Big difference for Portra 400 between 0.3 and 0.4

[risbac]

Hello !

I'm testing the new 0.4 versions, and there is a big difference at least on the Portra 400VC that I'm using very very often. The new version is really much more saturated. I have uploaded three different versions of the same photos with 0.3, 0.4 and no style applied : https://www.flickr.com/photos/smooth-bokeh/ (see at the top).

I was often adding some saturation to my Portra photos, so I guess maybe I'm biased and too used to the 0.3 version and not to the real film. But I thought it could be useful to let you know my feeling about it, as it's a big change and it's kinda my default film style.

Thanks for the new version in any case, I'll be happy to test all the other ones slowly.

best regards,

François

[t3mujin]

Hi François, I think I should've been more clear on the blog/release with the portras, I've done a lot of work on them and changed them a bit. You telling me you had to add saturation kind of tells me you felt the same, I was missing the richer colors, every time I used it I felt it had to correct it every time. Your v0.4 sample has your saturation corrections?

[risbac]

Hi João

My v0.4 sample has zero saturation correction, just like my v0.3 sample was also just with your style applied. The gamut alert on Darktable is also saying that the v0.4 version is really pushing the colors. It feels a bit like going from one extreme to the other one. I need to test on more screens, but it really feels like I will have to desaturate a bit to avoid too flashy colors. I really don't mean to sound like criticizing, I use your themes a LOT (basically on every wedding set for instance), so I wanted to share my feelings. I had already downloaded the v0.4 version when it was a work in progress, and it gave me the same feeling : the colors were too vivid, so I thought you would tame them a bit when finalizing. But again, I'm not familiar with the film itself, so my reference is just the v0.3 and my eyes :)

On Thu, Aug 2, 2018 at 5:30 PM João Almeida notifications@github.com wrote:

Hi François, I think I should've been more clear on the blog/release with the portras, I've done a lot of work on them. You telling me you had to add saturation kind of tells me you felt the same, I was missing the richer colors, every time I used it I felt it had to correct it every time. Your v0.4 sample has your saturation corrections?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/t3mujin/t3mujinpack/issues/6#issuecomment-409967680, or mute the thread https://github.com/notifications/unsubscribe-auth/AICYngqOYs9j5Ki7lpBwsi7SQD1RcJnBks5uMxsygaJpZM4Vn3Z7 .

[t3mujin]

Can you send a corrected 0.3 or 0.4? So I can get a better understanding on what's the expected result for you.

...But those greens really look weird. I'll save your photo to my test samples collection.

[risbac]

Maybe I'm doing something wrong, or I have a wrong setting in my Darktable ? Not to rely only on my eye and my use of your style (since I don't know the film), I have use two photos with the same settings : my default "flat" style in Darktable, then your V0.4, then from the "flat" jpg, I have used 2 different Portra 400VC simulation : the gmic Gimp plugin, and the online 29a.ch tool. Both gives a pretty similar result, so I guess they must be fine. Then the v0.4 version is too saturated, and with a red tint apparently. Even the 0.3 version seems too red.

I'm trying to attach the two series of photos here :

[risbac]

Shit, the image names are changed. So in the right order : Flat Jpg / GMIC / 29a.ch / V0.4

then second photo : Flat Jpg / GMIC / 29a.ch / V0.3 / V0.4

[t3mujin]

Thanks François, I'll check with your images (and add them to my pool of test samples). I think I haven't tested the Portra VC/NC/UC with enough sample images, maybe I should explore darktable-cli to try to automate that.

[t3mujin]

Want to give a try with the presets I committed this weekend?

[risbac]

Hello !

Just did with the Portra 400VC and I don't see any difference. Or it's incredibly subtle. I compared the two tones curves in the lab, and I can't see any visual difference on the curves, are you sure you commited the right file ? Or my Darktable has a problem maybe ? I will test again on my main computer tonight.

On Mon, Aug 6, 2018 at 2:22 PM João Almeida notifications@github.com wrote:

Want to give a try with the presets I committed this weekend?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/t3mujin/t3mujinpack/issues/6#issuecomment-410689637, or mute the thread https://github.com/notifications/unsubscribe-auth/AICYnmYwMrADWf43x59XXXz8ZAYvO_Ftks5uODUDgaJpZM4Vn3Z7 .

[t3mujin]

Have you checked if the style is turned on? Sometimes I save it with the curves set but the module disabled.

[risbac]

yes I did. I duplicated the same photo I used for my comparion, I deleted any editing, applied your style and used the snapshot option to compare it with v0.4. I see virtually no difference. I'm wondering if your 400 UC is not what the VC should be. UC is REALLY very saturated. VC is more saturated than NC, but still not that saturated. I need to compare more on my main computer, sometimes I'm wondering which version I'm really using, it's easy to mix them if I don't name them with the version number in the photo tag.

On Mon, Aug 6, 2018 at 2:54 PM João Almeida notifications@github.com wrote:

Have you checked if the style is turned on? Sometimes I save it with the curves set but the modules disabled.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/t3mujin/t3mujinpack/issues/6#issuecomment-410698057, or mute the thread https://github.com/notifications/unsubscribe-auth/AICYntYTw0YHad316apvJ8h8gRuojzmVks5uODyQgaJpZM4Vn3Z7 .

[t3mujin]

I wonder if you mixed some of your files, because I just downloaded the current ones styles from github and the release ones and the results are clearly different. Did you download them directly from github? I haven't released any new version yet.

current 400NC / 0,4,0 400NC / current 400UC / 0,4,0 400UC / current 400VC / 0,4,0 400VC

[risbac]

indeed, your files look much better ! I must be mixing something. Or lacking coffee. Or maybe both. It's bloody hot these days here, my brain must lack some oxygen. I can't imagine how it must be in Portugal ;)

I will test again tonight with more time. Sorry for the failed test.

On Mon, Aug 6, 2018 at 3:30 PM João Almeida notifications@github.com wrote:

I wonder if you mixed some of your files, because I just downloaded the current ones styles from github and the release ones and the results are clearly different. Did you download them directly from github? I haven't released any new version yet.

current 400NC / 0,4,0 400NC / current 400UC / 0,4,0 400UC / current 400VC / 0,4,0 400VC

[image: 43611188-a353ce48-96a8-11e8-8523-1ae5e5eed6bb] https://user-images.githubusercontent.com/331767/43719018-69b9e396-9984-11e8-945c-7870dec775da.jpg [image: 43611188-a353ce48-96a8-11e8-8523-1ae5e5eed6bb_01] https://user-images.githubusercontent.com/331767/43719026-6dc35d6e-9984-11e8-8f77-d27f130e3106.jpg [image: 43611188-a353ce48-96a8-11e8-8523-1ae5e5eed6bb_02] https://user-images.githubusercontent.com/331767/43719028-70725d6c-9984-11e8-8cca-d16161fa6a42.jpg [image: 43611188-a353ce48-96a8-11e8-8523-1ae5e5eed6bb_03] https://user-images.githubusercontent.com/331767/43719032-721815e4-9984-11e8-9e13-1bf9f0033028.jpg [image: 43611188-a353ce48-96a8-11e8-8523-1ae5e5eed6bb_04] https://user-images.githubusercontent.com/331767/43719035-73d6d528-9984-11e8-850e-ef24c8115eab.jpg [image: 43611188-a353ce48-96a8-11e8-8523-1ae5e5eed6bb_05] https://user-images.githubusercontent.com/331767/43719041-75f6bf6c-9984-11e8-8c1a-c5920e86128a.jpg

[image: 43611193-a3b70c56-96a8-11e8-90de-67f762a56835] https://user-images.githubusercontent.com/331767/43719050-7ad5ad86-9984-11e8-97a3-db8b43ca8e9a.jpg [image: 43611193-a3b70c56-96a8-11e8-90de-67f762a56835_01] https://user-images.githubusercontent.com/331767/43719054-7c382014-9984-11e8-8238-17e5a8373014.jpg [image: 43611193-a3b70c56-96a8-11e8-90de-67f762a56835_02] https://user-images.githubusercontent.com/331767/43719058-7e2fa220-9984-11e8-9f53-e7970c6843d5.jpg [image: 43611193-a3b70c56-96a8-11e8-90de-67f762a56835_03] https://user-images.githubusercontent.com/331767/43719061-7fa03b92-9984-11e8-9ad6-96c1db521650.jpg [image: 43611193-a3b70c56-96a8-11e8-90de-67f762a56835_04] https://user-images.githubusercontent.com/331767/43719063-80fe78dc-9984-11e8-92be-290f1f76ee12.jpg [image: 43611193-a3b70c56-96a8-11e8-90de-67f762a56835_05] https://user-images.githubusercontent.com/331767/43719065-82f36ec2-9984-11e8-8492-b390b6c6c46d.jpg

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/t3mujin/t3mujinpack/issues/6#issuecomment-410708276, or mute the thread https://github.com/notifications/unsubscribe-auth/AICYnoNzmTEPhCs391HV9Q-M5bOXH5ZNks5uOET6gaJpZM4Vn3Z7 .

[t3mujin]

Ahahahahah! I can relate to that, the weekend was excruciatingly hot, today's 31º even feels like a mild summer day.

Those aren't the final versions of the styles yet, just a work in progress, I'll need to do some further testing with a broader ser of image samples, and then release a 0.4.1, with these fixes .

[risbac]

it looks very promising in any case. I will make sure to do better tests tonight. Keep up the good work !

On Mon, Aug 6, 2018 at 3:36 PM João Almeida notifications@github.com wrote:

Ahahahahah! I can relate to that.

Those aren't the final versions of the styles yet, just a work in progress, I'll need to do some further testing with a broader ser of image samples, and then release a 0.4.1, with these fixes .

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/t3mujin/t3mujinpack/issues/6#issuecomment-410710273, or mute the thread https://github.com/notifications/unsubscribe-auth/AICYnnsKnZk2Qs9GwblHDs09cm43Qp_2ks5uOEZpgaJpZM4Vn3Z7 .