Jump to content


Photo

Dumb Question


  • Please log in to reply
9 replies to this topic

#1 Craig Knowles

Craig Knowles
  • Basic Members
  • PipPipPip
  • 182 posts
  • Director
  • Cleveland

Posted 23 January 2006 - 11:41 PM

If I take a 35mm negative, scan it, and then invert it in something like Photoshop, why does it not become the positive?
  • 0

#2 Craig Knowles

Craig Knowles
  • Basic Members
  • PipPipPip
  • 182 posts
  • Director
  • Cleveland

Posted 24 January 2006 - 12:51 AM

Example:

Posted Image
  • 0

#3 Dominic Case

Dominic Case
  • Basic Members
  • PipPipPipPip
  • 1357 posts
  • Other
  • Sydney Australia

Posted 24 January 2006 - 01:24 AM

This is one of the classic problems of scanning. WHat range of tones to you capture.

However you have captured this negative, the scanner has taken the perf area as "white" and set that to the highest bit values. But really, the brightet bit of the negative ypu are interested in is the unexposed fim between the perfs - which reads at about 131R, 46G, 15B. So for the blue componet of the image for example, the etire tonal range of the image itself is jammed up betwen 15 and 0, instead of 255 and 0.

As a result, there is almost no blue information in the image for you to invert.

Another thing - the gamma of negative is only around 0.5, so you would need to increase the contrast considerably when inverting.

You woiuld have a better result if you are able to set black and white levels when you scan.
  • 0

#4 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19759 posts
  • Cinematographer
  • Los Angeles

Posted 24 January 2006 - 01:25 AM

http://www.forum4des...ssage27908.html
http://www.photograp...opic-87337.html
http://www.photo.net...g?msg_id=0005Q1
http://www.c-f-syste.../ColorNegs.html

The most common advice seems to be to get a scanner that will do the conversion before you take the file into PhotoShop, because it will have a more sophisticated look-up table to correct the color.
  • 0

#5 Craig Knowles

Craig Knowles
  • Basic Members
  • PipPipPip
  • 182 posts
  • Director
  • Cleveland

Posted 24 January 2006 - 03:28 AM

Thanks for the replies, guys. I was thinking I must have gone completly insane...if a negative isn't a "negative", than what the heck is it!?

Luckily, this is not for a project of any kind. I was just fooling around, testing out some ideas.

Thanks again.
  • 0

#6 K Borowski

K Borowski
  • Basic Members
  • PipPipPipPip
  • 3905 posts
  • Camera Operator
  • I.A.T.S.E. Local # 600 Eastern Region

Posted 24 January 2006 - 06:31 AM

First off, cute puppy. Be sure to post a final corrected positive for those here who can't "read" C-41 negatives. Also, you should be ashamed of yourself using Kodak Gold 200. I've been Gold free for almost a year now ;-)

Your quandary as to why a negative isn't a true opposite is justifiable. I was just as shocked when I saw m first B&W negative and wondered why it was so clear and didn't have a heavy layer of orange. Originally, color negatives were just as clear as B&W negs. The negative was an exact inversion of what the final print (film or paper) would look like. The problem was (and is) that there was some cross-talk between the layers, i.e. colors representing colors they had no business representing. Since layers cannot all be in the same physical piece of space time, there was difficulty in getting clean channels from the cyan yellow and magenta layers. I still do not understand this completely myself, but apparantly the orange mask that Agfa invented (and Kodak later improved upon) "hides" this color crosstalk and makes for a pretty clean color image. So the orange was an imperfect solution for this imperfect world of ours. With that in mind, color negative was never designed with scanning in mind. It has now been reconciled pretty well with the latest scanner software, but it still isn't something originally intended as an application. This is why many still photogs still insist that scanninng chromes is better. TV shows used to "finish on film" because the film chains were very primitive and couldn't handle color negatives (although they could do B&W negatives because they were maskless).

In recent years, there have been several maskless negative films designed for scanning only. The problem of color crossover can now be minimized or effectively eliminated in the scanning process. However, scanners do such a good job with masked C-41 or ECN-2 negs that these films really haven't caught on well. We will have to see how Kodak's latest incarnation does. Consider investing in a good film scanner. You can probably get a model adequate for your needs for less than $200. Also, college photo departments have them available, so you might be able to get access to one for free.

Regards.

Karl Borowski

One more thing: your negative is flipped. The letters and numbers on the side are going in the wrong direction.

~Karl
  • 0

#7 Filip Plesha

Filip Plesha
  • Basic Members
  • PipPipPipPip
  • 1267 posts
  • Other
  • Croatia

Posted 24 January 2006 - 07:23 AM

If I take a 35mm negative, scan it, and then invert it in something like Photoshop, why does it not become the positive?


The scanner software does exactly what print paper or print film does to the negative image, and this is also what you must do manually in photoshop when scanning like this

Here are the steps to the print image:

1.Invert
2. a) Find white points and black points (Dmax and Dmin) Find a clear part of film base (base between two images). This is your black point (Dmin).
B ) Find your white point (Dmax). Usually hotspots (light reflections on surfaces) are your white point.
To be sure you are using Dmax of a given film, it is better to expose a frame to maximum and use that as a reference for that roll. For example: go outside open the lens to maximum and expose the film for a few seconds waving all around with the camera. You will get a pure white frame. Use that as your white point.

c) Now that you have determined which is Dmin and Dmax, Use levels in photoshop and select the determined areas for Dmin and Dmax as your white point and black point. FOR EACH CHANEL SEPARATLY.
That will also remove the orange mask in shadows and highlights.

3. Further fix the orange mask by adjusting color balance or any other tool of wish.


If you've done the first 3 steps properly you now have a digital image that represents the entire tone range from the negative from Dmin to Dmax and has neutral color balance.
The image is now ok, but a little flat...

4. Simulate the S curve of print film. Go to curves and make an S curve of chosen strenght depending on how much contrast you like.


Now you have a basic print image.
The trick with the final stage is that you have not cliped any image information, you have just distorted it to follow a high contrast S curve. This is what happens in an ideal print.
So you have a regular punchy viewable image with all the advantages of latitude that the negative has provided you.

You see, print papers and print films, don't just invert the image, they do all of the above:
they aim for the dmax and dmin, they remove the orange mask and they increase the contrast to follow the S curve.

Edited by Filip Plesha, 24 January 2006 - 07:26 AM.

  • 0

#8 Filip Plesha

Filip Plesha
  • Basic Members
  • PipPipPipPip
  • 1267 posts
  • Other
  • Croatia

Posted 24 January 2006 - 10:25 AM

Ok, here is how it looks in practice.
Here are two pieces of film scanned together, on the top is the first piece which is an edge from a Dmax frame, which I exposed for the purpuse of scanning like this.
The second piece is the image.

Here is the raw scan:

Posted Image


now I've inverted it:

Posted Image


Now I've set the white point and black point to match base color (blank film) and a fully exposed piece of frame on the top (Dmax):

Posted Image

This takes care of a part of the orange mask problem

Since I overexposed this image a bit to get more saturation, I've now darkened it to match normal exposure, I've also used color balance to compleatly remove the orange mask:

Posted Image

This is more or less the image as it on the negative, containing all the light range that is recorded from clear base to Dmax

The image is far to linear to look natural, so we have to "twist" this linear range to match the way
print film/paper sees negative, and the way our eyes see nature (more or less).
I change the curve to look like S carefully adjusting every part of it (one or two points won't do it, use multiple points for a more complex and precise curve)
Here is the final image, this is how a print would look like on say Kodak Royal paper:

Posted Image


This one is a bit punchier, but you can do it milder. This example shows how to simulate a higher contrast print. On paper it would look much better, that's why for electronic displays it's better to go a bit softer (higher contrast images don't look so pretty on monitors as they do on paper)

Take a look at highlights. Even thought they seem bright, they are not cliped, the detail is not lost, it's just "gathered" around white level (the result of using curves instead of levels). This is where film beats digital capture. Linear digital capture would clip the highlights whereas a film print compresses them around white point, so even though they seem bright, when your eyes adjust you still see all the details that were in the negative.

The film is Kodak Portra 160VC

And here is a reference, here is how the scanner software thought it should look like:
Regular automatic scan
A bit softer, better for internet and monitors. But you can make it look like that too using the above techniques by adjusting the curve differently


P.S. Don't you just LOVE film?
  • 0

#9 Dominic Case

Dominic Case
  • Basic Members
  • PipPipPipPip
  • 1357 posts
  • Other
  • Sydney Australia

Posted 24 January 2006 - 09:15 PM

P.S. Don't you just LOVE film?

Yes, especially positive print film, which does all of the above for you without any fuss at all.
  • 0

#10 Craig Knowles

Craig Knowles
  • Basic Members
  • PipPipPip
  • 182 posts
  • Director
  • Cleveland

Posted 24 January 2006 - 09:22 PM

Wow. Thanks guys. This is one of the most interesting things I've learned here so far.

Yes, especially positive print film, which does all of the above for you without any fuss at all.


Haha! That's what I was thinking, too.
  • 0


Rig Wheels Passport

Abel Cine

CineLab

Paralinx LLC

FJS International, LLC

Willys Widgets

Visual Products

Broadcast Solutions Inc

Gamma Ray Digital Inc

Technodolly

Glidecam

Wooden Camera

Metropolis Post

Ritter Battery

Media Blackout - Custom Cables and AKS

Aerial Filmworks

rebotnix Technologies

Tai Audio

Opal

CineTape

The Slider

CineLab

Tai Audio

rebotnix Technologies

Technodolly

Visual Products

Metropolis Post

Media Blackout - Custom Cables and AKS

Glidecam

Opal

Paralinx LLC

Rig Wheels Passport

Gamma Ray Digital Inc

Abel Cine

Ritter Battery

Willys Widgets

Wooden Camera

FJS International, LLC

Aerial Filmworks

Broadcast Solutions Inc

The Slider

CineTape