Ad
Ad
Ad
Pages: [1] 2 »   Bottom of Page
Print
Author Topic: Unexpected Behavior of Adobe DNG Profile Editor  (Read 3598 times)
samueljohnchia
Sr. Member
****
Offline Offline

Posts: 327


« on: January 11, 2013, 10:58:42 PM »
ReplyReply

Hi everyone,

In my research of camera profiling, I discovered a rather puzzling behaviour with the Adobe DNG Profile Editor (I'm using 1.0.0.45). Before I proceed to describe this phenomenon, I would first like to acknowledge Eric Chan's and Sandy McGuffog's invaluable help in this journey. I am sure that I can speak for many of us, who are indebted to Eric for the monumental work he has contributed to improve the tools we use daily for creating better images. Eric also wrote the DNG PE program that I speak of, and although I will quote him a number of times in this post, it is with all due respect to Eric, and I only wish to clarify his statements and my own findings. I am not accusing Eric of anything - alas, we are all prone to mistakes, and it is my hope that I grossly misunderstood how all this works and that perhaps Eric can step in to clarify matters.

This issue I speak of concerns the illuminant used by the DNG PE. As seen in the screenshot attached, the PE allows one to build color table(s) for a custom camera profile in three different ways: "Both color tables", "2850 K only" and "6500 K only". Adobe's documentation states that to create a dual illuminant profile, one would have to take two photographs of the color checker, one taken under a 6500K illuminant and one under a 2850 K illuminant. One builds the dual illuminant profile in the PE by first selecting the 2850 K photograph and "2850 K only" color table, then the 6500K photograph and "6500 K only" color table or vice versa. The order of creating the tables do not affect the result. Choosing "Both color tables" in this situation is decidedly wrong! In fact, I am not exactly sure when the option of "Both color tables" would be useful...

To simplify matters, I decided to test the difference in the result of creating a camera profile using "Both color tables" against "2850 K only" followed by "6500 K only" to achieve theoratically identical profiles. I used the same photograph of the colorchecker (taken in noon daylight, sufficient exposure of the capture), and left the control "end points" in the PE in exactly the same position for all the profiles. After generating and exporting each profile, I went to Edit>clear All Adjustments to be absolutely sure that I started from a clean slate again to build the next profile. In total, four profiles were built:

1. Both color tables
2. 2850 K and the 6500 K color tables
3. Only the 2850 K color table
4. Only the 6500 K color table

From my understanding of how this works, 1. and 2. are dual illuminant profiles, while 3. and 4. are single illuminant profiles. I used Sandy's dcpTool to decompile the 4 camera profiles to xml format. I am making the decompiled versions of the profiles available here.

Observations:
3. profile has a HueSatDelta1 table that correcponds exactly to the HueSatDelta1 table of the 2. profile. 4. profile has a HueSatDelta2 table that corresponds exactly to the HueSatDelta2 table of the 2. profile. 3. profile contains a HueSatDelta2 table, but it's values are null. 4. profile contains a HueSatDelta1 table, but it's values are null. This is expected so all is good.

My expectation was for 1. and 2. to be identical. However, they were not. Applying profiles 1. and 2. in Adobe Camera Raw 7.1 (processed to Prophoto RGB and 16 bits/color channel) to the same image of the colorchecker used to build these profiles, I noted a differences in the color values of the colorchecker. None of the neutral patches shifted in hue or saturation, but all the 18 color patches did (verified in 16 bit precision). In 8 bit rounded integer values, a maximum of 2 values difference was observed in the blue channel for some patches, while the red and green channels mostly differed by one value. Decompiling the profiles, 1.  profile has HueSatDelta 1 & 2 tables different from the 2. profile. That explains the difference in application. This is unexpected behavior to me.

I initially put this down to the illuminant assumed for the calculations to generate the HueSatDelta tables, and chromatic adaptation to D50 if necessary. So I tried to verify this thinking against Eric's comments on how the DNG PE works:

http://www.luminous-landscape.com/forum/index.php?topic=59688.msg482472#msg482472
"Yes, that's right.  If you build a single-illuminant profile in DNG PE, the target values used by DNG PE are from a set of averaged ColorChecker charts with adopted illuminant of D50.  If you build a dual-illuminant profile in DNG PE, the illuminants used by DNG PE are A and D65."

http://www.luminous-landscape.com/forum/index.php?topic=70762.msg561604#msg561604
"For the single-illuminant profile, DNG PE assumes a D50 illuminant (which is what the normal published numbers assume, too), and for the dual-illuminant profile, DNG PE assumes Standard Illuminant A for the first table, and CIE D65 for the second table (same calibration illuminants that Adobe uses for its Adobe Standard profile)."

But I got confused here:
http://www.luminous-landscape.com/forum/index.php?topic=70762.msg562781#msg562781
"Yes, if you do the 6500k table only, it's D65.  If you do the 2850K table only, it's Std illuminant A.  Regardless of the order you choose to do them in individually, the mapping is always the same."

Questions:
1. I am confused as to what is really happening. Is D50 assumed for the "Both color tables" option and Std illuminant A for the "2850 K only", D65 for the "6500 K only" options?

2. Also, which resulting profile is more "accurate" in theory? Say if I was trying to create a profile for use under daylight (single illuminant), would I have a more acurate profile by creating a 6500K only profile, or a 6500K and 2850K profile from the same capture of the colorchecker, or a "Both color tables" profile?
Logged
Schewe
Sr. Member
****
Offline Offline

Posts: 5533


WWW
« Reply #1 on: January 11, 2013, 11:33:38 PM »
ReplyReply

Just to be sure I understand, you tried to do a dual-illuminate profile (Standard A and D-65) from a single capture? If I read what you wrote correctly, you can't do this. If you take a single capture at "daylight" all you can generate is a single illuminate profile.

Due to the potential metameric failure of a sensor under two very different special illumination, you would need to shoot a target under near Standard Illuminate A (or at least close to tungsten) and a target under D-50 to D65 (the exact K isn't required but you need to be somewhat close).

A single CC target can simply never create a useful dual-illuminate profile. Pretty sure Eric has been pretty specific about how to create such a profile. Not sure why you though you could create a dual-illuminate profile from a single target (that is NOT what the docs say).

You need to shoot two separate CC targets, one lit under D-65 (or close) and one under Standard Illuminate A (near 3200K). That's how a dual-illuminate profile is useful.
Logged
samueljohnchia
Sr. Member
****
Offline Offline

Posts: 327


« Reply #2 on: January 12, 2013, 02:34:51 AM »
ReplyReply

Hi Jeff, but I did manage to create a dual table single illuminate profile. I never tried to create a dual illuminate profile from a single capture. The DNG PE always creates a dual table profile, regardless of whether one is creating a single illuminate or dual illuminate profile. Sandy replicated this test with his own photograph of his colorchecker and arrived at similar results - it puzzled him too.

I was also wondering if there is a circumstance where the DNG PE would not only just adopt D65 and Std Illuminate A so that profiling one's camera under florescent or HMI lights for example would also be successful. Eric's own comments elsewhere do suggest that it is possible, and a successful profile can be created.

With all due respect, I'm pretty sure that a single CC target can create a useful dual illuminate profile - one would just have to photograph it on separate occasions under two different light sources (as recommended by Adobe and Eric, also yourself, something close to 6500K daylight and 2850K tungsten).

As a side note, I've tested different methods of creating a single illuminate camera profile for daylight, and not too surprisingly, using the "Both color tables" option produces the most accurate daylight profile. Relatively more accurate than simply using the "6500 K only" option, and only somewhat more accurate than a profile created with both "2850 K only" and "6500 K only" (for some colors that was slightly better). The accuracy was determined against two synthetic colorcheckers, one I customised myself, and another from Babelcolor. I recall that Eric mentioned elsewhere if one were to set the white balance value to something like 3500k in ACR, but the camera profile is "6500 K only" generated, colors will be less accurate from extrapolation than if the WB slider was at 6000 - 7000k. My thinking is that a dual table profile would be more accurate, because the "teeaning" between the 2850K and 6500K table will improve the accuracy for WB slider values between 2850K and 6500K. If my assumptions are wrong, I'll be happy to be pointed in the right direction.
Logged
opgr
Sr. Member
****
Offline Offline

Posts: 1125


WWW
« Reply #3 on: January 12, 2013, 02:40:19 AM »
ReplyReply

Questions:
1. I am confused as to what is really happening. Is D50 assumed for the "Both color tables" option and Std illuminant A for the "2850 K only", D65 for the "6500 K only" options?

D50 refers to the colorchecker values, not the profiles. A & D65 refer to the internal whitepoint references of the profiles.

So your Illuminant A table should generally be a tungsten capture, and your daylight capture should preferably be close to D65.

However, the white point is primarily relevant to the multipliers used for balancing the individual R,  G, and B channels of your camera, so the actual temperature of daylight is not vital.


2. Also, which resulting profile is more "accurate" in theory? Say if I was trying to create a profile for use under daylight (single illuminant), would I have a more acurate profile by creating a 6500K only profile, or a 6500K and 2850K profile from the same capture of the colorchecker, or a "Both color tables" profile?

You would have a more accurate result by creating a 6500K only profile.

However, you would make that profile by setting both tables to your 6500K capture. I recall that the single table option refers to adjusting one of the tables of an existing dual illuminant profile. Don't know if that is still the case.

Logged

Regards,
Oscar Rysdyk
theimagingfactory
opgr
Sr. Member
****
Offline Offline

Posts: 1125


WWW
« Reply #4 on: January 12, 2013, 02:46:51 AM »
ReplyReply

Hi Jeff, but I did manage to create a dual table single illuminate profile. I never tried to create a dual illuminate profile from a single capture. The DNG PE always creates a dual table profile, regardless of whether one is creating a single illuminate or dual illuminate profile. Sandy replicated this test with his own photograph of his colorchecker and arrived at similar results - it puzzled him too.

That is not puzzling, that is by design. You always get a dual illuminant profile. You might as well think of a dual illuminant profile as 2 separate profiles. One is normally based on the capture data for illuminant A circumstances, the other should be based on daylight capture circumstances. However, nothing stops you from using a single capture for both tables. This is what used to be referred as a "single-illuminant profile", i.e. a dual illuminant profile where the illuminant of both tables is the same, based on a single capture.
Logged

Regards,
Oscar Rysdyk
theimagingfactory
samueljohnchia
Sr. Member
****
Offline Offline

Posts: 327


« Reply #5 on: January 12, 2013, 02:52:31 AM »
ReplyReply

Quote
D50 refers to the colorchecker values, not the profiles. A & D65 refer to the internal whitepoint references of the profiles.

So your Illuminant A table should generally be a tungsten capture, and your daylight capture should preferably be close to D65.

The colorchecker values can be specified using any illuminant, not just D50. Eric said before that for a single illuminant profile, the DNG PE adopts D50 as the illuminant, and the colors need not be chromatically adapted to the internal reference values for the colorchecker that the PE uses. I guess the main question is, what is a single illuminant profile? 6500K only? 2850K only? Both of those? Both color tables generated from the same capture of a colorchecker?

Quote
You would have a more accurate result by creating a 6500K only profile.

This is not consistent with my own testing... especially when the WB sliders in ACR are set to less than 6000K...

Quote
However, you would make that profile by setting both tables to your 6500K capture. I recall that the single table option refers to adjusting one of the tables of an existing dual illuminant profile. Don't know if that is still the case.

One can copy the table values from the 6500K table to the 2850k tablein the PE. So in that case, is it still a single illuminate profile?
Logged
samueljohnchia
Sr. Member
****
Offline Offline

Posts: 327


« Reply #6 on: January 12, 2013, 02:55:55 AM »
ReplyReply

That is not puzzling, that is by design.

No, that was not what I meant. It was puzzling to both of us because we (and Eric) expects that "Both color tables" generation to be identical to "2850 K only" plus "6500 K only" generation. Eric said, "This sounds unexpected to me.  How different are the readouts?"
Logged
samueljohnchia
Sr. Member
****
Offline Offline

Posts: 327


« Reply #7 on: January 12, 2013, 03:00:43 AM »
ReplyReply

One can copy the table values from the 6500K table to the 2850k tablein the PE. So in that case, is it still a single illuminate profile?

Okay, I've just tested that, and no, it results in a relatively inaccurate profile.
Logged
Schewe
Sr. Member
****
Offline Offline

Posts: 5533


WWW
« Reply #8 on: January 12, 2013, 03:04:20 AM »
ReplyReply

With all due respect, I'm pretty sure that a single CC target can create a useful dual illuminate profile - one would just have to photograph it on separate occasions under two different light sources (as recommended by Adobe and Eric, also yourself, something close to 6500K daylight and 2850K tungsten).

Uh, correct me if I'm wrong, but a single target could NEVER produce an accurate (note the term accurate) DNG profile as a dual-illuminate profile with a single target...

Do you disagree?

The fact you can screw with various tables doesn't change the fact that if you want to create a dual-illuminate profile you'll need a target shot near D-65 and a target shot near near 3200K.

So, did you shoot two targets?

If not, what you are posting are a complete waste of time...

Exactly what did you do and what do you think your discovered?
Logged
samueljohnchia
Sr. Member
****
Offline Offline

Posts: 327


« Reply #9 on: January 12, 2013, 03:09:07 AM »
ReplyReply

That is not puzzling, that is by design. You always get a dual illuminant profile. You might as well think of a dual illuminant profile as 2 separate profiles. One is normally based on the capture data for illuminant A circumstances, the other should be based on daylight capture circumstances. However, nothing stops you from using a single capture for both tables. This is what used to be referred as a "single-illuminant profile", i.e. a dual illuminant profile where the illuminant of both tables is the same, based on a single capture.


I modified this post for accuracy.

I think I am getting muddled by the terminology - I'm pretty sure that the DNG PE can produce single illuminate profiles - Eric has used that expression before.

I do not doubt that the DNG PE generated profile has dual tables, and it seems to have the ability to create

1. A single illuminate profile with the HueSatDelta1 table (HueSatDelta2 table with null values)
2. A single illuminate profile with the HueSatDelta2 table (HueSatDelta1 table with null values)
3. A single illuminate profile with the HueSatDelta1 table and HueSatDelta2 table (built from the same capture, without moving the control points)
4. A dual illuminate profile with the HueSatDelta1 table and HueSatDelta2 table (from two separate captures)

And what it appears to be to me thus far:
5. A single illuminate profile with the HueSatDelta1 table and HueSatDelta2 table (built from the same capture, without moving the control points) - using "Both color tables" that arrives at slightly different HueSatDelta1/2 table values.

5. A single illuminate profile with just the HueSatDelta1 table (built from the same capture, without moving the control points) - using "Both color tables"

Why? I don't know... But since Eric mentioned that the adopted illuminant for single and dual illuminant profiles are different, I suspect that is at play here.

I highly doubt that the DNG PE would know that I'm creating 3. and that its not from the same capture (by right it is expecting two different captures).

But if I run it for the "Both color tables" option, surely it knows that its working from only one source (it certainly does not pause to let me load another capture of the colorchecker) - so it's gotta be single illuminant, hence no chromatic adaption to D50, and D50 is assumed? Is that how it works?
« Last Edit: January 12, 2013, 03:42:58 AM by samueljohnchia » Logged
opgr
Sr. Member
****
Offline Offline

Posts: 1125


WWW
« Reply #10 on: January 12, 2013, 03:10:16 AM »
ReplyReply

The colorchecker values can be specified using any illuminant, not just D50. Eric said before that for a single illuminant profile, the DNG PE adopts D50 as the illuminant, and the colors need not be chromatically adapted to the internal reference values for the colorchecker that the PE uses.

Exactly, so the D50 illuminant refers to interpreting the colorchecker values and are generally irrelevant to your profiling. IF you use your own measurements of a chart, you should convert to Lab using D50. That is what it means. It has no relevance to your profiling beyond that.

I guess the main question is, what is a single illuminant profile? 6500K only? 2850K only? Both of those? Both color tables generated from the same capture of a colorchecker?

It used to be this: "Both color tables generated from the same capture of a colorchecker?"

This is not consistent with my own testing... especially when the WB sliders in ACR are set to less than 6000K
One can copy the table values from the 6500K table to the 2850k tablein the PE. So in that case, is it still a single illuminate profile?

Yes, and it is important that the table values are copied. They need to be equal for the profile to act as a single illuminant profile.

And yes, that means that the profile is accurate only for the original capture conditions.

IF you want accurate slider control over the range of temperatures for a single light-source (to avoid the word illuminant), you need to capture the colorchecker in under that lightsource exhibiting the colortemperatures mentioned. That is: if you want a daylight response with accurate slider control you need to capture the colorchecker at sunset near 2800K and at noon near 6500K and use both those captures to generate two tables.

You then get an accurate "dual illuminant profile" with a single illuminant=lightsource. But I doubt that is what you want or need.
Logged

Regards,
Oscar Rysdyk
theimagingfactory
samueljohnchia
Sr. Member
****
Offline Offline

Posts: 327


« Reply #11 on: January 12, 2013, 03:18:36 AM »
ReplyReply

Uh, correct me if I'm wrong, but a single target could NEVER produce an accurate (note the term accurate) DNG profile as a dual-illuminate profile with a single target...

Do you disagree?

The fact you can screw with various tables doesn't change the fact that if you want to create a dual-illuminate profile you'll need a target shot near D-65 and a target shot near near 3200K.

So, did you shoot two targets?

If not, what you are posting are a complete waste of time...

Exactly what did you do and what do you think your discovered?

Jeff, there may be a misunderstanding on my part. I used only one physical target, one CC . I made one capture of the target (I "measured" the camera's spectral response to known color patches under a illuminant). I fed the DNG PE the "measurement" file and it generated the profile. I was not trying to create a dual illuminate profile. I know exactly how that is supposed to work. What I am doing is not futile. I'm sure you know, for daylight illuminated images, a dual illuminant profile is quite a bit less accurate when the WB slider is set to less than 6500K, because the tungsten end of the profile is coming into play. And daylight is certainly not identical to tungsten... all my testing proved that that was true.

I am trying to create an accurate single illuminant profile for daylight, because I only photograph in daylight.

I have had great success thus far with my single illuminant daylight profiles - I'm not trying to condemn the DNG PE and dual illuminant profile method! I actually think it is quite brilliant, and I love it more than any other method of creating custom camera profiles. In my research, I discovered that there is a difference in the method that the PE uses with regards to the "Both color tables" option. I just want to know if it is because a different illuminant is adopted, and if that results in a more accurate profile.
Logged
samueljohnchia
Sr. Member
****
Offline Offline

Posts: 327


« Reply #12 on: January 12, 2013, 03:31:04 AM »
ReplyReply

Quote
Yes, and it is important that the table values are copied. They need to be equal for the profile to act as a single illuminant profile.
I guess so. So letting the PE calculate both color tables at once is not a single illuminant profile? Surely it must be, because it knows that it's asking for one capture source only...

Quote
IF you want accurate slider control over the range of temperatures for a single light-source (to avoid the word illuminant), you need to capture the colorchecker in under that lightsource exhibiting the colortemperatures mentioned. That is: if you want a daylight response with accurate slider control you need to capture the colorchecker at sunset near 2800K and at noon near 6500K and use both those captures to generate two tables.

My own testing has proven to me that if one were to shoot in daylight around 5000 - 6500k, having "a sunset near 2800k and daylight 6500k dual illuminant profile" is also less accurate than the profile generated with "Both color tables", from a daylight 6500k capture.

I have separate profiles for daylight and sunset - and they are wonderfully accurate for describing those kinds of light. But one must have separate profiles or it becomes less accurate.

Quite often even for a sunset image, the WB slider in ACR is set to somewhere between 4500k to 5500k give or take, to preserve the "warmth" of the light. One would rarely set the WB to less than 2850k. This means that the daylight 6500k table is going to come into play, reducing the accuracy the of the "sunset profile"...
Logged
samueljohnchia
Sr. Member
****
Offline Offline

Posts: 327


« Reply #13 on: January 12, 2013, 03:36:53 AM »
ReplyReply

My apologies to have failed to observe and mention this earlier - I took a look again at the "Both color tables" profile, and it contains only one HueSatDelta table, not two. It still has two color matrix and forward matrix tables, which are identical to the 2850K plus 6500K profile, all from the same capture.

As only the HueSatDelta tables are different, I'm pretty sure that is what's causing the differences.

With only 1 HueSatDelta table, the "Both color tables" profile is actually a single illuminant profile in a certain sense! Wow, never realised that...
Logged
Tim Lookingbill
Sr. Member
****
Offline Offline

Posts: 1231



WWW
« Reply #14 on: January 12, 2013, 01:05:47 PM »
ReplyReply

The sample 2800K tungsten lit CC chart shot I posted in the linked thread below (reply #5) explains the advantages of using a dual illuminant profile. It all has to do with controlling the appearance through color table manipulation of hue/saturation adjusting WB from warm tungsten to cool D50 in ACR/LR shooting under non-conventional=(non-natural daylight/D50) lit scenes. Dual illuminant basically acts as a good enough one size fits all profile for shooting under a wide range of lighting scenarios.

http://www.luminous-landscape.com/forum/index.php?topic=43733.0

I still don't know from reading this current discussion what new information has been added to the above. Could someone enlighten me?
Logged
samueljohnchia
Sr. Member
****
Offline Offline

Posts: 327


« Reply #15 on: January 12, 2013, 07:34:42 PM »
ReplyReply

The sample 2800K tungsten lit CC chart shot I posted in the linked thread below (reply #5) explains the advantages of using a dual illuminant profile. It all has to do with controlling the appearance through color table manipulation of hue/saturation adjusting WB from warm tungsten to cool D50 in ACR/LR shooting under non-conventional=(non-natural daylight/D50) lit scenes. Dual illuminant basically acts as a good enough one size fits all profile for shooting under a wide range of lighting scenarios.

http://www.luminous-landscape.com/forum/index.php?topic=43733.0

I still don't know from reading this current discussion what new information has been added to the above. Could someone enlighten me?

Yes, I read your post while conducting my research about camera profiling. My own attempts at creating dual illuminant profiles (6500K table with a daylight shot of the colorchecker, 2850k table with a tungsten shot of the colorchecker) have proven to be less accurate for describing daylight and sunset scenes. As long as the ACR/LR WB slider is set between 2850k and 6500k, both ends of the profile come into play. If the scene photographed is illuminated by daylight, the tungsten end of the profile is going to influence the rendition of colors a bit or alot, depending on how close to 2850K the slider is set to. As we all know the way a camera sensor responds to tungsten is not quite the same as daylight, or even sunset light of the same correlated color temperature, so how could it be more accurate?

I wholly agree that a single table daylight 6500k profile is inadequate for the interior lighting conditions that your test was conducted in. I thought that was obvious?

Unfortunately what's acceptable to most may not be accurate enough for me, and I went on to develop profiles for two flavors of sunlight - one good for D50 - D75 (give or take) and another for intense reddish sunsets. The SPD of daylight does change quite a bit for sunsets, and a daylight illuminant (even a dual table profile) profile is too yellow-green. A dual illuminant profile with a tungsten end for sunset lit scenes is slightly too warm/too cool, depending on how close to 2850k the tungsten illumination of the CC is. They have proven to be extremely beautiful profiles to work with, and I rarely have to make localised color adjustments because natural surfaces under natural light are just rendered much more accurately than the Adobe Standard and other dual illuminant profiles, generic or custom made. This is both a visual personal preference, but also a scientifically tested one with shots of CCs against synthetic CC charts. I'm not going to elaborate further about this as this is getting OT.

I think a significant question that might point us in the right direction is What is the "Both color tables" option in the DNG PE good for? Based on my own testing, it is to build the kind of profiles that I'm after! I'm just trying to seek clarification on what the DNG PE is really doing in the bckground, as Eric's comments on how it works seem to contradict each other. Hopefully Eric will see this and point us in the right direction, as only he really knows what's going on.
Logged
Tim Lookingbill
Sr. Member
****
Offline Offline

Posts: 1231



WWW
« Reply #16 on: January 13, 2013, 02:26:17 AM »
ReplyReply

Quote
Unfortunately what's acceptable to most may not be accurate enough for me, and I went on to develop profiles for two flavors of sunlight - one good for D50 - D75 (give or take) and another for intense reddish sunsets. The SPD of daylight does change quite a bit for sunsets, and a daylight illuminant (even a dual table profile) profile is too yellow-green. A dual illuminant profile with a tungsten end for sunset lit scenes is slightly too warm/too cool, depending on how close to 2850k the tungsten illumination of the CC is. They have proven to be extremely beautiful profiles to work with, and I rarely have to make localised color adjustments because natural surfaces under natural light are just rendered much more accurately than the Adobe Standard and other dual illuminant profiles, generic or custom made. This is both a visual personal preference, but also a scientifically tested one with shots of CCs against synthetic CC charts. I'm not going to elaborate further about this as this is getting OT.


How about posting one of those "accurate" sunsets that demonstrates the effectiveness of those profiles you developed. Can you show us what's not accurate or show this unexpected behavior which this topic is based on?
Logged
samueljohnchia
Sr. Member
****
Offline Offline

Posts: 327


« Reply #17 on: January 13, 2013, 05:29:21 AM »
ReplyReply


How about posting one of those "accurate" sunsets that demonstrates the effectiveness of those profiles you developed. Can you show us what's not accurate or show this unexpected behavior which this topic is based on?

In summary, the "unexpected behavior" is that when building a DNG camera profile with the DNG PE, one would normally expect a profile built using the "Both color tables" option to be identical to a profile build using the "2850 K only" and "6500 K only" options from the same source. As I mentioned earlier, even Eric Chan, who wrote the DNG PE program, said that was unexpected. All I want to know is why is it different.

I have already provided a link to the decompiled xml files of the profiles I have built in my first post, that shows what the differences are in the profiles. Sandy, author of the dcpTool I used to decompile the dcp profiles, also arrived at the same conclusions as I did. Anybody can reproduce this experiment with a single shot of a CC target, with similar results. Showing examples here is not going to get us any closer to the answers...only Eric knows the math behind the DNG PE!

Any comment about accuracy was just me being curious, as the two profiles that should be identical are actually subtly different, I was curious as to which one is more accurate, albeit only slightly. As to accurate sunset profiles, I think that should be a separate topic of discussion. I am planning to publish an article eventually which will contain all the relevant examples, after I am sure I have covered every thing I can think of.
Logged
Tim Lookingbill
Sr. Member
****
Offline Offline

Posts: 1231



WWW
« Reply #18 on: January 13, 2013, 10:06:12 AM »
ReplyReply

And I'm sure that article you plan to write is going to be as useful to photographers as this discussion.

I'm betting on it.
Logged
Peter_DL
Sr. Member
****
Offline Offline

Posts: 421


« Reply #19 on: January 13, 2013, 01:16:47 PM »
ReplyReply

... when building a DNG camera profile with the DNG PE, one would normally expect a profile built using the "Both color tables" option to be identical to a profile build using the "2850 K only" and "6500 K only" options from the same source.

ACR/LR were reported to interpolate between, or extrapolate beyond, a pair of matrices ("profiles") built for illuminants D65 and A depending on white balance. I'd expect the DNG Profile Editor to do the same after auto-whitebal when running the Chart Wizard with the "Both Color Tables" option. Hence, the starting point (= interpolated matrix) for building the deltaHue/Sat. on top would be different compared to "2850 K only + 6500 K only" and starting with either this or that matrix. The latter approach does not seem to make sense to me.

Peter

--
Logged
Pages: [1] 2 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad