Film versus digital with backlit subjects?

[andyptak]

A photographer friend, who shoots mainly digital, told me that he always switches to film for backlit subjects because film is superior when the light source is also in the frame. Is this his prejudice, or is there any truth to this? Thanks.

[eleanorbrown]

I completely disagree.  I have been working on a new series called "beyond the forest" and most everything is shot looking into the sun, with the sun covered by say one branch or trunk".  very dynamic light this way. (these were shot with a Canon 5D II converted to IR)  eleanor
http://web.mac.com/eleanorbrown/ELEANOR_BR...the_forest.html

A photographer friend, who shoots mainly digital, told me that he always switches to film for backlit subjects because film is superior when the light source is also in the frame. Is this his prejudice, or is there any truth to this? Thanks.

[bjanes]

I completely disagree.  I have been working on a new series called "beyond the forest" and most everything is shot looking into the sun, with the sun covered by say one branch or trunk".  very dynamic light this way. (these were shot with a Canon 5D II converted to IR)  eleanor
http://web.mac.com/eleanorbrown/ELEANOR_BR...the_forest.html

I looked at Eleanor's web site and her work is impressive. I don't have the equipment or experience that Eleanor has, but her observations can be backed up with experimental data from Roger Clark's web site. Roger discusses Exposure Lattitude and Dynamic Range of digital and negative film. Contrary to popular wisdom, Roger finds that digital has better dynamic range than film.

However, negative film and digital must be exposed differently for optimum results. Digital clips abruptly in the highlights, whereas film is more resistant to overexposure. Digital must be exposed so that important highlights are not clipped, and the shadows can be brought to the proper level, which is often possible because digital has a better signal to noise ration than film in the shadows. If you expose for the midtones, then film would offer more highlight protection.

[ErikKaffehr]

Hi,

I'd just point out that slide film an print film are totally different animals. Slide film doesn't handle overexposure at all. But, fotochemical processess saturate and don't clip. Digital can have problems with bright highlights where we can get some artifacts.

DR on digital is impressive, around 9-12 stops while it's no more than 5 stops on slide film. Print film, I don't know...

Best regards
Erik

I looked at Eleanor's web site and her work is impressive. I don't have the equipment or experience that Eleanor has, but her observations can be backed up with experimental data from Roger Clark's web site. Roger discusses Exposure Lattitude and Dynamic Range of digital and negative film. Contrary to popular wisdom, Roger finds that digital has better dynamic range than film.

However, negative film and digital must be exposed differently for optimum results. Digital clips abruptly in the highlights, whereas film is more resistant to overexposure. Digital must be exposed so that important highlights are not clipped, and the shadows can be brought to the proper level, which is often possible because digital has a better signal to noise ration than film in the shadows. If you expose for the midtones, then film would offer more highlight protection.

[Tony Beach]

A photographer friend, who shoots mainly digital, told me that he always switches to film for backlit subjects because film is superior when the light source is also in the frame. Is this his prejudice, or is there any truth to this? Thanks.

The point missed in the above replies is that film does not reflect in the same way a digital sensor does.  On a digital sensor light bounces off the low pass filter and into the lens elements (which is then recorded in the image).  Given that, I see your friend's point.

[andyptak]

His general feeling was that he got less flare and could capture highlights better. I have limited experience in this area, but this guy shoots directly into the setting sun frequently.

[bjanes]

Hi,I'd just point out that slide film an print film are totally different animals. Slide film doesn't handle overexposure at all. But, fotochemical processess saturate and don't clip. Digital can have problems with bright highlights where we can get some artifacts.

DR on digital is impressive, around 9-12 stops while it's no more than 5 stops on slide film. Print film, I don't know...

Eric,

My discussion specifically addressed print film. For the DR comparisons, see Roger's articles.

[bjanes]

The point missed in the above replies is that film does not reflect in the same way a digital sensor does.  On a digital sensor light bounces off the low pass filter and into the lens elements (which is then recorded in the image).  Given that, I see your friend's point.

Digital sensors have specular reflections, whereas film behaves more like a Lambertian reflector as shown in the images below for a D3 at two angles and for a F100 with film.

[attachment=19244:D3_img1_small.jpg]

[attachment=19245:D3_img2_small.jpg]

[attachment=19246:F100_img_small.jpg]

With film, the reflection is diffuse and any flare would be of the veiling type. With digital, the sensor reflection can in turn be reflected off the back element of the lens back to the image plane. With multicoated lenses, the reflection would be small and whether it is visible or not would depend on the shape of the rear element of the lens. Such reflections appear as a spot in the central portion of the image as shown here and here.

Its effect is often negligible. Here is an image taken with my D3 of my light table with a lens cap. I don't see any reflection. Also, Eleanor has shot both film and digital and prefers digital. Your reasoning is likely specious.

[attachment=19243:D3_Senso...fl_small.jpg]

[Tony Beach]

Digital sensors have specular reflections, whereas film behaves more like a Lambertian reflector as shown in the images below for a D3 at two angles and for a F100 with film.

With film, the reflection is diffuse and any flare would be of the veiling type. With digital, the sensor reflection can in turn be reflected off the back element of the lens back to the image plane.

Your own links show the problem.

With multicoated lenses, the reflection would be small and whether it is visible or not would depend on the shape of the rear element of the lens. Such reflections appear as a spot in the central portion of the image

That is only one manifestation.  When the sun is shining directly on the sensor, and especially at any sort of oblique angle, we can get severe ghosting (or what Andy's friend called flare).

Its effect is often negligible. Here is an image taken with my D3 of my light table with a lens cap. I don't see any reflection. Also, Eleanor has shot both film and digital and prefers digital.

So what?  Shoot into the sun with film and digital and come back with comparisons.

Your reasoning is likely specious.

Like I said, I can see Andy's friend's point, and I don't think it is specious or reflects [pun intended] a prejudice.  Maybe he needs to buy a different lens.  Many argue that this is a reason to choose primes over zooms; and Bjorn Rorslett discusses this aspect in many of his reviews of lenses (and only a few ever really get high marks).  Nonetheless, with the lenses he is using and based on his experience, I wouldn't jump to the conclusion that he's some sort of knucklehead.

[bjanes]

That is only one manifestation.  When the sun is shining directly on the sensor, and especially at any sort of oblique angle, we can get severe ghosting (or what Andy's friend called flare).

So what?  Shoot into the sun with film and digital and come back with comparisons.

Shooting into the sun with either digital or film will cause severe problems. However, most backlit subjects do not include direct sun. You seem to be basing your conclusions on some theory and have no data to back up your assertion. I did shoot a severely backlit subject and saw no problem. Now the ball is in your court.

Like I said, I can see Andy's friend's point, and I don't think it is specious or reflects [pun intended] a prejudice.  Maybe he needs to buy a different lens.  Many argue that this is a reason to choose primes over zooms; and Bjorn Rorslett discusses this aspect in many of his reviews of lenses (and only a few ever really get high marks).  Nonetheless, with the lenses he is using and based on his experience, I wouldn't jump to the conclusion that he's some sort of knucklehead.

Now you are changing the subject from film vs digital to the subject of flare in lenses and mention Bjorn's reviews of different lenses with flare that could affect either medium. Your reasoning is very tangential and disconnected. Where does Bjorn discuss the problem of sensor ghosting?

[Tony Beach]

I did shoot a severely backlit subject and saw no problem.

There as no sun (or other genuinely bright light source) in the shot, so it doesn't address the issue that concerned Andy's friend.  The question is which medium does better under adverse conditions, so start with a shot taken with digital that is in fact impaired and then take the same shot with film and compare them.  My bet is that is what Andy's friend has done, and is why he uses film for those situations.

Funny thing is, I'm somewhat agnostic on this issue; but I am aware that there is an issue and I can see why Andy's friend does what he does.  Andy asks if his friend is just being prejudiced against digital, but his friend mostly uses digital, so that hardly makes sense.  You on the other hand, do seem prejudiced, since the definition of prejudice is to pre-judge something (i.e., without actually testing it).

Now you are changing the subject from film vs digital to the subject of flare in lenses and mention Bjorn's reviews of different lenses with flare that could affect either medium. Your reasoning is very tangential and disconnected. Where does Bjorn discuss the problem of sensor ghosting?

I'm not going back to read your previous post (I'm frankly tired of reading your posts, and have now put you on my Ignore User list), but you were the one that raised the lens issue -- I believe it was in one of your links -- something to do with lenses optimized for digital photography to minimize the reflection issue that is widely understood to be an issue with digital sensors.

[EricWHiss]

Probably the friend of the OP who uses film for backlit scenes is wanting to expose the subject properly without worry about ETTR and blown highlights.  I find film superior in this regard as there's always something still there and no weird purple sensor bloom.

[bjanes]

I'm not going back to read your previous post (I'm frankly tired of reading your posts, and have now put you on my Ignore User list), but you were the one that raised the lens issue -- I believe it was in one of your links -- something to do with lenses optimized for digital photography to minimize the reflection issue that is widely understood to be an issue with digital sensors.

Tony is ignoring my posts and I can understand why--he hates being proved wrong, such as was recently done with regard to setting a Linear Tone Curve in Nikon cameras. Really, this forum is for collegial exchange of information. but that is hard to do with someone who thinks he is the smartest person in the room and needs to prove it to everyone in that room. That dust off began when I agreed with him that WB does have a huge effect on the histogram, but added that one should not ignore the tone curve.  

[Peter_DL]

... Roger discusses Exposure Lattitude and Dynamic Range of digital and negative film. Contrary to popular wisdom, Roger finds that digital has better dynamic range than film.

However, negative film and digital must be exposed differently for optimum results. Digital clips abruptly in the highlights, whereas film is more resistant to overexposure. Digital must be exposed so that important highlights are not clipped, and the shadows can be brought to the proper level, which is often possible because digital has a better signal to noise ration than film in the shadows. If you expose for the midtones, then film would offer more highlight protection.

Bill,

Just out of interest,
would you have a yellow filter, to be placed in front of your Stouffer step chart?

I think that’s the sort of comparison which needs to be done in order to finally resolve this matter and to simulate backlit subjects or straight sunset shots.

With increasing exposure, and with classic transparency film, I’d expect that luminosity differences are getting lost soon while the yellow color is hold surprisingly long before turning to white. Whereas with digital capture, DR may be much higher to preserve luminosity differences between the steps, but I’m (still) not convinced if this is valid compared to film color clipping as well.

Admittedly, I did not address this matter since that time.

Best regards, Peter

--

[Daniel Browning]

Contrary to popular wisdom, Roger finds that digital has better dynamic range than film.

Roger provides lots of good data, but this is one area where his information incorrect by a wide margin due to experimental error:

http://forums.dpreview.com/forums/read.asp...essage=32912574

[bjanes]

Roger provides lots of good data, but this is one area where his information incorrect by a wide margin due to experimental error:

http://forums.dpreview.com/forums/read.asp...essage=32912574

I don't that that one critique of Roger's methods proves any experimental error. I am not familiar with the qualifications of that author. He seemed to assume that Roger exposed the negative film for the midtones. However, Roger did expose the film to the extreme and even scanned the film at two levels of gain in order to get the maximum detail in the highlights and then looked at the negative to make sure that all detail was accounted for. He also pointed out that the noise characteristics of film and digital are quite different and that the two media require different exposure strategies.

I note that there was no further discussion on the post in your link. A single post on the internet doesn't really prove much. If we really need an expert second opinion, we could look at Ken Rockwell's site.  

[ErikKaffehr]

Hi,


I don't really have much experience with print film but had a lot with BW. I'd suggest that the matter is a bit complex. Negative film saturates. With extreme overexposure we don't get much increase in density for each increase of exposure but don't have the same kind of clipping we see on digital. The area around the sun used to have posterisation/banding in sunset shots. On the low exposure side we get a lot of noise on print film. Print film has shallow gradient so a steep gradient is applied when scanning which enhances grain.

Best regards
Erik

I don't that that one critique of Roger's methods proves any experimental error. I am not familiar with the qualifications of that author. He seemed to assume that Roger exposed the negative film for the midtones. However, Roger did expose the film to the extreme and even scanned the film at two levels of gain in order to get the maximum detail in the highlights and then looked at the negative to make sure that all detail was accounted for. He also pointed out that the noise characteristics of film and digital are quite different and that the two media require different exposure strategies.

I note that there was no further discussion on the post in your link. A single post on the internet doesn't really prove much. If we really need an expert second opinion, we could look at Ken Rockwell's site.  

[Daniel Browning]

I note that there was no further discussion on the post in your link. A single post on the internet doesn't really prove much. If we really need an expert second opinion, we could look at Ken Rockwell's site.  

That's true, but the poster is Gordon Good, someone I know to be highly knowledgeable about the technical side of digital photography. He was also collaborating with Ed Leys and Jay Turverville, another person that I know to have a very high level of understanding. Roger's numbers also fly in the face of the practical experience of film photographers everywhere. Many notable cinematographers use 12 stops of dynamic range as a matter of routine with modern cine film stocks and Digital Intermediates. To me, all this indicates that Roger's test was an outlier.

[Jonathan Ratzlaff]

I can see the point of shooting film when shooting directly into the sun during sunsets and such.  The main difference is the lack of blooming.  To my eye slide film seems to hold detail around the sun a bit better than does a digital sensor.  With negative film it is better, however it seems just too much work   to get a good scan from negative film

[ziocan]

Shooting into the sun with either digital or film will cause severe problems. However, most backlit subjects do not include direct sun. You seem to be basing your conclusions on some theory and have no data to back up your assertion. I did shoot a severely backlit subject and saw no problem. Now the ball is in your court.


Now you are changing the subject from film vs digital to the subject of flare in lenses and mention Bjorn's reviews of different lenses with flare that could affect either medium. Your reasoning is very tangential and disconnected. Where does Bjorn discuss the problem of sensor ghosting?

Lady or Sir, you are a pain in the neck.