Denoir - I could not get a usable RAW file from your zip file (which says it only compressed 2%). I could get a TIFF and here is my result. Mainly some highlight and mid-tone adjustments. They help, but this type of light (and wet water reflection) is hard to get the sharp look:
rscheffler wrote:
I only have recent experience with Munich. The airport is fairly far outside the city center, but is easily reached by S-Bahn (train) from a station below Terminal 1. It takes about 45 minutes each way. A day pass for all zones, which you will need, is 10,40 €. I believe there is a short term luggage storage option at the airport, though you should confirm this.
Munich's central area offers a number of fairly popular sights, such as Marienplatz - the city's center - and nearby areas such as the Viktualienmarkt market, Odeonsplatz, the Residenz park, Englischer Garten park. You can climb the steps of St. Peter's church tower to get a nice overview of central Munich. Here's a panorama: http://www.panorama-cities.net/munich/frauenkirche_5c.html
Other Munich panoramas: http://www.panorama-cities.net/munich/munich_germany.html
Otherwise, the airport is pretty much in the middle of nowhere. The first stop outside the terminal is called Besucherpark - visitor's park - which might be something to check out, though I have never had the inclination to do so. In Terminal 2, outside the secure zone, there is a viewing platform accessed from the second floor for watching airplanes. Terminal 2 also offers a somewhat interesting setting itself for photos, though nothing extremely extraordinary. There are some examples here: http://www.flickr.com/groups/photowalkingmunich/pool/page21/
Another option would be to take the S1 S-Bahn to Oberschleißheim, which is about 15-20 minutes from the airport. There is a nice palace with extensive parklands about a 10 or so minute walk from the train station: http://www.schloesser-schleissheim.de/englisch/schleissheim/index.htm
WOW, pretty in-depth review Ron, thanks a bunch. So far Munich airport and surroundings seems like it'll be a decent area to blow a few hours. All those sites you sent me are fantastic, are some of your photos on the flickr stream taken with a Zeiss lens
As it turns out, it's cheaper then Paris stop-over but slightly more then Istanbul. Since Philber chimed in on Paris airport not being ideal, I think Munich so far is the winner. Nobody for Istanbul?
philber wrote:
Much as I love to promote Paris, Adam, it is not a good choice. First, because its airport is one of the worst rated in the developped world. Coming from Singapore with the A380, you would be in for a rude shock.
Second because, in 3 hours back to back, there is not much you could do, with traffic and all.
Sorry.
Darn, and I was going to have you show me around, so I could watch how you shoot all your beautiful photos. Well, Paris will have to wait then.
rscheffler wrote:
I see a difference between the RGB and LAB images, though not necessarily in sharpness.
The idea isn't about sharpness. Using LAB and sharpening only the L channel, you don't increase color saturation with the sharpening radius. It can be seen in denoirs examples, especially in grass and leaves that now doesn't show the CA-like artefacts.
To sharpen is to increase local contrast, and increasing contrast of an RGB image will also increase saturation.
Edit: I just downloaded the RAW of that bridge shot, and I think the sharpening problem is a combination of overexposure (of that particular area to the left), very high contrast and CA.
Adam, there is one more reason for which Paris is not a good choice. You are coming from Singapore and going to Serbia. Paris would just lengthen your flight time both ways, because you would be essentially overflying Serbia going West to Paris, and then backtracking East, adding easily 2 hours to your flight time.
Istanbul is a fabulous city, one of the world's jewels. The problem there is that I don't trust either the airport nor public transport on such a tight schedule, and traffic is unbelievably bad. Actually, for want you want to do, and without knowing if there are good rates and schedules, Vienna is a logical choice. A good smallish (relatively speaking) airport, close to Serbia and probably well connected, a great city not very far from the airport, with superb photo opportunities.
Hope this helps.
philber wrote:
Vienna is a logical choice. A good smallish (relatively speaking) airport, close to Serbia and probably well connected, a great city not very far from the airport, with superb photo opportunities.
+1
I've only been there once, but it did make a pretty good impression on me. They seemed to be quite punctual. The layout of the airport is slightly confusing though. The Austrian Airlines plane was the cleanest and neatest I've been on too.
@Luka: Very interesting comparison. In sets B and E I see an obvious difference in saturation, but more noticably, in the warmth of colours(!). The RGB sharpened versions consistently look warmer - in set B it's really evident in the tree's foliage. At first I thought that the two images from set E were processed with a different white balance, which I'm sure they are not. In set F the slightly cooler colors of the L-sharpened version make it look slightly flatter, perhaps.
Ron: Yes, fade luminosity is the same thing. Another alternative is to put a gauss blur on the a and b channels in LAB mode to smooth out any color artifacts. Again however, if you are using good glass the difference is very small.
Paul: Nope, I'm afraid that looks very oversharpened. I as well have only manged to process that image to get it either over or undersharpened.
Makten: Yes, increasing contrast in RGB space increases saturation. As for the image, overexposure is a problem, although the same applies for the other parts of the image. There is no CA in that shot. The 21/2.8 is completely free of regular CA. You can push it in very extreme situations to show some LoCA in OOF areas but it's not applicable in that shot. Regular CA is non-existent. That's why I said that it was perhaps a poor choice for this test - it has no CA that could be amplified through sharpening.
The real problem with that image however is I think that it already has very high local contrast to begin with. So when you smooth it through a bicubic resize or amplify it through sharpening it just becomes weird.
Here is a second shot of the same nature, and there are no overexposed parts here:
Histogram, just to show that there are no problems with exposure:
denoir wrote:
There is no CA in that shot. The 21/2.8 is completely free of regular CA.
Oh yes, there is. But perhaps not of the normal sort. Look at the borders to the overexposed patches. There are clearly visible "red spots", which I can't explain as something else than CA. In ACR it went away by applying the "defringe" option.
The real problem with that image however is I think that it already has very high local contrast to begin with. So when you smooth it through a bicubic resize or amplify it through sharpening it just becomes weird.
Here is a second shot of the same nature, and there are no overexposed parts here:
Oh, that makes my head ache! Very, very 3D though.
Makten wrote:
Oh yes, there is. But perhaps not of the normal sort. Look at the borders to the overexposed patches. There are clearly visible "red spots", which I can't explain as something else than CA. In ACR it went away by applying the "defringe" option.
Oh, that makes my head ache! Very, very 3D though.
I agree on the overexposure issue for the first shot, although part of the blame must go to how poorly the camera handles them. The red areas might just be sensor bloom. I also noticed the usual Canon red tinge in the overall color (especially the shadows). The second subject lacks contrast, so it a seperate issue. While the Zeiss 21 has great (micro)contrast and sharpness, not every scene taken using the lens will show its strength. I also believe a Canon DSLR (5DII in my case) is not the best match for this lens. The AA filter, poor highlight headroom, and redish shadows work against the lens. My Kodak SLR/c would bring out the edge detail and handle the overexposed areas much better.
denoir wrote:
Ryan: Thanks and interesting shots with the M8.
Ron: Yes, fade luminosity is the same thing. Another alternative is to put a gauss blur on the a and b channels in LAB mode to smooth out any color artifacts. Again however, if you are using good glass the difference is very small.
Paul: Nope, I'm afraid that looks very oversharpened. I as well have only manged to process that image to get it either over or undersharpened.
Makten: Yes, increasing contrast in RGB space increases saturation. As for the image, overexposure is a problem, although the same applies for the other parts of the image. There is no CA in that shot. The 21/2.8 is completely free of regular CA. You can push it in very extreme situations to show some LoCA in OOF areas but it's not applicable in that shot. Regular CA is non-existent. That's why I said that it was perhaps a poor choice for this test - it has no CA that could be amplified through sharpening.
The real problem with that image however is I think that it already has very high local contrast to begin with. So when you smooth it through a bicubic resize or amplify it through sharpening it just becomes weird.
Here is a second shot of the same nature, and there are no overexposed parts here:
Benjamin: The slight shift in color is probably due to that unlike the LAB mode the color channels are affected in RGB mode. ...Show more →
My two minute example (from TIFF) may be oversharpened, but my main point was to show the highlights and midtones could use some adjustments to improve the look of sharpness. Perceived sharpness is a combination of actual sharpness + contrast.
Makten wrote:
Oh yes, there is. But perhaps not of the normal sort. Look at the borders to the overexposed patches. There are clearly visible "red spots", which I can't explain as something else than CA. In ACR it went away by applying the "defringe" option.
I'm sorry but I can't see it at all.
Here is a 100% crop of an overexposed region:
Here it is pulled down four stops:
Here with four stops down plus max fill light:
Applying "defringe" just reduces the red hue overall but there are no differences in the edges:
By the way, if you are wondering why the shot isn't tack sharp - it was a longer exposure and my tripod wasn't on very steady ground. I should have probably picked a better example, but that particular image has annoyed me as I have not been able to sharpen it the way I want.
pdmphoto wrote:
I agree on the overexposure issue for the first shot, although part of the blame must go to how poorly the camera handles them. The red areas might just be sensor bloom. I also noticed the usual Canon red tinge in the overall color (especially the shadows). The second subject lacks contrast, so it a seperate issue. While the Zeiss 21 has great (micro)contrast and sharpness, not every scene taken using the lens will show its strength. I also believe a Canon DSLR (5DII in my case) is not the best match for this lens. The AA filter, poor highlight headroom, and redish shadows work against the lens. My Kodak SLR/c would bring out the edge detail and handle the overexposed areas much better. ...Show more →
I agree with the standard Canon red color cast but not about poor highlight latitude. It's the shadow latitude that sucks big time - you get noise banding when you push up the exposure. Highlight latitude on the other hand is excellent - as you can see in the example above - I could effortlessly pull down four stops without any problems. As for the Kodak SLR/c, as nice as it is, IIRC it has about 1.5-2 stops less dynamic range than a 5DII, so I'm not convinced that it could handle a scene like that better...
denoir wrote:
I'm sorry but I can't see it at all.
Good for you! It's definitely there. Look at the most abrupt transitions from dark to overexposed areas, mainly in the upper left part of the crop.
This is not very problematic, but it'll be enhanced when sharpening the picture (especially in RGB mode). The 35/2 is much worse, and that's why I always use LAB mode myself (I don't have the 21).
Edit: I noticed that your XMP had +25 of color noise reduction. That takes away some of it (I saw it more clearly when trying to PP the CR2 file myself). Are you using ACR/Lightroom or something else?
Makten:
I'm using Lightroom 3 (i.e ACR 6.1) and the shots were developed with all default settings except one - camera profile was set to "Camera Standard".
If it was in the transition of high contrast areas, it could only be LoCA and not regular CA. The problem with that is that LoCA at f/8 is a theoretical impossibility. I have looked through Lloyd's review of the 21 and regarding CA he says that there is "little or no evidence" of chromatic aberration existing at all. So the thing that you can see and I can't must be something else - sensor blooming is of course a possibility. Do you see it as well in the underexposed crop?
As for the 35/2, there it's easy enough to spot (I've turned off all the default noise reduction here):
In my experience with 3 zeisses, the 21 has almost no CA, the 35 has a little, and the 28 quite a bit more. Sensor blooming should be a constant, regardless of the lens.
Very, very 3D though. 
