PDA

View Full Version : Dynamic Range of 1D Mk2 or 20D


Radtech1
17th of October 2004 (Sun), 20:26
One limitation that I have a difficult time with the the not too wide dynamic range on my 10D. I believe that it is 4 stops, but I may be wrong.

Does anyone know what the dynamic range is on the Mk2 (either one, but with the 1Ds not being out yet, I dont expect the info to be available) or the 20D?

The three most compelling reasons - in order - that I can rationalize upgrading are:

1) Decreased noise levels at high asa's
2) Increased dynamic range
3) Less than the 1.6 crop

If anyone knows, thanks in advance.

Rad

unners
17th of October 2004 (Sun), 20:44
yeah, its something i'd really like to know too.
i basically run a reverse S-curve on all of my 10d raw files to try and increase the dynamic range. i do a bit of stuff for real estate guides and it can be very frustrating trying to make the house look appealing without blowing out the sky.
i'd particularly like to hear the comparison between the 1dII and the 20d as i would expect the 1dII to be slightly better because of the larger sensor. does anyone know about the original 1ds as well?
thanks
dave

Jesper
18th of October 2004 (Mon), 00:00
Running an S-curve on the photos from your 10D does not increase the dynamic range. It increases the contrast of the photo. The dynamic range (i.e. what range of brightness values it can record before recording pure black or pure white) is a property of the sensor; you can't change it in Photoshop.

Read Tonal quality and dynamic range in digital cameras (http://www.normankoren.com/digital_tonality.html) for an in-depth explanation.

Olegis
18th of October 2004 (Mon), 00:33
Also read this article (http://www.luminous-landscape.com/columns/determining-exposure.shtml), which explains how to determine the dynamic range of the digital sensor / film (paragraph 4).

chris.bailey
18th of October 2004 (Mon), 00:46
I havn't measurbated it but I would say my MkII has at least one more stop of DR than my 10D. I am finding far less tendency to blow out highlights and retain more detail in the dark areas.

drisley
18th of October 2004 (Mon), 00:49
"According to Canon's Chuck Westfall, the Canon EOS 1D Mark II provides 9 stops of dynamic range - more than any other Canon EOS DSLR at the time of introduction. Chuck indicated that the 10D provided 8 stops. Realized dynamic range may be a bit different, but the number should be somewhat useful for comparison purposes"

http://www.the-digital-picture.com/Reviews/Canon-EOS-1D-Mark-II-Digital-Camera-Review.aspx

J___
18th of October 2004 (Mon), 01:39
whats dynamic range?

Olegis
18th of October 2004 (Mon), 01:57
http://www.dcviews.com/tutors/tt53413.htm

J___
18th of October 2004 (Mon), 02:37
thnx.

unners
18th of October 2004 (Mon), 23:13
Running an S-curve on the photos from your 10D does not increase the dynamic range. It increases the contrast of the photo. The dynamic range (i.e. what range of brightness values it can record before recording pure black or pure white) is a property of the sensor; you can't change it in Photoshop.

Read Tonal quality and dynamic range in digital cameras (http://www.normankoren.com/digital_tonality.html) for an in-depth explanation.

thanks for the link jesper
i guess what i was referring to was more like apparent dynamic range and is actually reducing the contrast of the image (its the opposite of the normal s-curve). for example if a wall of a house is in the sun and the shade, the reverse s-curve will brighten the darker areas without overexposing the lighter areas making the difference between the light and dark parts less (reducing contrast). to my way of thinking this would be achieved automatically on a camera with higher dynamic range as the larger range must still fit into the same grayscale thus reducing contrast.
i'm not sure if i'm making sense but i'm interested in hearing your thoughts.
thanks
dave

Jesper
18th of October 2004 (Mon), 23:45
Yes - with an S-curve, you can increase the contrast (the difference in brightness between the dark and light areas in an image), and with a reverse S-curve you can decrease the contrast.

All equipment that can record images can record only a limited range of brightness values. For example, with the wall partly in the sun and partly in the shade, if your camera doesn't have enough dynamic range, the contrast (= brightness difference) between the dark and light part could be too large for your camera to handle - what happens is that either the part in the shadow will become completely black or the part in the sun will become completely white (depending on how you expose the image).

A disadvantage of (almost) all digital cameras is that they have a limited dynamic range compared to print film. The best digital cameras (excluding medium format digital backs) have maybe 6 stops of dynamic range. Print film (negatives) have a much larger dynamic range. Slide film also has a limited dynamic range - that's why it's very important to expose properly when using slide film.

Note: The numbers "8 stops" for the 10D and "9 stops" for the 1D Mark II that drisley mentions above are theoretical, ideal values. In practice you won't get 8 stops of useful dynamic range from a 10D.

Also note that if you want to use the full dynamic range of your camera, you need to shoot RAW. Because JPEG limits you to 8 bits per channel (instead of the 12 bits per channel the sensor records), you're throwing away part of the information in the image.

drisley
19th of October 2004 (Tue), 20:39
I find the "shadow/highlight" tool in PS CS is a great to to "increase" dynamic range in an image. It's not perfect, but it does work really well.

chris.bailey
20th of October 2004 (Wed), 00:40
I find the "shadow/highlight" tool in PS CS is a great to to "increase" dynamic range in an image. It's not perfect, but it does work really well.

surely that has nothing to do with dynamic range, its a selective contrast/brightness tool in effect. If detail is not there in the original picture data (at the dark and bright end) there is nothing you can do to bring it back again.

drisley
20th of October 2004 (Wed), 02:34
No, it's not really increasing the dynamic range, but it can give the appearance of increasing the dynamic range.

Jesper
20th of October 2004 (Wed), 03:40
No, it's not really increasing the dynamic range, but it can give the appearance of increasing the dynamic range.

Then why do you call it "increasing the dynamic range"..... :? You're not increasing the dynamic range, but the contrast. It is not possible to increase the dynamic range using Photoshop. If part of your image is completely black or blown out to completely white, information in that part of the image is lost and you can't magically re-create that information.

drisley
20th of October 2004 (Wed), 08:43
Actually, if you read more carefully, it's "increase" dynamic range (note the quotes). That means it's not literal.
You're not increasing the dynamic range, but the contrast
That is incorrect, you are NOT increasing the contrast with the shad/highlight tool. If anything, you are decreasing contrast,but even that is not a correct assesment.
Using this tool you can often bring detail to areas that you could not see before. This gives appearance of increasing dynamic range.

Hatem Eldoronki
20th of October 2004 (Wed), 10:11
The dynamic range is how many colors the camera can see, and record.
Adjusting contrast may help expose those colors that were not showing in the original capture, but will not increase the dynamic range. It will not affect it at all. Dynamic range of that particular picture will change, like say you could only see 256 at first but now you can see 512 in that particular picture.
But Dynamic Range as a property of the camera/sensor will not be affected at all. Meaning that for example if the 1D can see 2 million colors, and the 10D 1.5 million, then by changing contrast, you will not make any more colors available for the 10D's dynamic range to see than the 1.5 million. Your original capture might have less than what the camera fully could record, and you could expose more of it by using software, and that might make you think that you've added to the dynamic range, but you really didn't...

(The numbers I used are just made up, to better explain; AND because I really don't know the real ones :lol: )

BigRed450
20th of October 2004 (Wed), 16:54
mcouper ... DR has nothing to do with the color. DR is the range of light a particular sensor or film can capture which is usually measured in stops, but is sometimes also measured or referred to in EV.

In comparing the 10D to the MkII or 20D the 10D tends to lean to hotter(higher) exposures (Firmware setting) then the others and so by presetting exposure compensation to -1/3 -1/2 you can protect your highlights. Shadow detail (still much better then most film BTW) is still there and can easily be brought out if required.

To simplify... If you put a 10D and a MkII side by side and took two identicle images at the same time the MkII image would be darker then the 10D image.

drisley
21st of October 2004 (Thu), 00:10
I understand what dynamic range is.
I know that software wont really increase dynamic range.
However, there is data in the highlights and shadows of the image that the naked eye can't see. Some people like to output 2 images with different exposures to "increase the dynamic range of the image" by combining them.
However, I find the shadow/highlight tool is a quicker way of doing this (perhaps not quite as much detail is gained) as you can pull alot of detail out of the shadows/highlights of a 16bit image.
You are not "creating data", but rather shifting data from the edge of oblivion back into the visible spectrum.
This is just my method of dealing with the limited dynamic range of digicams.

ron chappel
21st of October 2004 (Thu), 00:30
"According to Canon's Chuck Westfall, the Canon EOS 1D Mark II provides 9 stops of dynamic range - more than any other Canon EOS DSLR at the time of introduction. Chuck indicated that the 10D provided 8 stops.

Realized dynamic range may be a bit different, but the number should be somewhat useful for comparison purposes"

I've seen this quoted often and find it very disapointing.
Chuck is stretched the truth so it's pretty much a lie.
Very disapointing from someone who has such a good reputation otherwise :?

I've tested my 300D at 6.5 stops absolute maximum :shock:
Anybody who's getting 8 stops out of a 10D is counting single off coloured pixels :lol: :lol: :lol:
5.5 is a much more sensible limit and even then it's a quite a hard limit


The only way to really imrove dynamic range is to proccess two seperate images from a single raw file-one bright and one dark-then selectively merge them.Luminous landscape has a tutorial i think


As can be seen from Mcouper's quote,dynamic range can be used to describe many different things.In this case though we are definitely talking about brightness range