oscar_van_der_velde

March 16, 2012

Depends on what you consider distortion. An ultrawide will create dramatic perspective with strongly converging lines creating a false perception of how big the place is while objects near the edges look warped unnaturally (unless you look around the image with your eye at the screen).

A stereographic fisheye (there is only one: Samyang 8mm) will bend straight lines, does not exaggerate the size of a room as much, and small objects, for example circular plant pots, keep their natural shape very well:

March 12, 2012

Interesting that DPP also gets the ability to do HDR, a relief because I have not seen any in-camera JPEGs from the 5D III that actually looked sharp (they really look like 10 megapixel images resized to 22 megapixels). It would be nice if DPP could also do that based on files from earlier cameras. The Digital Lens Optimizer sounds interesting, but will not work with raw files from the old 5D! What information does the 30D include that the 5D does not? Shooting distance information? Too bad, because the current Lens aberration correction function simply allows you to drag the shooting distance slider. Anyway, I'll be out of luck with my 3rd party lenses. Of course, if this DLO is indeed a great improvement it will be more attractive to buy those ever more expensive Canon EF lenses instead of Sigma, Tamron, Tokina, Zeiss, Cosina-Voigtlander or Samyang. An interesting note is that CR2 files will increase 2-3 times in size after applying DLO. Wonder what happens there...

March 6, 2012

Bob Atkins wrote:

But the pixel density of the D800 is only the same as that of a 15MP APS-C sensor. I don't see EOS 7D users lamenting that they need to use manual focus and live view in order to get the best results out of their 18MP sensor.

</blockquote>

Those 18 MP sample a certain field of view, so you have x pixels/angular degree. Use 36 MP to sample the same field of view and you will have 1.414x pixels/degree. If your technique is the limiting factor, you will produce the same image quality with either camera, e.g. your camera shake will remain a fixed percentage of the total angle of view, though you waste more potential of the higher resolution camera.

One could always shoot a medium format camera hand-held and get the same image quality as 35mm (finer grained though), but the point of spending more money and lugging it around was to get more details in the images, not just more grains/pixels in the blur.

February 28, 2012

DPP has a high quality / moiré reduction option for raw images. It could be worth a try.

February 16, 2012

If these are 100% crops without upsizing then this does not look sharp at all for a 5D and those are not single pixels but groups of 4 pixels. If your raw converter produces that at 100% then think about using DPP. I have plenty of examples of high contrast edges not showing this at 100%.

However, it strongly looks like a 200% crop (or 100% crop magnified 2x) to me. Viewed at half size would then be the 100% and then this image looks perfectly fine.

February 14, 2012

May depend also on the raw converter. I have not seen this for quite some time in DPP with my 5D.

February 13, 2012

correction for my post above: "... that people who claim IS is "not useful on a wide angle lens" don't realize it's about the 1/30 - 0.5 second exposure range rather than the 1/200-1/60th second range where telephotos start to become shaky handheld where wides are sharp."

February 13, 2012

About image stabilization on a wide angle: Nikon has the 16-35mm f/4 VR. In DPreview's test of the stabilization effectivity,

http://www.dpreview.com/lensreviews/nikon_16-35_2p8_vr_n15/page5.asp

they found 40-50% of shots were sharp at the pixel level at 0.4 seconds shutter speed. To me that sounds very useful in situations where the tripod was left at home. Can't help thinking that people who claim IS is "not useful on a wide angle lens" realize it's about the 1/30 - 0.5 second exposure range, not the 1/200-1/60th second where telephotos start to become shaky handheld where wides are sharp.

February 9, 2012

Well, the small primes have these advantages over any other (zoom) lens:

<ul>

<li>compact and light - 280 grams</li>

<li>no flare even when sun in frame (my experience with 28mm f/2.8)</li>

<li>hard infinity stop for landscape/night</li>

<li>very sharp even wide open (see Photozone.de for the 28mm - APS format though)</li>

<li>minimum focus distance 20-23 cm gives much higher magnification than on any zoom, also keeping good working distance without shadowing the subject</li>

<li>nicer bokeh than old versions due to revised aperture blades</li>

<li>4-stops IS can be used at any f-stop like f/11 for a landscape, in theory f/2.8+4 stops = two stops better than f/1.4</li>

</ul>

But I'd like to see a test of IS answering the following questions:

<ul>

<li>if say 50% of images is critically sharp (some laboratory definition) without IS at 1/60 second, is this indeed also 50% at 1/4 second? (4 stops)</li>

<li>which shutter speed is necessary to eliminate shake in 90% of shots for a wide angle on a still hypothethical 35 MP-ish camera (for sure not 1/30 sec, it might be 1/500 sec), and does IS indeed help here at the faster shutter speeds?</li>

</ul>

But these lenses should be sold for under $600. The temptation to grab a Samyang 24mm f/1.4 may otherwise be too large.

November 27, 2011

I once had long exposure noise reduction on while shooting lightning with a 5D. If you do a series of images with say 10 second exposures and keep the cable release locked, it will shoot continuously for a number of shots, then pauses to do the dark frame, and continues with more shots, and so on. This would cause interruptions in a star trail series. In super wide angle star trail shots perhaps so small it would not be noticed. But if long exposure noise reduction is off, the hot pixels can be overwhelming. In cases when I forgot to turn it on, the Pixel Fixer program (http://www.pixelfixer.org) saved me, but for optimal results it is better to take a dark frame during the shoot for ingesting into this program.

November 23, 2011

Sigma 20mm f/1.8 is also a good option for landscape. You will often be using it stopped down to f/9 or more, by which the (full frame) corners are sharp. It's also easy to (pre)focus manually. It does have some moustache distortion, but I read that the Zeiss 21mm also has this. I am now using Tokina 16-28mm which is a relief since I tended to switch a lot between 28mm and 20mm primes. Also the hood is built in so no time lost searching in the bag. But makes it a little difficult to reach everything with a lens cleaning pen and if you need to use filters the 16-28mm is no option. It is even sharper than the Sigma around 20mm. Corners can be very sharp at mid apertures at any focal length if focused well (there is curvature of field which shows up if focused too far into the scene). Trying to manually prefocus is more difficult since beyond one meter distance the focus throw is very short. And a well known downside: it's heavy.

September 16, 2011

Last Saturday night, thunderstorms seen from the Dutch island of Texel:

Relatively few visible lightning channels, but I was fortunate to capture this <a href=" crawling lightning flash</a> with a fisheye lens from inside the car.

July 1, 2011

"Which wide angle is sharp wide open?"

The Tokina 16-28mm f/2.8 which I use on a 5D is very sharp and contrasty wide open, and stopped very sharp into the far corners. Sharpness falls off towards the edges, at 28mm this starts to happen closer to the center, at 16-20mm you will find a very wide area sharp with soft corners. There is curvature of field: at 16mm f/2.8, aiming down 45° and focusing at a subject, there is very good sharpness in a parabolic area from close to far (upper corners), a little less good inside this area, and rather fuzzy outside, including lower corners. However for portraits this wouldn't matter much. Bokeh on this lens is pleasant, and there is no vignetting. Focus throw is very short at longer distances, which can make manual focus tricky. You can't use filters. Be aware of sun flares. It's also a heavy lens. I experienced rather obvious decentering issues in two samples of the lens, so be prepared to test it and return if necessary.

I used also the Sigma 20mm. At f/1.8 my copy is sharp but with poor contrast and vignetting. At f/2.8 it is very sharp in the central area, good for portraits. Stopped down beyond f/8, this lens is very sharp and contrasty into the corners and easy to work with for landscape. You just manual focus halfway or two-thirds between 1 and infinity and you won't have to think about it for most subjects.

June 10, 2011

Perhaps it works with a different USB port or after a computer reboot. Occasionally EOS Utility is somehow not able to find the camera (in my case 5D on Windows 7) until repeated plugging/on-off switching, and shows only "Monitor folder".

June 2, 2011

By pixel level I meant 100% view, i.e. when the displaying medium is able to reveal differences between adjacent pixels. By perfect sharpness I of course refer to the maximum level which can be obtained by proper equipment and use: good lens, in focus, at best aperture, steady.

May 29, 2011

Increasing resolution will only be more favorable for image quality, but it will be increasingly difficult to reach perfect sharpness at the pixel level (or viewing large prints close-up): the laws of depth of field, lens aberrations and alignment, and camera shake do not change, so any flaws will be revealed more easily. One can predict that if you're not using a tripod with a 35 MP camera, even during the day, camera shake will show at the pixel level. You might then as well shoot SRAW or MRAW. The advantage comes down to how disciplined you are. Issues of lens imperfections and shutter slap could also become obstacles. Perhaps Canon will move away from mechanical shutters. That could also offer benefits as extending dynamic range by reading multiple exposures in one shot. Finally, in landscapes you won't gain pixel level sharpness all over the distance range, it's a choice between DOF blur and diffraction blur. It's nothing new, it's just like shooting medium format (but scaled-down).

May 7, 2011

@ Andrew and Hadi: thank you for your demonstrations, appears to work more effectively than I thought was possible - though the heal and clone approach adds more and fragmented flowers ;) For the dodging approach by Andrew I think only a little contrast and some warmth could be added locally. I will experiment with this. Thanks again! (and apologies for initiating off-topic discussion)

May 6, 2011

April 25th, EOS 5D, Samyang 8mm (cropped on this camera). I love the combination fisheye/wide angle lens and low sunlight, but is there a trick to make my own shadow invisible?

April 10, 2011

John Crowe wrote:

P.S. Calling the 16-35 a walk-around lens is a bit of a misnomer....do you know how big it is?! You certainly don't want to bang it on anything either!

</blockquote>

It's 11.1 cm, only 0.5 cm larger than the EF 24-105mm f/4L and 50 grams lighter. I'm sure people walk around with 70-200mm lenses of 17 cm! Alternatively you could take three small EF primes for the same weight but you'll be walking less and swapping more. You've got to walk around with something, no? ;) "Walk-around" means versatility in one lens to most, not so much the size.

April 4, 2011

My bets are a 16-40mm f/4L IS, introduced as kit lens for the 5D mkIII. I wish a third party would make tilt-shift alternatives that cost under €1000.

I like the idea of a 40mm f/2 pancake with autofocus as Canon's cheapest lens instead of the EF 50mm f/1.8 II or EF 35mm f/2.

April 2, 2011

Sometimes this occurs when too much RAM is used by other programs. But you have only one other program open so it seems unlikely here.

March 13, 2011

Maybe there is a huge piece of "dust" somewhere in the optical path? Does the dark spot become smaller when stopping down? What happens when you shoot something else of a uniform color. This looks a bit too circular (and odd at the top?) to be polarized sky.

However, even without filter, in shots taken with a Samyang 8mm fisheye at a mountain top I noticed the sky was darker in a band at 90° from the sun as if the lens had a built-in polarizer.

February 26, 2011

Robin,

Not sure you read this because this thread has shifted off the main forum page, but I received the sample this week and I've put up test shots:

http://forums.dpreview.com/forums/readflat.asp?forum=1029&thread=37833118

I think this isn't a very good sample. At longer lengths it is decentered. It is well described by Photozone's side note. So, as such, it seems a typical sample...

February 19, 2011

Cosina Voigtländer offers 20mm and 40mm pancake lenses for EOS (manual focus, digital aperture). Also Canon's EF 35mm f/2 is reasonably small (4.2 cm vs 2.5 cm) and practically the same weight.

February 16, 2011

Yes, the sample variation is a bit worrisome. On the Fred Miranda forum someone received a bad and an okay sample before finally obtaining a really good one. I received one last month, but it was slightly decentered, noticeable at f/8 on a 5D. For my balcony test scene, I got much sharper images when the camera was held upside down. I sent it back for an exchange (Tokina does not appear to fix this lens in service centers yet) and I am waiting for a new one from Japan which could take about 4 weeks. Take into account a testing and return period before the time you actually hope to use the lens.

I think anyone who buys this will look critically for a good quality sample of this lens, as there is simply no point in lugging around 1 kg of unsharp glass limited to 28mm on the long end with no filter option. Also because the lens will be used on near-future 30 MP cameras which will be less forgiving in terms of resolution.

Sign In

Profiles

Forums

Blogs

Events

Downloads

Gallery

Store

Posts posted by oscar_van_der_velde