Tuesday, October 1, 2024

What is "Unsharp Mask", anyway?

Also in this issue:

  • Just read it, OK?

What in the World is "Unsharp Mask"?

I just finished giving the seminar in Kansas City, KS - what fun that was!!  People came from several states away to attend, which is always gratifying.  A few days before that I gave a lecture to a local photo club on Wireless Flash, with the goal of getting people to be less intimidated by this very useful technique.

Gave a talk at the Digital Dimensions and Beyond photo club in Kansas City, KS.

During the seminar's segment on Photoshop I received an unusual amount of good questions regarding the most unintuitive name ever given to an editing tool: "Unsharp Mask".  

What is it??  How was its name derived?  And how does it work?

I explain all in this excerpt from the streaming version of The Friedman Archives Seminars:

Historic note: I got all those Rolex watches when I was teaching English in China.  I was able to get the price down to $20 each.  Here's the full story about how I was able to negotiate such a low price.


Upcoming Projects

The Photo Safari in Kenya had been sold out but one family had to drop out, resulting in two openings (one tent) becoming available.  But you'll need to act quickly - the actual event goes from November 8 - 15th.  Lots of details at https://FriedmanArchives.com/safari, and you can sign up on that website as well.  You've already got a great camera; now you can give it something to do!

After the Kenya trip I'll be traveling to a neighboring country for a week to take on a most unusual photography assignment.  I'll tell you all about it when we return!

Also, this month I'll be working on a free supplement for the Sony A7R V, covering the new features in the latest firmware update, plus an updated Chapter 4 to cover the new Creators' App software.


Real Email From Real Seminar Attendees

Hello Gary;

I can’t express adequately how helpful/ informative/insightful your two day workshop was for me. As you said yourself, the materials you covered were, in effect, a fire hose of info. But your artistic/ engineering approach, tinged with your wit, helped me get a much better handle on my camera and my photography. And the first day on Light and composition were terrific. 

I couldn’t have spent my time more wisely, even coming in from Pittsburgh.  Thank you and I look forward to seeing you in Kenya!!

and this from another attendee:

Hi Gary,

It was totally worthwhile for me to travel from New Orleans to Kansas City for your seminar last week. The way that you present ideas speaks to everyone in a manner that is completely understandable to all. And you augment your concepts by assigning tasks that reinforce what you are teaching. 

Feel like you're missing out?  Your local photo club can hire me to either give zoom lectures, or to bring the seminar out to where you are.  Not a member of a photo club?  The streaming version of the Friedman Archives Seminars is available for you to watch in the comfort of your home.  So many options.  :-)


Upcoming Lectures and Workshops

Oct. 7th – Lynn Photo Club 6 PM – Talk on Computational Photography (hey, it’s a popular topic!)  564 Boston Street, Lynn, MA 01905.

Oct. 9th – 7 PM talk at Duxbury Camera Club – also Computational Photography.  77 Alden St.  Duxbury, MA 02332  Ask Mari Ryan Photos mari@mariryanphotos.com permission to attend since you’re not a member.  

Oct 15th – Computational Photography talk over Zoom (for a club in the UK; but you can attend – who’s going to know? 😊 )  Email me for details.

Nov. 3rd – Lynn Photo Club 1 PM – Wireless Flash Workshop.  564 Boston Street, Lynn, MA 01905.  There’s a fee.  Contact 978 664-2620  Susan Mosser  cell 978 337 3258  S.JMosser@comcast.net if you’d like to attend.


Next Time in CameraCraft Magazine

In the upcoming issue I interview Mary Doo, an accomplished wildlife photographer who seems to find perfect light wherever in the world she goes.  Subscribe to CameraCraft Magazine today for content and insights you just won't find anywhere else!

Update on MIT Project

Many of you may remember my blog post from a couple of years ago, talking about my volunteer work at MIT and how I was totally impressed with this high-school program that got kids to do things you wouldn't expect high school kids to be able to do: programming a quantum computer, building an automous air vehicle with ground-penetrating radar, or building a prototype cubesat.  (Entire blog post is here - it's long but worth a read.)  

So let me bring you up to date with my MIT activity by sharing a letter I wrote to a fellow volunteer who was equally enamored with the program, so much so that he vowed to put together a pitch deck and take it around to all the aerospace companies to help fundraise, like I wanted to do.

===

Hi, Michael!

I was elated to hear that you were able to see first-hand how BWSI was able to positively impact your son this summer.  I started out at BWSI as a volunteer mentor (I'm still a volunteer, actually) for the Build a Cubesat course.  Like you, I was amazed and impressed at the program, starting from the ambitiousness of the material and the giftedness (that's a word!) of the course designers and instructors.  This isn't just some vanilla STEM course!!

And, like you, my enthusiasm for the courses was so high that I went to Bob [Shin, the founder and director] and asked how I can help the organization.  Can I help design more classes, or coordinate with professors and work on courseware?  Can do boring administrative things to allow more students to participate?   15 minutes into that conversation it became clear that Bob has access to all the smart people he needs.  What he really needed was more funding so they could scale.  I have no background in fundraising (I'm a NASA veteran, truth be told) but I told him, without making any promises, that I would try my best.

That was about 3 years ago, and like you, I had visions of visiting various companies, giving enthusiastic talks about how great the program was and soliciting funding.  I have approached maybe 40 major aerospace companies, sometimes having live phone calls with representatives.  So far, I've been unsuccessful, but in the process I've discovered a glut of invisible roadblocks which has soured me on this approach.

My first thought was to approach the HR departments of aerospace firms: "Would it benefit your company to have more systems-level thinkers apply for your advertised positions (or as summer hires)?"   It turns out that HR departments are more concerned about avoiding lawsuits from disgruntled employees than about hiring quality (that's up to the manager interviewing the candadite), so they don't have budgets for this sort of thing.  And they were not interested in putting me in touch with those who did.  (Just as well, the folks who advertise job openings are bureaucrats who don't think about competitive edges for the company.)

Then I pivoted to the corporate Aerospace foundations.  Surely they would see the value in this program and enthusiastically fund it, for that elusive competitive edge!!  Nope.  All of them have specific goals for that money, earmaked to address a specific, predefined, unfulfilled needs: Not enough women engineers entering college.  Not enough people from a specific geographic region going into engineering.  Not enough LGBTQ interest in engineering.  Things like that.

The BWSI program doesn't get so easily pigeonholed.  YES we can accept students from all of those backgrounds, but what we do probably better than anyone else is we turn good engineers into GREAT systems engineers, regardless of their backgrounds, geographic location, or DEI status.  These foundation administrators get unsolicited proposals all day long; if it doesn't fit like a glove it gets rejected, regardless of merit and regardless of the potential benefit to the founding company.  Again, no systems-level thinking (which is kind of ironic, since this course promotes systems-level thinking.)

And consider this dismaying observation: Let's say you're in the position to grant funds to an organization, and you have three proposals sitting on your desk:

  1. "If you don't fund our animal shelter program, thousands of innocent animals will have to be euthenized!"
  2. "If you don't fund our homeless veterans program, thousands of veterans who selflessly served their country will go hungry!"
  3. "If you don't help us scale this outstanding STEM program, hundreds of talented high school students will go to college and gratuate and go on to have great aerospace careers anyway". 

Which proposal will you fund??  Rightly or wrongly, this program is seen as help for privledged kids that doesn't solve a pressing pain point.  (And the fact that it's from MIT, which people think is the equivalent to Harvard in terms of well-endowededness (that's a word!), doesn't help.)

You and I know from the ground level how great BWSI courses are.  But from the 30,000 foot level (the administrators who grant funds), it's just another STEM program in a literal sea of STEM programs.  You have to experience it from the ground level to appreciate just how unique and powerful it is.

So that's been my experience; and I've now shifted (yet again) to concentrating on government grants like this one since it so clearly addresses the customer's stated pain point of not enough experienced engineers in the employment pipeline.  Statistics say the more I try the more likely I'll be successful.  But so far, nothing.

===

So now you know.  Anyone know of a generous and forward-thinking funding source for such a dramatically transformational program??

Until next time,

Yours Truly, Gary Friedman

Carol and I were invited by two members of the Digital dimensions photo club to shoot some interesting landmarks in Kansas City.  You can see the entire collection here.


9 comments:

  1. Thanks for explaining the antecedents of unsharp-mask. When post-processing images with GIMP, I use it more often than not after applying a white balance and stretching the HSV contrast. Usually the sequence improves the image, but if not then it's easy to undo. Having taken the default values, I will now try your recommended settings, keeping in mind that I'm using GIMP not Photoshop with the possibility of different results.

    ReplyDelete
  2. I can now understand why some of my stereo pairs can make a good, maybe better, virtual image when one of the real images has a slightly different focus. The same principle is operating as when using Unsharp Mask in PhotoShop. Regards, Harold

    ReplyDelete
    Replies
    1. I never thought of improving the "3D effect" by applying unsharp mask to only one of the two 3D images. Thanks for the tip!!

      Delete
  3. Thank you for being the the photography guru that you are. With the increase in resolution of video cameras, the ability to extract photos from videos, and the ability to set shutter speed and or the aperture in video or movie mode, I am wondering if I should consider shooting in movie mode more that images on my DSLR . I am also curious about what cine cameras do that video cameras don't do.

    ReplyDelete
    Replies
    1. There's no correct answer there; however I can tell you that I've been using the "Shoot an 8K video and grab a freeze frame later" technique a lot recently. I used it back in May to get an impossible shot at my granddaughter's Bat Mitzvah (https://friedmanarchives.blogspot.com/2024/06/a-month-on-road.html , toward the end), and in Kansas City I used it to try to capture frames with lightning bolts (unsuccessfully - but I was able to get some cool cloud illuminations ( https://photos.app.goo.gl/QZqe7ZWK8HxhfT3Y6 ). Cine cameras give the videographer significantly more control over the variables; and a full list would require more space than a blog comment would allow.

      Delete
  4. Why does an unsharp mask work? Simple. It emulates our visual system heuristics for sharpening dots, lines, and object boundaries. That happens in the retina and the bundle of nerves that sends the jittery video into our brains for further processing. For every point (really photoreceptors, but you can think pixels), send a negative signal to surrounding points. This is the same as subtracting a blurred image from the original. The result is a darkening of points surrounded by bright points and vice-versa. So, if a dark object is superposed over a light background, its boundary will be darker, and vice-versa. It is called center-surround. It works primarily for luminance but also for color. Painters have used this technique forever. Check the area next to the hair and collar for The Laughing Cavalier by Franz Hals [1964, Wallace Collection.] DaVinci refers to it in his Notebooks.

    ReplyDelete
  5. With the advent of AI sharpening in apps such as DXO Photo Lab and Topaz Photo AI Unsharp mask is obsolete.

    ReplyDelete
    Replies
    1. Well it's true that there are new tools that can make the job easier, but I've learned that you get the best results when you taylor the sharpening to the images final use - different values depending on whether they'll be printed or viewed on the screen.

      Delete

Thank you for your comment! All comments must be approved by a moderator before they will appear.