Category Archives: Tests

Post-processing Workflow, Software, Tests

Does Adobe Dream of the Moon?

Image Leave a comment

Yesterday while I was editing a raw file in Adobe Camera RAW (ACR) I used the Generative AI Remove feature to deal with a shadow created by some dust on my sensor. It removed the spot from the sky in the image but inserted a roundish object in its place. I remember thinking, “That’s odd? Did ACR just add a 3/4 moon in place of the dust spot?” But I then forgot about it and continued editing.

Today I again used the Generative AI Remove function to remove another faint dust shadow in the sky of another photograph from the same shoot. It happened again! I think Adobe Generative AI may be hallucinating!

Above are two tiny, highly magnified sections of the original image. In the upper one, if you look carefully, you can see a sort of faint, darker smudge behind the individual light-colored bird above the line of darker birds. I used the brush to outline the darker area, then had Generative AI Remove take it out.

The lower image was the result. It is faint, but I swear that it replaced the light colored bird with… the moon!

That’s just plain weird.

G Dan Mitchell is a California photographer and visual opportunist. His book, “California’s Fall Color: A Photographer’s Guide to Autumn in the Sierra” is available from Heyday Books, Amazon, and directly from G Dan Mitchell.

Blog | About | Instagram | Flickr | Facebook | Threads | BlueSky | Email

Links: Articles, Sales and Licensing, my Sierra Nevada Fall Color book, Contact Info.

Scroll down to share comments or questions. (Click post title first if viewing on the home page.)

All media © Copyright G Dan Mitchell and others as indicated. Any use requires advance permission from G Dan Mitchell.

Equipment, Tests

Format Size Comparison Data

Image Leave a comment

The following charts summarize various data that are useful when comparing formats, either film or digital. (Not all possible comparisons are shown.) Some additional background regarding some of the traditional film formats known as “medium format” is found at the end of this post.

Basic Formats Data

Comparisons Normalized to Full Frame Format

Comparisons Normalized to miniMF 33×44 Format

Comparisons Normalized to 645 Medium Format Film

Comparisons Normalized to Full Frame at 4:3 Crop

Comparisons of Square Formats Normalized to 6×6 Medium Format Film

Comparions Normalized to 6×17 “Panoramic” Medium Format Film

Visual Comparison

The following illustration shows the relative sizes of some common digital and film formats. (Not all variations are shown.) The leftmost group includes common digital formats. The next group — indicated in yellow — includes common traditional formats collectively known as “medium format” film. At the far right is 4×5 “large format” film.

A chart showing the relative sizes of several digital and film photography formats ranging from micro-four-thirds to 4x5 film
A chart showing the relative sizes of several digital and film photography formats ranging from micro-four-thirds to 4×5 film

Notes on this illustration:

  • 645 film is shown in both the “landscape” and “portrait” orientations in order to make clear that it shares the 56mm dimension with the other types of medium format film. (The portrait version is partially hidden the landscape version.)
  • There are multiple variations on “large format film.” Shown here is the most common 4″ x 5″ size, the smallest of the large format film formats typically still in use.

About Medium Format Film

While those of us who have “done photography” for a long time are familiar with the traditional “medium formats,” those without that perspective may be less (or not at all) familiar with the meaning of the term. Medium format” traditionally (for many decades) referred to film formats using (primarily) 120/220 film with a 6cm (60mm) physical width. While details of these formats are included in the tables above, here is a brief contextual overview:

  • 645 format (1) is the smallest of the traditional common film medium formats. It is named based on “6cm x 4.5cm.” In actual use the frame size is slightly smaller than the 6cm film width at about 56mm. (This is typical of film medium formats — the “6” always refers to the physical film width, not the smaller actual image size.)
  • 6×6 format produces a square image of approximately 56mm x 56mm, and is the next larger film medium format above 645.
  • 6×7 format uses the 56mm width-determined dimension for its shorter side, and thus has a larger area than 6×6. (It approximately replicates the 5:4 aspect ratio of LF film.)
  • 6×9 format also keeps the 56mm dimension its shorter side, but expands the longer dimension to produce a 3:2 aspect ratio – like 35mm film and full frame digital.
  • 6×17 format (also known as “panoramic format”) is the largest common medium format. It also uses the 56mm dimension for its short side but greatly extends the long dimension to produce a 3:1 aspect ratio.

(1) While Pentax names its miniMF digital cameras using the “645” term, they do not use 645 format sensors. They use 33mm x 44mm sensors, just like Fujifilm, Hasselblad, and others. Additional note: On a few occasions, Fujifilm has referred to 33mm x 44mm format as “super full frame.”

(For comparison purposes, note that the dimension that is 56mm on 645 film format is 44mm on miniMF. When comparing to the other medium formats, the analogous comparison is between 56mm and 33mm.)

G Dan Mitchell is a California photographer and visual opportunist. His book, “California’s Fall Color: A Photographer’s Guide to Autumn in the Sierra” is available from Heyday Books and Amazon.

Blog | About | Flickr | Facebook | Email

Links to Articles, Sales and Licensing, my Sierra Nevada Fall Color book, Contact Information.

All media © Copyright G Dan Mitchell and others as indicated. Any use requires advance permission from G Dan Mitchell.

Equipment, Tests

Some Camera Format Numbers

Leave a comment

I created the following little spreadsheet last night. I think it provides an interesting way to think about comparisons among various camera formats. It uses the familiar concept of “crop factor,” but it moves the baseline to two other format besides just full frame.

I won’t say more at this point, but you might be able to draw some inferences about various format options if you spend some time considering these numbers. (If the chart is hard to read, click it to open full size.)

Formats Compared Via Normalized Crop Factors
Film and digital camera formats using crop factors normalize to full frame, mini MF, and 645 medium format film.

See top of this page for Articles, Sales and Licensing, my Sierra Nevada Fall Color book, Contact Information and more.

G Dan Mitchell is a California photographer and visual opportunist. His book, “California’s Fall Color: A Photographer’s Guide to Autumn in the Sierra” is available from Heyday Books and Amazon.
Blog | About | Flickr | Twitter | Facebook | LinkedIn | Email

All media © Copyright G Dan Mitchell and others as indicated. Any use requires advance permission from G Dan Mitchell.

Equipment, Tests

Fujifilm GFX and Canon 5DsR part II

2 Comments

Earlier I shared the beginning of an article that compares the Fujifilm GFX 50S and the Canon 5DsR, two cameras with different formats but approximately the same sensor resolution. That article (“Fujifilm GFX and Canon 5DsR” wasn’t complete … and now I’m back with one more point of reference.

When it comes to resolution, making judgments based on screen images has its limits. For example, 100% magnification crops “show” you things that won’t be visible even in  a very large print, and in many cases you must interpret what you see in the screen image in order to speculate about what the print will look like. As we say, “The proof is in the print.”

With that in mind, I have prepared some files that may be useful for comparing print resolution from Fujifilm GFX 50S and the Canon 5DsR. These files each hold three “test strips” labeled A, B, and C. The files have been formatted for producing prints with a print area of 10″ wide and 6″ tall (use letter-size paper) when printed at a resolution of 360. They are sRGB color files in the highest quality (Photoshop: 12) jpg format.

There are four files. Let’s call them:

(Click on the links to see and download the files.)

IMPORTANT NOTE: When first posted, these links went to downsized files rather than the full size print-ready files. This was corrected on 1/21/18. The print areas of these files should be 3600 pixels on the long side, for printing 10″ long at 360 ppi.

Viewing them on screen is not the object here — in fact, for reasons I won’t go into here, trying to evaluate that way is of very limited value and may mislead you. (For example, the files have been optimized for print in ways that may make them look poor on screen at 100% magnification.)

You will need to print.

Since there are four samples, you’ll need to make four letter-size prints. The correct settings will probably be entered automatically, but check that the print area is 10″ wide and 6″ tall and that the resolution is set to 360.  As you download and print the files you may wish to make a note on each (in pencil or ink) indicating which file it is — sample 1, sample 2, sample 3, and sample 4.

Observe each print carefully. Note anything you can see related to the resolution of each printed strip A, B, and C. (Note that this is only a resolution test — don’t worry about things like image size, brightness, color, and so forth.)

In order to avoid confirmation bias I’m not going to reveal anything else about the files at this time. Once a few people (or a lot of people?) have weighed in with their observations — please use the comments below — I’ll reveal more important information about what you are looking at.

G Dan Mitchell is a California photographer and visual opportunist. His book, “California’s Fall Color: A Photographer’s Guide to Autumn in the Sierra” is available from Heyday Books and Amazon.
Blog | About | Flickr | Twitter | Facebook | Google+ | LinkedIn | Email

All media © Copyright G Dan Mitchell and others as indicated. Any use requires advance permission from G Dan Mitchell.