Printing black and white images with dark tones

[fatoddsun]

Hi everyone,

recently I sent a couple of black and white pictures to be printed to one of the most common UK online printing services, in small format. The idea was to get an idea of how those pics look on paper, and maybe select a couple of them for printing in larger format (20cmx30cm) to mount on forex and hang on the walls.

When printing in large format and/or for hanging on the walls or similar, I typically order from a very good quality service that also serves some of the most famous photography exhibition of London, and they normally do an amazing job at a very reasonable price.

What I've noticed with my last batch is that all the shadows and darker tones of my images were basically printed in pure black, hence most of the details were completely lost, and the pictures look very dull.

You can see some of these pictures here , for instance.

I am now hesitant to send those that I like the most to the "good" service, because I'm afraid that the same could happen. I have done something similar in the past, but never with pictures that were so biased towards the darker side of the grey scale.

Do you normally take particular precautions in post-production before printing pictures that are heavy on shadows? How do you "trust" what you see on screen? Do you raise a bit the exposure before exporting in jpeg?

As an additional detail, the "cheap" service where I ordered my small pics prints with inkjet technology on photo paper, whereas the "good" service actually works with c-prints, that is properly impressed photographic paper. Would this make any difference?

Thanks!

Damiano

[Manfred M]

I'm heading out in a few minutes, but the long answer is that small prints tend to look darker than large ones; that's just visual adaptation at work.

If the prints are too dark, the likelihood is that your computer screen is set too bright. Normally we are looking at output in the range of 80 - 120 candela / square meter. Most screens when they come straight out of the box are about double that if not more (people run their screens far too bright for photographic work).

The problem with labs is we don't know if they adjust for brightness or assume that the photographer has gotten it right; your test prints would likely help there. The other issue is that images tend to be over-lit in shows and galleries, so if the "good service" is printing that way, the problem could be the light you are viewing them under. These images are printed dark to look good under spot lights.

Is there some way of scanning an original at full size as well as a jpeg that you sent to the lab?

I do a lot of printing myself so understand the prep work that needs to be done. It takes a lot more time an finesse to prepare a print versus an electronic image.

[Manfred M]

I've gone over to your website (I missed the link in my original read of your posting) and had a close look at the images. I suspect I am looking at downsized images, so this might be part of the issue.

I don't know how large these images actually are, but from what I see they appear to be unprintable. There is a very high level of digital noise in most of them and when I look at the shadow details, there is generally no workable data there at all and the areas are all blocked up.

What camera and lens are you using?

[DanK]

I did go to your gallery when I first read your post, but I delayed answering because Manfred is more experienced printer than I am, and I wanted to see what he would say.

I agree with what he posted. The first thing that struck me was the very strong digital noise. That doesn't stand in the way of printing, however. I think the problem with printing is that you have a lot of areas that are highly compressed toward the dark end of the distribution. Your site uses an image format photoshop can't read, so I took a screen shot of the swan photo, saved it as a JPEG, and opened that in photoshop. Most of the image is crammed into the darkest 20%. For example, the sky, which is lighter than some of the rest, gives me RGB values mostly below 50, most far below.

Some time ago, I took a good test image and printed it on a variety of papers, using a good (Prograf 1000) printer. You can find the test image here. Even some luster and baryta papers, which have a relatively large tonal range, lost some detail at the darkest end, failing to show distinctions at the left-hand edge of the bar at the bottom.

So, if you want prints to show shadow detail, I think the solution is to have somewhat brighter shadows.

Concerning the noise: it's all a matter of personal preference, of course, but I do a fair amount of night photography and know quite a number of other people who do, and most of us avoid high ISOs if at all possible. Most of my urban night photography is shot at ISO 100 or 200 and use long exposures to compensate. If you have people in the scene, it's a different matter, of course, because you need faster shutter speeds. In addition, I generally expose to the right and then darken the image in postprocessing if necessary. That way, the shadows have a lot of detail, and noise is minimized.

In addition, many urban night scenes have dynamic range that exceeds that of the camera, so if you take a single exposure, you have to choose: clip the shadows, or clip the highlights. An alternative is to bracket shots and merge them. Or you can compromise, using a merge (HDR, exposure fusion, or a manual blending) to avoid some clipping but allow some to be present. I've used the image below as an example. This is a merge of two or three bracketed shots (probably 2, but I don't recall), merged with the 'merge to HDR' function in Lightroom. That allowed me to keep shadow detail, but I allowed the lights to blow out. You'll see that the image isn't noisy. I applied no noise reduction in postprocessing.

[fatoddsun]

Thank you both for your very helpful comments!

I am aware there is high noise in my pictures, but I am fine with it, as I consider it part of the "texture" of the image.
Still, thanks for the suggestions about possible alternatives that one might consider to reduce that noise, if desired.

The pictures were taken with a Panasonic Lumix TZ70 at ISO 1600, with the original files all being roughly 11 megapixels.
The sensor of this camera is of small format, hence 1600 ISO are already enough to create a considerable amount of digital noise.

My impression is that the "cheap" service where I sent the pictures to applied some sort of standard filter that - among other things - also tried to reduce the noise. The result is that the lower tones are even more compressed and everything became very "smooth", making almost impossible to distinguish edges between objects.

I do not want to reduce the noise, but I'd like to keep the details in the shadows. If I interpret correctly your comments, my only choice is to brighten up the shadows a little bit, and to hope that the "good" service will develop a faithful representation of my digital image. On this note, do you know whether c-type vs printing could do any better or worse for this specific topic?

Thanks again,

Damiano

[DanK]

If I interpret correctly your comments, my only choice is to brighten up the shadows a little bit,

You really can't do that very well in some of these photos. When you expose as dark as you did, the amount of signal is very low, and as you note, the amount of noise as you have shot the images is high. When you try to lighten the shadows, you will be amplifying both signal and noise, so you will get mostly a lighter noisy area, with very little visible detail.

To illustrate this, I took your swan image and lightened the shadows using Photoshop. I lightened them more than you would probably want to, but the extreme adjustment makes my point clearer. Look at the water in the bottom left. If I stare hard, I think I can see a little more detail than in the original, but mostly, the brightening has made the noise more apparent.

The key to this is exposing shadows so that you have a good bit of shadow detail. The noisier the image, the lower the signal-to-noise ratio, and the more shadow detail you will need to offset the noise.

[Manfred M]

Thanks for the information - I understand the style of camera well. My first two digital camera were the Canon S40 and Panasonic FX9. The problem with this type of camera, for night photography, is the 1/2.3" sensor at the 11MP image size. The individual sensels are tiny and do not pick up a lot of light. In daylight shooting conditions, that is fine, but at night, this means you are getting a lot of digital noise in the shadow areas. In a digital camera, little or no data is found in areas of dark, so these are extremely susceptible to showing noise. While we see this issue in large open areas especially the sky, if you carefully examine the dark areas of the image; the shadows, the noise is even worse there because these are the darkest parts of the image. While you may not want to reduce the noise in your image, I suspect you are writing about the "look". That's fine, but the problem is that there are areas of the image where you need detail in the shadow areas for night time images to be effective. Unfortunately you get both the noise that you are okay with and the noise that is causing the problems you are complaining about in the prints.

When you write "I do not want to reduce the noise, but I'd like to keep the details in the shadows", sorry that can't be done. They are both an outcome of the camera sensor and shooting conditions you are working under. If you want shadow detail, you have to reduce the noise. With your camera, you can only do this in brighter conditions. If you want to take shots at night with acceptable levels of shadow detail, you need a camera with larger sensels.

Did the lab use a noise reduction filter? I don't know, I'd have to examine the full-size images that you submitted and the prints that came back. What I can tell you, certainly in ink jet processes is that any value between 0 - 15 will be printed as pure black, so even if there are "details" in the data, the process has its limits. The C print process, unfortunately, does not have any published data that I can refer to and the resolution is dependent on the paper itself as well as the printer capabilities. You would need the spec sheets on both to get an understanding of this. Generally both these processes use 8-bit JPEG input, so you have a fairly limited gamut.

P.S. - People that spend the money on cameras with good low light performance do not do so because they like spending money, they do so because this technology is expensive. As an example, the current top of the line full-frame Nikon D5 has a 20MP sensor and is considered a low light camera. The D850, the next model down (same sensor size) has a 45MP sensor, but does not perform as well in low light conditions due to the smaller sensels.

I generally do not do B&W night photography, but I do a lot of night shooting. The spots you see in the sky are not noise, they are stars. Lots of shadow detail here and no blocked up areas.

[DanK]

That is indeed a really small sensor, but I suspect one could get major improvements by shooting at a lower ISO and exposing more to the right. There is simply too little data relative to the amount of noise in the shadows.

I have a larger Lumix with a much larger sensor, a micro 4/3, but even with that, there is noticeable degradation once I go above ISO 400.

[Manfred M]

That is indeed a really small sensor, but I suspect one could get major improvements by shooting at a lower ISO and exposing more to the right. There is simply too little data relative to the amount of noise in the shadows.

I have a larger Lumix with a much larger sensor, a micro 4/3, but even with that, there is noticeable degradation once I go above ISO 400.

Agree 100%. This is one of those cases where the equipment is limiting what the photographer would like to do.

In my experience, there are only two solutions:

1. Don't bother with this genre / aspect of photography and move on; or

2. Upgrade to equipment that is capable of this type of work.

[xpatUSA]

Agree 100%. This is one of those cases where the equipment is limiting what the photographer would like to do.

In my experience, there are only two solutions:

1. Don't bother with this genre / aspect of photography and move on; or

2. Upgrade to equipment that is capable of this type of work.

Manfred, a few threads ago, we were all discussing blending multiple images in the context of night photography IIRC - namely, a computational approach.

Is that not a solution in your experience ... apart from the usual caveats, movement, etc?

[Manfred M]

Manfred, a few threads ago, we were all discussing blending multiple images in the context of night photography IIRC - namely, a computational approach.

Is that not a solution in your experience ... apart from the usual caveats, movement, etc?

It is a solution for the technically savvy / advanced photographer which means:

a. Having the software tools to do this and knowing how to use them;

b. Knowing / having the tools to hold the camera steady (usually this means using a tripod, although I have hand-held on occasion); and

c. Having a camera that can be operated in manual mode, ensuring that the aperture, focal length and focus point do change between exposures. Small point and shoot camera is not going to deliver that functionality.

As well five of the nine posted shots have a movement component and would not work with this technique; something that you have already pointed out.

Bottom line is that we have a technique that can be used in some circumstances but is not a generalized solution and certainly not one I would suggest to Damiano. I've gone through all the images he has posted on his website and they suggest a photographer who takes an intuitive / impulsing approach rather than a considered approach in his work (although I could be wrong). There is nothing wrong with either approach, but the technique works well for landscape photography, not street photography.

[xpatUSA]

It is a solution for the technically savvy / advanced photographer which means:

a. Having the software tools to do this and knowing how to use them;

b. Knowing / having the tools to hold the camera steady (usually this means using a tripod, although I have hand-held on occasion); and

c. Having a camera that can be operated in manual mode, ensuring that the aperture, focal length and focus point do change between exposures. Small point and shoot camera is not going to deliver that functionality.

As well five of the nine posted shots have a movement component and would not work with this technique; something that you have already pointed out.

Bottom line is that we have a technique that can be used in some circumstances but is not a generalized solution and certainly not one I would suggest to Damiano. I've gone through all the images he has posted on his website and they suggest a photographer who takes an intuitive / impulsing approach rather than a considered approach in his work (although I could be wrong). There is nothing wrong with either approach, but the technique works well for landscape photography, not street photography.

Thanks for the clarification.

Damiano's camera does have a Manual Mode, for what that's worth.

I have no idea about his software, nor his ability use it, so no comment about that ...

[Manfred M]

Thanks for the clarification.

Damiano's camera does have a Manual Mode, for what that's worth.

I have no idea about his software, nor his ability use it, so no comment about that ...

I suspect manual mode on the Panny = manual exposure only. I'm not convinced that the focus by wire can be disengaged and that it can be locked properly. I know some of the older Panasonic cameras had some limitations due to how the focus was implemented. I don't know his specific model.

[xpatUSA]

I suspect manual mode on the Panny = manual exposure only. I'm not convinced that the focus by wire can be disengaged and that it can be locked properly. I know some of the older Panasonic cameras had some limitations due to how the focus was implemented. I don't know his specific model.

TZ70 aka ZS50.

"The camera has full manual controls, Raw support, focus peaking, plus a control ring."

https://www.dpreview.com/products/pa...asonic_dmczs50

Perhaps Damiano might knock something while impulsively shooting stuff ...

[fatoddsun]

oh wow guys, you are amazing! : ) Thanks so much!

As for the various points discussed above:

More on how I shot and processed this image
As someone mentioned above, my tiny little camera at least allows me to shoot in raw, so perhaps starting from there instead of the compressed image could help keeping a little more detail. That's what I thought, at least, so I tried to export to jpeg straight from the raw without touching any tone curve or level clipping. I only applied the same (soft) noise reduction with the exact same settings as before. You can see the old photo here and the new one here. It is quite clear that in the first post-production process I deliberately clipped even more part of the shadows because of a personal preference. I wanted a picture with lots of "true black", and little light defining the scene only. I know you might disagree on a personal taste level, but that's the look I was researching, and I have to say that I was quite happy with how the picture was looking on my screen. That said, the problem clearly arose at a second stage, that is when I realised that the printed version of the image was actually collapsing most of the remaining shadows to pure black too, hence my disappointment. How do you normally prevent this? Do you usually use different processing settings for the copies that will be printed and those who will live on screens? Also, I can select both sRGB and Adobe as colour profile. Does it make any difference at all for black and white pictures?

Longer exposures
I like the idea of trying with longer exposures, perhaps from one of those mini-tripods, at least when there are no people in the shots. I am not a fan of longer exposures necessarily, but I can very well see like in these cases they would greatly help to have a neater final product.

Camera modes/blending multiple images technique
With the right fiddling my camera should allow for this technique. It is a little inexpensive tool, but I have to say that it defends itself quite well for its tiny dimensions! For post-processing I typically use Darktable only, but in the past I gave a shot to Gimp to try a bit of "HDR" techniques with reasonable success. I don't exclude I will play around again with it

Shooting approach
I think this is the main point for me, as noted earlier by Manfred. I am indeed an "impulsive" photographer, and I love a very spontaneous approach to my photographic experience. Walking around, "stealing" the glimpses that catch my eye, trying to capture a feeling, or a certain mood in a picture is what makes me happy to shoot. This is why I like carrying a tiny camera with me. I am not a purist of the image, and I don't look for perfection in my shoots, neither artistically (if such a thing exists) nor technically. Nevertheless, I appreciate the limitations that come from a specific type of camera in difficult settings like street night shooting. You have rightly indicated several flaws in my current setup and production, and highlighted just as many solutions that could work for me. I am seriously considering the possibility of getting a better camera in the near future, possibly a compact with a bigger sensor, like a Ricoh Gr ii or a Fujifilm XF10.

Thanks so much for your help guys!

[Manfred M]

oh wow guys, you are amazing! : ) Thanks so much!

As for the various points discussed above:

More on how I shot and processed this image
As someone mentioned above, my tiny little camera at least allows me to shoot in raw, so perhaps starting from there instead of the compressed image could help keeping a little more detail. That's what I thought, at least, so I tried to export to jpeg straight from the raw without touching any tone curve or level clipping. I only applied the same (soft) noise reduction with the exact same settings as before. You can see the old photo here and the new one here. It is quite clear that in the first post-production process I deliberately clipped even more part of the shadows because of a personal preference. I wanted a picture with lots of "true black", and little light defining the scene only. I know you might disagree on a personal taste level, but that's the look I was researching, and I have to say that I was quite happy with how the picture was looking on my screen. That said, the problem clearly arose at a second stage, that is when I realised that the printed version of the image was actually collapsing most of the remaining shadows to pure black too, hence my disappointment. How do you normally prevent this? Do you usually use different processing settings for the copies that will be printed and those who will live on screens? Also, I can select both sRGB and Adobe as colour profile. Does it make any difference at all for black and white pictures?

I'm going to break up my answer into the sections you have used to try to keep the read to manageable length:

1. Raw versus JPEG - I don't know the native bit-depth of the raw files on your camera; typically this is 14-bits but 12-bit might be used as well. I had a look through the camera manual but could not find this information. Regardless, this is significantly more than an 8-bit JPEG. More data is generally better than less...

The only issue is that additional processing is required by the photographer - sharpening, contrast, noise reduction area all done when the JPEG is created and the photographer has to make those adjustments when using raw data.

2. As a general rule, clipping highlights and shadows is something that needs to be handled very cautiously. One is throwing out data and unless one is using a non-destructive workflow with a pixel-based editor or with a parametric editor. Data loss is usually not desirable. There are times when it makes sense to do so in order to improve global contrast.

In B&W work, it is generally desirable to have a small amount of pure black and a small amount of pure white in an image to ensure a full tonal range in the image. This can generally be achieved in a B&W image even if those values are not recorded by your camera by setting the black point and white point appropriately. If your original image shows pure black in the histogram then there is generally an underexposure issue. If there is pure white, other than specular highlights or light from a light source (which is common in night time photography), this can be a sign of overexposure. If the histogram shows both crushed shadows and clipped highlights, you have exceeded the dynamic range of your camera's sensor.

Large areas of pure black or pure white are generally not considered good practice. In my wet darkroom days, the rule I was taught for B&W work was every B&W print needed to go from pure black to pure white with no more than a 90% range; i.e no more than 5% of the image should be pure black and no more than 5% of the image should be pure white. In those days, unless images met that spec, a photo editor would reject the print. Opinions these days are more varied, but all the printers that I hold in high regard still adhere to that rule in the digital world. In competitions, trained judges will deduct points (major and minor) when they see crushed shadow detail or clipped highlights in both digital images and prints.

3. When it comes to preparing an image for printing, there are steps taken after the image editing process has been completed to prepare the image for printing. These will be specific to the image and paper used and are largely just about sharpening and resizing. In some cases there are adjustments made for differences in print brightness to compensate for computer screen brightness and the illumination level that the print will be displayed under. In my case, my screen is set low enough that I don't need that brightness compensation step. If you are finding issues with areas blocking up on you, that is related to your processing.

One thing to remember is that if you work on a computer screen in a reasonably dimly lit workspace, you should be getting a dynamic range of around 10-stops. A typical print under gallery lighting runs at around 8-stops for coated luster or glossy papers and around 7.5 stops for uncoated matte fine art papers.

[Manfred M]

Longer exposures


Longer exposures are sometimes a necessity (it's hard to capture fireworks without this technique) and sometimes a creative approach (as an example of the light streaks from cars and buses I posted earlier in this thread). It is generally not used in street photography and without question it requires some level of planning as well as the tight tools.

There is no real "clear" definition of what long exposure photography is and exposures that run from fractions of a second to minutes and hours all fall into that category. Most photographers will shoot with a shutter speed that ensures that they get a sharp image without any motion blur, but there are creative techniques. One of the first decisions I make before I take a picture is what to do with movement; do I want some in the image or do I want to freeze it?

Long exposure photography generally requires some method of stabilizing the camera with a tripod and ball head being the most common tool. Shorter exposures can be made using a monopod. Some techniques require the use of neutral density filters mounted to the camera lens (something that would be challenging with your current camera). Most cameras are limited to 30 second exposures, so a remote cable release and "bulb mode" is required to keep the shutter open (again a feature I suspect is missing from your current camera. I sometimes combine long exposure photography with flash photography (a technique referred to as "dragging the shutter") that uses a mix of ambient light and flash for creative purposes.

Long exposure photography is also a technique that is used to blend multiple images together; smoothing water can be done this way as well as with neutral density filters, but it requires software that can stack and manipulate multiple images at once. I sometimes take shots over a number of hours - from the beginning of golden hour to the end of blue hour and blend these images in post.

[Manfred M]

Camera modes/blending multiple images technique

I never use in-camera tools to modify my images.

The images my camera outputs are the raw material I use to create images, so I want the cleanest, least processed images possible to feed my creative process. Yes, I can do things like HDR in camera, but I generally don't like the output and prefer to hand tune my image.

Most of the time I shoot raw + JPEG. A lot of the photographers on this site shoot raw only.

[Manfred M]

Shooting approach

To each his or her own...

I'm a touch surprised why you didn't ask me how I recognized your approach...

I do use your technique and wander around taking random shots of things that interest me. Sometimes I take my phone, sometimes my little Panasonic camera and sometimes my large camera. These are my planning shots and when I go through them, the ones that look interesting and intriguing are the places a revisit and shoot again, this time shoot them deliberately and with a plan. Try heading back to some of these locations after you have had a good look at your work and reshoot the same scene (as much as possible). See which image works better.

I have run into many photographers that use your approach and a number of them have approached me to ask how they can improve their photography (I teach photography at a number of the local camera clubs). I generally tell them to buy a decent tripod and use it for virtually all their work. The ones that do have come back to me to tell me well this has worked and it is something you might want to consider.

The reason the tripod works for most people is that it slows them down and turns photography from a quick intuitive snap into a more considered process. They generally find fewer blown shots and better compositions. It also helps you recognize potential shots a lot faster and even your "grab shots" will come out being stronger because to the additional skills you have picked up.

[DanK]

Agree 100%. This is one of those cases where the equipment is limiting what the photographer would like to do.

In my experience, there are only two solutions:

1. Don't bother with this genre / aspect of photography and move on; or

2. Upgrade to equipment that is capable of this type of work.

I've thought about this, and I think it may be the wrong advice, although one would have to try things with Damiano's camera to know, and I think it could unnecessarily discourage him from continuing.

I agree that his camera is likely not sufficient for really high-quality images of this sort. However, that shouldn't obscure the fact that he is doing things that he can fix--using an ISO higher than his camera can really handle and badly underexposing. My advice would be to start by fixing what he can with the equipment he has. I would drop to ISO 800 or 400 to start (my guess is that 800 will still give him noise, if he wants that), use longer exposures (not just to compensate for the lower ISO, but more than that), and then darken the image either globally or locally in postprocessing if the result of the longer exposure is a brighter image than he wants. I am guessing that these steps would substantially reduce the problems that led to his original posting. that is, I think if the last step were done carefully enough, he could have very dark shadow areas with printable detail.

If he then wants to upgrade his equipment, the question would be how good that equipment has to be. One doesn't need high-end cameras to do night photography successfully, although it does help. I have done some with a Canon 50D, an old and quite noise-prone camera that you can now pick up used for about $200 or $250 US. I'll post a couple below. It is true that that camera could not handle very long exposures, e.g., more than 10 minutes, without problems, but for some wilderness work and certainly for urban night photography, it's adequate if you expose to the right and keep ISO down. The images from my 5D III are better, but the ones from my 50D are serviceable. It's not just because of the low resolution here that the photos below look reasonable. I have a 13 x 19 (A3+) print of the first that has gotten consistently very favorable responses and that is in a portfolio I displayed at a local photography museum not long ago.