Processing "dull", gray-sky images in PhotoLab 5

BHAYT · January 27, 2022, 10:49pm

@mikemyers I could not process the image in DxOPL so I “attacked” it in SageLight (it would have been way better in PL5) and cropped in FastStone and wound up with an image similar to that of @Wolfgang. I have chopped off the top of the crane and would probably have tried to remove some of the remnant but …

I like the overall picture and when cropped the boats appear to be “travelling” in an arc, albeit “Margaritaville” is at anchor I believe. Plus we have the rule of 3, 3 liners (O.K. 1 and 2 bits) and three smaller boats (excluding the painter).

To me the real value is in the original photo because it captured a moment in time when you might have been in the wrong place at the wrong time (I don’ actually believe that) but that was where you were when you took the picture!

I am mostly a “philistine” and I “hate” shallow depth of field and Bokeh!! I favour smaller sensor cameras for a number of reasons but the biggest reason is the greater depth of field, I mostly take pictures of gardens and the plants in them and want the whole of the head of a plant in focus not just the tip of one petal!!

With respect to the fault I was being mildly sarcastic and will try the second picture later. The fault occurs as PL5 is trying to render the image for display I believe and the “crash” was a message about an Internal Error where the rendered picture should have been, the program didn’t hang but could not be persuaded to go any further with the picture even when the DOP was removed?

Here is my image,

I wonder if there should have been a bit more water to the right but…

Having said I hate shallow DOF, this picture actually works I believe, taken with a 12-200 lens (24 - 400) on an Olympus EM1 Mkii @ full zoom (so as not to spook the flower)! Not as dramatic as your skyline but just in my front garden!

EDIT:- The second photo also caused an “Internal Error ()”, to be honest I feel that it would have been more useful to have a value between one ( and the other ) to identify the error but to the developers the chances are that an error at the point this occurred is probably fairly easy to locate!

mikemyers · January 28, 2022, 12:30am

Since I don’t have Windows, I can’t do this. Maybe you can create a new thread someplace in this forum for posting errors, link to both images, and copy what you wrote before in as much detail as possible, so the DxO tech people can replicate the error.

I’m completely at a loss to explain this - it might be another leftover “bug” from the Apple OS, but you’re using Windows, so that’s out. Maybe re-start your computer, delete these images from your file system, and re-open the image as new. It should load, as your system won’t recognize anything wrong. When it crashes, take a “screen capture” either from your computer or phone, and post that along with everything else.

If I’m saying anything incorrect, @Joanna will correct me. Of course, they will want to know everything about your Windows PC.

Anyone else involved here, while using Windows? Does the image cause the same problems for you?

Joanna · January 28, 2022, 10:52am

This is definitely not a problem on my Mac and I am guessing it might be a Windows only problem, although I can’t imagine why.

The best thing for Bryan (@BHAYT) to do would be to download the other one taken by the same camera and see if that faults. If if doesn’t, the only thing left is that the file got corrupted somehow when he downloaded it. (or it could just be a Bryan thing )

The lens you are using isn’t getting written to the EXIF consistently.

With the first image…

[EXIF]          Lens Info                       : 0mm f/0
[EXIF]          Lens Make                       : Leica Camera AG
[EXIF]          Lens Model                      :

With the second image…

[EXIF]          Lens Info                       : 21mm f/2.8
[EXIF]          Lens Make                       : Leica Camera AG
[EXIF]          Lens Model                      : Elmarit-M 1:2.8/21

In any case, this lens is around 60 years old (!!!) and might possibly be great optically but, from a point of view of relating to PL lens modules, I would say you have very little chance. It’s certainly not on the compatibility list. Since it’s the camera that writes the EXIF, something really screwy is going on in your M10 if it can write two different models.

As regards this image, if you are ever going to hope to get a reduced depth of field at this kind of distances you are going to have to use a much longer lens.

Here is a shot, taken from one of our bedroom windows, with the focus set to the chimney on our neighbour’s house, about 50 metres away, both at f/5.6, which is maximum for my 28-300mm lens at the lengths used.

200mm focal length…

300mm focal length…

At 200mm, I am just starting to get a reasonable separation from the background, but it takes 300mm to really start to make the foreground subject stand out from such a complex background.

BTW, both shots are full frame, which highlights the problem of having to be far enough away from a distant subject to get it to fit in the frame, but the further away you are, the longer the focal length has to be to restrict the depth of field enough to separate it from its background.

Now try using a decent camera with a recognisable lens to do the same thing

Wolfgang · January 28, 2022, 11:20am

Bryan,
you have to remove (deactivate) Mike’s watermark.

I ALWAYS have to do so.

Then, Mike explained somewhere, that his famous cam doesn’t communicate with any of his lenses. So, no DxO lens profile …

platypus · January 28, 2022, 11:28am

…newer Leicas (2006ff) use the following: Leica M-Objektivkodierung – Leica Wiki (deutsch)

Older lenses can be “upgraded” by Leica. This page says it’s 250 Euros/lens, ouch.
Black marker can do the trick too, at least for a while, until you have to repaint the code.

mikemyers · January 28, 2022, 11:51am

None of the Leica M cameras have any “connection” to the lenses used. The lenses mount electronically. Leica added “coding” to their new lenses, so the camera will know which lens is which, but I don’t have any of those lenses.

I had left lens detection “on”, so it picked something, which I ignored. I have just turned lens detection OFF so the camera shouldn’t be adding any lens information to the EXIF data.

I’ll take another similar photo later today, and we’ll see if that makes a difference for people using Windows computers.

Joanna, I think the DoF calculations are for a full-frame 35mm image. Since I’m using only around half of my image because of cropping, my 135mm lens is behaving like perhaps a 200 or 220mm lens, I would guess. That should also have an effect on the apparent DoF, since the image is “blown up” more. Not sure how to apply this though, since a “normal” photo from a LF camera has minimal DOF compared to a similar photo from a camera using very small film, or maybe a mobile phone… I have to think this through when I’m more awake.

mikemyers · January 28, 2022, 12:07pm

Sounds to me like a bug in PL5 - if it doesn’t recognize the watermark, it should ignore the watermark, not crash.

I hope one of you can send me a Windows file, with a watermark, so I can try to open it on my Mac. I’d like to know if this also causes a crash.

BHAYT · January 28, 2022, 12:54pm

@joanna my FZ330 would make mincemeat of that (I think) tiny sensor 600mm zoom constant f2.8 mostly useless RAW, O.K. to good JPGs with reasonable light, “cold” images as soon as the light fades! But most of the time I don’t want separation, regardless of how heretical that may be. I hate “misty” water (and yes I do understand why it happens) it has been made into a “fashion” because of the issues with long exposure. I want to see the water the way that my eyes see it (plus I have a balance problem so tend not to go out photographing when the light is getting really low!)

I bought a GX80 (on offer) to keep it company because with my Olympus f1.7 17mm (34mm) I can fit both into my camera case and now with my Pixel 4a I can roam around and cover most situations (at least with JPGs), except I tend to favour the G9 with the Olympus 12-200 (24-400), which needs more space (Christmas present from my wife was a bigger case that can take the G9 with the 24-400 and the GX800 with the f1.7 17mm, now all I need are the “muscles” to carry the load)!!

JoJu · January 28, 2022, 1:01pm

Are you using text or another picture as watermark? If text, it will likely be a font which is not available or differently named on Windows.

In this case one could say, a couple of things are still not compatible between Mac and Windows. And one could also say, PL should use a fallback font. But then, which photog will give his delicate and most evolved watermark together with a DOP file to other photogs? Alright, now I know one About the photo I think Joanna summarized pretty well the other “problems” or issues of it. Postponing the composition work to the post process usually leads to a bit (or two bits) frustration like “why couldn’t I see it when I had the camera pointed at it?” Simple answer: because I didn’t take the time I took looking at it lateron. You picture could have been done with any smartphone and some would add a bit fake bokeh.

Joanna · January 29, 2022, 9:59am

This is not the case. DoF is calculated based on the physical focal length, aperture and sensor size, not things like a virtual sensor size.

It also correlates with factors like CoC (circle of confusion) to govern when diffraction starts to kick in, which starts to make everything slightly softer as the aperture gets small enough to exceed the diffraction limit for a given aperture.

This is something TruDoF-Pro takes into account and limits what apertures you can use to avoid diffraction. By default, the blur spot diameter is set to 30µ, which gives a diffraction limit of f/14 for a full frame sensor. In actual fact, minimum blur spot diameter can be customised to better suit high resolution sensors as it is calculated as a function of the distance between pixels. Having done this calculation for my D810 and D850, I use 20µ, which ensures sharper images but at the expense of a more limited depth of field. It’s all about the end use for the image and its intended viewing distance. Because I print big and at 240ppi, I need all the sharpness I can get but if your only audience is going to be computer screens at 110ppi, 30µ is perfectly fine.

But if you are planning on cropping and want things a tad sharper, your D750 will play very nicely at 20µ, so you might like to see what effect that has. Although, for minimum DoF, you want the opposite end of the scale with as wide open aperture as possible.

I did this series of shots of guitar strings, with the focus on the strings, to show the effect of aperture on diffraction. You can see that the smaller the aperture, the more the reflections on the strings become “fuzzier”…

mikemyers · January 29, 2022, 3:56pm

Before I read further, I now understand that I have had some inaccurate, or maybe inappropriate, thoughts stuck in my mind over time, without really thinking about them.

First, wide angle lenses have more DoF than “normal” lenses, and telephoto lenses have even less DoF. This fits into my previous thoughts - irregardless of anything else, the DoF for lenses that are closer to the film or sensor is more than the DoF for lenses further from the film or sensor, for the simple fact that the focal length is shorter or longer.

Without having thought about it in detail, I know the DoF for large format cameras is less than for 35mm which typically have more DoF - for the simple fact about LF cameras have longer focal length lenses. Again, small-frame cameras typically have even more DoF, for the simple fact that their lens have even less in focal length.

This explains why a zoom lens on any given camera has much more DoF in wide-angle (using a small focal length) compared to telephoto, where the lens is usually zoomed out, to a longer focal length. So my hypothetical 25-200mm lens would have more DoF at 25mm, and much less at 200mm.

Moving on…

What happens to DoF as the aperture size is changed? Since the opening in the lens, the aperture, is decreased or increased, it seems logical that the DoF would decrease with the lens wide open, let’s say f/2, but increase as the aperture was closed down, perhaps to f/22. I understand this, but I’m still struggling to say exactly why this happens. It seems logical to me, but that’s probably based on years of experience, not something I can clearly explain.

Yes, I also know about diffraction - once the lens aperture is closed to a minimum, light scatters around inside the camera, which we call “diffraction”. That’s a good reason to stay away from extremely small lens apertures. I need to read up on it, to remember the “why”.

Moving on, let’s take my example photo, with a 135mm lens mounted on a standard size 35mm camera shooting a 24 x 36mm piece of film or sensor, at an aperture of f/4. I can look this up online, or by viewing the DoF scale on the lens. Now, what happens if my lens couldn’t capture the only part of the image I wanted, and I needed to crop the image? Let’s say I cropped the image to half its normal size. It is still a 135mm lens, at still at an aperture of f/4. My question - is the DoF the same?

**At first, I think it’s obviously yes, the same, BUT then I consider that my image is so small, I need to enlarge it more to get the same size print. The “circles of confusion” which are used in calculating the DoF will be different, as the larger image will naturally seem to be less sharp. This probably also need to be part of the equation for calculating DoF for cameras with larger or smaller film/sensors. (I need to think this through, but I think I’ve been missing something - and the film/sensor size needs to be included in the equation for calculating DoF.)

More later - need to study what you wrote again, and to do some more reading…**

This clearly demonstrates why stopping the lens down too far can lead to a decrease in apparent sharpness. I don’t fully understand it, but I do understand it well enough to not do this. If there is too much light, an ND filter would be better than going to f/32, for example. Thanks for making this so obvious!

Joanna · January 29, 2022, 4:55pm

Which explains why it is very difficult to get anything blurred on an iPhone

Let me give you a concrete example: with LF, for portraits, I use a 400mm T lens (the T means I don’t need 400mm of bellows) at f/9, to give me the front of the face sharp but not much else.

On the D850, I use a 85mm f/1.4 prime lens at f/1.8 to give me roughly the same DoF.

400mm / 85mm = 4.7

f/9 / 4.7 = f/1 but that would only give me ¾" DoF (if I opted for the smaller dot size of 20µ) hence choosing f/1.8, which gives me just under 1½" tack sharp (for the eyes), fading gently away so that the rest of the head is softly focused without being too blurred.

Laws of optics plus you’ve had it drilled into your brain since you were nowt but a lad

As I asserted earlier, absolutely! Cropping after the fact doesn’t change the DoF that is already inherent in the captured image.

Indeed, but that has nothing to do with DoF or even diffraction, it’s the simple fact of the the more you enlarge an image, the more pixels have to be “invented” to fill the available space - it’s called interpolation.

Not from a DoF point of view. It’s only going to affect the aperture at which diffraction kicks in.

Nailed it! The problem is how many people nowadays think of using ND filters? Hint - I do

mikemyers · January 29, 2022, 5:39pm

Eventually this too sank into my brain. So when I select a small part of an image, and crop, I haven’t in any way affected the “depth of field”. The sensor/film size doesn’t matter, cropping, or not, doesn’t matter, the DoF is a function of the focal length and the aperture.

So, were you to shoot a portrait using your LF camera, with a correspondingly long focal length lens, you would get less DoF than were you to switch to your D850 - assuming you selected a focal length to keep the portrait looking natural, rather than huge noses and tiny ears. The other factors have nothing to do with DoF…

You wrote if you use an 400mm lens at f/9 (to get the front of the face sharp, and nothing else) were you to use your D850 this would mean using an 85mm lens to get a similar effect.

400mm at f/9 compared to 85mm at f/1.4

So, your LF camera, with large film fills the frame with your subject with the 400mm lens, and
Your D850 camera similarly fills the frame with your subject - yes?

What my brain is coming up with, is your 85mm lens has too much DoF to get a similar result, so you need to cut the DoF down by using a wider lens aperture, f/1.4 …yes?

If I’m right about this, the “reality” of my brain is correct, even if I never really thought it through until just now.

I think this is all starting to come together in my mind, but before I think about it any more, I’d like your thoughts on what I just wrote. (I feel like I’m talking to a professor in a advanced level photography class, but I would really like to understand this.)

OXiDant · January 29, 2022, 5:40pm

Because they think they are Non Digital filters… ?

mikemyers · January 29, 2022, 5:53pm

My Fuji X100f has a built-in neutral-density filter, which I use quite often to force my lens to open more, giving me reduced DoF. I wish my other cameras had this. Leica never will, but I’d like to think Nikon could do so.

I usually don’t think about it, because I don’t have any ND filters. I also dislike adding yet another lens surface in front of my lens.

Both Nikon and Leica are now offering electronic shutters, with extremely high speeds. Are these an acceptable substitute for ND filters? My M10 goes up to 1/4000th of a second. I think my D750 goes even higher, and the new Nikons go even higher, much higher. Is there still a need for ND filters?

platypus · January 29, 2022, 6:03pm

Shallow DOF + Motion Blur + Bright Light = ND Filter

If you want to capture a shallow DOF image that shows some motion blur in bright light, you need longer exposure times than 1/8000. In such occasions, an ND filter can be quite handy.

BHAYT · January 29, 2022, 7:25pm

@mikemyers I also need to check what has been written about DOF but what I certainly know is that when I moved from my FZ38 (18 x Bridge camera with a 1/2.3" sensor) to my LX7 (3.5x f1.4 with a 1/1.7" sensor) I noticed the change in DOF partly caused by the much faster lens but also the lens size.

To get deep DOF I need to sacrifice speed for depth (decrease in aperture/increase in f) which will hit either shutter speed (blurry pictures of the grandchildren) or higher ISO which is “poison” on an old small sensor camera (the LX7 is only a 10mpixel sensor to help keep noise down). Things have changed with smartphones with high megapixel counts and pixel binning and computational photography (my pixel 4a takes acceptable pictures but they are over sharpened and it struggles to get the colour blue right (it attempts to enhance the “sky” and it plays havoc with blue flower heads!)

A comparison of sensor sizes

My current cameras are LX7 (1/1.7" with 77,000 on the clock), FZ330 (600mm zoom with 1/2.3 sensor), GX80 (4/3rds sensor @ 16 Mp), G9 also 4/3rds. (20Mp) and EM1 Mkii (20Mp) another 4/3rds) . I have only 1 prime, a 17(34)mm f1.8 (the other prime I got free, a 17mm(34mm) f1.2 is currently with my youngest son). All my other lenses are zooms that came attached to the various 4/3rds cameras I bought and then eventually sold the bodies to part finance the upgrades, plus an Olympus 12-50 with a pseudo macro function.

I got used to zooms with my earlier bridge cameras; zooms reach the back of deep garden beds, they allow shots to be framed to avoid as many visitors to the gardens as possible (how dare they come at the same time as me!). I want as much of what I am taking to be in focus for the maximum amount of the time.

The following was a test I did some time ago taking pictures of Skimmia berries with a G7 and the Olympus f1.8 17(34) lens. While I like @joanna’s examples with the guitar strings the objects I frequently photograph are deeper (more 3D) and irregular shaped e.g the bell of a flower, the cluster of berries etc. (some/most of which will insist on moving with the wind!!)

According to the book of words “at ƒ/8, diffraction limiting begins to set in, but there isn’t a significant impact on sharpness until ƒ/11, and even then it’s just a slightly overall decrease. Images start to lose their sharpness at ƒ/16, and become moderately soft at ƒ/22.”

From my perspective having some of the berries on the same cluster out of focus at f1.8 is why I want greater depth of field, going all the way up to f22 brings other penalties and was only possible because it was a very sunny day, even then speed had dropped (a lot) and ISO had started to climb.

The picture I put in a post earlier in this topic, of a self-sown hebe, resulted from me not wanting to grovel on the floor to get close to the flower head so I use the zoom!

I do have trouble with the pseudo macro lens and wind up with the following

Heavily sharpened but with some of the petals clearly out of focus (Olympus 12-50 locked in “macro” mode) and

mikemyers · January 29, 2022, 8:45pm

One thing at a time - for now, DoF.
If I take a photo with my 50mm lens, say at f/8, I will end up with a certain amount of DoF.

Now, let me cut that negative of file in half, so instead of one image measuring 24 x 36, I have two images measuring 24 x 18.

I know both of these have the same DoF, as they were taken at the same time, at the same settings. To me, this proves @Joanna right - image size does not affect DoF.

Everything else you mentioned is “something else”, but I suggest we consider them one at a time. On all your cameras, if you shoot at the same focal length, at the same aperture, you will get the same DoF.

But other things change - especially when you change to a different focal length which is perhaps more appropriate for a camera with a smaller sensor.

To consider:
Blurry photos
Shutter Speed
Different ISO
Sharpening
Composition

This will likely involve more things than simply changing the f/stop to say, f/22.
Regardless of what you do, there will be only ONE distance at which the image will be in focus.
What you are trying to achieve, is getting more of the image to be in acceptable focus (DoF).

Everything is a compromise, and you likely need to think of additional changes in your settings to get a good image.

Suggestion - enter just ONE photo here, from ONE camera, and ONE lens, and identify what you see as a problem, and others here may be able to help. PhotoLab can only do so much, and you always need to be aware of GiGo. (Garbage In, Garbage Out). You have to find the best compromise in your camera, such that PhotoLab can give you the best possible result from that image.

OXiDant · January 29, 2022, 9:14pm

Effecting dof.
1 Aperture
2 sensor size
3 Focal length and focus point
so larger hole, less DoF -1, bigger sensor, again less DoF -2, longer focal length. -3.
A FF 600mm f2.8 has a razorthin DoF. 0.02m 2cm!
A m43 600mm efl thus 300mm f2.8 and object at 10m distance has 0,09m DoF
So i shoot around f5.6 is 18cm till f9 29cm at 10m. And longer distance say 20 it’s 1,19m at f9
.

Source photographers companion app.

mikemyers · January 29, 2022, 9:51pm

I see nothing exotic about my watermark - it is just “text”, in a font that I found on PhotoLab 5.

If anyone wants to copy it, it’s “Chalk Duster”.

Yes, this is on a Mac, so if Windows has something similar, it probably is listed with a different name.

My suggestion - PhotoLab should not crash if it sees a font it doesn’t recognize, or in this case Windows doesn’t recognize - better to use some “default font”.