D-Range - A Technical View

bonneteau · 11-26-2008, 03:11 PM

hum,
the D-Range is only effective on jpeg files, no ?
like the Noise Reduction, no ?

Das Boot · 11-26-2008, 03:20 PM

Quote:

the D-Range is only effective on jpeg files, no ?
like the Noise Reduction, no ?

Yeah that's right.

-Brian

**DAZ** · 11-26-2008, 03:40 PM

So if I shoot in PEF or DNG I will not see anything different using D-range?

DAZ

Das Boot · 11-26-2008, 03:59 PM

As far as I can tell, the d-range info is passed along to the raw converter and it's interpreted by the program. Somewhat like how white balance, hue, saturation, etc. is passed along, but easy to change. It was discovered that Nikon's similar function does the same thing. Surely if you like the results and shoot raw only, it can save you some time in PP and eliminate a step or two.

Thanks-
Brian

**DAZ** · 11-26-2008, 04:05 PM

OK, what setting in the RAW converter tells is to not use the D-range info?

DAZ

falconeye · 11-27-2008, 06:02 AM

This topic keeps coming back

Much could be said. Maybe the most helpful addition to the OP is a link to some measurements done by DPR:
Pentax K20D Review: 1. Introduction: Digital Photography Review
Go to page 21, section "D-Range mode" and consult the curve comparison for ISO100 vs. ISO200+DR. The difference in the highlights is clear. The difference in the shadows is less clear (it is -5.8 vs. -5.4 EV as seen from the table on page 20). So, the overall DR without DRange is 9.0 EV and with DRange 9.6 EV (1 EV gain in the highlights vs. 0.4 EV loss in the shadows).

Also note that you must compare ISO100 vs. ISO200+DR. Otherwise, your tests misleading as the first thing which DR does is changing the ISO...

All this applies to the K20D. I don't know anything about the K200D...

One last remark: In your tests, I don't see that you have enough DR in your subject. You must at least have ~10EV DR in the subject to start with.

And a clarification: JPEG does not store logarithmic values for captured luminance. They are linear digitized values in the range 0 ... 255, after applying the color spaces gamma correction (which is 2.2 for sRGB).

ukbluetooth · 11-27-2008, 06:55 AM

Originally posted by DAZ

OK, what setting in the RAW converter tells is to not use the D-range info?

DAZ

DAZ - on a previous thread :

https://www.pentaxforums.com/forums/pentax-dslr-discussion/22265-what-differe...ly-make-2.html

you convinced me with your shots of dishes that D-Range affects the RAW file directly. This makes sense to me if a different analogue amplification mapping of sensor signal is made prior to the A/D converter.

Are you now less convinced or is the jury still out on this? I really would like to know for certain if RAW is modified or not.

**DAZ** · 11-27-2008, 09:42 AM

No ukbluetooth I think that whatever the camera is doing it has made a fundamental change to the RAW file. I also thing D-range adds ½ to 1 stop DR. I think it takes a little from the dark end and adds a little more to the light end. To me the dark end often has less information that I care about then the top end so it is usually a good tradeoff. Saying that I will be the first to say I don’t know everything and if more facts come in I am willing to change my mind.

It was stated that D-range only effected JPG files and I was trying to get more facts to see if they could support that claim with something I did not know. I find engaging in conversation and asking for more info gets me better results then saying “I think your wrong prove your point”.

This subject keeps coming up. I usually just try and point others to the previous threads and let them make up there own mind. There is a lot of info in those threads but this time the int21h was adding some good new info. That is always a good thing and should not be discouraged.

I thing that the int21h was trying to say is that the camera adds a tag to the RAW file that tells the RAW converter to use some special conversion curve to the output file to get the additional DR. I don’t thing this is the case but the only way to know this is for someone poke a new code into the file and see what the RAW converter output looks like.

All RAW converters use some kind of conversion curve. If all you had to do was under expose one stop and run some special conversion curve to get more DR the RAW converter people would be yelling that from the rooftops.

I think part of the problem with this is some people thing that D-range is just Pentax’s version of Nikon’s D-lighting. AFAIK D-lighting is an output digital filter that doesn’t increase the DR. So some think that D-range must be the same thing. I don’t see why if that were the case the camera would have to give up 100 ISO if it was just a digital filter.

DAZ

Das Boot · 11-27-2008, 04:53 PM

I didn't mean to put you off with a reply. I never intended to look deep into how a raw converter does what it does. I did try - I spent a good deal of time on it. I tried to see how ACR converts the two different files - no different. I then tried it in DNG - there was a difference in how it applied the DR compared to the PEF files... there was a lot less difference in the histogram, but still a difference. I used PhotoME and saw a few changes in the white points and the flag for DR. I tried to manipulate them in a hex editor (the flags only) and only ended up with a corupted file. The only thing that let me know that DR didn't affect raw was using an outdated raw converter for dng - no difference. I even tried to compare files directly in a hex editor, but PEF files are compressed, and something is up with DNG files... more so than I am willing to put into this.

I had always thought that DR had an effect on jpeg only and really it does. It does play a part in newer raw converters like saturation, hue, etc. and can help reduce PP, but in my heart I know it doesn't effect the pure raw data. If someone wants to persue it further, have at it, but it was never my intention to do so - I was just looking to find out if I should keep it on while shoot jpeg shots... that's all.

-Brian

PS - For what it's worth, I took a raw PEF @ ISO100 and in ACR increased the exposure by +1.0 and saw no difference in histogram or visual output compared to a shot taken at ISO200 with DR on. It's interesting. I wonder if ACR automatically applies a curve or increases linear. There are a lot more variables concerning raw than jpeg. Maybe someone else wants to tackle that beast.

falconeye · 11-28-2008, 07:48 AM

Originally posted by the_int21h

Maybe someone else wants to tackle that beast.

I abandonned this issue a long time ago when sort of a consensus was reached:

For RAW files, DRange means the following: Switch the camera internally to half the published ISO value AND expose according to the published ISO value (i.e., underexpose by 1 stop according to the internal ISO value). As a result, the raw values are darker than usual. Some raw converters, however, read the DRange flag and cope with this. BTW, this is the reason why DRange cannot be combined with ISO100.

I don't want to discuss this topic once again. But also, I didn't want that the results of discussions we had 6 months ago would be forgotten.

bonneteau · 11-28-2008, 07:52 AM

Originally posted by falconeye

I don't want to discuss this topic once again. But also, I didn't want that the results of discussions we had 6 months ago would be forgotten.

could you give us the link please ?

Das Boot · 11-28-2008, 07:55 AM

I didn't want to beat a dead horse, I just couldn't find any technical data. I looked all over and all I could find were comparison pictures and conflicting reports - nothing concrete. I looked for the answer myself and decided to share what I found - that's all.

Thanks -
Brian

falconeye · 11-28-2008, 08:08 AM

Originally posted by bonneteau

could you give us the link please ?

I would have done so if I could remember it. I now used Google on "falconeye" and "EDR" and found this resource which is one of the original works, I believe to remember:
Re: K20D: EDR *does* affect RAW images.: Pentax SLR Talk Forum: Digital Photography Review
You may be able to find prior discussions here on pentaxforums using the same Google search.

Originally posted by the_int21h

I looked for the answer myself and decided to share what I found

IMHO, your work is a great addition to the topic. Note however, that a difference of 9.0 vs. 9.5 EV DR is at the edge of measurability for non-professionals...

Mountain Vision · 12-17-2008, 03:07 AM

Unless I'm missing something you tried to replicate the DR-on and off both at ISO 200.

The problem is the camera actually appears to drop down to ISO 100 with it on, intentionally UNDER EXPOSING the image. Then it does bump the levels.

So you need to retest with DR-on ISO 200, and DR Off ISO 100. Leave exposure the same in both shots!

Then bump your levels.

11-27-2008, 04:53 PM	#24
Das Boot Veteran Member Join Date: Feb 2008 Location: Sparkle City, South Cackalacky Photos: Gallery Posts: 689 Original Poster	I didn't mean to put you off with a reply. I never intended to look deep into how a raw converter does what it does. I did try - I spent a good deal of time on it. I tried to see how ACR converts the two different files - no different. I then tried it in DNG - there was a difference in how it applied the DR compared to the PEF files... there was a lot less difference in the histogram, but still a difference. I used PhotoME and saw a few changes in the white points and the flag for DR. I tried to manipulate them in a hex editor (the flags only) and only ended up with a corupted file. The only thing that let me know that DR didn't affect raw was using an outdated raw converter for dng - no difference. I even tried to compare files directly in a hex editor, but PEF files are compressed, and something is up with DNG files... more so than I am willing to put into this. I had always thought that DR had an effect on jpeg only and really it does. It does play a part in newer raw converters like saturation, hue, etc. and can help reduce PP, but in my heart I know it doesn't effect the pure raw data. If someone wants to persue it further, have at it, but it was never my intention to do so - I was just looking to find out if I should keep it on while shoot jpeg shots... that's all. -Brian PS - For what it's worth, I took a raw PEF @ ISO100 and in ACR increased the exposure by +1.0 and saw no difference in histogram or visual output compared to a shot taken at ISO200 with DR on. It's interesting. I wonder if ACR automatically applies a curve or increases linear. There are a lot more variables concerning raw than jpeg. Maybe someone else wants to tackle that beast. Last edited by Das Boot; 11-27-2008 at 05:01 PM.

11-28-2008, 08:08 AM	#28
falconeye Veteran Member Join Date: Jan 2008 Location: Munich, Alps, Germany Photos: Gallery Posts: 6,871	Originally posted by bonneteau could you give us the link please ? I would have done so if I could remember it. I now used Google on "falconeye" and "EDR" and found this resource which is one of the original works, I believe to remember: Re: K20D: EDR does affect RAW images.: Pentax SLR Talk Forum: Digital Photography Review You may be able to find prior discussions here on pentaxforums using the same Google search. Originally posted by the_int21h I looked for the answer myself and decided to share what I found IMHO, your work is a great addition to the topic. Note however, that a difference of 9.0 vs. 9.5 EV DR is at the edge of measurability for non-professionals... Last edited by falconeye; 11-28-2008 at 08:14 AM.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Just how does the Pentax DA* range fare against the Canon L range?	Reportage	Pentax SLR Lens Discussion	10	03-11-2009 10:56 AM
Technical Jargen help	jbrowning	Pentax Camera and Field Accessories	5	09-28-2008 01:40 PM
Suggestion Technical Section	Gadget_Guy	Site Suggestions and Help	2	05-29-2008 03:56 PM
Technical Question	AmdamJayne	Post Your Photos!	7	05-16-2007 12:37 AM
Non-technical tips	JJJPhoto	Digital Processing, Software, and Printing	1	12-21-2006 01:56 PM

11-26-2008, 03:11 PM	#16
bonneteau Senior Member Join Date: Dec 2007 Location: Europe Posts: 116	hum, the D-Range is only effective on jpeg files, no ? like the Noise Reduction, no ?

11-26-2008, 03:20 PM	#17
Das Boot Veteran Member Join Date: Feb 2008 Location: Sparkle City, South Cackalacky Photos: Gallery Posts: 689 Original Poster	Quote: the D-Range is only effective on jpeg files, no ? like the Noise Reduction, no ? Yeah that's right. -Brian

11-26-2008, 03:40 PM	#18
DAZ Loyal Site Supporter Join Date: Sep 2006 Location: Everett, WA USA Photos: Gallery Posts: 744	So if I shoot in PEF or DNG I will not see anything different using D-range? DAZ

11-26-2008, 03:59 PM	#19
Das Boot Veteran Member Join Date: Feb 2008 Location: Sparkle City, South Cackalacky Photos: Gallery Posts: 689 Original Poster	As far as I can tell, the d-range info is passed along to the raw converter and it's interpreted by the program. Somewhat like how white balance, hue, saturation, etc. is passed along, but easy to change. It was discovered that Nikon's similar function does the same thing. Surely if you like the results and shoot raw only, it can save you some time in PP and eliminate a step or two. Thanks- Brian

11-26-2008, 04:05 PM	#20
DAZ Loyal Site Supporter Join Date: Sep 2006 Location: Everett, WA USA Photos: Gallery Posts: 744	OK, what setting in the RAW converter tells is to not use the D-range info? DAZ

11-27-2008, 06:02 AM	#21
falconeye Veteran Member Join Date: Jan 2008 Location: Munich, Alps, Germany Photos: Gallery Posts: 6,871	This topic keeps coming back Much could be said. Maybe the most helpful addition to the OP is a link to some measurements done by DPR: Pentax K20D Review: 1. Introduction: Digital Photography Review Go to page 21, section "D-Range mode" and consult the curve comparison for ISO100 vs. ISO200+DR. The difference in the highlights is clear. The difference in the shadows is less clear (it is -5.8 vs. -5.4 EV as seen from the table on page 20). So, the overall DR without DRange is 9.0 EV and with DRange 9.6 EV (1 EV gain in the highlights vs. 0.4 EV loss in the shadows). Also note that you must compare ISO100 vs. ISO200+DR. Otherwise, your tests misleading as the first thing which DR does is changing the ISO... All this applies to the K20D. I don't know anything about the K200D... One last remark: In your tests, I don't see that you have enough DR in your subject. You must at least have ~10EV DR in the subject to start with. And a clarification: JPEG does not store logarithmic values for captured luminance. They are linear digitized values in the range 0 ... 255, after applying the color spaces gamma correction (which is 2.2 for sRGB).

11-27-2008, 06:55 AM	#22
ukbluetooth Senior Member Join Date: Aug 2007 Location: Reading, UK Posts: 189	Originally posted by DAZ OK, what setting in the RAW converter tells is to not use the D-range info? DAZ DAZ - on a previous thread : https://www.pentaxforums.com/forums/pentax-dslr-discussion/22265-what-differe...ly-make-2.html you convinced me with your shots of dishes that D-Range affects the RAW file directly. This makes sense to me if a different analogue amplification mapping of sensor signal is made prior to the A/D converter. Are you now less convinced or is the jury still out on this? I really would like to know for certain if RAW is modified or not.

11-27-2008, 09:42 AM	#23
DAZ Loyal Site Supporter Join Date: Sep 2006 Location: Everett, WA USA Photos: Gallery Posts: 744	No ukbluetooth I think that whatever the camera is doing it has made a fundamental change to the RAW file. I also thing D-range adds ½ to 1 stop DR. I think it takes a little from the dark end and adds a little more to the light end. To me the dark end often has less information that I care about then the top end so it is usually a good tradeoff. Saying that I will be the first to say I don’t know everything and if more facts come in I am willing to change my mind. It was stated that D-range only effected JPG files and I was trying to get more facts to see if they could support that claim with something I did not know. I find engaging in conversation and asking for more info gets me better results then saying “I think your wrong prove your point”. This subject keeps coming up. I usually just try and point others to the previous threads and let them make up there own mind. There is a lot of info in those threads but this time the int21h was adding some good new info. That is always a good thing and should not be discouraged. I thing that the int21h was trying to say is that the camera adds a tag to the RAW file that tells the RAW converter to use some special conversion curve to the output file to get the additional DR. I don’t thing this is the case but the only way to know this is for someone poke a new code into the file and see what the RAW converter output looks like. All RAW converters use some kind of conversion curve. If all you had to do was under expose one stop and run some special conversion curve to get more DR the RAW converter people would be yelling that from the rooftops. I think part of the problem with this is some people thing that D-range is just Pentax’s version of Nikon’s D-lighting. AFAIK D-lighting is an output digital filter that doesn’t increase the DR. So some think that D-range must be the same thing. I don’t see why if that were the case the camera would have to give up 100 ISO if it was just a digital filter. DAZ

11-28-2008, 07:48 AM	#25
falconeye Veteran Member Join Date: Jan 2008 Location: Munich, Alps, Germany Photos: Gallery Posts: 6,871	Originally posted by the_int21h Maybe someone else wants to tackle that beast. I abandonned this issue a long time ago when sort of a consensus was reached: For RAW files, DRange means the following: Switch the camera internally to half the published ISO value AND expose according to the published ISO value (i.e., underexpose by 1 stop according to the internal ISO value). As a result, the raw values are darker than usual. Some raw converters, however, read the DRange flag and cope with this. BTW, this is the reason why DRange cannot be combined with ISO100. I don't want to discuss this topic once again. But also, I didn't want that the results of discussions we had 6 months ago would be forgotten.

11-28-2008, 07:52 AM	#26
bonneteau Senior Member Join Date: Dec 2007 Location: Europe Posts: 116	Originally posted by falconeye I don't want to discuss this topic once again. But also, I didn't want that the results of discussions we had 6 months ago would be forgotten. could you give us the link please ?

11-28-2008, 07:55 AM	#27
Das Boot Veteran Member Join Date: Feb 2008 Location: Sparkle City, South Cackalacky Photos: Gallery Posts: 689 Original Poster	I didn't want to beat a dead horse, I just couldn't find any technical data. I looked all over and all I could find were comparison pictures and conflicting reports - nothing concrete. I looked for the answer myself and decided to share what I found - that's all. Thanks - Brian

12-17-2008, 03:07 AM	#29
Mountain Vision Veteran Member Join Date: Nov 2006 Location: America's First Wilderness Posts: 529	Unless I'm missing something you tried to replicate the DR-on and off both at ISO 200. The problem is the camera actually appears to drop down to ISO 100 with it on, intentionally UNDER EXPOSING the image. Then it does bump the levels. So you need to retest with DR-on ISO 200, and DR Off ISO 100. Leave exposure the same in both shots! Then bump your levels.