Something Old and Something New

By Ctein

Here's a couple of choice items that don't merit full columns of their own.

Post Exposure is free
After 15 years, my book Post Exposure has finally gone out of print—an excellent run. Both I and Focal Press are very happy to have created this book.

The rights have reverted to me. At this time I haven't decided whether to massively rewrite the book to incorporate digital photography and printing, or to write an entirely new book on that subject.

In any case, such a book is a good year off.

In the meantime, I've decided to give Post Exposure away, in PDF form. The PDF is identical to the published book, save that it's missing the 16-page color signature (I didn't do the layout files for that). Sorry, you'll just have to use your imagination. Also, it's just a "dumb" Acrobat file—no internal links or any cool stuff like that.

But, hey, you can't beat the price.

You can get it by going to my website and clicking on the book cover image at the lower right.

I only ask one thing of you: Please don't redistribute the file. Don't post it to your own site, don't torrent it, don't give a copy to friends. If someone you know would like a copy, send them a link to http://ctein.com.

That's not asking too much, is it?

Enjoy!

Now we know how high ISOs can go
About a month back I wrote "Photography at the Speed of Light" in which I asked the question of how fast digital sensors could really get. When would we hit the brick wall of absolute physical limits of photon counting statistics that can't be surpassed?

I wasn't up to the challenge of doing the photometry, but one of our readers, Nick Condon, was. He's posted his answer online. 

For those of you who don't want to wade through the technical work, the conclusion is that image quality comparable to the Nikon D3S at ISO 6400 is theoretically achievable at ISO 100,000. We've got about four more stops to go before we hit the wall. But, please, read his analysis before commenting on the answer here. Of course any first-principles analysis assumes an absolutely perfect system, so the real world answer might be a stop lower. Conversely, it assumes no nonclassical tricks played with detection or statistics, so the real world answer might be a stop higher. But at least we know what the ballpark looks like.

While such high ISOs would never be what you'd use routinely, they will have one advantage over more conventional ones. You're collecting so few photons (and, hence, photoelectrons) that you never have to worry about saturating the pixel in any decent sized sensor. Blown-out highlights become a non-problem.

One can turn the problem on its head to answer another interesting question: how small a sensor could the ideal camera have and deliver current D3S image quality? The answer is about 1/4-scale, comparable to the sensors in compact digital cameras.

Important caveat: while Nick is a real optical scientist and doesn't just "play one on TV" like me, and while he has shown the analysis to his colleagues, none of us are experts in the specific area of photometry. It is possible there is some mistake buried in the math or the assumptions. If one of you readers does photometry professionally and would care to take the time to give Nick's work a review, we'd both really appreciate it.

If you do have any technical criticisms of his analysis, please do not post them here as comments; TOP cannot serve as a technical debate forum. Instead, e-mail your remarks directly to Nick (condon [dot] photography [at] gmail [dot] com) if you want them to be private, or post them to his blog if you want them to be public. (And, if you care to, cc them to me—ctein [at] pobox [dot] com.) If any conclusions change substantially, I'll write a follow-up to this follow-up.

Ctein

Ctein's regular weekly column appears on Wednesdays.

Send this post to a friend

Please help support TOP by patronizing our sponsors B&H Photo and Amazon

Note: Links in this post may be to our affiliates; sales through affiliate links may benefit this site. More...
Original contents copyright 2011 by Michael C. Johnston and/or the bylined author. All Rights Reserved.

Featured Comment by Ben: "Just dipped into the PDF for the first time and found it to be very interesting and well written. Perhaps this is the justification I need to buy a Kindle...but where's the TOP link to Amazon to buy one? ;) I have some reading to do...."

Mike replies: "Uh, same place it's always been—in the right-hand sidebar, under "PORTALS," second item from the top, where it says "Please Help Support This Site—Use Our AMAZON LINKS." If you can't find it, please do let me know.

Featured Comment by Semilog: "Ctein, wow. After 15 years out of the darkroom, last month I bought ('stole' is only slightly too strong a descriptor) a used Leitz V35 enlarger and I've just re-started in the darkroom. I've read about half of your PDF now, and it's just incredibly useful, refreshing things I used to know and filled with useful bits that I did not know. There really should be a PayPal donation link on the download page. Thank you so much for sharing this."

Comments

[Mrten]

I'm now curious for the borrowed and blue ;)

[Edd Fuller]

"...don't torrent it..." What does that mean??

[Bernard Scharp]

Hah, so my copy of Post Exposure just became part of a limited edition? ;-)

But seriously, I'm glad I have the `real book', since the images are a quite important of the book, and I think that without them, it wouldn't have been quite so useful to me.

Still, no images still beats no book, so I highly recommend people who haven't read it to check out the PDF.

[Jim Bullard]

Ctein, Thank you very much for the PDF version of your book. I'm sure I will enjoy reading it.

[Tom McNamara]

Thank you, thank you, thank you for that Ctein.
It really is appreciated by those of us with church-mouse-funds.

Tom

[Alex Vesey]

To Ctein.
The offering of a out of print book like this is very generous.
Thank you.

Alex Vesey

[ctein]

Dear Folks,

A reader has alerted me that there's a minor bit of font garbage on page 3 of the PDF. I'm on holiday and can't fix it right now. It doesn't cause any of my PDF readers to crash, but it caused his to.

So, if you run into problems reading it, try opening it with a different PDF reader, try skipping over page 3 (the garbage is only there) and/or try checking your option preferences for how your PDF reader handles unknown fonts.

Sorry for the minor inconvenience, to anyone who is inconvenienced.

pax / Ctein

[Ctein]

Dear Bernard,

Just to be clear, the PDF has all the B&W figures, it is just missing the color ones, which are a minor part of the book.

Pax / Ctein

[David Bostedo]

Edd - "Torrent" refers to BitTorrent which is a means of distributing files using a lot of hosts at once. From Wikipeddia :

"The BitTorrent protocol can distribute a large file without the heavy load on the source computer and network. Rather than downloading a file from a single source, the BitTorrent protocol allows users to join a "swarm" of hosts to download and upload from each other simultaneously . The protocol works as an alternative method to distribute data and can work over networks with low bandwidth so even small computers, like mobile phones, are able to distribute files to many recipients."

[Geoff Wittig]

Thank you for your generosity.

Post Exposure is, bar none, my favorite book on the subject of photographic printing. It doesn't have lots of info on the latest inkjet printers, which would make it a perishable commodity. Instead it has deep thought and solid analysis of the interaction beween the surprisingly quirky human visual/perceptual system and photographic images on paper. It then goes into great detail on how to optimize the second for the first. Great stuff, and most of the principles, if not the details, are eminently transferrable to the digital darkroom.

I breathlessly await a digital printing equivalent.

[ctein]

Dear Geoff,

Thanks for those extremely kind words.

You've gone to the heart of my indecision over how to proceed. There's a lot to be said for starting afresh on a book on digital printing and not being bound by the strictures of POST EXPOSURE. On the other hand, the principles and intellectual logic of POST EXPOSURE naturally extend to encompass digital photography and printing.

I'm leaning towards the latter, because many of the "old baggage" issues that would come up around producing and marketing POST EXPOSURE III disappear with a self-published electronic book.

Really, though, what's holding this up is that I simply don't know what the narrative flow should be to properly convey the new information. I've got a whole lot of words, and a whole lot of facts, but no "book" in my head, if you get my drift.

I can say that whatever way I go, the book will be cheap. $15, max, and I may very well price it at $10. Dead electrons are sooooo much cheaper than dead trees.

pax / Ctein

[Ben]

Just dipped into the PDF for the first time and found it to be very interesting and well written. Perhaps this is the justification I need to buy a Kindle... but where's the TOP link to Amazon to buy one? ;)

I have some reading to do...

[Gijs]

hello Ctein,

Without having read Richard's analysis yet, there is an additional advantage to such high-speed sensors. When you get down to single photon counting, you no longer need the Bayer array since the charge deposited in the sensor pixel will be proportional to the energy (i.e. the colour) of the photon. Software writers will probably have to invent a new colour space to handle that?

regards and thanks also from me for putting your book online!

Gijs

[Darr Almeda]

Thanks Ctein for writing and generously giving us your Post Exposure pdf file. I would certainly purchase POST EXPOSURE III in any format.

Darr

[Dan Berry]

Ctein;

Thank you for "POST EXPOSURE" both in its bookstore form and now to give it away for free, a gift that keeps on giving. Anyone who has not read this book has likely not reached their potential.

The top shelf of my photography bookcase has very few authors,
Ansel Adams
Leslie Strobel
Norman McGrath
Ctein

Thank You
Dan

[hugh crawford]

I read the post at Nick Condon's site and it seemed to me that he was describing a monochrome sensor and left a more detailed comment there, but something I've been wondering about for a long while is rather than measuring the number of photons that strike the receptor in a given length of time, why not measure the length of time it takes for an arbitrary number of photons to strike the receptor? Bright parts of the scene would effectively get a short exposure and dim areas would get a long exposure.

[Avi Joshi]

Just a thought. Maybe the amazon links would be a little more visible at a quick glance if you placed an image of their logo there instead of just the text. I think that would help quite a bit. Not that I don't have them bookmarked, but just sayin' :)

[Riley]

Re: dead electrons.

I think that they are not dead, but rather well used.

[amcananey]

Mike,

Don't ask me why, but I always seem to miss the Amazon links as well. (Don't get me wrong, I usually order though your links since I have them bookmarked, but whenever I actually LOOK for them (e.g., when using someone else's computer), I somehow seem to gloss over them. Maybe it's because I'm looking for the Amazon logo? In any case, you might want to have someone better versed in such matters look into it...the "Tip Jar" is hard to find (or easy to overlook), too...

Best regards,
Adam

[ctein]

Dear Hugh,

Cristoph*, in a private email, also correctly pointed out that this hypothetical perfect camera would not use a Bayer array. We did not make any assumptions, though, about what kind of sensor it would be-- this is an entirely theoretical calculation.

There are full color photon detectors that don't use spectral rejection filters to extract a subset the photons. I've alluded to some of those technologies in previous future-tech columns. It's not really central to this discussion.

As I described in my previous column on this, a modified Bayer array camera could get within about a stop of theoretical perfection.

pax / Ctein

(* Cristoph also correctly observed that Nick treated the entire solar flux as if it were in the visible band, which is obviously not the case. I guesstimated that costs about half a stop. Of more concern to both Nick and me, though, is being sure that there's not some more fundamental error in the analysis that makes it total hogwash.)

[Semilog]

Ctein, wow. After 15 years out of the darkroom, last month I bought ("stole" is only slightly too strong a descriptor) a used Leitz V35 enlarger and I've just re-started in the darkroom. I've read about half of your PDF now, and it's just *incredibly* useful, refreshing things I used to know and filled with useful bits that I did not know. There *really* should be a PayPal donation link on the download page.

Thank you so much for sharing this.

[Charles]

The book explained a situation I just faced. I photographed a yellow rose many times using different lenses and cameras, in shade and sunlight, at different times of day. In the computer one shot displays a red tinge on some petals, in other shots a green tinge. Both are difficult to change, and the flower has to be selected in order to avoid side effects on the plant leaves and other background.

Started reading and there on page 10 Ctein explains "the two different kinds of yellow. There’s the spectrally pure yellow that corresponds to a small slice of the spectrum between orange and green, and there’s the yellow we get if we mix equal parts of red and green light. Both kinds of yellow stimulate the retina in the same way, and we perceive them as the same color." (p. 10)

Doesn't tell me exactly what to do (he didn't write a book on yellow roses), but it probably explains why probing a petal and vectoring the color there along a line between full red and full green is extraordinarily sensitive to the amount of change. (This is done with the Color Correction tool in Picture Window Pro.)

[ctein]

Dear Semilog,

My pleasure.

Hey, I'll never turn down anyone who wants to give me free money, but if you really feel the need you don't need a Paypal button. You just log into your account, tell PP you're sending some money as a gift to my email address (which is my account), and it's done.

But I ain't gonna go asking for it. This is for free, and I mean it. With two editions and four printings, I made a living wage on the time I put into Post Exposure. Not much more than that, but that's enough.

pax / Ctein