The GetDPI Photography Forum
Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!
DLDMT Leica M question
- Thread starter gogopix
- Start date
robertwright
New member
My belief is that the science of colour reproduction in film is more mature than the science of colour reproduction in digital. I could be wrong. Also since film is an analogue process the colour will always be more differentiated than in digital.
In terms of our perception and the S curve, why is it then that no one feels that HDR effects are more realistic than old school compressed film renderings- at least I don't.
All reproduction is compression- but we humans tend to prefer compression I think since our perception is adaptive- I think instantaneously we see blocked shadows and rolled off hightlights but then we look again and see into shadows. So S curve impression with linear adaptation?
You are right that both film and digital colour is engineered and that the S curve is inherent not designed. My oversimplification.
I guess Im well off topic here.
In terms of our perception and the S curve, why is it then that no one feels that HDR effects are more realistic than old school compressed film renderings- at least I don't.
All reproduction is compression- but we humans tend to prefer compression I think since our perception is adaptive- I think instantaneously we see blocked shadows and rolled off hightlights but then we look again and see into shadows. So S curve impression with linear adaptation?
You are right that both film and digital colour is engineered and that the S curve is inherent not designed. My oversimplification.
I guess Im well off topic here.
Shashin
Well-known member
Robert:
Color science is as advanced in digital as in film--where do you think the digital color scientists come from and the foundation they built on? Both media use basically the same RGB scheme, OK film uses CMY, but the parts of the spectrum are the same. And both have the same contamination problem between color channels.
About the S-curve. The S-curve used in digital processing is just a contrast curve. More contrast and more pleasing images. The reason for the S is not to lose detail in the shadows and highlights, but if the image is low contrast, a linear adjustment will work as well.
The film curve has a toe and shoulder, but it is not an S-Curve simply because the entire response is not needed. In fact, you want to adjust your exposure in order to put the most important part of your image on the straight-line portion of the curve. The richest prints, film of digital, come from low-contrast scenes where there are no dark shadows nor brilliant highlights and so has good contrast throughout the image.
The HDR thing is something completely different. That is taking an unnatural DR and compressing it. The idea that it reproduced the human visual system is simply false. What they are trying to achieve is the reproduction of their experience, which is a psychological perceptual problem. And as you can see in the results, does not work. And this is about the extreme use of HDR.
Compression is not a problem. An image is an illusion. All it needs to do is "look" "real." It does not have to reproduce the scene in any absolute sense. The fact that complex curve can be used to make natural appearing images shows that no particular curve is ideal.
I agree that more DR will give a flatter image. A camera with infinite DR will make horrible images. but there is no magic in making a flat color image look good. I think the problem comes from either a lack of experience or a cooking problem. The flavor by taste method of cooking where you add spice or sugar as you go tends to result in food either too spicy or sweet. You do the same with contrast and saturation in processing, and you overcook the image. Many good photographers spent a great deal of time learning how to see. Learning what is natural. Folks starting out process for what they "think" it should look like, and they invariably get it wrong. I was at a forum where a member asked the community to processes her image. The result was a disaster. The color, contrast, and sharpness was all wrong, all overcooked. A golden hour image is more than just warm. If the picture was not taken during the golden hour, making it yellow just makes it look yellow. Take an image made on a foggy day and set the black and white points in the histogram to where the pixels start, it will be ugly and unnatural. We all know what over sharpening looks like.
I know when I get a new camera, it takes a while for me to understand how to process the images--they are not the same as the ones from my other cameras. Folks with the M will have to go through this process to. An M9 user should not expect to open an M file and process it the old way and expect the same result. They need to learn how the camera sees and how that file represents that. It seems with every new camera there is this nostalgia for the old model. But it does not persist. People learn.
Color science is as advanced in digital as in film--where do you think the digital color scientists come from and the foundation they built on? Both media use basically the same RGB scheme, OK film uses CMY, but the parts of the spectrum are the same. And both have the same contamination problem between color channels.
About the S-curve. The S-curve used in digital processing is just a contrast curve. More contrast and more pleasing images. The reason for the S is not to lose detail in the shadows and highlights, but if the image is low contrast, a linear adjustment will work as well.
The film curve has a toe and shoulder, but it is not an S-Curve simply because the entire response is not needed. In fact, you want to adjust your exposure in order to put the most important part of your image on the straight-line portion of the curve. The richest prints, film of digital, come from low-contrast scenes where there are no dark shadows nor brilliant highlights and so has good contrast throughout the image.
The HDR thing is something completely different. That is taking an unnatural DR and compressing it. The idea that it reproduced the human visual system is simply false. What they are trying to achieve is the reproduction of their experience, which is a psychological perceptual problem. And as you can see in the results, does not work. And this is about the extreme use of HDR.
Compression is not a problem. An image is an illusion. All it needs to do is "look" "real." It does not have to reproduce the scene in any absolute sense. The fact that complex curve can be used to make natural appearing images shows that no particular curve is ideal.
I agree that more DR will give a flatter image. A camera with infinite DR will make horrible images. but there is no magic in making a flat color image look good. I think the problem comes from either a lack of experience or a cooking problem. The flavor by taste method of cooking where you add spice or sugar as you go tends to result in food either too spicy or sweet. You do the same with contrast and saturation in processing, and you overcook the image. Many good photographers spent a great deal of time learning how to see. Learning what is natural. Folks starting out process for what they "think" it should look like, and they invariably get it wrong. I was at a forum where a member asked the community to processes her image. The result was a disaster. The color, contrast, and sharpness was all wrong, all overcooked. A golden hour image is more than just warm. If the picture was not taken during the golden hour, making it yellow just makes it look yellow. Take an image made on a foggy day and set the black and white points in the histogram to where the pixels start, it will be ugly and unnatural. We all know what over sharpening looks like.
I know when I get a new camera, it takes a while for me to understand how to process the images--they are not the same as the ones from my other cameras. Folks with the M will have to go through this process to. An M9 user should not expect to open an M file and process it the old way and expect the same result. They need to learn how the camera sees and how that file represents that. It seems with every new camera there is this nostalgia for the old model. But it does not persist. People learn.
The M8 and M9 files came out of the camera with more punch and presence. Btw, M8 and M9 files have the same resolution; the M9 sensor is just larger.
The M240 files, while not as contrasty are a lot more robust wrt processing compared with M9 files. The M240 files can be made to look like M9 files, but not the other way around. I'll take the M240 files any day.
I've profiled the cameras with the Colorchecker Passport, and the differences from out-of-camera to profiles are very small for the M240 compared with the differences for the M9. I do very little studio stuff, so my take on this might be different from someone who does, but I find the colour rendering of the M240 to be very good. It's more accurate than that of the Canon 5D MkIII which I also use.
Henning
The M240 files, while not as contrasty are a lot more robust wrt processing compared with M9 files. The M240 files can be made to look like M9 files, but not the other way around. I'll take the M240 files any day.
I've profiled the cameras with the Colorchecker Passport, and the differences from out-of-camera to profiles are very small for the M240 compared with the differences for the M9. I do very little studio stuff, so my take on this might be different from someone who does, but I find the colour rendering of the M240 to be very good. It's more accurate than that of the Canon 5D MkIII which I also use.
Henning
gogopix
Subscriber
I've seen others comment on the robustness of the M240 files. I do remember the mess with earlier M8 profiles, until someone here on the forum produced a very workable one.
BTW, I generally use Capture One (6, I thried 7 and was SO slow I went back, but maybe will upgrade with the M)
In C1 you can create and save profiles with a lot of flexibility...that can drive you crazy if you are not skilled. I wound up modifying he one I received.
Has any M user here tried a custom profile? David Farkas has some great ones for the "S". Maybe he will try his hand at one for he M.
regards
Victor
PS If you look at the M images on the LUF site they have a lot more punch and look much closer to the M9 rendering. Many use LR4 as here, so I am confused (I know, as usual!! LOL) . But I must say, the pix there are show more what I expect from a Leica!
Last edited: