[wim]

Thanks for quick replay Brightcolour.

See my comments bellow

Ooops you aleredy make an assumption and forget for the second one.

 

The second assumtions is - How do you get the saturation point? 

 

My point is that above metioned absolute Fstops are meaningless unless you specify the conditions. 

This values can be used only to compare two cameras - same as photozone's LW/PH for lenses

 

Some test methodes defines DR fot low quality Normal quality and high quality and the difference is between 2.....5stops

 

PS: My personal taste is 2 stops lower than Dxo

 

Do you mean to say you overexpose by 2 stops? Lose two stops? How? Where?

 

As to comparing cameras: you can't, unless you have specialised equipement to measure such stuff. DxO uses a specific noise floor to measure DR. The funny thing is that Nikon and Sony manipulate Raw (I'd have to find the scientific article again), artificially lowering the noise floor by blacking out (parts of) the noise at certain low DR levels. This allows them to reach the minimum level at a lower point in the measured DR range, because the remaining noise on average remains below the DxO noise floor a bit longer.

 

There is in itself nothing wrong with this, other than that it potentially throws away information. Canon does not do this, resulting in more noise at higher DR levels, and hence also retaining more info. However, it is not easy to extract the information which others effectively throw away.

In practice, the lower DR levels are so low that it is almost impossible to see anything but black anyway, so it is a bit of a moot point whether a camera with advanced preprocessing prior to saving the Raw and hence getting a higher DR range than another which doesn't, is better than the other. It is a matter of preference.

 

In the past, with film, black was black, the lower part of the tonal range did not show any detail at all, whereas there still was some detail in the high ranges. With digital it is the opposite. Info is lost when overexposure happens, but more can be obtained from the lower tonal ranges than ever was the case with film.

 

Generally, I like contrasty images when processing my own stuff, and therefore I will lose the lowest and/or highest end of an original tonal scale anyway. Funnily enough it does not make a difference in that case whether a DR is 12 or 14 stops; I only need 8 to 10 at the most, and with proper exposure I generally do not need much if any leeway. I am quite capable of processing a jpeg to the exact view I want and had previsualized without even bothering with Raw;  I happen to be capable of ending up with the same end result in most cases - when viewed on screen or in print .

 

Kind regards, Wim

[JJ_SO]

I wonder why one would want to measure a bird's eye? One does not have to meter a subject, one can meter for the light conditions. One can even rely on the camera to make the judgement when needed (called "evaluative metering" in Canon land, "matrix metering" in Nikon land).

 

As normal tonal curves with nice contrast span around 7 stops of DR, that would leave 5 stops of headroom, what is the obsession with that 14 number about, exactly?

 

I too had an interest in"HDR", many years ago, I shot scenes with 3 bracketed shots with a 6 stop spread. Luckily it only lasted a short while, finding the unrealistic, tasteless cheap scifi movie results as unattractive as they are.

 

It is very interesting to me that I never see your examples of 14 stops of DR?

 

A bit wrong assumptions and the good old children reply "why do I never see your examples" - childish at first look, but on the other hand I obviously did a not so bad job in showing huge DR pictures not looking like those plasticky "HDR"-thingies we share at least the same dislike to   Let's just agree I show more pictures than you did lately, but amongst the ones I could find in PZ's archives were nice ones. Like the silhouette pictures - here one can say making the best out of a handicap (working with low DR sensors) but one can as well say, the idea and the realisation is really great.

 

Each nightshot has a bigger range than only 14 stops. Snow landscapes, deserts, beach scenes, a simple evening sky, inside a church, front light situations but I don't want to flash against the shadows. 

 

The birds eye was my reply to Wim who suggested to put the birds eye on zone V. That would be the tone of a standard grey card. Which is far too bright for a bird's eye and we still have the problem to measure it. Using zone system is measuring the reflected light to get an exposure which gets the best start for the final, but pre-visualized picture - volume light metering (you call it "metering the light conditions" which might be the better term) is not the way zone system works. I say, when using Evaluative/Matrix metering it can happen that important details become too dark.

 

High DR is not only about headroom - also about "foot room". Not only some kind of noisy shadows with big grains, but punchy and clear low tones, as well as great highlight recovery.

 

Look, I don't want to convince you, that 14 instead 12 stops would be better for your pictures - for my pictures these two extra stops DO matter.

[Brightcolours]

Handicapped sensors are only used by Sigma these days (low DR).

 

Nightshots need 14 stops of DR? Maybe if you want to see the detail of the filament of lights, or the whiskers of the cat hiding under that Mercedes 190... 

 

Snowlandscapes 14 stop? No, not that I have noticed. Deserts? I wonder what would make them such dynamic scenes. And I wonder how all those desert movies and photos have been taken in the film days. Beach scenes? My beach scenes fit well within 8 stops, and yes on sunny days. Sunsets? I have shot those in JPEG with good results. And I guess you mean backlight instead of frontlight.

 

Here a 2005 (EOS 350D JPEG) example just to show that claims about high DR in scenes usually is just nonsense:



 

Yeah I know, that sun is blown out, isn't it!  :lol:

 

Of course, if one wants to lift shadows 4 stops whenever there are shadows, one will see less noise with a Sony sensor than with my 6D. No doubt about that. But why someone thinks it is a good idea to lift shadows 4 stops... puzzles me.

 

Another example, here is one of those silhouette scenes, the standard tonal curve OOC image, exposed for the silhouette idea:

 

[ATTACHMENT NOT FOUND]

 

If one wants to be silly and lift those shadows 3 stops?

[ATTACHMENT NOT FOUND]

 

There is more than enough room in 12 stops of DR anyway.

Oh and I forgot, that was an ISO 400 image, I believe.

The Nikon D5 is a great pro tool and can handle snow landscapes, sunsets, night scapes and so on too. And all these things were being photographed on film too, and all these things get photographed nowadays at higher than base ISO ISO settings too.

 

Still have to see a 14 stops of DR image that shows you need 14 stops of DR from you?

[JJ_SO]

Replying to you is just such a waste of time.

[Brightcolours]

Hmm, you want examples and give statements that seem to not based on much (the 14 DR scene statements). I give examples, and you don't want to show how, for your images, the two stops (from 12 to 14 stops of DR) do matter. Your reply speaks volumes...

[miro]

examples

1. many wedddigs shoots - not given examples due to privacy

2. landscapes

taken with canon 400d/50d +-2EV bracketing - CP pol filter to avoid reflections /reduce highliths/

http://forum.photozone.de/index.php?/gal...-bulgaria/

 

http://forum.photozone.de/index.php?/gal...0d-2836pz/

 

 

Sinngle exposure but CPL used to reduce DR of canon 400D

 

http://forum.photozone.de/index.php?/gal...nostalgia/

 

Do we need high DR always?

The answer is NO but is nice to have when need it. 

 

Low DR example .Just played with low key.

 

http://forum.photozone.de/index.php?/gal...tion-time/

 

http://forum.photozone.de/index.php?/gal...-50d-6694/

[wim]

examples

1. many wedddigs shoots - not given examples due to privacy

2. landscapes

taken with canon 400d/50d +-2EV bracketing - CP pol filter to avoid reflections /reduce highliths/

http://forum.photozone.de/index.php?/gal...-bulgaria/

 

http://forum.photozone.de/index.php?/gal...0d-2836pz/

 

 

Sinngle exposure but CPL used to reduce DR of canon 400D

 

http://forum.photozone.de/index.php?/gal...nostalgia/

 

Do we need high DR always?

The answer is NO but is nice to have when need it. 

 

Low DR example .Just played with low key.

 

http://forum.photozone.de/index.php?/gal...tion-time/

 

http://forum.photozone.de/index.php?/gal...-50d-6694/

 

Actually, I think with many of these images, if you haven't done so, they would benefit from slight overexposure to get the most detail, effectively putting the main subject(s) in Zone V or VI, and then pulling back in PP.

 

I think they are excellent pics, BTW.

 

Kind regards, Wim

[wim]

 

+1. Different people different approaches. I have always been using reflectors, screens, fill flash, gradient and polarization filters.

 

Wim, I have one question - how do you exposure the grey dutch sky without gradient ND filter and limited DR? - so called "stapelbewolking" the hallmark of the Netherlands

 

====================

<div>I guess we just have different approaches, and that is all there is to it.
 


 

</div>

 

Sorry, I just realized I had not replied to your question yet.

 

Essentially, it depends on what I'd want to achieve. Detail in the clouds, I tend to expose as the light meter indicates, or even a little more, effectively underexposing, up to 2 1/2 stops or thereabouts. If they are not the main subject, I will likely overexpose for the clouds.

 

I generally have not really encountered any problems yet with clouds, as long as the sun is not in the picture. If it is, and it is important to get the most out of the vision I had, I will create an HDR set and take it from there.

 

Kindest regards, Wim

[wim]

BTW, interesting blog post on DPReview on this subject:

 

https://www.dpreview.com/articles/465344...to-capture

 

For a change, since BC and I were treated as a group, I read his posts in this topic as well, and for a positive change, I do agree with BC .

 

Actually, the article I linked to above, also does, to a degree anyway <ROFL>

 

Kind regards, Wim

[JJ_SO]

Have fun with <ROFL>ing 

 

We were talking about how much F-stops are nice to have - you two kept sticking to the "more than 12 is useless, learn to expose, whimps"-theory. Others were saying, 14.8 could save your ass sometimes - not always.

 

The article is about color depth...  ^_^ which somehow is related to dynamic range like everything in this world is related to every other thing. Basically I read the author saying if you get more DR, let the highlights clip at the same point as with 12 stops DR, but get less noisy shadows and midtones. An acceptable decision - but not the only right one. As always, it depends. To be honest, being spoilt with a huge DR of the current camera (and the next as well) I'm no expert in this kind of exposure tricks. So far, I got my shots the way I liked it.

 

Out of interest: I googled a bit but could not find a way to tell by checking the histogram, how much DR the actual RAW needs to start at left and end at right border of the histogram window. When I tried to verify what 14.8 shops mean, I had to lift exposure and shadows. Apparently the exposure slider shows f-stops, but what are +100% shadows? Anyway, +4 stops from an underexposed to the correct exposed sample are no problem. An overexposed picture with -4 is more a problem, because a lot bright tones are then already clipped. Personally, I don't agree with overexposing and lateron lower exposure - I see the sensor closer to positive than to negativ film.

 

As for the 14 vs 12 bit thing: Forgive me, but I'm not a big fan of "everything set on manual and long dives in the menus before taking one picture"- I like to keep complicated things simple, otherwise I would not need to employ a fat CPU doing all the needed steps for me.

 

I'd save 9 MB by going from 14 to 12 bit. 40 MB to 49 MB - sounds much, but deleting every 5th pictures saves the same disk space - and how could I know in advance which color depth would be sufficient for a certain picture? And bigger color depth is good, no?

 

Well... with two quick and dirty comparisons, pulling up shadows and dulling down highlights, for both exactly the same values I have to say, the bigger color depth gets more colourful noise  So, here I will change settings for this kind of available light shots.

 

The next comparison was a bit more curious - if there's actually a bit of improvement (less visible color noise, less disk space) what would happen if I take these shot with ISO 64? Long story short: I don't know. First I didn't see anything, then thought "yes, you're looking at a 8 bit JPG preview. Render it as 16 bit TIF!" Did so, just to find again no difference I would feel comfortable to call it a difference. The  I thought again "Yes and what is the color depth of Mac OS? In short - no 14 bit at all, not even 10 bit/channel" At least with my Late 2010 iMac. The new big one with Retina display finally got a more professional color depth.

 

So, I ask for patience until I made top my mind to go and throw a couple of grands after Apple and get an iMac Pro once it surfaces in webshops, but so far the differences in color depth became visible at high ISO and not in favor for the bigger color depth. It was more colors, yes, but I would try to get rid of them by using the noise>color slider.

 

I learnt something new (to me, but obvious, now as I think about) However, it still better to have a bigger DR - because I don't only take pictures at "most of the time 12 bits / f-stops are enough)