twenty year olds don't have the same deeply ingrained biases. The film "look" will not be the standard in another twenty tears, any more than there are very many people are left who argue the superiority of vinyl. Film is at the peak of a 100 year evolution. Digital is in its infancy with an unlimited future.

I profoundly disagree. Perhaps we should label this the argument to familiarity (if anyone knows Latin can you please translate that?). I grew up with both interlaced video and projected film. And when I became aware of the difference as I grew up, I found that the film (or at least the telecine) was superior. In every case. Of course, telecine transfers are almost always 'clean' so that had nothing to do with it.