It is difficult to tell with this image how stretched the corners are. I am using lensfun like this, FZ330 automatic Profiled Lens Correction help - #3 by gaaned92, so that may not be helping. The image from Rawtherapee is 8 pixels bigger.
Out of camera jpeg.
Rawtherapee export with Lensfun automatic lens corrections applied. Vignetting is not automatically corrected for this lens and I have not tried to manually correct it.
Yech, that’s annoying unless the camera recorded some sort of metadata (some cameras will indicate if the JPEG is the result of the crop) telling how much it cropped.
It is definitely more than just distortion and vignetting correction. Look at the hay bales near the edge of the right side of the image. There’s a single hay bale in the mid-rear, with two bales to the right of it at the treeline in the JPEG. In the distortion-corrected image derived from the RAW, there are an additional two or three bales visible. 
In the demosaic section you can change the border (just increase the value by 4) then you should have the same setting as the OOC jpeg.
Hallo Stampede,
do you mean with “tone curve in the DCP file” this:
micha
No. A DCP file is camera-specific and describes the unique color properties of your camera sensor. I installed Adobe DNG converter on a Windows machine, because it comes with a lot of DCP files. Then I found the DCPs for the cameras that I own and saved them where I could find them easily.
The first step in me developing a raw, is that I go to the Color tab in RT, then choose “color management” and select the input profile.
Hit the “tone curve” checkbox, and it gives you a very good starting point. You still have the two tone curves in the exposure tab that you can use to fine tune your image.
Hope that helps.
Edit, I should add, that before doing this, I apply the “neutral” processing profile. Because RT will default to applying its own tone curve. If you then add the DCP’s tone curve on top of that, it will look awful. I suppose you could also just manually revert the auto-matched tone curve also.
Hallo Stampede,
oh, yeah, that helps me understand your approach. I also used to work with Adobe’s dcp. It produces practically the same colors as the jpg of the camera. But the profile of RT makes the colors more realistic. At least that’s how it is with my my Panasonic DC-G9. I think this RT dcp was made by Mrogan Hardwood, with shots of a colorpicker of mine. Strangely enough, RT’s dcp is four times bigger than that of Adobe (=Panasonic DC-G9 Camera Standard.dcp).
I have to see what will give me better results in the future. I first decided on the RT version because it reproduces the colors better, especially in the blue area.
I want to summarize what I understood: first select the right camera color profile (dcp), then select Tone curve and then work with the curves. Yes, that sounds good, but lately I’ve left out the “Tone curve” checkbox because I thought I could work more freely with the curves. Now I see that the Adobe dcp works fine, but you should definitely turn on “Tone curve” and “Look table”. Perhaps also “Baseline exposure”.
When I select this setting, I have almost no more work setting the correct tone values, even without Auto-Matched Tone Curve.
thanks
micha
@micha Iirc, the tone curve in all dcp files shipped with RT is the same as the Standard Tone curve in the Standard Film curve profiles.
But that does not mean, using auto-matched makes no sense, as it gives a better match to the embedded jpeg in case you used different in-camera settings to produce the embedded jpeg if you want that behaviour.
The DCP is camera-specific, but the tone curve in the DCP is not. It’s generic.
http://rawpedia.rawtherapee.com/Color_Management#Use_DCP.27s_Tone_Curve
Disabling the DCP tone curve when using the auto-matched tone curve is the sane thing to do (that’s what the bundled PP3 profiles do), but I wanted to mention that the auto-matched tone curve is calculated in the pipeline after the DCP curve, so it mitigates the effects of the DCP curve, though not entirely.
Happy to hear that.
Hallo Morgan_Hardwood,
is the dcp from Panasonic DC-G9 (and the Panasonic FZ1000 and Fujifilm X-T10) in RT (Auto-matched camera profile) really made by you with my photos? That pleases me as well, of course.
Yes, the colors are always more natural, more realistic and better than those of the Adobe dcp. I’ve tested this several times. Especially the blue colors, also the faces at Adobe become unpleasantly redish, similar to the jpg of the camera.
The tonal values of Adobe’s dcp are very good, even without A-MTC, but with Tone curve + Loot tabel, but without Baseline exposure.
If I want to achieve the same good tonal values with the RT dcp, I have to switch on A-MTC or adjust it manually.
Do you perhaps have the possibility to adjust the tone curve of the RT dcp in this sense, or perhaps to read it out of Adobe’s dcp without losing the better colors?
If you need new shots of a colorpicker from me, I can do some. If necessary, tell me which colorpicker is best for you, I would get it from somewhere or maybe buy it. If you need the old recordings again, I still have them.
And another question: Do you know why RT’s dcp is four times bigger than Adobe’s?
micha
Hallo Ingo,
kann “Auto-Matched Tone Curve” die Tonwerten kaputt machen, was man dann manuell kaum noch korrigieren kann?
Oft macht A-MTC die Tonwerte sehr gut, manchmal aber zu hell. Ist es dann sinnvoll, diese mit der zweiten Kurve manuel wieder herunterzuregeln, oder ist es besser, in diesem Falle A-MTC einfach aubzuschalten, Neutral zu wählen und dann manuel mit den Kurven die passende Helligkeit zu schaffen?
can “Auto-Matched Tone Curve” break the tonal values, which can then hardly be corrected manually?
Often A-MTC makes the tonal values very good, but sometimes too bright. Does it make sense to manually adjust them down again with the second curve, or is it better in this case to simply switch off A-MTC, select Neutral and then manually create the appropriate brightness with the curves?
micha
In this cases I try to reduce the exposure. I often underexposure intentionally (to not blow the highlights). The AMTC can not correct for this underexposure as the embedded jpeg in this case is also underexposed. In this case I just increase the exposure in RT by the amount I underexposed in camera (* -1). That should also work for overexposed images.
wie ich Dich verstehe, hast Du keine Bedenken mit A-MTC anstatt mit Neutral zu arbeiten? Ist es sinnvoll, wenn man A-MTC wählt, bei DCP auch noch “Tone curve” zu wählen? Beide machen ja wohl sehr ähnliches, wobei aber A-MTC stärker korrigiert.
as I understand you, have you no reservations about working with A-MTC instead of Neutral? If you choose A-MTC, does it make sense to choose “Tone curve” for DCP? Both do very similar things, but A-MTC corrects more.
micha
If you want to be independend on the style you chose in camera for the jpeg, you should not use AMTC. If you want to mimic that style, AMTC is closer to the jpeg than the dcp tonecurve
ach, wunderbar. Jetzt verstehe ich, A-MTC ist eine große Hilfe, wenn man den Style des Kamera Jpeg erreichen will. Das ist bei mir aber nicht der Fall. Dann braucht man es nicht wirklich.
Ich danke Dir sehr.
Oh, wonderful. Now I understand, A-MTC is a great help if you want to achieve the style of the camera Jpeg. But that’s not the case with me. Then you don’t really need it.
Thank you very much.
micha
@micha Let’s don’t waste time to write the same things in german and english.
Your english definitely is good enough to understand 
I will answer in english anyway, because that’s the language most users understand.
it’s hard for me to imagine your camera really tends to overexpose. The Panasonic G9 photos are pretty underexposed by default. You can see that very well if you set RT to Neutral and select only Base table in DCP. Or click “Show raw histogram”.
Surely you know the theory: ettr, so as not to lose tonal quality unnecessarily in the shade.
You can upload a normal exposed photo, I’m curious whether it really tends to eroded lights.
micha
Hallo Ingo,
Your photo is truly not overexposed. If you select Neutral and select RT Atuo-matched camera profile and in DCP only Base tabel, there is still a lot of space in the histogram on the right.
If you additionally select DCP-Tone curve, the image will be brighter, but still good.
And if you choose A-MTC instead of Normal, it will be a bit brighter, but also ok.
But of course you know all that and I, as a RT beginner, don’t have to tell you.
However, if you upload a test shot of a normal motif (please without white sky), we could talk about it more precisely if you like.
And this is how a correctly exposed photo of a Panasonic G9 looks like.
RT: Normal, Auto-matched camera Profile, DCP only Base line.
I can’t imagine that a Nikon exposes that much differently than Panasonic or others.
micha
The sky is blown in raw without any doubt.
Yes, the sky is without drawing. You can see it in the histogram - but isn’t a white sky against relatively dark walls unavoidably too bright? Is it really necessary to get drawing into it? How much would you have to underexpose there?
