Intel X25-M firmware updatePERMALINK

I sucessfully updated the firmware on the 80GB Intel X25-M that I normally use for my boot drive.

Prices have just dropped— if you have a Mac Pro, the X25-M is a great way to make it feel lightning fast as your boot and applications drive (always separate system and data, which is easiliy done on a Mac Pro).

You can get the updated firmware at Intel.

Intel X25-M solid state drive (SSD)

I’m currently booting off a 64GB Intel X25-E (“Extreme”) on both my Mac Pro and MacBook Pro (units on loan from Intel). The X25-E is really fast, but not noticeably different than the X25-M, because nearly all access to the system disk is read access. The X25-E is the same speed for reads, but shines at write speed (170MB/sec), about twice as fast as the 80MB/sec or so of the X25-M.

Color uniformity of the Voigtlander 20/3.5 Color-SkoparPERMALINK

I’ve added a color uniformity assessment of the Voigtlander 20/3.5 Color-Skopar to my review.

Uneven color renditionPERMALINK

Why can’t simple tests be simple? Like checking for uniform color across the frame.

Mouse over the image below. Both variants show color non-uniformity, a fairly common phenomenon with wide angle lenses, and not specific to the Voigtlander 20/3.5 Color-Skopar used here. Symmetry is my expectation, but it’s asymmetric.

One image shows a too-blue right side, and one image shows a too-blue left side. The difference? I shot the first image with the camera right-side-up, and the second one with the camera upside down, to rule out something funny with my target or the lighting. Well, the not so funny business is with the lens and camera!

Nikon D3x + Voigtlander 20/3.5 Color-Skopar @ f/8

Is my D3x sensor still off? It seems not, since the same lens on the Canon 5D Mark II shows the same effect. And so do other lenses! I don’t have a good explanation offhand, but this is why a 30-minute test takes a few hours to shoot and shoot again when oddball findings present themselves. My reviews don’t just test, they teach, because I take the time to present and explain such things.

Focus shift articlePERMALINK

Look for my article on focus shift in Photo Techniques magazine, highly recommended. Previous issues have articles on Optimizing Photoshop, Diffraction, Digital Infrared and more. More articles are coming to future issues.

Want to subscribe? It’s easy at Amazon.com.

May/June 2009

Update to Zeiss ZF Lenses for ZF 21/2.8 Distagon now onlinePERMALINK

Just published are my additions to Zeiss ZF Lenses for the Zeiss ZF 21/2.8 Distagon, including 4 comparisons, one of which is using the Canon EOS 5D Mark II, the others the Nikon D3x.

Direct link for subscribers: revision history.

Nikon D3x + Zeiss ZF 21/2.8 Distagon @ f/5.6

Voigtlander 20mm f/3.5 Color Skopar SL ASPH comparedPERMALINK

Hot off the heels of today’s earlier update, I’ve published an in-depth comparison of the Voigtlander Color-Skopar 20mm f/3.5 SL ASPH to the Nikon 20/3.5 AI-S and Nikon 14-24/2.8G zoom.

Direct link for DAP subscribers: Voigtlander page, and the comparison page.

Nikon D3x + Voigtlander 20/3.5 Color-Skopar @ f/10

Voigtlander 20mm f/3.5 Color Skopar SL ASPH reviewPERMALINK

Just published is my review of the Cosina Voigtlander Color-Skopar 20mm f/3.5 SL ASPH.

Direct link for DAP subscribers: Voigtlander page, and the 20/3.5 main page.

Nikon D3x + Voigtlander 20/3.5 Color-Skopar @ f/10

Update to Zeiss ZF Lenses for ZF 21/2.8 Distagon now onlinePERMALINK

Just published are my additions to Zeiss ZF Lenses for the Zeiss ZF 21/2.8 Distagon, including 4 comparisons, one of which is using the Canon EOS 5D Mark II, the others the Nikon D3x.

Direct link for subscribers: revision history.

Nikon D3x + Zeiss ZF 21/2.8 Distagon @ f/5.6

Update to Zeiss ZF 21/Distagon tomorrowPERMALINK

I worked diligently yesterday and today to prepare 3 comparisons with the Zeiss ZF 21/2.8 Distagon, two on the Nikon D3x and one on the Canon 5D Mark II (and that’s just the writing part, making a solid comparison is hard work, and often requires shooting many subjects, plus careful cross-checking).

I’ll be publishing the comparisons tomorrow in Zeiss ZF Lenses, as well as a supplementary page Alignment Woes. The comparisons will include the ZF 21/2.8 Distagon, as well as zooms and the Voigtlander 20/3.5 Color Skopar.

The Voigtlander 20/3.5 Color Skopar will also get attention with comparisons soon in DAP, but the head-to-head results with the ZF 21/2.8 Distagon will be found only in Zeiss ZF Lenses. In brief, if you’re looking for top sharpness, you’ll have to pay for it.

Voigtlander, Zeiss and computer meltdownsPERMALINK

I’ve prepared both Voigtlander 20/3.5 and Zeiss ZF 21/2.8 comparison pages to be published soon. The comparison with the ZF 21/2.8 Distagon will be in Zeiss ZF Lenses (only), but the bulk of the coverage of the Voigtlander 20/3.5 will be in DAP. I had planned to publish today, but cross-checking results took priority.

To wit, f/8 and be there, and you get what you pay for.

A computer meltdown last week threw a monkey wrench into my schedule, though I learned a great deal (paid for in hair), much of which will make it into Macintosh Performance Guide over time. Let’s just say I won’t be recommending a certain brand of data-destroying RAID card anytime soon, and that I’m very, very pleased and relieved that Sonnet has their venerable Tempo E4P working on the Mac Pro Nehalem—two of them are installed in mine now.

Separately, an iMac can work as a photographer’s workstation—but with a lot of screwing around to achieve the level of reliability and backup and speed I require for my data. I now have a lot of experience with the iMac 3.06Ghz, and I can state unequivocally that unless it’s a space constraint issue, run screaming from the room and go straight to the Mac Pro. And if you want help, I offer consulting on how to set up the Mac Pro for performance and reliability and backup.

I’ll also be reporting on how the screaming-fast Intel X25-E solid state drive works in a Mac Pro and MacBook Pro. The X25-E is a proxy for what we can expect in the next 18 months from solid state drives, which is to say that hard drives are dinosaurs whose fate is sealed. I’ve been booting off the Intel X25-M for some time now, and the X25-E only improves matters.

Voigtlander Color-Skopar 20mm f/3.5 SL ASPHPERMALINK

Just arrived, the new Voigtlander Color-Skopar 20mm f/3.5 SL ASPH, courtesy of a reader who traded a loan of 3 weeks for my subscription offerings—a win-win for both of us

I’ll be reporting on the 20/3.5 in DAP as well as here. I’ve already done some shooting and comparisons, results to follow within a few days.

My initial impressions: what a gem! Build quality like the Zeiss ZF line, and it’s diminutive, about 1/3 the size of the Zeiss ZF 21/2.8 Distagon, and I like that a lot—it makes me want to carry it, just like the Voigtlander 40mm f/2 Ultron ASPH.

Optically, I can’t say just yet, but I will soon figure that out, and it should prove interesting next to both the Zeiss ZF 21/2.8 Distagon and the old Nikon 20/3.5 AI-S.

Voigtlander Color-Skopar 20mm f/3.5 ASPH

The 20/3.5 Color-Skopar is available at CameraQuest, along with the 40/2 Ultron and 58/1.4 Nokton.

Lines of confusionPERMALINK

I shot this image seeking a result to confuse and disorient. It annoys the heck out of me — I can’t rest my eye anywhere — so I think I succeeded. Would the IRS be interested in a large print for the audit room?

Line of Confusion
Nikon D3x + Zeiss ZF 21/2.8 Distagon

Nikon D3x happierPERMALINK

As I mentioned a few days ago, Nikon turned around my D3x in record time.

Preliminary field shots suggest that the issue I suspected might now be gone, but I have to do the tedious work of checking this rigorously—only then am I prepared to publish my comparisons with the Zeiss ZF 21/2.8 Distagon in Zeiss ZF Lenses. PITA.

Single CPU Mac Pro NehalemPERMALINK

I’ve updated my review of the Mac Pro Nehalem to include test results comparing the single CPU to the dual CPU model (both 2.93GHz).

The results are generally what one might expect, but with some perverse results for Photoshop CS4, which I explain in When More is Less. We’ll probably have to wait for a (hopefully) 64-bit CS5 to see this problem go away, as milking the cows seems to be the priority for now.

As shown below, a single-CPU Mac Pro handily beats a dual-CPU Mac Pro on the diglloydMedium benchmark by 45 seconds to 61 seconds. Anyone working with large file(s) that use the scratch volume should sit up and pay attention here.

Dual CPUs are slower than a single CPU!
"1/2 cores" means that half the CPU cores were disabled

I also intensively tested the OWC 4GB modules in the single-core Mac Pro and they show no issues whatsoever—rock solid, just as in my dual-core Mac Pro. At $1749 for 32GB versus $6100 from Apple, the OWC product is the right choice. Single-CPU Mac Pro Nehalem users have only 4 slots, so four 4GB modules are the only way to get to 16GB (or 12GB).

Adding 8GB memory to the MacBook ProPERMALINK

See also: my review of the 2008 unibody 15" and 17" 2009 MacBook Pro as well as Pimp my MacBook Pro.

I just got word (tonight) from OWC that several models of the latest incarnation of the 15" MacBook Pro now support 8GB of memory, just like the latest 17" model.

This is a big deal for me: I do not want to lug a 17" laptop when traveling, but I also want 8GB in it. So now that’s possible, and there’s good reason to want 8GB, not the least of which is working with many files or large files in Photoshop. See Optimizing Photoshop as well as my previous review of the OWC 6GB/8GB kits.

MacBook Pro models that support 8GB memory: the MacBook Pro March 2009 2.66GHz and 2.93GHz models (15" or 17"). The OWC memory page is not yet updated to reflect this finding, but will be soon.

The current and former 2.4GHz models, and slightly older 2.53GHz and 2.83GHz models are limited to 6GB, a real shame for anyone who bought a little too early—remember that when Apple introduces a new machine, waiting 4-6 months to buy is wise. Of course, the 1

How does OWC know this? Because they go to the considerable expense of actually buying the Macs for which they sell memory—remember that the next time you buy memory for your Mac—support vendors that do it right.

Reader comment on Zeiss ZF 21/2.8 Distagon reviewPERMALINK

I cannot do any better than providing clarity as to whether the 21/2.8 Distagon (or any ZF lens) is the right lens for you (or not). That’s my goal in Zeiss ZF Lenses, and why so many people have found it a great investment (which BTW actually offers value well beyond the ZF/ZE line).

Rabbit Heaven
Nikon D3x + Zeiss ZF 21/2.8 Distagon

Reader Saad G writes about the recent additions for the ZF 21/2.8 Distagon:

I read your review and must say that it is fantastic, because it is enlightening, because it is clearly understood and because now I ___ buy one... Looking forward to your resolution tests.

Saad’s buy/not-to-buy decision is not the point; the point is that I detail the strengths and weaknesses of the entire ZF line so you can decide which lenses are best for your individual photographic goals.

Review of the Zeiss ZF 21/2.8 Distagon is in progress, with my focus now turning to comparisons with the Nikon 14-24/2.8G and Canon EF 16-35/2.8L II. And, if I can dig one up, the new Voigtlander 20mm f/3.5 (any reader out there is welcome to loan me one, but I’ll get an eval unit sooner or later). I do have the Nikon 20/3.5 AI-S, and might show that also.

Canon users should remember that the 21/2.8 Distagon is eminently usable on Canon with a lens adapter.

Fantastic turnaround from Nikon servicePERMALINK

I sent my Nikon D3x into Nikon on Monday, overnight, arriving in New York (from California) on Tuesday). Today, Thursday, I received it back. Wow! Now that’s service. Thank you Nikon.

I sent the D3x in over concerns about some right-side softness I was seeing. Only shooting the serviced body will tell me if that’s changed, and unfortunately the computer printout doesn’t go into detail on what was done in that regard. Shooting will prove it out. With a 24MP camera, tiny deviations in the lens or camera alignment become important, and that’s not a Nikon thing, it applies to any high-res DSLR.

I also had some trouble with my 14-24 zoom (one reason comparisons have been delayed), and Nikon replaced it. Wow! That goes way beyond the call of duty there, but it does mean I have to test and “certify” the replacement for my testing purposes—some tedium.

Zeiss’ Dr. Nasse on MTF, part IIPERMALINK

Dr. H H Nasse has published How to Read MTF Curves part 2 white paper. This follows part 1 in December. Well worth reading.

Zeiss ZE Lenses on Canon EOSPERMALINK

In Zeiss ZF/ZE Lenses, I’ve added a page describing the differences between the Zeiss ZF and ZE lines as well as why even Canon users might want to use the ZF (Nikon mount) versions.

Zeiss ZE 85/1.4 Planar (left), ZF 85/1.4 Planar (right)

Zeiss ZF 21/2.8 Distagon review postedPERMALINK

I’ve posted the first installment of my Zeiss ZF 21mm f/2.8 Distagon review, part of Zeiss ZF Lenses.

The first installment includes images from the Nikon D3x and Canon 1Ds Mark III.

The 21/2.8 Distagon is one heck of a nice lens, and I think it lives up to its reputation (from the Contax world), though Zeiss says it has been recomputed and improved eg better.

I usually shoot my lenses without a filter, but in “iffy” places like the ocean or sand dunes or sticky-fingered and snot-nosed children, be sure to get an 82mm filter for the 21/2.8 Distagon.

The ZF 21/2.8 Distagon is in stock at B&H Photo. And while you’re at it, bite the bullet and get the one lens you’ll never regret: the 100mm f/2 Makro-Planar.

Deck Planks
Canon 1Ds Mark III + Zeiss ZF 21/2.8 Distagon

Nikon Capture mysteriously behavesPERMALINK

I’ve been fighting a severe speed problem with Nikon Capture NX2.

On a hunch, I shut down my Mac Pro Nehalem, and swapped 32GB of memory for 16GB.

Guess how much faster Capture NX2 ran? About 15X faster— yes at least fifteen times faster. What took 15 minutes now takes 1 minute.

The only puzzler: after swapping back to the 32GB configuration, the speed problem remains banished (NX2 still makes use of only about 1/5 of the available processing power).

Well, i’ll be happy if it stays that way so I’m crossing my fingers.

Zeiss ZF 21/.28 Distagon — progress updatePERMALINK

A tremendous amount of shooting and analysis and writing has now gone into my report on the new Zeiss ZF 21/2.8 Distagon, and I’ve gained some solid insights in using it over the past week.

I had hoped to publish a first pass today in Zeiss ZF Lenses (subscriber access here), but I’ve made the decision to defer that until this Wednesday, in order to complete a key portion on vignetting and color. Your patience is appreciated,

On thing the ZF 21/2.8 Distagon does not suffer from is color fringing, which is as well controlled as anything I’ve ever seen in a ~21mm lens. It’s so well corrected I don’t feel any need to use Nikon’s Auto Color Aberration feature.

The ZF 21/2.8 Distagon is in stock at B&H Photo, but order before the April 8 holiday closure (Passover). Although it’s Nikon mount, with a lens adapter Canon users can shoot any ZF lens, and the 21/2.8 is likely to become a favorite for Canon EOS, barring some outstanding new effort from Canon.

Shopping Carts
Nikon D3x + Zeiss ZF 21/2.8 Distagon @ F/5.6, ISO 800

Nikon Capture NX2 ProblemsPERMALINK

I struggled to use Nikon Crapture yesterday with the same slow mode problems I described in Dec 2008.

As I write this, NX2 is on file 8/11 in batch mode, having taken ~20 minutes to get that far on my 2.93GHz 16-core Mac Pro with 32GB memory.

Since my April 1 post about Nikon Capture NX2, Nikon PR contacted me. In my email response to Nikon PR regarding Capture NX2, I detailed the following (partial) list of NX2 problems:

• “Slow mode” a state the program enters without fail within a minute or two, in which operations that should take a few seconds take minutes and/or never complete (the program hangs). Sometimes it's just ultra slow, and sometimes it just hangs and never recovers.Both interactive and batch mode are affected.
• With batch processing in 2.2.0, I observe CPU usage in the range of 7% of one CPU core, with 16 cores on my Mac Pro Nehalem 2.93GHz. Alternately, sometimes I observe 100% of one core in use, with half the CPU cycles being system (red, indicating a likely busy-wait scenario). Most of the threads are blocked on a pthread semaphore.
• Very poor cpu usage even when things are working normally, typically 1.5-2 cores out of 16 available. There is ample free memory, ~600MB used by NX2.
• opening more than one NEF can confuse the program; clicking the close box closes the wrong window.
• Opening the first NEF might take a second or two to draw the screen, opening more, particularly 3 or more can cause drawing to take minutes or hang the program. Closing the front window frequently closes a window behind the one whose close box was clicked!
• The batch feature doesn't allow queuing more files once a batch is in progress. Defeats the whole point.
• A crash 100% of the time of the NkMc process when launching NX2 or ViewNX.
• "Unsupported file type" message in View NX. D3x files right out of the camera generally cannot be viewed.
• Many more bugs than this.

For Capture NX2, I requested contact with the lead engineer for NX2 (being a veteran software engineer myself). I don’t know why Nikon couldn’t just do that and I’ve had gladly spent time on the weekend if necessary. Where is the sense of urgency to fix customer problems? We shall see what this week brings.

I’ve uninstalled and reinstalled both Capture NX2 2.2.0 and ViewNX, thinking this might reset the brain damage somehow. All I got for my trouble was loss of my NX2 license (I couldn't locate my serial number), so I’m now running in trial mode.

A few readers have written to say that Nikon Capture NX2 is working for them (mostly PowerPC users), and just as many have written to express their dissatisfaction.

I’ve sent multiple program dumps to Nikon PR (why am I working with PR on a technical problem instead of an engineer?!!!). The geeks among you can view the thread dump — nearly all the threads are blocked on a semaphore.

Hopefully Nikon PR will actually route these to an NX2 engineer who will fix the problem. Actually, are there engineers working on NX2?

Zeiss ZF 21/2.8 DistagonPERMALINK

Like the rest of the Zeiss ZF line, contrast is exceptional with the ZF 21/2.8 Distagon. Canon users are going to want to look hard at the 21/2.8, even though it means shooting with a lens adapter, manual focus and manual stop-down (on EOS).

How many points on a ZF 21/2.8 Distagon diffraction star? Hint: it has 9 aperture blades.

This image is straight out of Capture NX2, no special processing, just great image quality that can easily be enhanced. At full screen on a 30" monitor, the effect is engaging

Wide Dawn
Nikon D3x + Zeiss ZF 21/2.8 Distagon at f/8

Readers of Zeiss ZF Lenses enjoy a wide variety of high-res examples, chosen to show off the lenses technically, and (when fortune smiles) artistically.

Gitzo GT0530 carbon fiber tripod failurePERMALINK

So far my Gitzo tripods have held up well over years of use, though I’ve had several reader emails griping about their own Gitzo experiences.

Now I have my own bad experience to add: my very lightly used GT0530 has developed quite a bit of “play” in one of the legs where it connects to the metal ferrule at top, as if it’s coming unglued. It absolutely needs repair.

Though my favorite is the GT3541XLS, I like the GT0530 (or GT0531) not because it is robust, but because it weighs next to nothing, can be hooked over a belt, and works OK with mirror lockup + remote release with even a D3x (with some care). I use it with the Really Right Stuff BH-25 Pro mini head.

Nikon makes contactPERMALINK

Nikon public relations made contact with me regarding both Nikon Capture NX2 (see prior comments) and D3x sensor/lens mount alignment tolerances, which concern me based on seeing how two D3x bodies perform with both the Nikon and Zeiss wide angle offerings.

I’ll be sending my D3x in for a health check and so the ZF 21/2.8 Distagon will see some “Canon time” which is how I’ve shot the ZF line extensively in the past.

Zeiss ZF 21mm f/2.8 DistagonPERMALINK

Previous comments.

I’m working on my full report on the Zeiss ZF 21mm f/2.8 Distagon, having shot the tedious comparison stuff with 4 samples and the Nikon 14-24 and Canon 16-35/2.8L II (some long days of effort already).

My writeup in Zeiss ZF Lenses will follow as quickly as possible, which means this weekend for initial coverage, but that’s not a firm deadline—quality comes first. Properly assessing a lens takes at least several weeks of shooting under a wide variety of conditions, to yield insights not accessible with limited usage. I will be focusing at first on some comparisons, distortion and initial impressions, with much more to follow.

There are some tolerance issues around short focal lengths (not specific to Zeiss) having to do with the camera body lens mount and sensor alignment, but it appears that this is just Life With 20+ Megapixel DSLRs today, not particular to any brand lens. But it does mean that any particular ultra-wide lens of any brand and your DSLR might not be perfect friends across the frame, something you can explore yourself and see with proper care and choice of subject. Details will be forthcoming in my full report in Zeiss ZF Lenses, well worth your money.

What’s clear in both resolution chart and field shooting of the ZF 21/2.8 Distagon is that there is simply no comparison. Which is to say that neither the Nikon 14-24/2.8 or the Canon 16-35/2.8L II can match the contrast and sparkle of the ZF 21/2.8 Distagon.

The Canon 16-35/2.8L II looks positively murky compared to the ZF 21/2.8 Distagon (shot on EOS with a lens adapter).

The Nikon 14-24 handily wins on the distortion front, and fares better than the Canon 16-35, but otherwise lacks punch compared to the ZF 21/2.8 Distagon.

You can get the Zeiss ZF 21/2.8 Distagon at B&H Photo.

Is Nikon Capture NX2 an April Fool’s joke?PERMALINK

I downloaded and installed the latest Nikon Capture NX2 version 2.2.0 today.

If there’s one reason not to shoot Nikon, it’s Capture NX2; I use it because it offers excellent image quality as well as correction of lateral chromatic aberration and (now with 2.2.0) axial chromatic aberration. It also offers vignetting correction and other goodies. A very nice feature set, world’s crappiest user interface.

And in terms of performance and bugs, it still remains the biggest pile of steaming shit ever foisted on digital shooters. On a 16-core 2.93GHz Mac Pro Nehalem, it’s still slow as mud, and still full of bugs that cause it to take take minutes to process a single image—if it doesn’t hang forever. As a programmer I can see there’s something very, very wrong with it internally (programming bugs), but my emails to various Nikon representatives go unanswered—maybe because they took a dislike to reality in The Good, the Bad, The Ugly.

If you know the head of NIK (Nikon’s software group), have him contact me please, these guys are oblivious to what manure they’re shipping to hapless customers (and charging for it).

After 3 minutes or so on a single file, one does lose patience

Helper application NkMC crashes every time when launching Capture NX2 or NikonView

But wait, it gets worse. During batch processing, Nikon Capture NX 2.2.0 now uses (on average) ~7% of one of the 16 available CPU cores eg 0.4% of the available processing power of the Mac Pro, taking half an hour or so to process 15 Nikon D3x NEF files. That’s not a typo: zero point four percent. Unbelievable.

Nikon Capture NX2 CPU usage during batch processing, not an April Fool’s joke