Case Studies start next week


Dan Margulis
 

Whether color correction, color grading, or retouching, our 2021 series of case studies will begin next Monday, with the posting of an image for which entries will be due Monday, 1 February. Assuming that we continue to get at least 15 entries per study, this will continue for eleven weeks.

As with our 2020 set, this group was chosen to offer a variety of challenges. The ordering is deliberate in that the lessons of one study can often be applied to a subsequent one.

On the whole I’d say this grouping is slightly more difficult than the last one, although I don’t think that any one image is as tough as A Toast to Greece was. My prediction is that six of the ten will draw several really poor entries, as opposed to four in the last set. The usual cause is challenging lighting conditions for the photographer.

Two of those problem children (in the last set) were Veiled Bride and Niagara Spray, which happened to be the first two we did. That was a bad idea in retrospect. This time, the first image we work on is the easiest of all, IMHO.

The initial group of ten consists of
2 portraits, in both of which the garments are at least as important as the faces.
2 building interiors.
2 images where half is grossly too dark, one from an amateur using an iPad, the other from a famous photographer.
1 shot of an animal.
1 exterior shot of a renowned hotel at night.
1 outdoor concert with less than optimal weather conditions.
1 scenic with an unusual challenge.

In seven cases the raw capture is available. The others: one was scanned film, a second is the one from an iPad, and a third is from a research project where there was a specific reason that the competitors were denied access to the raw.

Four of the images come from the MIT series that constituted half of our 2020 set. Two were previously used in my advanced ACT classes. The other four have never been submitted to a group before AFAIK.

The 2020 set started out with an eleventh challenge, but one rarely seen by group members. It featured a brilliantly red costume, which annoyingly had to be prepared for CMYK, where matching such reds is not possible so we had to make the best of a bad situation. In honor of that exercise, one of the images in this set also features a brilliantly red costume, although it doesn’t have to go to CMYK. And in honor of that exercise, there will be an eleventh, voluntary exercise after we finish the first ten. Let’s just say it’s a type of work that you would be most unlikely to encounter, but you never know.

The rules will be the same as last time, except there is no longer any limitation on sharpening. To remind you of some of the traps:

*The files will be posted each Monday and are due back the following Monday morning at 06:00 Eastern US time. If the file is available in raw format, I will pull a default and a flat version and post them. If you want the raw file itself, I’ll be making a .zip that contains it and the two default versions.

*If you are using the raw, the first thing is to make sure you’ve acquired it at the correct size, which is specified in the text message announcing the study, or you can open the default versions and check Image Size. If you are opening it with any module capable of a lens correction, before launching a correction apply it to one of the default versions to be sure they are perfectly in alignment.

*When I receive all images, I will quickly choose what look like five of the best (excluding those with highly eccentric treatments) and average them to create a “par” version. Please make sure that you don’t do anything that would prevent your image from being averaged with the others: no cropping, rotating, or deletion of any objects, please.

*Whether color correction, color grading, or retouching, our 2021 series of case studies will begin next Monday, with the posting of an image for which entries will be due Monday, 1 February. Assuming that we continue to get at least 15 entries per study, this will continue for eleven weeks.

As with our 2020 set, this group was chosen to offer a variety of challenges. The ordering is deliberate in that the lessons of one study can often be applied to a subsequent one.

On the whole I’d say this grouping is slightly more difficult than the last one, although I don’t think that any one image is as difficult as A Toast to Greece was. My prediction is that six of the ten will draw several really poor entries, as opposed to four in the last set. The usual cause is challenging lighting conditions for the photographer.

Two of those problem children (in the last set) were Veiled Bride and Niagara Spray, which happened to be the first two we did. That was a bad idea in retrospect. This time, the first image we work on is the easiest of all, IMHO.

The initial group of ten consists of
2 portraits, in both of which the garments are at least as important as the faces.
2 building interiors.
2 images where half is grossly too dark, one from an amateur using an iPad, the other from a famous photographer.
1 portrait of an animal.
1 exterior shot of a renowned hotel at night.
1 outdoor concert with less than optimal weather conditions.
1 scenic with an unusual challenge.

In seven cases the raw capture is available. The others: one was scanned film, a second is the one from an iPad, and the third is from a research project where there was a specific reason that the competitors were denied access to the raw.

Four of the images come from the MIT series that constituted half of our 2020 set. Two were previously used in my advanced ACT classes. The other four have never been submitted to a group before AFAIK.

The 2020 set started out with an eleventh challenge, but one rarely seen by group members. It featured a brilliantly red costume, which annoyingly had to be prepared for CMYK, where matching such reds is not possible so we had to make the best of a bad situation. In honor of that exercise, one of the images in this set also features a brilliantly red costume, although it doesn’t have to go to CMYK. And in honor of that exercise, there will be an eleventh, voluntary exercise after we finish the first ten. Let’s just say it’s a type of work that you would be most unlikely to encounter, but you never know.

The rules will be the same as last time, except there is no longer any limitation on sharpening. To remind you of some of the traps:

*The files will be posted each Monday and are due back the following Monday morning at 06:00 Eastern US time. If the file is available in raw format, I will pull a default and a flat version of each and post them. If you want the raw file itself, I’ll be making a .zip that contains it and the two default versions.

*If you are using the raw, the first thing is to make sure you’ve acquired it at the correct size, which is specified in the text message announcing the study, or you can open the default versions and check Image Size. If you are opening it with any module capable of a lens correction, before launching a correction apply it to one of the default versions to be sure they are perfectly in alignment.

*When I receive all images, I will quickly choose what look like five of the best (excluding those with highly eccentric treatments) and average them to create a “par” version. Please make sure that you don’t do anything that would prevent your image from being averaged with the others: no cropping, rotating, or deletion of any objects, please.

*DO NOT POST YOUR ENTRY TO THE LIST. Send it to me personally, at DMargulis (at) aol.com. Be sure that they are Level 9 JPEGs, tagged sRGB (you may work on them in any space you like, but must convert to sRGB when finished) and that your e-mail client isn’t configured to downsize them automatically.

*Immediately before posting I will identify each version by random-number generator. If you have put your initials or name in the title of your file, I will delete them. If you wish to be put a few extra letters or whatever so that you can quickly find your file once it is posted, those will remain.

*Each Monday morning I will post a .zip containing all the entrants. I will then open a thread for discussion by list members. I won’t join it for about 48 hours. The discussion will be by number; names won’t be associated with the entrants unless participants in the thread decide to identify their own versions publicly.

*Shortly after posting the results each Monday morning, I will post the next case study, due the following week.

*MOST IMPORTANT OF ALL: when you submit your version, include a step-by-step of how you achieved it and, if appropriate, why you did things that way. If it turns out that you have one of the best results, we are all going to want to hear your reasoning.

The 2020 studies were a good learning experience for all. I hope this set will be at least as interesting.

Dan


Join colortheory@groups.io to automatically receive all group messages.