Dynamic Range of Film still better than Digital Cameras?

[PhotoHobbyist]

Saw a youtube video showing that film cameras have far better dynamic range than today's digital cameras.
They still use high speed film cameras to capture rocket launches for engineering purposes because of that.
I was surprised. Is this the case in ordinary cameras too?

 

[Ysarex]

Saw a youtube video showing that film cameras have far better dynamic range than today's digital cameras.
They still use high speed film cameras to capture rocket launches for engineering purposes because of that.
I was surprised. Is this the case in ordinary cameras too?

No. One of the reasons I shoot digital now and won't go back to shooting film is dynamic range -- digital is so much better.

Now there's a lot of qualifiers to that. If you're willing to jump hoops -- lots of big hoops, and you don't care what it costs, you can do well with film. Better than digital? Unlikely. You got it when you said "ordinary cameras."

A huge part of the issue is practice. How are you going to practically handle the film? Are you going to scan it? Scan it with what scanner? The scanner you need in order to get anywhere close to the DR of a modern digital camera they don't make any more. The few still in operation are being horded and if you want one it's going to cost you (think 10s of thousands). And if you get it you won't be able to run it because the software is so outdated it won't work on a modern computer. How are you at writing scanner driver software? You can have someone who has the right scanner do the scan for you -- think $75.00 per negative. Not practical -- then get a desktop scanner and that will be the weak link in your chain and you won't come close to what digital can do effortlessly.

Going to just make prints in the darkroom from film? Again practically speaking if you go that route you lose all the advances in software processing that make actually using lots of DR practical. Hey, a couple decades of devotion and practice and you can become a wizard darkroom printer capable of managing maybe 40% of what modern software can do.

The photo below is from a B&W film negative. I shot it back in the early 1980s. That's a big DR range from shadows to highlights. I was a good darkroom technician but printing that image pretty much beat me. I had to switch filters during the multiple exposures in the enlarger -- extensive dodging and burning was required to get detail in the water as well as the dark shadows. If I did finally make a print I thought was pretty good odds are I wasn't going to do exactly the same to get another print.

I still have the 6x7 negative and I have Epson 700 - 800 class scanners today. To get the image you see below what I had to do was scan the negative twice, once for the shadows and again for the highlights. Then I had to combine the two scans with a mask in PS before I could continue processing the image. We're talking major PIA. A single pass scan with a drum scanner would be nice but I don't have one. That's the scanner I mentioned above that they stopped making.

With a modern APS class sensor digital camera that photo would be no problem for me today.

Look at the side by side illustration below. I processed the raw file on your left and the camera JPEG is on your right. The camera JPEG gives you an idea of the DR for the scene. The highlights in the camera JPEG are blown -- they're not blown in my processed raw image. The camera used in that example was a Canon G7 -- a 1" sensor compact. A modern FF camera will deliver 3 more stops of DR over that compact. That photo was easy to do with the G7. Try it with film and you're going to sweat blood and hemorrhage cash.

 

[vintagesnaps]

Then don't shoot JPEGs... For the two on the bottom, I'd aim toward the desk and meter, then reframe to get a proper exposure. Or bracket shots so out of three shots you'd have one decent one.

All I know is I've had photos accepted into juried exhibits that I shot on film, or digitally, or on Polaroid pack film (and the exhibit had nothing to do with Polaroids or film).

Good is good. You need to know what you're doing with whatever medium you're using. I find using my digital camera that shooting Raw usually is best, then I can save a JPEG copy as needed. I also find proper exposure is key.

I've shot sports (hockey) where things happen fast and they expected results. Didn't really care how you got the shots, just that you did. There's often no time to edit later; these days I see photos on websites before a game's even done.

If someone can't get at least decent photos in camera most of the time witout needing a lot of editing, then it would probably be a good idea to bring up the skill level. I found doing wet darkroom work that if I had a good exposure then I could get a good print without much adjusting. It's like anything, the more you practice over time the better you get.

 

[vintagesnaps]

And I looked to see current pricing at Dwayne's Photo who I've used. I can get a roll of 24 exp. color film developed, scanned hi-res to a CD and/or uploaded to their website, a set of 4x6 prints - for $22. Other options are available.

I do scan hi res copies of my lumen prints (sun prints made using expired midcentury B&W photo paper). I have a couple of times had the image disintegrate so I scan the ones I want to save in a digital format, then I can print the copy as needed.

 

[Strodav]

This is for B&W photography and shows that film does indeed have a larger useable dynamic range: Film vs. Digital: This is How Dynamic Range Compares My takeaway is for film, expose for shadow detail, but for digital expose for the highlights.

 

[Ysarex]

And I looked to see current pricing at Dwayne's Photo who I've used. I can get a roll of 24 exp. color film developed, scanned hi-res to a CD and/or uploaded to their website, a set of 4x6 prints - for $22. Other options are available.

Hi-res is not high DR. Here's a link to a modest priced film scanning service that provides drum scans: Drum Scanning Service - Michael Strickland Images

I do scan hi res copies of my lumen prints (sun prints made using expired midcentury B&W photo paper). I have a couple of times had the image disintegrate so I scan the ones I want to save in a digital format, then I can print the copy as needed.

 

[Ysarex]

This is for B&W photography and shows that film does indeed have a larger useable dynamic range: Film vs. Digital: This is How Dynamic Range Compares My takeaway is for film, expose for shadow detail, but for digital expose for the highlights.

I found the video basically of no value right from about 10 secs. in where he says; "To do this video I went out and bracketed 21 exposures. 10 under, 10 overexposed and of course 1 properly exposed image. I want to see what happens when you drift away from the correct exposure." I don't know because he didn't say but I suspect his understanding of that term, "correct exposure" has little to do with my understanding of that term --- and off he went.

So he concluded film has a 12 stop dynamic range, but said nothing about whether it is usable. I doubt it is. My digital cameras have a 12 stop usable dynamic range.

 

[Ysarex]

Then don't shoot JPEGs... For the two on the bottom, I'd aim toward the desk and meter, then reframe to get a proper exposure. Or bracket shots so out of three shots you'd have one decent one.

I never do shoot JPEGs. I just put that one there as a reference. The raw file was exposed by placing the brightest diffuse highlight (out the window) at the sensor saturation threshold -- I expose all digital photos the same way. For me correct exposure is full utilization of the sensor's recording capacity -- my exposure goal for every photo I take. If circumstances prevent me from achieving that goal then I fall back to exposing the sensor as much as possible.

All I know is I've had photos accepted into juried exhibits that I shot on film, or digitally, or on Polaroid pack film (and the exhibit had nothing to do with Polaroids or film).

Good is good. You need to know what you're doing with whatever medium you're using. I find using my digital camera that shooting Raw usually is best, then I can save a JPEG copy as needed. I also find proper exposure is key.

I've shot sports (hockey) where things happen fast and they expected results. Didn't really care how you got the shots, just that you did. There's often no time to edit later; these days I see photos on websites before a game's even done.

If someone can't get at least decent photos in camera most of the time witout needing a lot of editing,

"...decent photos in camera" would suggest the camera JPEG. For me a proper exposure disregards the camera JPEG -- I only care about the sensor exposure.

then it would probably be a good idea to bring up the skill level. I found doing wet darkroom work that if I had a good exposure then I could get a good print without much adjusting. It's like anything, the more you practice over time the better you get.

 

[snowbear]

Are all (B&W) films equal, or can the DR vary?

Note: though we normally do not allow Film vs Digital type threads, the mods are allowing this one to go forward, as long as posts remain informative and stay civil.

 

[Ysarex]

Are all (B&W) films equal, or can the DR vary?

Note: though we normally do not allow Film vs Digital type threads, the mods are allowing this one to go forward, as long as posts remain informative and stay civil.

DR for films varies substantially by film type (5 or more stops) -- and don't forget color. Higher ISO films have the widest DR. King of the Hill in what's available now would be Kodak Vision3 500T color neg film. Here's a link where cine folks discuss it: Vision3 real/usable dynamic range

Complicating the issue especially in B&W is processing. Tri-X for example can be processed to record a DR of 10 to 11 stops (typical) or if you prefer 20 stops as measurable by a densitometer and that often makes these types of threads go south. A Tri-X neg that has recorded a measurable 20 stops of DR isn't going to make any kind of usable image and so the 20 stops becomes a moot point.

DR for digital cameras likewise varies one camera to the next and tends to correlate with sensor size -- bigger sensor = more DR. Again the issue of usability versus measurability plays a big role and what's measurable is again moot. For example DXO Mark says the sensor in my Nikon Z7 has a DR range of 14.6 stops. Wow! but that's a case of what their machines can measure -- nobody can actually make use of that with a Z7.

Then there's the question of what's enough and what are you going to do with 10 or more stops of DR assuming you can record it? Are you going to make a print or display it on your phone or computer? The real DR of a piece of printing paper hasn't changed and isn't going to -- you can only squeeze in so much info.

So for most of our normal day in day out photography 6 to 7 stops of DR is plenty and both film and digital provide that. When do we really need more -- the photo below is an example, backlighting. That scene below presented a DR of 10 plus stops. Back in the film days a scene like that would snap me to attention. Could I do it? Good chance I'd fail. No chance using color transparency film. I'd really need a 120 camera with color neg stock like Fuji PRO400H and assuming I capture the scene then comes the real hard part -- getting the image from the film to a final output of print or electronic display. Roll up your sleeves this is going to take the weekend.

But today with digital that photo was unplanned, taken on a walk with my wife. I recognized the backlighting which now I treat as no problem. I set my little Canon G7 (1" sensor) to expose to the full capacity of the sensor and click -- easy processing when I got home. That compact G7 has a usable DR of just over 9 stops which was plenty for that scene. Using film, capturing a scene like this is still a daunting exercise. Digital provides the advantage of easy usability and that makes all the difference.

 

[snowbear]

Thank you.

I'm a bit reminded of some car options and specs: 120 or 140/160 MPH speedometer. Since most of us generally don't drive over 70 or 80, it doesn't really matter, though I had a brother-in-law that sold us a Dodge Intrepid and mentioned that it wouldn't go past 112. I didn't ask how he knew.

 

[vintagesnaps]

Oh, a drum scan... well yeah, those are pretty pricey! not an option for me (unless it was a one time thing for some particular reason, then maybe). Joe I think that's getting into more specialties than run of the mill everday options.

And that Kodak film (which I looked up and found that FPP has some respooled) I thought sounded familiar, cinema film with a remjet layer... not something that many places develop.

Charlie, B&W films still being made as far as I know are Kodak TMAX and TriX, and Ilford films, etc. And there are a number of color films still made, Kodak Portra is nice (athough it's gotten more pricey, but what hasn't since the pandemic? just going to use it more sparingly.) Using 120 medium format gives you larger negatives so you don't have to enlarge as much., but I've gotten good 8x10 prints from 135mm B&W films. I don't know that you could go too big before you'd be getting more grainy enlargements, depends on the sharpness/quality of the image.

With B&W I learned to look for a 'black' black and a 'white' white somewhere in the image (in a college grad art class many years ago). That's what I still do, and look for nice shades of gray in between. And often I'm noticing the light and dark in a scene before I even look thru the viewfinder. Even if shooting color, I look for a nice range of light to dark (but I don't know how many stops it is, I guess I think of it in those terms).

That last photo posted is what I'm talking about that I'd likely aim the camera downward to meter the scene in the foreground, then raise the camera and reframe the shot. I don't want the meter to just read the light coming in from the background. Then I usually go up a stop and down a stop to make sure I got a good exposure, especially if I'm shooting B&W film. With color film the lab is going to adjust anyway!

I guess you need to know/learn how to use whatever medium you choose. (And if NASA thinks film is good, then I guess I'd take their word for it! lol)

 

[f64_or_bust]

Saw a youtube video showing that film cameras have far better dynamic range than today's digital cameras.
They still use high speed film cameras to capture rocket launches for engineering purposes because of that.
I was surprised. Is this the case in ordinary cameras too?

There are several fatal flaws in reasoning in that video. The big comparison between the old NASA film cameras and the current video camreas is also a comparison of a day launch and a night launch. It's bad enough that the engines are incredibly bright, but if you also impose a 20-22 stop difference in ambient light between the film shot in sunlight and the video shot at night, there will be some issues with either capture medium. How about compairing the two in identical situations? Duh.

Also, if you look at the way NASA mounted cameras, and used massive tracking camera rigs, it's clear that they're not going to that lenght (expense) today. Government money vs private industry? Hmm. But the installation, setup, and operation of any camera is what maximizes the end result. The video skips conveniently over that part, point at how much worse the modern footage is without actually noticing why.

No, film does not have better DR. But digital DR is more linear, film is not, which works slightly in films favor in the highlight end.

Then there this one, almost an elephant in the room: Film? Which film? Color reversal films were between 6 and 12 stops. Negative films range from 15 to 18 stops. ISO drives DR too, with higher ISO films having higher DR at the expense of grain.

And then, it gets worse. People throw DR numbers around easily, as if images with similar DR were somehow comparable. Simply not true. The definition of DR is the range from the brightest to the darkest usable tone. "Usable"...there's a subjective term, right? And what makes a dark tone unusuable is usually noise or grain. As we all know by now, there are some huge differences there, as well as significant differences in image quality between cameras with the same specified Dr.

So DR numbers are a nice start, but hardly worth much in determining image quality in a high DR shot.

Film is not, and never was, a free ride. The same exposure challenges exist in either medium. Like so many things of the past that seem better than their modern counterpart, the real secret lies in the skill of the users. Those NASA guys were brilliant, and there was emphasis put on getting everything exactly right. Not to say things are more slip-shod today, necessarily, but clearly in the videos shown, they didn't try nearly as hard to get it all right. That's not a film vs digital thing, that's a training and professionalism thing.

But it may not matter anyway. The fastest film cameras were limited to 50,000 frames per second, and that over a very short interval (far less than a second). Digital high speed cameras have smashed that barrier, with 200,000fps being fairly routine, and 10M fps and higher now being possible, and with memory buffers that can take care of any practical shot lenght. Its just $$:Gb. Or is that $$=Gb? Sounds fun, right?

And, I suspect, it is.

 

[Ysarex]

Thank you.

I'm a bit reminded of some car options and specs: 120 or 140/160 MPH speedometer. Since most of us generally don't drive over 70 or 80, it doesn't really matter, though I had a brother-in-law that sold us a Dodge Intrepid and mentioned that it wouldn't go past 112. I didn't ask how he knew.

Oh, a drum scan... well yeah, those are pretty pricey! not an option for me (unless it was a one time thing for some particular reason, then maybe). Joe I think that's getting into more specialties than run of the mill everday options.

And right now new drum scanners are off the market. They're not being made. The point I wanted to get across to the OP who asked if film DR was still better is; so what if you can't access it. Wow I shot a backlit scene on 120 Tri-X and processed it to record 14 stops of DR! Now I just have to scan it with my desktop scanner that delivers a maximum 7 stops of DR, Oh no. So I'll send it to the lab and have it scanned using their Noritsu scanner which can wring out 9 stops of DR in a 16 bit scan and Oh wait.... It's about usability and usability is determined by the weakest link in the chain. You've got 14 stops of recorded DR in a 120 Tri-X negative but you can't scan it or afford to scan it so you can't use it.

And that Kodak film (which I looked up and found that FPP has some respooled) I thought sounded familiar, cinema film with a remjet layer... not something that many places develop.

Charlie, B&W films still being made as far as I know are Kodak TMAX and TriX, and Ilford films, etc. And there are a number of color films still made, Kodak Portra is nice (athough it's gotten more pricey, but what hasn't since the pandemic? just going to use it more sparingly.) Using 120 medium format gives you larger negatives so you don't have to enlarge as much., but I've gotten good 8x10 prints from 135mm B&W films. I don't know that you could go too big before you'd be getting more grainy enlargements, depends on the sharpness/quality of the image.

With B&W I learned to look for a 'black' black and a 'white' white somewhere in the image (in a college grad art class many years ago). That's what I still do, and look for nice shades of gray in between. And often I'm noticing the light and dark in a scene before I even look thru the viewfinder. Even if shooting color, I look for a nice range of light to dark (but I don't know how many stops it is, I guess I think of it in those terms).

That last photo posted is what I'm talking about that I'd likely aim the camera downward to meter the scene in the foreground, then raise the camera and reframe the shot. I don't want the meter to just read the light coming in from the background.

You're identifying an exposure practice common to film photography. Film records the total range of light/dark in a scene non-linearly. When we graph film's density response to exposure we get a curve. Working with film you have to expose to that curve. You still work with film and so you think about exposure in film terms.

A digital sensor's response to exposure is linear and graphs as a straight line. Using a digital camera I identified the brightest diffuse highlight in the scene which was the lit side edges of the grave markers. Then set exposure to place that highlight at the sensor's saturation threshold and click. With digital I think of exposure as utilizing the recording capacity of the sensor.

Then I usually go up a stop and down a stop to make sure I got a good exposure, especially if I'm shooting B&W film. With color film the lab is going to adjust anyway!

I guess you need to know/learn how to use whatever medium you choose. (And if NASA thinks film is good, then I guess I'd take their word for it! lol)

 

[santino]

All I can say is „who cares?“. Both are different mediums and both create a different output. Courses for horses.