kodak's portra 400 vs 5d mark 2: latitude (in english)

 After hearing all the excitement about Kodak's "new" (2010) Portra 400 stock, especially it's exposure latitude, I thought it would be best, to do a test on my own.
 You couldn't say the methodology was over complicated: I just threw a set of items in the frame, that all represent different luminosity values: the gray card, the white porcelain and the black lens of the light meter. I used the incident metering mode, placing the lumisphere right in front of the gray card. I scanned with a Minolta SD 4, with Vuescan in raw, than inverted with Colorperfect plug-in. For comparison, i used my 5d mk2. (Same lens, ISO, that goes without saying.)
 Before we begin, there are some things to be noted: first and foremost, this test is in no way scientific. It all happened because my curiosity overcame my laziness on a very gloomy Sunday afternoon. You will see some color banding on the scanned images. That is my fault, it's a scanning problem, and has got nothing to do with the film itself. I was just to lazy to rescan. Regarding the quality of the images, a pleasing look was not the goal here, I just maxed out recovery, fill light, and zeroed blacks and contrast, so that we could see everything that is in the raw file.
 So here it is, from dark to bright:


This is 7 stops under. Not much point in taking a picture under such conditions, but still.

portra

5d mk2

  6 stops under

portra

5d mk2

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 5 stops under

portra

 

5dmk2

 

 

  

 4 stops under

portra

 

5d mk2

  

 3 stops under

portra

5d mk2

 2 stops under  

portra

 

5d mk2

  1 stop under

portra

 

 

5d mk2

  Correct exposure

portra

 

5d mk2

  and here is where it gets interesting...

 1 stop over

portra

5d mk2

 
  2 stops over

portra

5d mk2

            

  3 stops over

portra

5d mk2

  4 stops over

portra

5d mk2

 

 5 stops over

portra

5d mk2

  It was at this point, where I just stopped shooting. By 5 stops, the LCD on the 5D was almost pure white (and I ran out of film too...get the irony...?). Anyway, I thought, that the white wall on the outside has got to burn out by now on the Portra too. It is at least 4 stops over the middle gray. And how wrong i was... Because this test doesn't cover the full range of Kodak Portra 400, it is only to give the reader, and myself, a general idea of the film's capabilities. Maybe one day I'm going to do a decent test, in a fully controlled environment. But, until then, here are some of my subjective conclusions:
 
  When it comes to highlights, film is...well, film. How about shadows? At about 4 stops under, the gray card is barely distinguishable from the black point. Ergo, if I wanted to be on the safe side, I'd put important shadow details not more then 3 stops under middle gray. It is remarkable though, how the overall image integrity is kept all the way down the 3 stops: that is 3200 ISO, without any compensation. Of course, you loose a lot of shadow detail...but that is still a very unusable image.
Digital clearly has more detail when it comes to shadows, but most of this detail is rendered pretty much unusable by color noise patterns, visible in even in the less extremely underexposed areas of the frame. Still, digital has the edge here, with about a stop or two. As I mentioned above, this test doesn't cover a very large range of measured luminosity values, but I'd put the latitude of the 5D at about 12 stops. The Portra? The wall on the outside still has plenty of detail, I have no way of knowing, how far from burning out it actually is. But it is 4 stops over gray, which suggests the portra has 13 stops of range, at minimum. Pretty damn impressive, if you ask me.

 So, what's the point lesson learned...? I'll let you decide for yourself. In any way, what people have been saying the past 10 years is still very much true: you should overexpose film, underexpose digital.

 Thanks for reading. Now, let's all just get up, go out, and shoot some film (or digital...whatever gets your bells ringing).