Added "Travels" section and Grand Canyon gallery

So I've added a new section called "Travels" (because Sara complained that I didn't have a place to put pictures up when I travel) and a gallery for our raft trip down the Grand Canyon.

There will undoubtedly be rearrangement as time goes on.  But these changes make it much more likely that I will do regular updates on the photo section.

So yay!

Pure White Backgrounds

There's a school of nature photography exemplified by Meet Your Neighbors that is natural objects (plants, insects) against a pure white background.  I've been interested for a while and I thought I'd give it a try.

Nikon D810 with Sigma 180mm f/2.8 macro; 0.4 sec, f/29, ISO 200, +2 EV, lighting described in main text, no crop

This is my third serious attempt (that is, one image taken from my third serious attempt to shoot like this).

The first was a proof of concept shot against my Wacom tablet with a pure white background on it.  Because the tablet requires power and lives on my desk, I literally picked a flower, tossed it onto the tablet, and took a shot sitting at my desk.

It kind of worked.  There was a bit of work involved in Photoshop selecting the entire background and blowing it out (because the Wacom wasn't bright enough to blow the background out in the original exposure); the Wacom isn't easily taken into the field; the office light front lighting the flower was too harsh; and so on.

The second attempt used a commercial tracing tablet with LED backlighting for a background. This particular tablet is USB powered, and has an internal battery.  I took it "into the field" (by which I mean my front yard) and attempted to use it as a backdrop.  I had better luck with the rear lighting, but without a strong mount for the tablet (ideally to a tripod), it wasn't going to work.

Today's third attempt was again done inside, using the same tablet hooked up to a USB charger battery pack and supported by cardboard boxes indoors.  I also added a couple of plug-in LED lamps I had lying around for front light.  It's important to place these lights as close to the subject as possible to diffuse and minimize shadows cast on the background.

I experimented with an exposure range from +0 EV to +3 EV, and eventually settled on +2 EV for this example.  Ideally the background is completely blown out to 255,255,255, but only just. The background on this image was probably 95% properly exposed, but I had to use color range selection in Photoshop to select almost the entire background and then touch it up using the lasso to get the last few pixels.  Not more than a couple of minutes work.  Levels then pulled the detail back into the subject and that was all I needed to do.

I spent some time playing around and it's really easy to use Contrast or Color layers to make a lot of interesting variants.

I've got some more lights on order with soft boxes to try this technique in the field and I'm looking forward to seeing what I can get!

Printing 2 of X: Renaming (a lot of) Color Profiles on the Mac

Previous post in this series                                                                            Next post in this series


With paper comes profiles.  With sample packs of paper come LOTS of profiles.  Profiles which invariably have unpleasant names, like "HFAPhoto_CanPro1000_PK_HahnemuehlePhotoSilkBaryta310" which neither rolls trippingly off the tongue nor makes it easy to pick "Hahnemuehle Fine Art Photo Silk Baryta 310 for the Canon PRO-1000" off of a menu.

Renaming the files is easy using Better Finder Rename, a mac utility I've been using for at least a decade.  It brings a step-by-step interface to simple and complex renaming of multiple files.  Well worth it.

But renaming the profiles isn't as simple as renaming the files.  ColorSync utility will let you do it, and there is a script provided to do the renames, but renaming dozens of profiles at once is a pain.  To use ColorSync directly, just double-click on the profile file, and then in ColorSync click on the 'dscm' Apple multi-localized description strings field if present and edit it, or click on the 'desc' field and edit it.  So far as I can tell, the 'dscm' field will take priority if it is present.  If you prefer, there is a 'Rename' script in the 'ColorSync' portion of the menubar script menu, which you turn on using the 'Preferences...' menu item of the 'Script Editor' app in the Utilities folder of the Applications folder.  But using either to rename a bunch of profiles remains a big pain.

So I wrote a shell script which renames profiles to their base filename.  It's crude and simple, and will probably break for other people, but here it is:


for file in ./H*
    BASENAME="sips -s description '$(basename -s .icc $file)' '$file' "
    echo 'executing $BASENAME...'
    eval $BASENAME

Hope someone finds it useful.

I rename the files to roughly PaperManufacturer PaperName PaperWeight BlackType Printer, e.g. "Hahn Photo Pearl PK Canon PRO-1000" which still isn't a brilliant name to sing out at night, but at least tells me what I'm looking at.


(Originally posted to my personal blog @ Mischievous Ramblings II)

Printing 1 of X: Testing (and Testing (and Testing (and Testing...)))

Previous post in this series                                                                            Next post in this series

In which our Hero (ahem) seeks out the work of earlier giants, absorbs mentorship, tries and fails, learns that obvious things are obvious even to him (after failing), and muses on the proper order of tasks...

So, learning to print...

It's a big topic.  Here's how I'm doing it.

First, obviously, I bought a printer.  You can find that event in a previous post, already linked above.

Second, I started reading.  The second link, above, is an example of the kind of web search I'm doing, although they are getting more and more specific as time goes on.  I also bought a bunch of books.  Among them are:

I haven't read them all -- hell, I haven't even cracked them all.  You will notice that I prefer ebooks.  They have less mass and volume than physical books.

Third, I started printing.  A manager of mine once said "you don't have time to make all the mistakes other people have made", so I gather other people's stories in order to make my own.  But in order to make my own, I've gotta actually try doing it.

Case in point: I'm starting with mostly black and white printing.  Not because it's easier but because I'm exploring my monochrome side :-).  I've got a variety of papers and a variety of images, I'm trying workflow improvements in Lightroom, testing control surfaces and keyboard overlays, adding color management (because when the images are only appearing on my displays, color is what I say it is, so why manage it?), and so on.  Eventually I observe that my sample prints are consistently far warmer (brownish, basically) than what I see I on the screen.  I can hold a print up to the screen and see the difference.  I find several different ways (eight? nine?) to characterize color management through the workflow path, using Lightroom, Photoshop, ColorSync, and multiple third party solutions.  I try them all.  I re-calibrate my monitors (3 of them!) and my printer.  I improve my physical set up so that I have allegedly D65 lighting above the printer.  I add an allegedly D65 task lamp to my desk.

And that is the magic.  Seen under the task lamp, images match the screen.  Seen under the overhead LED bulbs in my track lighting, images are warmer.

Lessons learned:

  1. Paper alters the visual appearance of a print (duh -- but while I had considered surface reflectance and texture, I hadn't really grokked that white paper has a color -- because I'm an idiot);
  2. Viewing light alters the visual appearance of a print (again, duh -- but I understand that in my gut now);
  3. Perfection is a rathole -- I spent days researching, experimenting, and tracking this down.  The paragraph above does not do justice to the amount of work involved.  I have a new understanding that, while I have lighting in place above my printer and on my desk that is allegedly D65, I am not yet prepared to create spaces and spend money to virtually guarantee consistency of lighting in creating and viewing prints, nor am I yet knowledgable enough to select which of several possible standards (D50, D65) I should choose; so...
  4. For now, good enough will have to do.

The number of details is overwhelming.  I have white cotton gloves to handle paper before it's used (to avoid getting skin oils on it, which may damage the print).  I have sticky notes all over one screen detailing my current (and some previous) procedures to go from (notionally) final image in Lightroom to image on paper.  I am in the process of creating somewhere between 12 and 30 proof copies of the same black and white image for paper tests.  And I will do the same for at least one color image (probably more -- a set of color charts won't suit me in this, so I will probably print images with several color sets).  Those proof prints will have to be protected and stored, because while I expect to settle on a couple of paper types for common use, there will undoubtedly be images I want to treat specially in the future, and I don't want to do sets of tens of proofs over again if I can help it.

Why am I in such a hurry to pick a paper?  Because a friend has indicated he would like a print (actually, he offered to buy it, but I'm giving it to him) and I don't want to lose this opportunity to share my art with someone else.  But I'm geeky enough to want to get it right.  And there is no management chain standing over me telling me how I have to do something, or setting a schedule.  I'm eager to get it done, but I want it right before I release it :-).

Soft Proofing for Printing

A huge element of this process is going from (allegedly) final image to print.  Working the image is a whole thing in itself (and I'm working hard on that too) but going from image to print is a new thing, not an adjustment of an existing thing.

I work in Lightroom wherever possible to do what I want.  In this case, that means I'm working in the Develop module, I'm using Before/After Left/Right Split as my View mode (I set it in the toolbar at the lower left corner of the working window), checking Soft Proofing (also down in the tool bar), and make sure the Before image is the "Master Photo".


 My preferred soft proofing screen, cut down to about 1/4 of my monitor.

My preferred soft proofing screen, cut down to about 1/4 of my monitor.

First thing to do is make a Proof Copy and select the paper's ICC profile in the "HIstogram/Soft Proofing" panel of the Develop module.  I see whether Perceptual or Relative intent better serves my purposes (almost always, for now, I'm ending up with Perceptual) and check what Schewe calls the "look like crap" box -- the "Simulate Paper & Ink" checkbox.  Then zoom to 100%

Why do I check that box?  There are places on the web suggesting you not bother, because Soft Proofing in Lightroom "overdoes" the loss of contrast moving to a print.  Yeah, that's true, and I often find myself looking at the difference between original black and simulated on-paper black and trying to figure out a way to make on-paper black blacker.  But I can't.  But the simulation gives a decent look at the overall color cast of the paper, which matters a LOT to me, relatively speaking.  The difference between white white and warm white (or cold white) in a black and white print is something I want to see.

Generally I find that relatively minor Basic adjustments are all I'm doing here.  Maybe up to 0.3 exposure change, less than 10 Temp change.  Sometimes a bit of Contrast.  Schewe advocates a bit of Clarity to boost the lost contrast, but I'm not there yet.

When I'm satisfied, I jump over to the Print module.  From there, I check the Page Setup... to make sure the Paper Size is correct.  When I have more than one printer, I imagine I'll be checking which printer I'm talking to, too.  If I haven't changed anything, I'm careful to hit the Cancel button because sometimes the OK button makes my panels go away.  No, I don't know why.  I just avoid it if possible.

Printing via Lightroom Print module

From top to bottom, I make sure:

  • I'm on Single Image;
  • Rotate to Fit is checked;
  • Margins are 0.25" each;
  • Cell Size is appropriate for page size;
  • all Guides are shown.

Print Resolution gets unchecked momentarily to check the actual print PPI (upper left corner of the image, but it only appears when Print Resolution is unchecked).  Since I'm working on a Canon printer, I want my resolution to be a multiple of 300.  If the resolution is between 300 and 600, I set it to 600.  I haven't gone higher than that yet.  Then make sure Print Resolution is checked so that Lightroom will resize the image to a multiple of the printer's native resolution.  If it were an Epson, it would be 360 (720) instead.  I should note that Schewe says this is only necessary when working with high-gloss paper and not when dealing with "watercolor" papers.  I infer that smooth surfaced papers are more likely to show a difference here, but don't worry too much about doing the upsampling when unnecessary.  It's just CPU cycles and RAM, and I've got enough of both.  When the new iMac Pro or Mac Pro gets bought, I'll have even more :-).

Print Sharpening gets set according to what I've done to the image before calling it a master.  If I did no sharpening, I set it to High.  If I did capture sharpening, I set it to Standard.  I haven't set to Low yet because I trust Lightroom to do my target sharpening for me (so far).

Media type is set to Gloss or Matte as appropriate.  I call Satine or semi-gloss glossy.

Make sure the profile is set correctly (why doesn't Lightroom make this profile match the soft proof profile if present?  I don't know, but it should). Check the Intent (ditto).

Now the Printer... dialog.  If I'm working off a commonly used paper, I have a preset.  If not, I start with defaults and check to make sure that:

  • Color Matching is disabled and set to ColorSync;
  • Paper Handling does NOT have "Scale to fit paper size" checked and that Destination Paper Size is correct;
  • Quality & Media has the right Media Type and the right Paper Source.  I rarely play with Print Quality so far, and am not yet testing the Black & White Photo Print mode;
  • Paper Detailed Settings has "Cancel Margin Regulation" checked in case I'm doing borderless printing or, really, just to make sure I'm in control of the margins.

At some point, I may set color options for one or more papers, in which case there will be more saved presets, but I haven't seen any reason to do this yet.

Hit the "Print" button in the dialog, and wait...

After Printing

It doesn't really take that long to do a print.  But a print should be let sit for minutes to hours after the ink hits the paper.  ColorMunki software says a minimum of ten minutes for color to stabilize.  Schewe's Digital Print notes that prints are susceptible to scratching for several hours and to avoid stacking prints without interleaf sheets and further advocates waiting 24-36 hours before framing to allow ink to outgas.  Since I am doing test prints right now, this isn't really relevant to me.  If I scratch a print I just have to compare it somewhere it's not scratched.  When I am doing production prints I will search out a source for interleaf sheets and use them.

Many of my prints go upstairs and get put on a table in the living room that gets lots of natural light in order to see them in an alternative viewing setting.  It also gives my wife a chance to peek at them and point out problems she sees that I do not.  My color discrimination is not great, so it's nice to have someone else take a look at them.

And there it is.  The current Testing process.


(Originally posted to my personal blog @ Mischievous Ramblings II)

Printing 0 of X: Introduction to a series.

 First Print on Canon PRO-1000, 4x6, Evan Robinson, scanned 2017-05-11

First Print on Canon PRO-1000, 4x6, Evan Robinson, scanned 2017-05-11

As part of a two-and-a-half week road trip to the Southwest, this year I attended the penultimate Moab Photography Symposium.  The symposium itself (and the Terrific Trio workshop immediately before it) are worth at least one post by themselves, but I want to concentrate on the upshot, which is that I committed printer when I got home.

And in my usual fashion, I went pretty big.  When I walked into Glazer's Camera Supply in South Lake Union (just blocks from my old job at Amazon) looking for a printer, ink, paper samples, and advice, I was 95% sure I was picking between an Epson P800 and a Canon PRO-1000, with the initial edge going to the Epson because of its paper roll support (important for printing panoramas).

After 20 minutes or so of conversation with Dana (the reason I support my local camera shop and try to buy a minimum of 1/3 - 1/2 of my gear from them, in addition to anything I need significant advice about, is this kind of knowledgeable support), the benefits of the Canon were convincing me.  In no specific order, they included: current deals on the printer plus about $400 in free paper, a $300 rebate, separate channels for photo black and matte black, higher build quality, hack to use bigger paper, and a previous lifetime's worth of Canon loyalty.  On the down side, there is no straight media path, the thickest media I can put into the Canon is about half the thickness of the Epson's ability, and I need a hack to print a large pano.

Regardless, I wasn't being steered to something I hadn't considered, and the reviews on the Canon have been good, so I went for it.  It came with five boxes (125 sheets) of 17x22 Canon paper, and I threw in a couple of sample backs of Moab and Canon 8x10 sheets.  I screwed up by not getting a pile of 4x6 pages for initial testing.  Getting the whole thing into the car alongside a cooler full of Costco meat (see previous posting on new pellet smoker) and the dog wasn't trivial, but part of the Glazer's service included a hand truck and Dana's help lifting and pushing.  I left the printer box in the car until Kiernan got home, and he helped me get it inside.

Next problem: this thing is NOT small.  The basic printer footprint is about 18"x30", but the 18" expands to nearly three times that when the paper feed tray and paper catching tray are open.  And it weighs 70 pounds, so it requires its own piece of furniture.  Plus storage for paper boxes and archival boxes for prints.  Without much hope, I visited Storables at U-Village when I went to the Apple Store.  And they surprised me by having a technology that seemed to do the trick: their Industrial Post Steel Wire Shelving, which is an easily assembled, more attractive and flexible version of Gorilla Shelves.  With each shelf rated for 300 pounds and each rack rated for 2000, all I needed to do was figure out my needs.  I sketched a few ideas, measured a space in my office, and spent about a day throwing shit out and packing shit more efficiently to get the space I needed next to my desk.

The shelving went together as easily as described (perhaps more so) and I wrestled the printer box (which could easily hold at least one body, in my inexperienced estimation) downstairs into my office, lifted the printer out onto the top shelf (easier than it should have been -- Canon ships the printer in a bag with handles which is designed with handles to lift the printer out), removed the 20+ bits of orange and yellow tape, plastic, foam, and so on which protected the printer during transit, installed the print head and so many ink cartridges I lost track (apparently they are individually keyed so that you can't put them in the wrong place, but I never tried because the labeling was so clear so I can't confirm that), and ran the setup program.  Which, contrary to my previous experience with Canon software, was reasonably clear and did a good job, even avoiding me having to enter my wifi password via a four direction keypad on the printer.  It takes quite a while for the printer to set itself up, and the process is rather noisy as the print head moves around, the vacuum feed system blows air here and there, and the ink cartridges are milked to fill in the lines to the head.  There's also apparently some kind of ink agitation system built in (I think Dana mentioned ball bearings in the cartridges) that rattles a bit.  The whole process of unpacking and setup probably took a couple of hours, but I was in no hurry and wanted to get it right.

Upon realizing that the bulk of the paper I had was 17x22, and the smallest was 8x10, I stopped at Office Depot and bought a box of 100 sheets of 4x6 HP semi-gloss photo paper.  I figured that if there were going to be big problems, starting small was a good idea.

Remarkably, the process is pretty straightforward at my current level of understanding.  It's annoying to have to print via the "Plug-in Extras" menu, and I wish Canon's Print Studio Plus did a better job of understanding borders, but it's OK.  I'm unclear as yet whether the printer's idea of what kind of paper overrules PSPs or vice versa, and I'm pretty sure I'm going to try out an additional printing manager (possibly PrinTao, possibly something else) at least to implement that pano hack and probably to manage paper types.  B&W prints on the HP paper are a bit warm, and I need to work on that, but I haven't even gotten a custom profile for this printer yet.  It may also be the paper, not the ink.  Things to investigate, things to learn.  Things to become (or resist becoming) obsessed about :-).

It will be an adventure.  I shall remain steadfast in memory of Mark Hamburg's immortal statement (Mark was the architect on Photoshop when I was at Adobe):

"Color Is Hard"


(Originally posted to my personal blog @ Mischievous Ramblings II)

Photography, Storage, and Backups

Thoughts on storage and backup requirements for a photographer, including why my previous plan failed, how I thought through a new plan, what that new plan is, and how long I have before I have to re-address the problem.

Read More

Complete Genetic Makeover

I joke that I am a Canon shooter because of genetics: my father shot Canon gear, so I shoot Canon gear.  My first SLR was my father's Canon F-1.  The first one I picked out was a Canon A-1 in college.  My first digital SLR was a Canon Digital Rebel.  A couple of years ago I upgraded that body with a Canon 60D.

Last Tuesday I dug up all the Canon gear I had around the house, packed it into as much original packaging as I could find, loaded it all into an Ikea bag, and drove to Glazer's Camera in South Lake Union.  When they asked me what they could do for me, I said "I have a bag full of used Canon gear.  I'd like to walk out of here with some Nikon gear."

Why did I do that?

I've looked at every digital photo I've got in the last two months and stripped my collection down from over 43K to under 2K worth keeping.  Some of those pictures were taken on digital cameras with 640x400 resolution, some on phones, many on the Digital Rebel (6.3 MP), and the vast majority on the 60D (18 MP). Aside from a lot of swearing at myself about getting shots wrong, there's only one thing I said more than once about a photo.  It wasn't "I need more pixels".  It wasn't "I need a better flash".  It wasn't "I need a longer lens" or "I need a faster lens".

It was "I want more Dynamic Range".

I've been planning to upgrade from the 60D to a 5D Mk IV (if it ever comes out) because a full frame camera system is better for landscape photography (which I'm doing more of) and bigger pixel sites are better for taking star photos (which I really enjoy).  As the delay went on, I questioned whether I should consider a 5D Mk III or a 5DS instead.  Or a 6D.  In either case, I was talking about a minimum of $1400 and probably more  like $3K or $4K.  Plus new lenses.  Major investment.

Once I knew I was looking for Dynamic Range, I started Googling.  There's this site called DXOMark that has allegedly objective measurements of camera bodies and lenses.  It's sortable by measurement.  Go check it out and sort the list of camera bodies by Landscape (which is largely a proxy for Dynamic Range).  Go ahead.  I'll wait.

Six of the top seven (and seven of the top nine) are Nikon bodies.  The odd ones out are from Sony, which is the company that Nikon buys sensors from.  Nikon, Sony, and Pentax dominate the top ranks with cameras going back five years or so.  This is not a flash in the pan.  You have to go to the 68th slot to find a Canon, and it's not a DSLR.  The Canon 5DS is 94th. The Nikon D810 has nearly 2 1/2 EVs more dynamic range than the 5DS and more than 3 more than my 60D.

If I wanted more dynamic range, I was buying a Nikon.  Which meant getting rid of about $6K worth of Canon gear.  Which I might get $2500 for if I was lucky.

While we were in NY to see George Takei's Allegiance musical, I visited B&H Photo (which was amazing, BTW -- like Fry's Electronics for photo and optical geeks with the entire staff wearing yarmulke and speaking English) and Adorama and spent perhaps 60 minutes playing with the D810.  I read the manual.  I researched lenses.  I concluded that I was going to be out of pocket over $6K to rebuild my kit, including getting between $2K and $2.5K for my old kit.

But it would be better kit, objectively, as measured by DXOMark.

Here's the planned kit:

  • Nikon D810 Body
  • Nikon 20mm f/1.8
  • Nikon 24-85mm f/3.5-4.5 VR
  • Nikon 18-35mm f/3.5-4.5
  • Nikon 70-200mm f/4 VR
  • Tamron 90mm f/2.8 Macro Di VC
  • Sigma 150-600 F/5-6.3 DG OS HSM Sport or Contemporary

I ordered the body, 18-35, and 24-85 from B&H with overnight delivery for Wednesday 10/28.  My Tekniq modular camera bag (from a Kickstarter I backed in May) happened to arrive.  I visited Glazer's on 10/27 and got trade-in money for my Canon gear, which bought the 20mm and the 70-200.  The kit packs into the Tekniq like a charm.  This is the first time I've ever planned kit instead of buying what I seemed to need as I went along.  There were what seemed like a million little details: UV filters, extra lens caps, lens caps retainers (I'm horrible about losing caps and hoods), batteries, chargers, new CF cards (the 60D used SD) and a high speed SD card, card readers, new cables for the remotes, polarizers, step-up rings, monitor protection, and so on.

Here's the first image I was willing to put up on Facebook:  




Here's the one that blew me away:


Which is a detail from: 


Which is itself about a 50% crop from the original.  What blows me away is the sharp detail on the crow where it is in focus (which isn't all of it, not at f/4, but that's not the point).  You can see dust on the feathers.  There are places where you can see each individual barb!  It's not that this is a great image, but the level of detail is, to me, nearly unbelievable.

Which brings me back to those DXOMark numbers.  I turned in lenses for my Canon 60D with sharpness values of 7, 8, 8, 9, and 12.  I bought lenses which, combined with the D810, have sharpness values of 17, 23, 25, 26, and 29.  The crow images are from the 70-200 with a 29 sharpness.  I would not have expected to get kit capable of 3x the detail, but I think I may have 2x or more (the average of each new lens' sharpness improvement over the equivalent lens in the old kit is about 2.6x, with the worst being 2.2x and the best being 3.2x).

I'm gobsmacked.  And I have no excuses that my equipment can't get a shot.  If I don't get a shot, it's my fault -- intent, design, execution, understanding of conditions, understanding of the equipment.  The gear can get the shot, if I can figure out what the shot should be.

Isn't genetic surgery amazing?


(Originally published on my personal blog @ Mischievous Ramblings II)