Dynamic Range Measure
/forum/topic/605245/0

1
2 3 end

Pondria
Registered: Jan 11, 2002
Total Posts: 11873
Country: United States

It's been more than 6 months since we updated teh FM DR database. We need contributions from the new camera owners.

The procedure I'm proposing is simple and easy. Constructive comments are Welcome. The essence of the procedure is to measure the upper and lower exposure limits of the camera with common s/w that most of us have.

1. Find a off-white color wall. Just make sure that the area has no feature or shadows. Naturally lighted In-door is preferred.
2. Set the Camera ISO @100 ( or whatever the Native ISO )
3. Set the Exposure stop increment to 1/3.
4. Set the Camera at the manual setting. Fix the Aperture @ f/8 or so ( Not critical ). Now, we are going to play with only the shutter speeds.
( If you take more than 2 min through this point, you would not be a photographer )

5. Click the shutter speed dial to push up your exposure meter to +2.
6. Take a shot, and increment the shutter speed dial by one click.
7. Repeat the step 6 above about 6 times to cover +2 to +4 range with 1/3 increment.

8. Click the shutter speed dial to opposite direction to push down your exposure to -2.
9. Take a shot, and decrement the speed dial by one click.
10. Repeat the step 9 above about 9 times to cover -2 to -5 range with 1/3 decrement.

Now you have bunch of raw files. Good news is that you don't need to process them. You only look at them with Adobe ACR.
11. Open Photoshop. Make sure the Brightness setting in the ACR is 50.

Let's decide the upper limit.
12. Open the files one by one from +2. The peak will gradually move toward the right . Eventually you will see something like the one on the left below.
13. IMPORTANT: Slide the Exposure adjust slider to the left to see if you can move the ENTIRE peak to the left. ( See the one on the right hand side ). If you can, the shot is within the range. If you cannot, it is blown-out.


This image is copyrighted by the owner




Now, let's decide the lower limit. This is not as obvious as the upper limit. This can be very subjective, too.
14. Open the files one by one from -2. The peak will gradually move toward the left. Eventually you will see something like the one on the left below.
15. The ENTIRE peak should be still contained. ( Check the left tail hitting the ground )
16. IMPORTANT: Slide the Exposure adjust slider to the right to see if you can move the ENTIRE peak to the right. ( See the one on the right hand side ). You will see that the left tail is dragging over long. You should be able to separate it cleanly. You will get to the point where you cannot separate the peak out of the noise ( or left wall ). ( The Blue tail will drag longer )


This image is copyrighted by the owner




In my case with D30, the upper limit was 1/5 sec and the lower limit was 1/500 sec. With the camera in hand, I counted the clicks from 1/5 to 1/500 to find that there were 20 clicks. So, the DR is 6 2/3 stops.

The whole procedure is much simpler than it looks. Once you understand how the lower and upper limits are determined, you can do this quickly.

==
Please, post your result with your Canon DSLR. I volunteer to collect the data.
The DR may look greater than you think. It all depends on how much noise you can take. It is important that everyone measures the lower limit the same way consistently.


This image is copyrighted by the owner





Edited by Pondria on Feb 18, 2008 at 10:16 PM GMT (Reason: Data update)


Tom_W
Registered: Jan 21, 2004
Total Posts: 5160
Country: United States

Glad to see this make a comeback. Perhaps it's not as precise as a technical lab could produce, but it definately shows the relative range of those cameras tested.

I look forward to seeing the test results of a few samples of the 1D3, 40D, and 1Ds3.

If I get the chance, I'll run the XTi/400D through it's paces and report my results. I hope that I can duplicate the setup I used with the old XT, 5D, and 1D2. Maybe I'll re-test the 5D as a gauge of my testing procedure.

BTW, 10DFT is now known as Tom_W.



cgardner
Registered: Nov 18, 2002
Total Posts: 7939
Country: United States

Another means to the same end...

Set custom WB off a gray card to equalize the channels..
Put card in consistent light, camera on tripod, card filling viewfinder
Open fast lens to widest aperture..
Find shutter speed which makes card near white (RGB = 240,240,240)
Bracket aperture in - 1 stop increments

What you get with a bit of cut and paste from sections the files on a facsimile of the camera histogram:


This image is copyrighted by the owner




The red lines are the position of the card spike on the histogram, numbers the RGB value. In practical terms I consider RGB values from about 30 to 240 represent the range of useful detail which can be seen on screen or the print. Doing that test helped me visualize what tone each spot on the horizontal scale of the histogram represents. I used 1-stop increments because I don't obsess about this type of stuff, I just like to be aware of what the histogram is actually telling me. I can interpolate the 1/3 stops in-between.

I selected aperture to control the exposure for the test because there is less possibility of mechanical variation: shutter speeds are not always as marked. Also it would allow the same lens to be used on several cameras, eliminating the variable of shutter speed camera to camera.

When exposing in the camera I try to put my values as shown below in this meter calibration test (there's glare on some areas but it didn't affect the test I was performing):



This image is copyrighted by the owner




The white towel is my standard exposure evaluation tool. I judge optimum exposure via its texture. Paper white winds up around 240, where it should be. The black thing in the background with the 4-6 reading is my hairlight SB which just happened to be back there when I did the test. By coincidence the distance from card to white muslin background made them nearly the same tone.

When shooting portraits I have the subject hold the towel, I adjust until it just starts to black out in my OEW, then back down 1/3 stop. Perfect exposure every time.

To calibrate my L-358 meter I shoot a bracket test around the indicated reading, then looking at the towel texture in DPP and Photoshop determine which is actually the best exposed. The difference between it and the meter reading gets dialed into the meter as the EC factor. My 20D ISO speed is about 1/3 stop off from the meter.

Chuck





Edited by cgardner on Jan 13, 2008 at 07:38 AM GMT


stanj
Registered: Aug 05, 2003
Total Posts: 8000
Country: United States

Following the instructions I got 9 2/3 for the 40D. I have to re-do the test of my 1Ds3, didn't get to it today as I was busy cutting down a tree



Roy Pertchik
Registered: Dec 05, 2004
Total Posts: 3341
Country: United States

Very cool. I'll do my 5D and 20D when I get back to NYC, late in the week



Pondria
Registered: Jan 11, 2002
Total Posts: 11873
Country: United States

stanj wrote:
Following the instructions I got 9 2/3 for the 40D. I have to re-do the test of my 1Ds3, didn't get to it today as I was busy cutting down a tree


Thank you !
It will be included in the next update.



stanj
Registered: Aug 05, 2003
Total Posts: 8000
Country: United States

My re-test today once again showed the same result:

1Ds2: 9 2/3
40D: 9 2/3
1Ds3: 9

My gut is telling me that there has to be something wrong with this, since I too "feel" that highlights recover easier. For instance, today I shot both my 40D and 1Ds3 back to back, and the 1Ds3 images were far easier to process, and far easier to recover any sort of flaw, including overexposure.

But at the same time I agree that shadows have less detail, which is an aspect of DR. I can't think of anything obvious wrong with my test - at least I executed every test the same way, three times, on two different days, and got the same results. So they all should be flawed / skewed the same way.

Getting more data would obviously be good.



BogongBreeze
Registered: Oct 28, 2005
Total Posts: 442
Country: Australia

Not sure if these results are worthwhile as I still consider myself a novice. Howver, I've just tested the 40D and it seems to get between 10 and 11 stops of DR, depending where I judge the end of the ranges.

Camera settings were 100 ISO and f8. Limits were reached at 3.2 secs and 1/320 secs (10 stops) but I'm finding it particularly difficult to judge the low end of the range where it peaks within the range but close to the limit. It could be up to one stop lower (11 stops total).

At the bright end, using the recovery slider makes it easier to judge the limit because it shows a smoother gradient (OK) or sharper cutoff (beyond the limit) to the histogram.



stanj
Registered: Aug 05, 2003
Total Posts: 8000
Country: United States

Yeah, I had the same problems, which is why I gave the raw files to Pondria and let him himself decide. That way, it was consistent with the other measurements. Plus I measured all three of my cameras the same way...



apdieb
Registered: May 29, 2006
Total Posts: 1405
Country: United States

stanj wrote:
My re-test today once again showed the same result:

1Ds2: 9 2/3
40D: 9 2/3
1Ds3: 9

My gut is telling me that there has to be something wrong with this, since I too "feel" that highlights recover easier. For instance, today I shot both my 40D and 1Ds3 back to back, and the 1Ds3 images were far easier to process, and far easier to recover any sort of flaw, including overexposure.

But at the same time I agree that shadows have less detail, which is an aspect of DR. I can't think of anything obvious wrong with my test - at least I executed every test the same way, three times, on two different days, and got the same results. So they all should be flawed / skewed the same way.

Getting more data would obviously be good.


I have not done these tests on my 5D and MKIII, but agree with your opinion about recovery ability (especially overexposure). Very noticeable to me.



Pondria
Registered: Jan 11, 2002
Total Posts: 11873
Country: United States

Any more new data please ?



Tom_W
Registered: Jan 21, 2004
Total Posts: 5160
Country: United States

I still owe you my XTi measurements as well as a retake of the 5D (as a sanity check). Hopefully, I'll have some light tomorrow. At least I have a day off!



Tom_W
Registered: Jan 21, 2004
Total Posts: 5160
Country: United States

OK, I tested the XTi (400D) and I came up with 7 1/3 to 7 2/3, depending on judgement. It could go either way.



Pondria
Registered: Jan 11, 2002
Total Posts: 11873
Country: United States

The data base has been updated with the new inputs. Thank you !



Tom_W
Registered: Jan 21, 2004
Total Posts: 5160
Country: United States

Thank you for your efforts.

Now, if a few more 40D's would post. I'd like to see the results.



Richard Steer
Registered: Sep 06, 2006
Total Posts: 123
Country: New Zealand

For my 40D at ISO 100 and f/8, the range was from 1.6" to 1/500". By my count that's 9 2/3 stops.

Also tested my 350D but had a terrible time trying to pick the shadow limit. Will try again tomorrow.

Richard



Pondria
Registered: Jan 11, 2002
Total Posts: 11873
Country: United States

40D seems to be a silent winner !



JasonJ
Registered: Oct 02, 2005
Total Posts: 2594
Country: United States

Wow this is interesting. I would participate, but I've made the switch to analog!

I'm interested in seeing comparisons between a Rebel XT and 20D though. I'm thinking about adding a digital body for some color work and this should be helpful.

Thanks for starting this Pondria and thanks to those who participated!



Richard Steer
Registered: Sep 06, 2006
Total Posts: 123
Country: New Zealand

I retested both cameras today, with two tests each.

350D: 7 1/3 stops
40D: 8 2/3 stops

Was possibly a bit generous at the dark end yesterday. The fact that I'm in about the right ballpark with the 350D makes me feel a bit happier about these results.

Richard



mfurman
Registered: Jan 16, 2005
Total Posts: 2849
Country: Canada


Richard Steer:
retested both cameras today, with two tests each.

350D: 7 1/3 stops
40D: 8 2/3 stops


I am wondering what could explain the discrepancy between these tests and the results found in this article:

http://www.clarkvision.com/imagedetail/digital.sensor.performance.summary/index.html

Please see Figure 4, for instance.



Harvey Moore
Registered: Jan 20, 2005
Total Posts: 1243
Country: United States

Probably the method used. See "Data Processing" on this page from the Clarkvision site. click here

mfurman wrote:

Richard Steer:
retested both cameras today, with two tests each.

350D: 7 1/3 stops
40D: 8 2/3 stops


I am wondering what could explain the discrepancy between these tests and the results found in this article:

http://www.clarkvision.com/imagedetail/digital.sensor.performance.summary/index.html

Please see Figure 4, for instance.



mfurman
Registered: Jan 16, 2005
Total Posts: 2849
Country: Canada

Harvey Moore:
Probably the method used.


Which one is more analytic and "correct" then?



Harvey Moore
Registered: Jan 20, 2005
Total Posts: 1243
Country: United States

Don't know the answer to this Michael.

The Clarkvision is a lot more detailed and number related, Pondria's method is simpler and ACR related.

From Clarkvision:
The procedure below is based on using ImagesPlus 2.5, hereafter called IP. The image processing system you use must be able to convert the raw data with a linear scale (Photoshop will not do this) and process the data with at least 16-bit math (IP uses 64-bit floating point math).

I think that any judgement I would make, for instance, is prejudiced by my use of the 5D, CS3 w/ACR, and what my eyes see on the screen and in in print.

The relative dr performance should be the same for all valid test methods.



ejmartin
Registered: Oct 27, 2005
Total Posts: 312
Country: United States

mfurman wrote:

Richard Steer:
retested both cameras today, with two tests each.

350D: 7 1/3 stops
40D: 8 2/3 stops


I am wondering what could explain the discrepancy between these tests and the results found in this article:

http://www.clarkvision.com/imagedetail/digital.sensor.performance.summary/index.html

Please see Figure 4, for instance.


You have to carefully read the text of Clark's webpage to see what is being plotted in Fig 4, which has nothing to do directly with any camera's DR. The plot is the number of stops between pixel saturation at the lowest ISO (full well capacity, as it is called), and the lowest read noise at high ISO. In other words, it is the ratio of the maximum possible signal that the camera can record to the smallest possible signal that the camera can record. It is not dynamic range because the former occurs at the lowest ISO and the latter at the highest ISO; but of course when one takes pictures one takes them at fixed ISO -- you don't get to choose one ISO for the part of the frame with highlights and another for shadows!

The figure in that article which has to do with actual DR is figure 5. DR plotted there is the so-called "engineering" DR, which is raw saturation level (max possible raw value of a pixel patch) over read noise (min possible raw value of a pixel patch), both measured at a fixed ISO. This is a rather liberal definition of DR, but has the advantage of being defined in terms of raw sensor data and properties, unfiltered through any raw converter.

Pondria's DR measurement filters the raw data through a particular raw converter, and while of practical use since it tells one how much latitude there is for extracting DR in conversion, it depends strongly on the programming of the converter -- how good the highlight recovery algorithms are, how the converter sets the black and white points, etc. It will be different from converter to converter, and even version to version of a given converter (for instance highlight recovery is much better in ACR 4.x than in earlier versions). The fact that it is a few stops less than the engineering DR could be interpreted as showing that ACR leaves a few stops of DR on the table; though I suspect some of it has to do with the fact that there is a distribution of luminance values being pushed around by the exposure slider, and the criterion being used to decide saturation uses the left edge of the shadow histogram and the right edge of the highlight histogram, so at the vary least it chops out the width of the histogram from inclusion in the DR.



John Power
Registered: Jul 03, 2003
Total Posts: 9386
Country: United States

Pondria, can't you come up with a testing procedure that involves sex, drugs or guns so I can participate too dammit....

Edited by John Power on Feb 11, 2008 at 03:28 PM GMT


1
2 3 end