Pages: [1] 2   Go Down

Author Topic: Dynamic range test  (Read 16419 times)

evgeny

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 495
Dynamic range test
« on: October 05, 2012, 08:06:33 pm »

Hi, does this unprocessed photo illustrate dynamic range of Hasselblad H4D-40?

Thanks
Evgeny
« Last Edit: October 05, 2012, 08:09:28 pm by evgeny »
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: Dynamic range test
« Reply #1 on: October 05, 2012, 11:15:01 pm »

Hi,

In now way.


DR is hard to measure. The best way to do it is to use a technique developed by Arri: http://www.arri.com/camera/tutorials/dynamic_range_test_charts.html

The second best is to use a "Stouffer Wedge", expose correctly to the right and try to recover as many dark steps as possible. Problem is that you need to reduce lens flare and light leaks.

Best regards
Erik
Logged
Erik Kaffehr
 

EricWHiss

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2639
    • Rolleiflex USA
Re: Dynamic range test
« Reply #2 on: October 06, 2012, 01:25:50 am »

If you had a couple cameras and took the same picture and then set the input curves to linear and used neutral input profiles you could use a picture like this to compare DR of the cameras.    Probably there are people with enough experience to be able to make a broad judgement on an image such as yours. 

Definitely in the bright mid day sun if you can catch everything without blocking the highlights or shadows, it seems you have enough DR!
Logged
Rolleiflex USA

MrSmith

  • Guest
Re: Dynamic range test
« Reply #3 on: October 06, 2012, 02:09:54 am »

It shows you used a fill in flash.
Logged

EricWHiss

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2639
    • Rolleiflex USA
Re: Dynamic range test
« Reply #4 on: October 06, 2012, 03:28:48 am »

It shows you used a fill in flash.

Good catch! Looking at the red eyes now how could I have missed that!   And I think I missed the joke too, correct Evgeny?   
Logged
Rolleiflex USA

Dick Roadnight

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1730
Re: Dynamic range test
« Reply #5 on: October 06, 2012, 04:02:03 am »

There is a difference between a practical "real world" illustration and a scientific test?

...and, if fill-flash was not used, and the was a white building in direct sunlight in the picture, it would have been a good illustration... except that sunlight is very inconsistent... depending on the cloud-fill.

For real-world assessment of a camera, post=processing would be used, and Hasselblad phocus is very useful.

This was Hasselblad H4D-50 (before I Upgraded to the 60) with cloud fill and pp with phocus... I thought that I might need HDR, so I took several pictures, but phocus managed.
Logged
Hasselblad H4, Sinar P3 monorail view camera, Schneider Apo-digitar lenses

evgeny

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 495
Re: Dynamic range test
« Reply #6 on: October 06, 2012, 04:07:10 am »

I kept it simple.
Used a built-in -0.7 flash for fill.
I don't see red eyes at 100% magnification, but my eyes are not perfect. ;)

I probably need to grab a 35mm camera to make a better comparison.  ;D

Thanks
Evgeny
Logged

michele

  • Full Member
  • ***
  • Offline Offline
  • Posts: 230
Re: Dynamic range test
« Reply #7 on: October 06, 2012, 05:52:25 am »

I think a better test can be trying to open the shadows... Here is an image taken with a 5d2... Exposed for the highlights, then processed 2 times, 1 for the sky, then for the shadows (+around 2 stops in capture one) and then blended together in photoshop...

FredBGG

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1630
Re: Dynamic range test
« Reply #8 on: October 06, 2012, 04:44:25 pm »

Hi, does this unprocessed photo illustrate dynamic range of Hasselblad H4D-40?

Thanks
Evgeny

Mom in a white silk embroidered shirt next to the lovely kid would have been better example.....
Logged

Wayne Fox

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4237
    • waynefox.com
Re: Dynamic range test
« Reply #9 on: October 06, 2012, 06:49:18 pm »

Hi, does this unprocessed photo illustrate dynamic range of Hasselblad H4D-40?

How did you insure it was "unprocessed"?  By looking at the image, I'm pretty sure some "default" processing was applied by whatever program created the file.  Just because you made no additional changes yourself, it doesn't mean it was "unprocessed".

It's not that easy to create an image without any processing from the raw file, and the resulting image doesn't look like much.
« Last Edit: October 07, 2012, 05:44:30 pm by Wayne Fox »
Logged

BernardLanguillier

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 13983
    • http://www.flickr.com/photos/bernardlanguillier/sets/
Re: Dynamic range test
« Reply #10 on: October 06, 2012, 08:04:44 pm »

This was Hasselblad H4D-50 (before I Upgraded to the 60) with cloud fill and pp with phocus... I thought that I might need HDR, so I took several pictures, but phocus managed.

If I may ask, why did you have concerns about the DR of this scene?

Cheers,
Bernard

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: Dynamic range test
« Reply #11 on: October 07, 2012, 01:51:41 am »

Hi,

If a subject has wide dynamic range and its is processed into an image it will either look flat or harsh, because screen or print cannot reproduce a wide dynamic range. So you need to apply some kind of tone mapping to an image to liven it up.

Some programs may do it on their own.

The image below is not extremely wide in tonal range, is just very boring:


The image has been reprocessed with tone mapping and a technique for darkening the sky:


The third image has been processed in LR4, with the sky darkened in Photoshop using a duplicate layer, multiply and a luminosity mask based on the blue channel.



Now these adjustment may go over the top, see it more as an illustration than a good example.

Best regards
Erik




How did you insure it was "unprocessed"?  By looking at the image, I'm pretty sure some "default" processing was applied by whatever program created the file.  Just because you made no additional changes yourself, it doesn't mean it wasn't unprocessed.

It's not that easy to create an image without any processing from the raw file, and the resulting image doesn't look like much.
« Last Edit: October 07, 2012, 02:57:18 am by ErikKaffehr »
Logged
Erik Kaffehr
 

Dick Roadnight

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1730
Re: Dynamic range test
« Reply #12 on: October 07, 2012, 08:40:45 am »

If I may ask, why did you have concerns about the DR of this scene?

Cheers,
Bernard

Hi, Bernard...

Of course you can ask.

There is a combination of deep shadows and dark foliage... and white painted buildings (and white van) in direct sunlight... I think that without the cloud fill it would have been difficult. (fill flash tends not to work too well at distances between 100 and 500 meters).

¿Has anyone got an app to wirelessly control their cloud fill reflectors from and iPad?
Logged
Hasselblad H4, Sinar P3 monorail view camera, Schneider Apo-digitar lenses

Guillermo Luijk

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2005
    • http://www.guillermoluijk.com
Re: Dynamic range test
« Reply #13 on: October 07, 2012, 08:54:58 am »

It's not that easy to create an image without any processing from the raw file, and the resulting image doesn't look like much.

I think it is very easy, just using a RAW developer that only applies white balance and demosaicing. E.g. DCRAW, or RAW Therapee and ACR with all settings to 0.

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: Dynamic range test
« Reply #14 on: October 07, 2012, 10:05:33 am »

Hi,

I don't think that scene was very challenging. Of course, you need to expose for the highlights, and that may leave some noise in the shadows. With LR, PV2012 seems to achieve some magic in highlight restoration.

Best regards
Erik


Hi, Bernard...

Of course you can ask.

There is a combination of deep shadows and dark foliage... and white painted buildings (and white van) in direct sunlight... I think that without the cloud fill it would have been difficult. (fill flash tends not to work too well at distances between 100 and 500 meters).

¿Has anyone got an app to wirelessly control their cloud fill reflectors from and iPad?
Logged
Erik Kaffehr
 

deejjjaaaa

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1170
Re: Dynamic range test
« Reply #15 on: October 07, 2012, 10:16:37 am »

I think it is very easy, just using a RAW developer that only applies white balance and demosaicing. E.g. DCRAW, or RAW Therapee and ACR with all settings to 0.

ACR applies color transformations (camera profile)... so you might want to cook a special profile then, right ?
Logged

Guillermo Luijk

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2005
    • http://www.guillermoluijk.com
Re: Dynamic range test
« Reply #16 on: October 07, 2012, 10:18:40 am »

ACR applies color transformations (camera profile)... so you might want to cook a special profile then, right ?

DCRAW applies them as well, in fact using Adobe's matrices. But this doesn't affect DR calculations since they are linear.

Doug Peterson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4210
    • http://www.doug-peterson.com
Re: Dynamic range test
« Reply #17 on: October 07, 2012, 10:26:38 am »

I think it is very easy, just using a RAW developer that only applies white balance and demosaicing. E.g. DCRAW, or RAW Therapee and ACR with all settings to 0.

There are no true "zeros" in a raw processor/viewer. Every "zero" or "default slider location" is an active choice made by an image-processing engineer. There are some established standards (e.g. what RGB value should "median gray" have) and some loosely followed conventions (e.g. how strong of an s-curve to apply to the linear output of the sensor). But both the goals and methods of getting there are entirely up to the developers of the processor/viewer and are proprietary, often patented, and vary in surprisingly large ways from each other.

What's worse is I feel this misconception hurts photographers. In the days of film, photographers would select a film and processor based on the subject, intention of the shoot, client needs, and intended style; i.e. they'd pick it for it's look (color, tonality, range, grain, etc) and technical attributes. Now, photographers often shoot digital raw files starting with the assumption that the "default" settings (whether the in-camera LCD/JPG or in their preferred raw processor) are a sort of "truth".

One negative impact of this mentality is that photographers assume different raw processors are just different interfaces to show you the one "true" image. Whereas in fact even the "default" look between raw processors can vary a lot. Everything about the image can vary between processors, from the amount of absolute detail, to the amount and structure of the grain, to the way tonal transitions are rendered, to the way highlights and shadows roll off (i.e. the placement/severity/shape of the toe and shoulder), to the color of the image. The fact is those "default" settings and "0" positions on the slider are largely arbitrary.

Moreover, the math in any given processor is varied from one camera to the next. Both the apparent-to-the-user slider positions and the fundamental behind-the-scenes math. e.g. Capture One does not apply the same underlying math to an IQ180 raw as it does to a Canon G10 raw - it would be stupid for it to do so; the sensors (and the characteristics of the data they put out) are radically different and demand to be treated differently.

I think cooking is a good analogy. Give two chefs the same ingredients, tell them to cook an Ohio-style beef casserole, and you'll get two different meals. Even if you give them a specific point of reference, say for instance your grandma's casserole (e.g. Fuji Astia with a 1/3rd stop pull), you won't get the same meal.

From unscientific surveys at raw-processing workshops I've taught I'd guess that roughly half of serious photographers have never bothered to open a raw file in more than one processor to compare the results. Compare this to the percentage of photographers who only ever shot one emulsion of film with one processor and I hope you'll agree that we've lost something; I attribute much of this loss to this misconception that "defaults = true image".  
« Last Edit: October 07, 2012, 10:43:47 am by Doug Peterson »
Logged

Doug Peterson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4210
    • http://www.doug-peterson.com
Re: Dynamic range test
« Reply #18 on: October 07, 2012, 10:48:26 am »

DCRAW applies them as well, in fact using Adobe's matrices. But this doesn't affect DR calculations since they are linear.

Dynamic range, as defined photographically rather than in engineering terms ("how much shadow and highlight detail can I render in a pleasing way") is very heavily dependent on the raw processor. There is a lot of development in this regard in the last few years, and you need only take a 1Ds II file and process it in Capture One 3 and Capture One 6 to see that difference.

In my experience the difference in how much range can be rendered in a photographically useful way can vary as much as two stops between raw processors.

But then again you could just use dXo's charts to tell you what the DR of a particular camera is - assuming that what you use cameras for is to generate charts and not pictures it will be perfectly accurate.
« Last Edit: October 07, 2012, 10:50:54 am by Doug Peterson »
Logged

Guillermo Luijk

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2005
    • http://www.guillermoluijk.com
Re: Dynamic range test
« Reply #19 on: October 07, 2012, 10:54:16 am »

There are no true "zeros" in a raw processor/viewer. Every "zero" or "default slider location" is an active choice made by an image-processing engineer.

Not really. DCRAW is a good reference for a true neutral RAW development (even the white balance and colour profiling can be avoided). It has not zeros, it simply only works with 0 processing in a linear workflow. By setting everthing to 0 in ACR (including the curve to linear), the result is identical to DCRAW. However if there is any doubt of what's going on under the hood, the solution is as simple as using DCRAW which is a universal and neutral RAW developer.


Dynamic range, as defined photographically rather than in engineering terms ("how much shadow and highlight detail can I render in a pleasing way") is very heavily dependent on the raw processor..

IMO DR is a camera sensor feature. What a RAW processor can or cannot do is somewhat arbitrary and depends on how the software was implemented. My camera sensor's DR won't be worse just because a particular RAW processor can't extract the most of my RAW files. DR is a hardware feature; how we manage and output it to the final image is a software process.

« Last Edit: October 07, 2012, 10:59:41 am by Guillermo Luijk »
Logged
Pages: [1] 2   Go Up