Help in understanding a camera's dynamic range lost when printing

[BarryE]

I'm new to this forum ... I trying to understand what I need to do to take my prints to the next level. Hopefully someone here can help or point me in a better direction. ...

I'm trying to get my head around how in practice differences in the dynamic range of camera sensors maps to the reduced dynamic range/DMAX of a final print.

I'm using a matte fine art cotton paper ( Hahnemule 308) and having it professionally printed.

So if a sensor offers, say, 1EV extra dynamic range (DR), then a greater tonal range can be captured. If this is processed, say in ProPhoto at 16-bit to ensure any processing losses are minimised, before mapping to Adobe 1998 and 8-bit for printing, then clearly some of the advantage of the greater initial DR from the sensor is going to be lost. This range is further reduced when printed.

So my question in simplified terms, to try and help me quantify what happens in practice: How much of the tonal range captured in a camera with a greater DR range is actually still visible when printed ? (I appreciate, that the framing glass used and viewing conditions will further affect/reduce the viewers experience and final processing needs to be done for specific paper characteristics.)

Any guidance will be greatly appreciated.

 

[maac]

No Title

look over this post and other like this keyword (printing gamut) in your Search


A question for a college project.

 

[DYP]

No Title

What is the output device. You can proof your photo in Photoshop using the output device paper profile.

Here is the gamut difference Hahnemule 308 and Adobe RGB on a Canon iPF8400. Large area is Adobe RGB.

One thing you want to do is keep your image at 16-bit. The Canon can except 16bit images and I am sure that Epson can as well.

 

[BarryE]

Thanks Michael. Is it a gamut question ? My printer accepts Adobe 1998 and I soft proof and adjust to stay in gamut. So an image captured in a camera with a wider DR is going to be shifted around and compressed into another colour space, then printed onto a paper with a reduced ability to produce the full range of the original capture. If the original capture records a greater DR on one sensor over another, I'm thinking that this DR advantage is going to be effectively lost when the image's colour space is re-mapped using whatever rendering intent I choose and then printed. Am I on the right lines ?

 

[BarryE]

Thanks DYP. The printing service I use, use either Epson Pro 9890s and 4880s. I use their 308 profile when I prepare my prints. I stay in ProPhoto and 16 bit until the end when I compress to the Adobe 1998 colour space and 8-bits, before making final adjustments using soft proofing.

Interesting what you say about the printers accepting 16-bit. The FAQ of the printing service I use, says to reduce it to 8 bits before sending the files to them. I'll check this is actually required.

 

[maac]

I'm really not the one who can answer your color gamut question. But if you were to put your key words into the search area you will be able to pull from many others who already post parts of your question.

 

[DYP]

Sounds like their really not quite up to date on the latest technology, but the quality you get will probably be fine and you probably wouldn't notice if there is any difference between 8bit and 16bit. You most likely will only see it in gradients.

 

[Correct Color]

So my question in simplified terms, to try and help me quantify what happens in practice: How much of the tonal range captured in a camera with a greater DR range is actually still visible when printed ?

...and the answer, put simply, is an absolute and unqualified...

It depends.

It depends on the image; it depends on the media; it depends on the printer; and it depends on the printer profile.

And it does help to think of this as gamut and not as dynamic range. The gamut of any color space is the entire range of colors that exist in that space, as opposed to a camera dynamic range which is typically defined as the range between its maximum black and maximum white.

So, let's say you have a camera though with a very high dynamic range but you shoot an image that has very muted colors and not very deep blacks...

Well, odds are that if all your color management is done correctly, you might get that image onto a soft water-color type stock with no loss of color at all.

Or you might take the same camera and shoot a very vivid-color image with a lot of detail in deep rich shadows and not be able to reproduce all of the image even on a really hard photo-gloss type material.

The rule of thumb being that the softer and more toothy the media, the more light will hit it and refract before getting to your eyes, thus the less gamut it will have; and the converse then being true: The harder and smoother the media, the more light will bounce directly from it back to your eyes, thus the larger the gamut it will have.

Of course, this assumes that the media profile used to describe the creation of the printing dots for either media is accurate and well made.

Not necessarily a given.

Every printer profile out there was made by some individual using some combination of color-reading device, profile-making software, and profile-making philosophy. And it's more an art than a science. Profiles you download can run the gamut from great, to terrible.

If you want to send it to me, I'll take a look at the profile you're using and break it down a little for you.



Mike Adams
Correct Color

 

[BarryE]

...and the answer, put simply, is an absolute and unqualified...

It depends.

It depends on the image; it depends on the media; it depends on the printer; and it depends on the printer profile.

And it does help to think of this as gamut and not as dynamic range. The gamut of any color space is the entire range of colors that exist in that space, as opposed to a camera dynamic range which is typically defined as the range between its maximum black and maximum white.

So, let's say you have a camera though with a very high dynamic range but you shoot an image that has very muted colors and not very deep blacks...

Well, odds are that if all your color management is done correctly, you might get that image onto a soft water-color type stock with no loss of color at all.

Or you might take the same camera and shoot a very vivid-color image with a lot of detail in deep rich shadows and not be able to reproduce all of the image even on a really hard photo-gloss type material.

The rule of thumb being that the softer and more toothy the media, the more light will hit it and refract before getting to your eyes, thus the less gamut it will have; and the converse then being true: The harder and smoother the media, the more light will bounce directly from it back to your eyes, thus the larger the gamut it will have.

Of course, this assumes that the media profile used to describe the creation of the printing dots for either media is accurate and well made.

Not necessarily a given.

Every printer profile out there was made by some individual using some combination of color-reading device, profile-making software, and profile-making philosophy. And it's more an art than a science. Profiles you download can run the gamut from great, to terrible.

If you want to send it to me, I'll take a look at the profile you're using and break it down a little for you.



Mike Adams
Correct Color

Thank you Mike, my understanding (maybe) is slowly improving from what you say ...

As background: I am thinking of upgrading my camera equipment to one that has a tested 1EV greater dynamic range than my present kit. All things being equal in my workflow and the professional printing service I use, I'm trying to understand if an improved camera would offer more when my images are printed.

I will be sticking to this paper stock for now as I like the gentle tonal graduations and feel of this paper, and importantly it sells for me as fine art photography.

The narrow tonal range shots is a style I prefer, so I am OK with these shots and the print output as you explain. However, I've been recently working more into highly contrasty light and processing these images is obviously harder than with a more muted palette.

I start using 16-bit and ProPhoto. At this point I assume I would be able to discern differences on screen between two camera systems that have different DR capabilities when looking at high contrast images. I then use a rendering intent to map the images to the output colour space (1998) and proof using the profiles my printer supplied to me for their printer and the paper I'm using. Then make a few adjustments to try and correct the images as seen in the soft proofing, and send to the printer ... and wait a couple of days.

They do their magic and a print returns. The soft processing appears to be only a guide to the final image, as you hint at, depending on a range of factors. It is, however, generally a good match for the most prints (test charts included).

I view under D50 lighting, my monitor is a calibrated graphics grade device etc, and my soft proofed images on screen do map well to my received prints.

** If I understand right from what you say: if my monitor can display the colour space I'm using (1998) and I'm properly colour managed, then the image data will have been mapped into the new colour space and I should be able get back from the printer what I see at the soft proofing stage. The two camera's ability to capture different dynamic ranges is only relevant to the point of proofing in my workflow.

However, the question still arises: will the differing abilities of the two cameras concerned show any discernible differences, apparent in the early stages of processing, when the images are viewed as soft proofs ?

I've re-jigged my understanding a little from I think you said. Hopefully, I'm nearly now.

Many thanks

 

[Stephen Marsh]

The brightest white in a digital file can be 255r255g255b or say L*100 in Lab colour, with the darkest black being 0r0g0b or L*0 – however an output device can’t reach these values, the paper will always be darker and the deepest shadow will be lighter than the digital file. It will depend on the media and inks/dyes etc. It is all about maximising the available dmin/dmax of the output and allocating contrast in-between these tones. Concentrate on the output, as the input way exceeds the output and it is all about milking the best tonal range and colour from the output as one can do. This often means exaggerating final tones to make up for the output being lesser. This can often be more of an art than science, with Ansel Adams and Dan Margulis being considered by many as being masters in their respective pursuits.


Stephen Marsh