Gandhi Indian Cuisine, Toronto

Concept: 3 out of 5
Execution: 4 out of 5
Yeah, but: Take a half-day from work. You'll need it.

The Long Version: Gandhi is one of the gems of Toronto - unquestionably the best roti in Toronto, I've been getting the same thing from them at least once a month for the last fifteen years. Chicken roti, usually medium (spicy); once or twice a year, I'll mix it up and get a medium chicken curry on rice. (On my first visit I ordered the beef roti, but it wasn't very good and has been dropped from the menu.) But that's a reflection on my taste, not a limitation of their culinary skills: the mixed vegetable roti, cauliflower roti, and chicken jalfrizi roti are all endorsed by people I've brought there, and I've been told by several authorities on lamb curries that theirs is one of the best. I don't doubt it, but lamb has never been my thing.

One unusual thing about Gandhi is that it's an East-Indian roti shop, so it's very different from the Caribbean roti that is much more common in Toronto. Unless you order the Channa, there's not a trace of chick peas to be found. Instead, potatoes are the other ingredient, and they're a great vehicle for the curries. The different dishes also have different curries, so it is worth trying different things, even though I've never managed to veer away from my favourite. Be warned, though, that this is not McDonalds, and there's some variety no matter what you do. Some days the Medium spiciness is nothing remarkable, other days it's so hot I can't eat it all in one sitting. Not that that's a bad thing, since the portion size is huge. Getting roti is the sign of a one-meal day.

Gandhi is at 554 Queen Street West, on the north side just east of Bathurst. It's worth ordering ahead of time if it's likely to be busy - like during daylight, or in the evening - since the waiting time for a friday lunch can be over an hour. Call them at 416-504-8155; if nothing else, it's a great way to skip the crush of people trying to get to the counter or fighting for one of the four tables. Their menu can be found on their website by either following the link to the menu or their chief's daily specials - they both go to the same page. They're open from 11:30 to 10pm.

It's also worth mentioning that Gandhi is closed on the weekends. That's either a sign that they're ridiculously successful, or that one of their chiefs is on an extended vacation. I suspect that those are both true; they've done this before. Regardless of the hassle of getting their food, it's absolutely worth it.

Stumpy Goes to Ottawa: The Nikon AF-S 105mm f/2.8 VR Micro

Concept: 2 out of 5
Execution: 3 out of 5
Yeah, but: It's a quick, stable, and available.

The Long Version: There are certain lenses that earn nicknames. Canon has The Brick, Olympus has the Big Tuna and Little Tuna, Sonica-Minolty has the Beercan, and we all have the Bigma. I don't usually name my lenses, even for my internal monolog - I think of them by their focal length - but I took a long look at the Nikon AF-S 105VR Micro lens on my D700, and thought: "Stumpy". So help me, I haven't been able to get it out of my head since then. Like the rest of this review, it's meant with affection, no matter how it sounds.

Physically, the 105mm f/2.8 is big: its length isn't that remarkable, but it's got massive girth on its side. It's also heavy, but considering its weight-to-volume ratio, it doesn't seem as heavy as it is - I had to put it on the kitchen scale before I believed that it actually was double the weight of the little 85mm f/1.8. The boxy-petal lens hood adds to its impression of volume, but I really like the aesthetics of the combination. On a DX or DXXX series camera, Stumpy looks big but balances well. On a DXX or DXXXX series, the smaller body might look better with the new micro-micro, the 85mm f/3.5 DX. I was able to play with one of them on a D300s as soon as they reached the store shelves - actually, slightly before that - and while it's a nice lens, it's not a replacement for Stumpy.

What I'm having a hard time deciding is what other lens, exactly, that Stumpy would replace. While Photozone and SLRGear disagree with each other, I'm going to side with the side that thinks that the 85mm f/1.8 is sharper. I'll also say that the VR makes a difference, giving more consistently sharp results overall. So compared to the lens that's half as expensive, it gains image stabilization in exchange for one-and-a-bit stops of light, and possibly loses peak sharpness. Incidentally, it's also twice as heavy and about a billion times the cubic volume.

Some people will look to Stumpy for portrait photography, but it's fifth out of four in my top recommendations for that particular task. Depending on the camera body and the working distance, for FX I'd be looking at the 85/1.4, 105/2.0 DC, or 135/2.0 DC as the top pick, with the 85/1.8 trailing behind as the budget choice. (For cameras that need motorized lenses to focus, the 50/1.4 AF-S is about the only reasonable choice. But, since I'm writing from the perspective of a full-frame camera, DX lenses are disqualified from the competition.)

All of those lenses will allow shallower depth of field and/or more control over the blur than Stumpy will. What Stumpy does offer is image stabilization, but that's not enough to ensure critical sharpness in the fuzzy-ears school of portraiture. Vibration Reduction isn't a universal replacement for tripods.

Since controlling the exact placement of the plane of focus is also absolutely critical for macro photography, camera shake isn't nearly as detrimental to sharpness as good old-fashioned front-to-back sway. Yes, VR will help run-and-gun close-up shots a little, but not as much as a twin or ring flash - and run-and-gun is an inherently bad way to get great photos. So when someone really cares about macro photography, Stumpy is never my first suggestion. The Nikkor 200/4, or the Sigma 180/3.5 and 150/2.8 all come before it, because there's no such thing as too much working distance for macros. The only exception is for doing copy-work, in which case you should look at the 60mm AF-S or AF-D lenses first.

So Stumpy isn't my first choice for portraits, although the focal length seems perfect, and it's not my first choice for macros, despite being capable of 1:1 reproduction. This must be seeming like a pretty grim review for a lens that I've already said that I like. The key to resolving this discrepancy is that while Stumpy may not be the best at either - or even in the top three or four lenses to use - it's because the others are more highly specialized. I probably wouldn't want the 200/4 Micro lens for taking portraits, and I definitely wouldn't get a decent macro shot with one of the exotic Defocus Control lenses. The 105VR is pretty good at both tasks in exchange for not being the best at either.

So I've come to think of Stumpy as a general purpose walk-around lens, and as that, it's one of the best that Nikon makes for full-frame cameras. It's a contest that it wins almost by default. No, it doesn't have the flexibility of a zoom, but the 24-70 is too heavy and expensive, and the lighter/cheaper 24-120 sucks. The 17-35 has a practical-to-wide range, but is still heavy and expensive; the 14-24 has the latter two traits and awesome optics without having the practical near-normal focal range. Neither wide zoom are great choices for the grab-it-and-go street photographers.

The story for primes is not much better; anything wider than 35mm has excessive 'character', and anything longer would benefit from image stabilization, which only Stumpy has. The photo of the clock tower was handheld at 1/13 of a second, taken before sunrise on three hours' sleep; either of those should rule out a sharp image, but Stumpy pulled it off. I shot it at iso800 f/2.8; with my 85/1.8 I would have been at iso3200/2.0. That photo also would have worked, but even on a D700, there are tradeoffs to boosting the sensitivity that high. It also would have lost the cool flutter in the flag.

Shooting at f/2.8 gives some control over the depth of field, and has a gentle transition from in-to-out. It's hard to say much about this; as I mentioned in my review of the 85/1.8, out-of-focus areas aren't something I see often in my photographs, and I'm not the most attuned to evaluating its appearance. So to make up for that, the next image is also shot wide-open, in this case at f/3.0. Apparently it's common for internal-focusing macro lenses to have their aperture gets narrower at closer focusing distances. At 1:1, Stumpy is at f/4.5; the non-IF 85PC-e is f/4.2 at 1:2 magnification. However, the Sigma 180/3.5 and 150/2.8 retain their widest apertures throughout their magnification ranges, as does my Olympus 50/2.0. Stumpy stops being an f/2.8 lens when it's nine feet from its subject, goes to f/3.2 at two feet, giving 1/4 magnification, and f/3.5 at half-life-sized.

This is the same tower from two photos ago, only this time I've focused on the fence just a few feet from the camera. Comparing the two images gives a little better idea of OOF rendering in a fairly tough setup. The lowest white disk is the clock, and it really does get dimmer at the edges. The triangle of yellow-orange above it are three windows that are lit by tungsten bulbs. Overall I would say that there's nothing to complain about here, and it might actually be doing a really good job. What I do notice is the vignetting: wide open, at whatever aperture that may be, it's fairly prominent. "Closed/Fermé" shows it clearly, and I can spot wide-open shots by their tumbnails in Lightroom.

This exceptional photo was taken facing away from Parliament Hill, with the early-morning sun still low in the sky and just out of the viewfinder at the top-left. I've cropped it to show only the lower-right quarter of the frame so that we can get a better look at Stumpy's flare characteristics. The shooting aperture was f/5.6 (iso200, 1/2000s) and only the default adjustments have been applied in Lightroom. I can't honestly say that the lens flares more easily than others, but when it does, it isn't pretty. With some wide lenses it can add an artistic flair, but with Stumpy, not so much.

On the long list of things that I'm not, a sports and action photographer is pretty high up there. Stumpy's autofocus is quick enough for street snapshots, but I haven't evaluated it for AF tracking with moving subjects, and probably never will. Compared to other macro lenses that I own of have used, I would put the 105's AF speed up in the top third; certainly better than average, and not significantly different from the 85/1.8. When I'm shooting with Stumpy, I don't need to make any extra allowances for it, and it doesn't frustrate me, which is increasingly the way I measure performance. On my D700, it's not as quick as the new Canon 100/2.8L IS on a 7D, but that's the fastest-focusing macro lens that I've ever seen. Not that fast autofocus or image stabilization are important for a macro lens, but they're great for a walk-around lens.

When I planned my day trip to Ottawa, I knew that I would be starting and ending in complete city-level darkness. Image stabilization would be very important, and Stumpy's Vibration Reduction let me hand-hold the camera down to about 1/15. That's about three stops, with good technique, which isn't as far away from Nikon's claimed performance as I would expect. (The MB-D10 might also have helped.) I also knew that the close-focus inabilities of the 85/1.8 would frustrate me during a day of shooting bilingual signs, making the 105VR a perfect choice. It's still heavier than I'd like, and after twelve hours I was getting tired of carrying it, but that's part of the inherent penalty of a full-frame sensor.

There were a couple of reasons to take Stumpy to Ottawa, and writing this review was one of them. A better reason was to test my ability to do medium-haul road trips, which is something that I plan on doing a couple of times a year. Six to ten hours overnight on the Greyhound, a day in a particular locale, and then back home that night. Packing the right gear into my T2 bag is as important as an iPod and a comfy pair of shoes - capability, complexity, and weight all have to be considered. For Ottawa, I carried Stumpy, my 50/1.8, SX20IS, and PCM-D50 audio recorder; far too much weight, especially considering that I was carrying a second set of batteries for each device. Next time I won't need the Canon camera - I used it for the gear shots here, as well as for sneaking photos in no-camera zones - but I might also consider using the 35/2 instead of the 50/1.8 and bringing the lighter 85/1.8 instead of Stumpy. The deciding factor would likely be the amount of daylight I have. With Ottawa, I arrived two hours before sunrise, and that was usable shooting time that I would have lost without the VR. Hopefully Nikon will invent In-Body stabilization some day, but for now, this is my best choice.

It might seem like I'm still on the fence about this lens, and to some extent, I am. It's an excellent lens, but it's not the best at anything; it has a useful capability that's unavailable in smaller lenses and/or shorter focal lengths, but it only compensates for Nikon deliberately crippling its cameras in order to leverage a lens technology that's only ten years old. I like Stumpy, and enjoy using it. It's great for general duty, and will often do a good enough job that I won't take the time to switch to my 85/2.8 or 180/3.5 when I need quick close-up photos. It suits my personal photography better than any other lens that I own, so it's usually the one that I reach for when I won't be far from home. I like it, I enjoy it, but when someone says to me, "Matthew, I need a lens for ________", Stumpy probably won't be the first one I suggest. But if you can't think of anything in particular that you need it for, and like short telephotos, the Nikon 105mm f/2.8 VR Micro is a great lens to have.

Instead of littering this with self-referential links, here's some of the other relevant stuff I've reviewed: Nikon 50/1.8, 85/1.8, and 85/2.8PC-e, Sigma 180mm and 150mm, Olympus 50mm f/2, Canon SX20IS, Sony PCM-D50 (case), and T2 messenger bag. For more photos, including some larger versions of these ones, you can use the tag Ottawa on my other blog.

Nikon 85mm f/1.8D AF Nikkor

Concept: 2 out of 5
Execution: 3 out of 5
Yeah, but: It's absolutely perfect, circa 1990.

The Long Version: The f/1.8 85mm lenses of the world don't seem to get much love. Overshadowed by their faster cousins, both the Canon and Nikon 85mm f/1.8have fantastic optics for a not-unreasonable price. But they aren't glamourous enough to attract the high-budget amateurs, and they don't zoom to attract the every-length-covered crowd. With a focal length that's a narrow-normal on a 135 format, or middle telephoto on a cropped sensor, it's not a remarkable way of seeing the world. It's a lens that will appeal most to practical people who just want to take photos.

So how I ended up with one is a massive mystery.

One of the nicest things about the 85/1.8 Nikkor is going to sound odd: it's a satisfyingly lens-sized lens. It's not a monster like the other 85's: Nikon's 1.4 and 2.8, or Canon's 1.2; it isn't an absurd zoom or super-tele, and it's not even too petite like some of the wide-angle lenses. With its solid metal screw-in hood - included with the lens - it suits a camera from the DXX or DXXX series perfectly. It has some heft to it, but it's not too much to carry around all day. Naturally, size is a personal thing, but from the first time I used the 85/1.8 it has felt like a classic. That's an impression that's let down a bit by the plastic lens barrel; but about the only modern lenses that are still made the old-fashioned way are the Pentax Limiteds. I won't hold that shortcoming against the Nikkors.

Optically, there are no significant compromises with the 85 f/1.8. It's never really soft, and it sharpens up across the entire frame once it's stopped down just a little bit. (In daylight, even the weak mid-winter stuff we get in Toronto, a 1/8000 shutter speed isn't always fast enough to shoot wide open.) The vignetting is essentially gone by f/2.8, and its complete absence of geometric distortion is enough to move me to tears. I feel very strongly about straight lines.

While its big f/1.4 brother is supposed to be the bokeh king of the family, the 1.8's not too shabby. I'm not a huge connoisseur of this aspect of lens design - I hear a lot of people pronounce it bouquet, as if photographer rhymes with sommelier - so it's not my greatest concern or something that I'm gifted at evaluating. It does have a bit of an annulus (donut) shape to its OOF highlights, but take a look at the headlights on the subway train above. Shot at f/5.6, the nine-bladed aperture does a great job of keeping the highlights circular. Fo what it's worth, the next photo is also shot at f/5.6.

The autofocus is a traditional screw-drive, so its speed will depend on the camera for both the momentum and the math. For example, on a D700 with its fancy MultiCam 3500 system, it focuses quickly enough to track a cyclist who's hopping his way up a flight of stairs. On a DXXXX series camera, the autofocus is going to pretty much suck - but maybe I can't hold that against the lens. It's not the zippiest, but for general use it shouldn't be an issue. If focusing speed is critical, then as always, try it yourself with the camera that you'll use it on. That will answer far more questions than reading any number of reviews will.

This is the first photo that doesn't include the colour red, so it's an apt introduction to my counterpoint: not everything is rosy with the 85/1.8. Photo 13, above, is the full frame at f/2.0. There's moderate falloff, and the image isn't as sharp as it could be. But nitpicking aside, it's a dated design, boasting neither image stabilization nor a speedy Silent Wave focusing motor. Adding those things to an optical formula that's essentially flawless is asking for trouble, can only increase the cost, and will essentially turn the 85/1.8 into a completely new lens. If they're doing all that anyway, I'd also want a minimum focusing distance that's much shorter than the three feet that the AF-D 85/1.8 has. In my usage, that's turned out to be my biggest and most consistent complaint with this short telephoto lens, and it's one that can't be overcome by shooting technique or a better camera body.

So: my ideal short telephoto would be medium-small, and have AF-S, VR, a high reproduction ratio, and a fast aperture. Now that I think about it, that sounds a bit like a lens that Nikon has already made.

Thewsreviews New Look

Concept: 2 out of 5
Execution: 3 out of 5
Yeah, but: I'm biased.

The Long Version: When I first decided to go back to writing for a hobby, I never really expected anyone but me to see it. As a result, I chose a template design that I like for my own reason, and not the one that's the simplest, most legible, or most appealing. Making matters worse, over the past eighteen months I've increasingly gravitated to simple monochrome layouts, and shot-on-white photography. The look of the blog was not working the way I wanted it to.

The new layout is a better match to my other websites, looks more like my photography, works better with my Blackberry, and uses Verdana throughout, which is a more legible humanist sans-serif and a better match for my usual Myriad. I like it, and hopefully it makes the site easier and more pleasant to read.

And a bit of thewsreviews trivia: I picked the old template as a tribute to the original blogger site for The Online Photographer. I've been playing with a new layout on my test site for a little while now, and finally made the switch in honour of TOP's fourth birthday. I've previously mentioned that the Luminous Landscape is one of my top three influences on this blog; TOP and Mike Johnson is one of the others.

Spinning Tokens - The 2010 TTC Fare Increase

Concept: 2 out of 5
Execution: 1 out of 5
Yeah, but: If only they'd done this before.

The Long Version: TTC tokens are great little things. Unlike the paper tickets that the Toronto Transit Commission phased out in 2008 (for adult fares), they don't get ruined if the go through the wash, they work in the automatic turnstiles that the TTC uses in all of its entrances, and they're more difficult to counterfeit - notwithstanding that the new bimetal design was introduced to get rid of five million fakes of the old aluminum design.

A TTC token has no face value - they're worth one ride, no matter how much that costs. The 2010 fare increase makes them the safest investment in town; their value is guaranteed to jump from $50 to $55 per troy ounce. According to information from the TTC, reprinted by the Globe and Mail newspaper: "Although there are about three million tokens in circulation, staff rely on about 400,000 tokens being recycled throughout the day in order to keep from running out." That means that there's only a seven or eight day supply in the system, and now people have a good reason to hang on to them. What could possibly go wrong?

The twenty-five cent fare increase was announced two months before it comes into effect, so it's no surprise that the TTC has had a surge in demand. They quickly - but not immediately - rationed token sales before eventually stopping them completely, which is the same thing they did the last time they increased the transit fares. But the TTC also has a corporate culture that emphasizes blaming their customers when things go wrong, which includes people actually using the transit system. The Toronto Star recently published an article "Hoarders Foiled as TTC Halts Token Sales" which is one of many articles that contains my favourite quote of the week: "TTC chair Adam Giambrone said the commission could have lost $5 million had it not suspended token sales."

Five MILLION Dollars! (and Curses, foiled again!)

To have their customers stockpiling tokens cost them five million dollars of lost revenue, then we would need to be holding on to every single token ever minted and have the fare increase by $1.67. Even the earlier unspecified loss of 'millions' in unearned revenue from customers buying their fares at a lower rate is implausible when the fare is only increasing by twenty-five cents. One million tokens, a third of the total in circulation, will increase in value by only $250,000 unless there's something seriously wrong with my math.

But token hoarding works both ways. The TTC is spending $50,000 to print temporary paper tickets that will only be valid for two months, and they'll need to be topped up by adding an extra 25¢ to the farebox each time they're used in January. The TTC won't be selling any more tokens until the fare increase comes into effect, which means that the TTC will also be hoarding tokens to its financial advantage. They'll make back the $50,000 investment - twenty-five cents at a time - by not selling a mere 200,000 tokens until after the price increase. By their own numbers, that's an average half-day's volume, and only one-fifteenth of the three million tokens in circulation.

More fun with numbers: inspired by information supplied by the TTC, the Globe and Mail writes in an article from November 22, "Ticket hoardings were costing the TTC an estimated $45,000 a day as frugal riders bought 20 per cent more tokens than they needed, stocking up in anticipation of the increase." [Note to the editors: Hoardings are temporary wooden fences. The correct phrasing should be 'Ticket hoarding was costing' - but I digress.]

A twenty percent increase on four hundred thousand is sales of an extra eighty thousand tokens a day, and at 25 cents each that's a potential loss of $20,000 - not quite half of the amount that the blame-the-riders movement is trying for. But even if the claim that the TTC makes of $45,000 a day is true, it's still less than half of what they'd need to reach their claimed loss of $5 million, even assuming that their losses started the day the fare increase was first proposed. And just for fun, let's remember that these costs are potential revenue that they won't get from the fare increase, which is valued at some $65 million dollars over the course of the year, and most of which will come from more expensive transit passes.

The TTC is using implausible numbers to villainize a huge number of its own customers, in what seems like a calculated effort to make themselves look like victims when people don't want to pay 10% more for the same intermittent-to-lousy service. This is nothing new, but it controls the discussion and intentionally directs it away from talk about how the system - transit and funding - is broken, and what can be done to fix it. Costs go up; the day-to-day and year-to-year fumbling of the status quo remains unchanged.

There's no question that the TTC is underfunded. There's also no question that the senior management hates its riders. Could those two things be related?