Follow
on TwitterI’ve updated my Headlamps and Flashlights article, with considerable material added to the Lupine Lighting Systems Wilma review.
The Wilma is without a doubt the best headlamp I’ve ever used, and if you’re in the market for a lighing system for nighttime photography and/or cycling/hiking/etc, then you’ll want to read it ASAP.
Since installing Adobe Photoshop CS3 and Adobe Dreamweaver CS3, I’ve experienced two system crashes coming out of sleep (screens don’t redraw, only pulling the power plug works). I’m used to Adobe products crashing (over the years Adobe gets its money on upgrades, and in return I receive a product with as many problems as the previous version). But a crash in Photoshop or DreamWeaver doesn’t annoy nearly as much as having to power-cycle the entire system.
After two months of nearly trouble-free operation, the purely circumstantial evidence is suggestive; the instability started the very next day after installing the Adobe products. That proves nothing, but it is the sequence of events. I’m using the very latest Mac O X 10.4.9, and minimal non-standard stuff, and a reasonably capable machine, a 3GHz Mac Pro with 16GB of memory and a few measly terabytes of disk space.
Adobe disavows any bugs in its products and spreads plenty of FUD with all the things you might be doing wrong in Troubleshoot system errors or freezes in Photoshop CS3 on Mac OS (there are clones of this article for XP and Vista, etc). On a modern “sandboxed” operating system such as Mac OS X, Adobe’s disclaimer just doesn’t ring true: “Photoshop may not necessarily be the cause--it may be the only application that uses enough memory or processor cycles to expose the problem”. Does this sound like a similar refrain? See An Experience with Software Technical Support. C'mon Adobe, Photoshop CS3 only sporadically uses more than one core of my quad-core, and gets sluggish using its maximum 3GB of the 16GB total memory.
If you read Adobe’s “Troubleshooting” guide carefully, you’ll find that it covers any and all conceivable computer problems, on Earth or on Mars, with a nightmarish litany of tasks that will keep the most ardent nerd busy for weeks. Eventually something will work (or you’ll give up, and trade your computer for a crayon), sparing Adobe any inconvenience.
Maybe Adobe could explain why one should have to delete carefully-chosen preferences (“Re-create the Photoshop preferences file to eliminate problems that a damaged preferences file might cause”). Are Adobe engineers so incompetent that invalid preferences can’t be detected and reset to valid defaults?
No, it’s easier (and less expensive) to foist the problem onto the customer, instead of making the engineers design a robust program. Or writing test tools that deliberately corrupt preferences so that the software can be tested for stability in situations which real users allegedly experience. This is software engineering 101. If preferences are “damaged”, guess what? Those preferences are maintained by the program itself! Asking the customer to pamper a finicky program that can’t detect its own invalid preferences is lame.
Probably not. But the premise of Adobe’s Troubleshooting guide is hard to swallow: that the only possible problems are caused by something other than the Adobe software. I wish Adobe would simply apply some honesty to their document—new software has problems. But there’s not a single link to Adobe forums in which to report problems, not a single suggestion as to how to proceed if the problem does indeed appear to be in Adobe software. In short, buy a support contract.
If you like dried fruit as I do, you might want to try some organic Lycium Eleagnus Barbarum, commonly known as Tibetan Goji Berry. It’s unclear if their considerable nutritional and anti-oxidant properties are preserved after drying, but they taste pretty good, especially in trail mix (or when covered with milk chocolate!). Unfortunately, after checking into them, I find that wikipedia claims that no such thing as a Tibetan Goji Berry exists, so I’m not sure whether they’re a good thing after all. For that matter, it’s unclear if Lycium Eleagnus Barbarum exists either.
Tibetan
Goji Berries Lycium Eleagnus Barbarum
I found “organic” ones at Sigona’s market in Redwood City, CA. I’ve called Sigona’s to find out from where they’re sourced, since the term “organic” would be meaningless if they come from outside the USA. I buy organic whenever possible, because our food today is contaminated with “safe” hormones, innumerable poisons, and pathogens (E Coli, mad-cow, etc).So I don’t want to eat Chinese Wolfberries laced with organochloride pesticides and their ilk.
In particular after seeing dead pets in the recent news due to Chinese-sourced pet food ingredients, one should be aware that Chinese-grown Goji berry products have been seized before by the FDA due to containing nasty stuff (search for “wolfberry” at fda.gov).
A few of the ten thousand flowering Yuccas
Big Bend National park, April 2007
As of April 20, 2007, Dagger Flat in Big Bend National Park was showing a glorious display of thousands if not tens of thousands of giant Yucca faxoniana or Yucca treculeana (not sure), with their 3-4 foot long flower stalks, each with up to several hundred individual flowers. They were in their prime while I visited, with soft, fresh flowers and some not even open yet.
The image above was taken in near darkness using the Lupine Lighting Systems Wilma LED headlamp [image] to “paint” the scene during a 30-second exposure using the Canon EOS 5D. The three brightness levels of the Wilma offer excellent control over exposure for this type of “painting”. The RAW (CR2) file was processed using Canon’s Digital Photo Professional with color balance (and picture style), showing the pleasing color rendition of the Wilma.
The Wilma is without a doubt the best headlamp I’ve ever used. For more on the Lupine Lighting Systems Wilma, see Headlamps and Flashlights.
Like Canon, SanDisk apparently is using the same crooks to process their rebates. The postcard I received (shown below) is a clone of the one I received from Canon, this time for a $75 rebate on an 8GB Extreme IV Compact Flash card.
The
same fraud as with Canon—claiming that required materials were not submitted
The scam consists of claiming that required materials are missing, while simultaneously professing ignorance of what was actually submitted when a customer calls to inquire. A short response deadline helps ensure that suckers (aka “customers”) have as little time as possible to comply. As with Canon, I photographed my submission prior to sending it in; the materials apparently all vaporized en-route, excepting the rebate form itself.
I have yet to make the phone call to see if there will be another mysterious “internal system error” that accounts for the “mistake”, but I’ll report on it in a future blog entry.
For those readers with a legal background: what is the appropriate governmental agency to contact regarding this sort of fraud?
I’ve returned from my Big Bend trip. All articles and software purchased during my absence have been delivered via email. Please contact me via email if for some reason you have not received a reply.
A reader of The Sharpest Image inquires whether I have compared the Burzynski ballhead to the Really Right Stuff BH-50/55 and/or the Arca Swiss B1 (now replaced by the Z1).
Indeed, I did swap heads between the G1325 and G1548 tripods, but found little evidence of any variance in sharpness caused by the ballhead, with image sharpness apparently being dominated by tripod-leg resonance. When appropriately tightened/clamped, all of them offered excellent stability with the camera (or lens) balanced over the head.
With unequal weight distribution, the story might be different; the outrageously sturdy Burzynski head inspires confidence, and its design is inherently strong. With really large lenses (300mm f/2.8 and larger), the Burzynski is unruffled whereas the other two ballheads might exhibit somewhat less stability, and simply are not built (or at least do not appear to be built) to accommodate such lenses. Of course, such huge lenses are often more easily handled with a gymballed head, such as the Wimberly.
A reader of The Sharpest Image inquires about the Sachtler ENG 2 CF carbon-fiber tripod as compared with the Gitzo models.
I used the Sachtler ENG 2 CF with the Burzynski head for a week in Norway last fall, kindly lent to me by Bjorn Rorslett of naturfotograf.com. The Burzynski head mounts into the 100mm bowl of the Sachtler, though a bolt/washer/nut are required to secure it. It’s a very solid low-resonance combination—as good as the best of the Gitzo line, albeit at a considerably higher price.
Portability is another matter. The Sachtler is bulkier due to the triple-tube design of its legs, and thus it is difficult to carry it using a small daypack, such as the North Face Recon. That makes it a poor fit with my needs, which involve hiking and easy portability. Even the moderately large Gitzo G1325 fits in my Recon daypack without trouble, although removing the Burzynski head is helpful.
The Sachtler legs also “flop” as compared to the Gitzo legs, which stay in position unless force is exerted on them. One can adjust to the “flop” characteristic, and some might prefer it, but I prefer the friction of the Gitzo legs, especially on unstable ground where I might be trying to keep my balance.
A photography backpack with its thick padding, dividers for lenses, etc is a necessity for protecting fragile photographic equipment during travel. But such packs are bulky, relatively heavy, and simply overkill for many situations. My all-time favorite daypack for photography is the North Face Recon.
My frequently-used older model of The North Face Recon
Though not designed for photography, it can easily accommodate a moderately large photographic kit. It has a mesh stash pocket on each side (use one for a tripod and the other for a water bottle), a large stash pocket on the top (jacket, etc), a large and medium zippered compartment, a small sealed zippered compartment, a waist belt and various compression straps. The current model weighs 35 ounces (992 grams). I prefer the earlier version (shown above, no longer available), which weighed about 25% less. However, the current model offers thicker padding and more durable construction, so the extra 8 ounces or so are probably worth it.
A few years ago, I used the Recon to carry a 4X5 kit: a Linhof Technikardan 4X5 view camera, six lenses, 4X5 sheet film and holders, filters, Gitzo G1325 tripod, water bottle etc. Neoprene pouches mitigate any concerns about items damaging each other, and cut down bulk considerably over a photo backpack. The Recon can easily accommodate two large DSLR bodies, several lenses, a tripod with head, filters, etc. Also, it’s an excellent 2nd carry-on bag; I carry a main photo backpack onboard an aircraft (a Lowepro DryZone 200), with the Recon taking an extra camera body and lenses that don’t fit into the main pack, personal sundries, etc.
At $79, it’s a must-have item for photographers who don’t want to carry a pack that weighs 6 pounds when empty. And it’s more comfortable too! I simply never leave on a photo shoot without it. And unlike a photo backpack, it’s versatility doesn’t end with photography.
If you order a review or software, I generally respond very quickly, usually within a few hours.
Very recently, two people emailed me wondering where their stuff was—in their spam folder no doubt. I do my best to work around the problem, but spam filters have become so aggressive that I often have to resort to plain text, and even a different email server in hopes of getting through.
Please keep this in mind if you expect a response to an email; in some cases I’ve taken the time to compose a detailed response to an email, sent it, then never heard back. We both lose—I’ve wasted my time responding, and the person inquiring doesn’t get an answer. So far, I’ve responded to every email sent to me, so if you’ve sent one and haven’t gotten an answer, now you know the reason.
For reasons explained previously in Goodbye Nikon?, this is my first trip ever on which the Nikon equipment is going to stay in the drawer at home.
I’ll be taking two Canon EOS 5D camera bodies with me—one unmodified color body and one modified for dedicated infrared use. Lenses will include the EF 16-35mm f/2.8L II, the EF 24-70 f/2.8L, the EF 100mm f/2.8 macro, the 300mm f/2.8L IS, and the Nikon 28-70 f/2.8 EDIF AF-S and 85mm f/2.8D PC-Micro Nikkor + whatever else fits into the bag. The latter two Nikon lenses are excellent performers in infrared on the EOS 5D-IR (using an adapter), hence their inclusion.
In a most dubious explanation, Mac OS X 10.5 is now allegedly delayed by the iPhone. Huh? Apple didn’t know this months ago? Yes, Leopard will be a nice improvement, but how about giving me a coupon for a free iPhone to make me feel better? Or a coupon for Leopard if I buy an iPhone?
Apple Statement
iPhone has already passed several of its required certification tests and is on schedule to ship in late June
as planned. We can’t wait until customers get their hands (and fingers) on it and experience what a revolutionary and
magical product it is. However, iPhone contains the most sophisticated software ever shipped on a mobile device, and
finishing it on time has not come without a price — we had to borrow some key software engineering and QA resources
from our Mac OS X team, and as a result we will not be able to release Leopard at our Worldwide Developers Conference
in early June as planned. While Leopard's features will be complete by then, we cannot deliver the quality
release that we and our customers expect from us. We now plan to show our developers a near final version of Leopard
at the conference, give them a beta copy to take home so they can do their final testing, and ship Leopard in October.
We think it will be well worth the wait. Life often presents tradeoffs, and in this case we're sure we've made the
right ones. [Apr 12, 2007]
Walter Mossberg’s column in today’s Wall Street Journal follows up on the reaction garnered by the prior week’s discussion of craplets (great name!). If the junk (trial software, ads and offers) that come installed on Windows Vista machines are craplets, does that make Windows Vista crap? Not really—it accomplishes that all by itself.
Some of the more interesting comments in Mossberg’s column include:
One ray of hope is a free program called PC Decrapifier...this software automates the process of uninstalling craplets.
Dell, Sony and others say they are moving towards a new scenario in which all of this stuff will be easily refused on all models.
Finally, an excellent way to avoid or minimize the craplet problem is to simply buy an Apple Macintosh computer...and you don’t need antivirus or antispyware programs because the Mac is essentially free from those menaces. So, even my year-old Mac laptop reboots roughly three times as fast as my three-week-old Sony.
Well, duh! You get what you deserve when you buy a Windows PC. Craplets, up your whazoo!
In a rather tepid conclusion, and after discussing the usual theoretical issues, the author finds that Mac OS X is more secure than Windows in Myth or Fact? Is Mac OS X Really More Secure than Windows? at informit.com. No sense offending anyone with the truth that millions of Windows PCs are compromised, with a huge hacker community and organized crime drooling at the rewards.
The fact remains that connecting a new PC to the internet without jumping through hoops to install anti-virus software and taking other precautions is electronic suicide, which is not the case with Mac OS X. My view is rather stronger on this issue—see my article PC or Mac—Making a Sensible Choice.
See the discussion below on why the 8-core Mac Pro is often no faster than the 4-core Mac Pro (or even slightly slower).
Rob at barefeats.com ran diglloydTools test-compute-speed, a CPU-intensive benchmark that assesses the scalability of multiple processor cores. The results show that the 8-core Mac Pro can be twice as fast as the 4-core Mac Pro with compute-intensive tasks:
| 4-core Mac Pro @ 3.0GHz | 8-core Mac Pro @ 3.0GHz |
|---|---|
Test size = 128MB... Testing using 4 threads...
|
Test size = 128MB... Testing using 8 threads... |
If you’re a scientist performing intensive calculations which don’t involve much memory or disk access, you’re
likely to see a huge benefit from the 8-core Mac Pro. But when working with images or video or sound, large amounts
of memory are accessed, and thus performance gains are likely to be modest, or even degraded in some cases.
At a 20% premium over the 3.0 GHz quad-core Mac Pro, and a 48% premium over the 2.66 GHz quad-core Mac Pro, most users will still find the 2.66 GHz model to be the “sweet spot”.
I predicted that an 8-core Mac Pro would be roughly equivalent to a 6-core Mac Pro as compared with the 4-core model. First test results at barefeats.com confirm this prediction, showing that the 8-core Mac Pro doesn’t scale in performance as might be assumed by a layman, but can be about 50% faster with certain tasks.
Many of the tests show no improvement in performance at all—even though all 8 cores are used. Why? As noted in April 5 blog entry, unless the task is compute-intensive, memory bandwidth will limit the performance to about that of a 4-core Mac Pro, because all 8 cores will be trying to access memory at the same time.
It’s not just memory bandwidth either—access to the hard disk and system services in effect require tasks to queue up single-file to wait for the needed resource (memory access, disk, exclusive lock, etc). This is called contention. Like a rush-hour freeway, doubling the number of cars can cut speed by more than half as cars (tasks) must jostle with one another for the same lane space (system resource).
Contention and overhead — Operating systems (eg Mac OS X ) have to juggle the outstanding tasks/applications/threads among the available processor cores as well as the operating system itself. As the number of processor cores increases, this overhead also increases.
In the real world, multi-core systems do not scale linearly and the performance gains of 8 cores over 4 cores are often modest, or even degraded.
Poor scheduling of tasks across processor cores is also responsible for performance degradation. In particular, Mac OS X is fond of switching a task (thread) between cores for apparently no reason at all—I’ve observed this myself on many occassions; with several idle cores, Mac OS X switches a task from one to another core at apparently arbitrary intervals. Each time this happens, the large on-chip processor cache of the “old” core must be flushed to main memory, which is very slow (relatively speaking). As the task resumes running, the on-chip cache must be populated from main memory, again a slow operation in relative terms.
The core-swapping behavior can be seen below; two runs of diglloydTools run-stress-test were done. The first run shows that the single-threaded portion of the test ran on the 4th core for its full duration. The second run shows that the single thread ran on the 1st core for about half its run, then was swapped to the 2nd core—in spite of the 3 other cores being idle! This had little effect under the circumstances, but can have a large effect if there is contention for the CPU by multiple threads.
Mac OS X core swapping
It seems clear that Mac OS X 10.4 has not been adequately optimized for even 4 cores; perhaps Mac OS X 10.5 will improve matters.
No, it’s not a book on smoking dope. Rather, it might be a book that changes your life for the better. It’s no cheesy self-help book, but rather a book on a philosophy of life, a deceptively easy read, but one that might require 10 iterations or more to fully appreciate. Paired with eccentric photos that subtly support the text, it’s definitely not just another pretty coffee table book.
Coincident with the release of the 8 core Mac Pro, Apple cut the memory prices by about 20% for build-to-order systems (see All About Mac Pro Memory). The prices on memory purchased separately have also dropped, but remain higher than the build-to-order prices. There were also price cuts on the Cinema Displays.
Following up on Tuesday’s comments on the new Canon EF 16-35 f/2.8L II color fringing problems, I made two prints on the Epson Stylus Pro 3800, each 10.8" wide. One print was made using the Canon Digital Photo Professional image, and another was made with the Adobe Photoshop Camera Raw image. The subject was carefully chosen to show any optical weaknesses.
Test image—Canon 5D + EF 16-35mm f/2.8L II
Digital Photo Professional / Adobe
Camera Raw (with correction)
A quick glance shows the print from each raw-file converter to be about the same, but by studying the DPP image for more than a second or two, the eye is forced to work very hard to reconcile detail that looks sharp, but isn’t. It’s rather like a newspaper color photo that is out of registration, a defect that rapidly causes eyestrain. The Adobe Camera Raw image does not suffer from this problem, because the color fringing has been eliminated.
In my view, images from the 16-35 II which have not been corrected for color fringing are unacceptable at 16mm, at least with some subjects, such as the white birches in the test image. I am sorely disappointed at Canon’s assertion that chromatic aberration has been “virtually eliminated”. The claim is so much at odds with the rendered images that one can only marvel at Canon’s brazenness. While we can’t expect too much from an ultra-wide zoom lens, we can expect realistic assessments of imaging performance from the manufacturer. Maybe the same Canon management is responsible for Canon’s rebate ripoff.
So what did I get for my $1599 (plus tax) over its predecessor? A disappointment. So far, it appears that the new “II” model might have more accurate focus, and might be sharper in the corners (or might not). Flare control is quite impressive and image contrast seems to be improved, but distortion is still very strong [example], making it unacceptable for architectural work without software correction.
Overall, the new 16-35mm f/2.8L II is an excellent performer, but it’s strong color fringing at 16mm mars an otherwise excellent performance. Color fringing can be corrected in software, so for some users this shortcoming will be of little concern. But at this juncture, I cannot recommend spending the money on the new 16-35mm f/2.8L II if you already own its predecessor.
I’ll be sending my 16-35mm f/2.8L (original version) to Canon service for a focusing mechanism adjustment so I can make a proper comparison between “I” and “II” (see April 3 blog entry).
I was a wee bit early with my “Murphy’s Law” prediction (it
took 5 weeks instead of 2), but the 8 core Mac Pro is now available at the 
.
If you are considering the Mac Pro 8 core option, be sure to read my March 13 blog entry discussing memory bandwidth—the 8 core Mac Pro is hampered by memory bandwidth (the rate at which memory can be read or written). Apple’s marketing claim of 21GB/sec memory bandwidth should be considered the drivel that it is—testing doesn’t support the claim, showing a maximum memory copy speed of 2.9GB / sec. As that includes reads and writes, we can conclude that for a real program, memory bandwidth is about 5.8GB/sec, and that’s if it’s not competing with any other cores for access to memory.
Ironically, any job that can actually make use of 8 cores (rare) will be hampered by the reduced memory bandwidth available. So in terms of real world performance, think of 8 cores as 6 cores when compared to the quad core Mac Pro—a step up worth paying for highly specialized tasks, but not for most users—I’ve yet to see all 4 cores of my quad-core Mac Pro at full utilization (except using diglloydTools, an artificial case).
See notes in my March 31 blog entry...I had planned to offer a detailed comparative review of the new Canon EF 16-35 f/2.8L II against the previous 16-35mm f/2.8L. I spent nearly 3 hours carefully choosing subjects and shooting the two lenses side-by-side, along with the EF 35mm f/1.4L and the Nikon 17-35 f/2.8. Lens testing is hard, really hard to do correctly, and critical focus accuracy is a prerequisite.
Analyzing the images today (which itself is more complicated than most people realize), I discovered to my chagrin that all my work is for naught because the older 16-35mm (apparently in perfect working order) front or back-focused on nearly every comparison, rendering the test images useless for comparing optical performance.
I might still offer the comparative review, but I’ll have to first send the 16-35mm f/2.8L to Canon for adjustment of its focusing mechanism.
When does “virtually eliminated” mean “still awful”?
Canon’s press
release for the new EF
16-35mm f/2.8L II states:
“Two UD lens elements virtually eliminate chromatic aberration”
One can only read and reread in disbelief such drivel when
confronted with the facts: the new 16-35mm f/2.8L II has some of the worst color fringing I’ve
ever observed in a professional lens, in spite of Canon’s bald-faced lie to the contrary (I am not happy
spending $1599 for a lens that in a direct comparison shows no improvement over its predecessor in this regard).
While it’s no worse than its predecessor in terms of color fringing at 16mm, it appears to be no better either. I’m puzzled how the image below (left) rises to Canon’s claim of “image quality that will meet the strict requirements of professional and high-end amateur photographers”.
Actual pixels crop, EOS 5D with Canon EF 16-35mm f/2.8L II @ 16mm, f8
Digital Photo Professional (left), Adobe Camera Raw (right) with correction
There is no evidence that the color fringing is due to misaligned optics, as I have obtained images from it with symmetric sharpness across the frame (and it is capable of quite sharp images). So I do not believe that I have an inferior sample. Also, the Nikon 17-35mm f/2.8 exhibits similar color fringing at 17mm. Apparently it’s very hard to eliminate color fringing at such wide angles, since both Nikon and Canon have failed badly, at least in the test images I’ve made.
But while Nikon’s Capture makes the issue irrelevant via outstanding software correction, Canon offers nothing in Digital Photo Professional to compensate. Maybe that’s what “virtual” means in the Canon press release—it’s not really there! While the color fringing can be corrected in Adobe Photoshop’s Camera Raw, one should not have to resort to an alternative raw-file converter to fix such problems; even an 8X12 print suggests a new and hitherto undiscovered birch species with a white center and brilliant red outer bark.
Indoor lighting can be really difficult to work with...or really interesting. Color-balancing skin tones under such lighting is next to impossible, so one just needs to go with the effect. And a color balanced image just wouldn’t capture the feel, so the “wrong” color balance is clearly the right one!
While the shot above was taken on a tripod with the Canon EF 16-35mm f/2.8L II, fat-moving handheld candids later in the evening were still a challenge even at ISO 3200 and f/1.4 using “fast” lenses. And I had to move fast, too. The Canon EF 35mm f/1.4L was indispensable, as was the Canon EF 85mm f/1.2L II. Only those who have not dealt with such lighting will express skepticism of the value of such “fast” lenses. It would have been nearly impossible to shoot at f/2.8, and very difficult at f/2. Perhaps the new Canon EOS 1D Mark III, with its ISO 6400 will offer more “reach” into the dark. Flash is simply not an option in such circumstances, being very distracting and downright annoying to wide-pupilled revelers.
Under such lighting, the Canon EOS 5D, unduly praised for its low noise performance, produces an uncountable number of bright speckles when an attempt is made to white balance. And a displeasing pattern noise (the worst kind!) develops as well. The upshot is that one must “go red” or use a monochrome rendition to avoid such effects.
Canon EOS 5D, ISO 3200
Canon EOS 5D, ISO 3200
A great place to take kids is the California Academy of Sciences in San Francisco. Near as I can tell, the species shown in my December 31, 2006 blog entry is the Monkeyface Eel, a species able to breath oxygen from the air. No wonder he (she?) didn’t look perturbed during extra low tide.
Monkeyface Eel Cebidichthys violaceus (Stichaeidae)
I don’t have any pretense against shooting through glass, as interesting images can result. And I stay a lot drier.
Snake Necked Turtle
Snapping Turtles—beware even of body-less heads
