add support for more fujifilm tags #1949

VJSchneid · 2021-10-06T10:52:04Z

I made some tests with the Fujifilm X-T4 and added/improved some missing makernote tags.

The following Fujifilm tags where added:

ColorTemperature
WhiteBalanceFineTune
HighIsoNoiseReduction
Clarity
FocusArea
FocusPoint
LensModulationOptimizer
GrainEffectRoughness
GrainEffectSize
ColorChromeEffect
ColorChromeFXBlue
ImageNumber

Additionally the print values for the following Fujifilm tags where extended:

Sharpness
WhiteBalance
Color
FlashMode
Shadow/Highlight Tone
Film Mode

kmilos · 2021-10-06T13:14:13Z

src/fujimn_int.cpp

-        { 4, N_("Hard mode 1") },
-        { 5, N_("Hard mode 2") }
+        {   0, N_("-4")             },
+        {   1, N_("Soft mode 1/-3") },


I wonder if there is nicer way to present this now, e.g. keep quantity consistently on the left like "-3 (Soft mode 1)" since +/-1 and +/-4 are now added without a label; or add "Soft/Hard mode (0)" and "Soft/Hard mode 3" labels... How is it presented in the camera menu?

Good point. The X-T4 (and probably all other recent Fujifilm cameras) display the sharpness as number (-4,-3,-2,-1,0,+1,+2,+3,+4). I think older cameras display this setting as text and support fewer options.

I would prefer something like "-3 (Soft mode 1)" because the labels might be a bit out of date instead of the numbers.

src/fujimn_int.cpp

kmilos · 2021-10-06T13:17:57Z

src/fujimn_int.cpp

-        { 771, N_("Monochrome + G Filter")  },
-        { 784, N_("Sepia")                  },
-        { 768, N_("Monochrome")             } // To silence compiler warning
+        {    0, N_("Normal/0")               },


Same as above, is there a more consistent way to present the value+label combo?

src/fujimn_int.cpp

kmilos · 2021-10-06T13:26:28Z

Thanks for your contribution!

postscript-dev · 2021-10-06T14:07:41Z

Thank you for taking time to contribute this.

Exiv2 collaborates with another open source project called ExifTool. Their website contains a more complete list of tags and values that have been discovered for the different camera types. ExifTool's FujiFilm tags can be found here and the source code (containing all the conditions, extra info and pretty printing) here. There is some additional information on the websites, that you may consider including (e.g., extra entries in the arrays).

VJSchneid · 2021-10-06T14:20:25Z

src/fujimn_int.cpp

        {  256, N_("F1/Studio Portrait")                           },
        {  272, N_("F1a/Studio Portrait Enhanced Saturation")      },
-        {  288, N_("F1b/Studio Portrait Smooth Skin Tone (Astia)") },
+        {  288, N_("F1b/Studio Portrait Smooth Skin Tone (ASTIA)") },


I am thinking about changing this notation too because the labels "F0/Standard", "F1b/Studio Portrait..." are also coming from older fujifilm cameras like the FinePix S5 Pro. Newer cameras display a name like PROVIA, ASTIA, etc.

Maybe this can be made consistently with the Color and Sharpness Tag.

VJSchneid · 2021-10-06T14:26:16Z

Thank you for taking time to contribute this.

Exiv2 collaborates with another open source project called ExifTool. Their website contains a more complete list of tags and values that have been discovered for the different camera types. ExifTool's FujiFilm tags can be found here and the source code (containing all the conditions, extra info and pretty printing) here. There is some additional information on the websites, that you may consider including (e.g., extra entries in the arrays).

Thanks for the information. I can extend the tags with the code from exiftool but I am currently unable to verify all these settings with my hardware...

kmilos · 2021-10-06T15:41:42Z

src/fujimn_int.cpp

+        {   1, N_("-3 (Soft mode 1)") },
+        {   2, N_("-2 (Soft mode 2)") },


This is a bit controversial (not symmetrical): soft 1 = -3 but hard 1 = +2; I would expect soft/hard 1 = -/+2 and soft/hard 2 = -/+3... any way to confirm these?

Or ditch the mode 1/2 altogether for a scheme similar to exiftool?

I am not sure where these labels "soft/hard modes" come from. Maybe I can find the origin of that with git. I will investigate that...

These modes are from here and where added 18 years ago to exiv2^^ Back then they where just called High and Low. I can not find any information on why the "mode 1/2" suffix was added...

Perhaps we can ditch "mode 1" and "mode 2" and stick with exiftool labels?

I used the labels and extended some more Tags from exiftool with my last commit

kmilos · 2021-10-06T15:42:55Z

src/fujimn_int.cpp

        {  128, N_("+1")                     },
        {  192, N_("+3")                     },
        {  224, N_("+4")                     },
-        {  256, N_("High/+2")                },
+        {  256, N_("+2 (High)")              },
        {  384, N_("-1")                     },
        {  512, N_("Low")                    },


Is this "-2 (Low)" now to anti-reflect "+2 (High)"?

Ah, I now see there is a -2 below where (Low) should be added (cf. exiftool)

In my understanding "Low" and "High" are color levels from older/different cameras. The X-T4 for example does not show "High" or "Low" labels, but numbers. Older cameras like the FinePix S5-Pro have the High and Low level as it is stated in the manual on page 109.
So "-2" and "Low" do not share a common tag value contrary to "+2" and "High" as it seems to be different implemented in these cameras. Therefore "High" is not an alternative representation of "+2" but a legacy value. Pretty printing dependent on the camera model might be better but I am currently unaware whether this is easy to implement. Do you know how this is handled with other manufacturers where old and new tags overlap?

Again, maybe just follow exiftool here as well?

I applied the labels from exiftool and additionally changed the label of value 512 to "-2 (low)" to unify it with the corresponding high value (256). In exiftool this was just "Low".

kmilos · 2021-10-06T15:54:10Z

src/fujimn_int.cpp

        {  304, N_("F1c/Studio Portrait Increased Sharpness")      },
-        {  512, N_("F2/Fujichrome (Velvia)")                       },
+        {  512, N_("Velvia (F2/Fujichrome)")                       },


Just out of curiosity, you know why there are 2 Velvia entries (F2 and F4)?

I don't know, but I think they come from here.

Value 512 translates on modern Fujifilm cameras to Velvia. The F1/2/... notation comes again from older cameras like the FinePix S5 Pro.

src/fujimn_int.cpp

piponazo · 2022-01-27T15:57:18Z

Hi everybody. Thanks @VJSchneid for the contribution. I am just taking a look to some open PRs and I was wondering if we could give the final push to this PR to get it merged. I am not familiar with the Fujifilm part of the code, but it looks like you added support for many new tags. It would be great to incorporate these changes into Exiv2.

Could you please rebase your branch in top of main so that the CI pipeline runs again? It looks like some of the CI jobs are in red because some hiccup in Github actions. Maybe we get those CI checks into "green" state just by rebasing your branch.

piponazo · 2022-02-27T18:28:51Z

Hi @VJSchneid again, and sorry for taking so long to come back to you. After you rebased the branch, most of the CI jobs failed. There has been lot of activity in the last month in main and there seems to be some conflicts now. Do you need any help with this? I just would not like to see this contribution getting stuck for not giving it enough attention (from our side).

Co-authored-by: Miloš Komarčević <[email protected]>

…ng signed values

VJSchneid · 2022-05-13T13:07:45Z

Hi @piponazo, sorry that this response took so long. I rebased my branch again and hopefully everything is in sync now.
Additionally some missing testfiles were also updated. Unfortunately I get some errors when runnung ctest with Exif.CanonCs.LensType but this should be independent of these changes. The fujifilm stuff seems to pass :)

VJSchneid · 2022-08-01T20:29:55Z

Hi @piponazo, I have merged the latest changes with this PR again. Is there anything more I can do to speed things up?

src/fujimn_int.cpp

+//! DriveSetting, tag 0x1103
+std::ostream& printFujiDriveSetting(std::ostream& os, const Value& value, const ExifData*) {
+  auto byte1 = value.toInt64() & 0xff;
+  auto byte2 = (value.toInt64() >> 8) & 0xff;


src/fujimn_int.cpp

+std::ostream& printFujiDriveSetting(std::ostream& os, const Value& value, const ExifData*) {
+  auto byte1 = value.toInt64() & 0xff;
+  auto byte2 = (value.toInt64() >> 8) & 0xff;
+  auto byte3 = (value.toInt64() >> 16) & 0xff;


src/fujimn_int.cpp

+  auto byte1 = value.toInt64() & 0xff;
+  auto byte2 = (value.toInt64() >> 8) & 0xff;
+  auto byte3 = (value.toInt64() >> 16) & 0xff;
+  auto fps = value.toInt64() >> 24;


codecov · 2022-08-01T20:36:09Z

Codecov Report

Merging #1949 (d62220a) into main (1d0530f) will decrease coverage by 0.09%.
The diff coverage is 26.92%.

@@            Coverage Diff             @@
##             main    #1949      +/-   ##
==========================================
- Coverage   63.53%   63.43%   -0.10%     
==========================================
  Files         118      118              
  Lines       19595    19647      +52     
  Branches     9576     9608      +32     
==========================================
+ Hits        12450    12464      +14     
- Misses       5079     5112      +33     
- Partials     2066     2071       +5

Impacted Files	Coverage Δ
src/fujimn_int.cpp	`29.62% <26.92%> (-70.38%)`	⬇️

Help us with your feedback. Take ten seconds to tell us how you rate us.

piponazo · 2022-08-03T06:34:12Z

Hi @piponazo, I have merged the latest changes with this PR again. Is there anything more I can do to speed things up?

Sorry for my late reply, I am not participating much in the project lately.

Everything looks green, let's merge it!

postscript-dev added enhancement feature / functionality enhancements makerNote Anything related to one of the various supported MakerNote formats prettyPrinter Anything related to the output formatting of a value labels Oct 6, 2021

kmilos reviewed Oct 6, 2021

View reviewed changes

VJSchneid commented Oct 6, 2021

View reviewed changes

kmilos reviewed Oct 6, 2021

View reviewed changes

kmilos reviewed Oct 13, 2021

View reviewed changes

src/fujimn_int.cpp Outdated Show resolved Hide resolved

kmilos reviewed Oct 13, 2021

View reviewed changes

src/fujimn_int.cpp Outdated Show resolved Hide resolved

src/fujimn_int.cpp Outdated Show resolved Hide resolved

VJSchneid force-pushed the main branch from c1efa8f to 42a313e Compare January 28, 2022 15:53

VJSchneid and others added 12 commits May 13, 2022 13:32

add support for more fujifilm tags

4a3cf6b

remove auto range for DynamicRangeAuto setting

c4ba70c

print newer labels first when interpreting Fujifilm tags

21d0763

remove redundant TagDetails and fix formatting

967f6a6

Co-authored-by: Miloš Komarčević <[email protected]>

add some fujifilm tags from exiftool and extend TagDetails

f9cbf24

remove TODO entries

040b106

fix typo in fujiSharpness and prevent undefined behaviour when shifti…

b8e378b

…ng signed values

add comment that fujiColor value 512 differs from exiftool

88bc575

update test cases to respect new fujifilm tags

2c41063

apply fujifilm tag changes to exiv2-test.out

fe0f0fb

fix some rebase issues

f325802

update test references to cover new and updated fujifilm tags

061a98a

VJSchneid force-pushed the main branch from 42a313e to 061a98a Compare May 13, 2022 12:53

apply clang-format rules to fujimn_int.cpp

6bbaa31

VJSchneid added 2 commits May 13, 2022 15:39

use correct type for element counter in printFujiFaceElementTypes

e185c58

update test/data/test_reference_files/exiv2-test.out

37dc2ee

VJSchneid requested a review from kmilos May 13, 2022 18:39

VJSchneid added 2 commits August 1, 2022 22:20

Merge remote-tracking branch 'origin'

73a1a0e

apply clang-format to src/fujimn_int.cpp

d62220a

github-advanced-security bot found potential problems Aug 1, 2022

View reviewed changes

piponazo approved these changes Aug 3, 2022

View reviewed changes

piponazo merged commit 0256775 into Exiv2:main Aug 3, 2022

nehaljwani mentioned this pull request Apr 30, 2023

Is a new release happening soon? #2406

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

add support for more fujifilm tags #1949

add support for more fujifilm tags #1949

VJSchneid commented Oct 6, 2021

kmilos Oct 6, 2021

VJSchneid Oct 6, 2021

VJSchneid Oct 6, 2021

kmilos Oct 6, 2021

kmilos commented Oct 6, 2021

postscript-dev commented Oct 6, 2021

VJSchneid Oct 6, 2021

VJSchneid commented Oct 6, 2021

kmilos Oct 6, 2021 •

edited

Loading

VJSchneid Oct 7, 2021

VJSchneid Oct 7, 2021

kmilos Oct 7, 2021

VJSchneid Oct 8, 2021

kmilos Oct 6, 2021

kmilos Oct 6, 2021

VJSchneid Oct 7, 2021

kmilos Oct 7, 2021

VJSchneid Oct 8, 2021 •

edited

Loading

kmilos Oct 6, 2021

VJSchneid Oct 7, 2021

VJSchneid Oct 7, 2021 •

edited

Loading

piponazo commented Jan 27, 2022

piponazo commented Feb 27, 2022

VJSchneid commented May 13, 2022

VJSchneid commented Aug 1, 2022

codecov bot commented Aug 1, 2022

piponazo commented Aug 3, 2022

		{ 1, N_("-3 (Soft mode 1)") },
		{ 2, N_("-2 (Soft mode 2)") },

add support for more fujifilm tags #1949

add support for more fujifilm tags #1949

Conversation

VJSchneid commented Oct 6, 2021

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

kmilos commented Oct 6, 2021

postscript-dev commented Oct 6, 2021

Choose a reason for hiding this comment

VJSchneid commented Oct 6, 2021

kmilos Oct 6, 2021 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

VJSchneid Oct 8, 2021 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

VJSchneid Oct 7, 2021 • edited Loading

Choose a reason for hiding this comment

piponazo commented Jan 27, 2022

piponazo commented Feb 27, 2022

VJSchneid commented May 13, 2022

VJSchneid commented Aug 1, 2022

codecov bot commented Aug 1, 2022

Codecov Report

piponazo commented Aug 3, 2022

kmilos Oct 6, 2021 •

edited

Loading

VJSchneid Oct 8, 2021 •

edited

Loading

VJSchneid Oct 7, 2021 •

edited

Loading