Jump to content

APO Summicron 50mm, too sharp!?


trond

Recommended Posts

Advertisement (gone after registration)

... I wasn't aware the M9's JPEG engine also applies the reduction as well.

It does. However, it comes at a price: The overall sharpness will be slightly reduced ... just as it would with a hardware AA filter. Even after having paid Leica prices, there's no free lunch.

 

So in some cases, it may be better to have full DNG resolution and to combat moirés individually, using Lightroom's anti-moiré brush locally, manually, and tediously. In other cases, it may be more appropriate to simply rely on the software AA filter applied to in-camera JPEG files. That's why I always use DNG+JPG format—this way, you always have both options.

 

If you're concerned with the additional memory space consumption on the SD card then your card is too small. Get a bigger one.

Link to post
Share on other sites

  • Replies 47
  • Created
  • Last Reply
That's an interesting "unintended consequence" of the high resolution of the lens. A good sign that your lens is performing as designed by in effect being able to out-resolve the sensor. I expect moire with this lens will be more common on the M8 and that both the MM and M240 will make it less common. Of course, the MM has the same number of pixels as the M9 but removing the need to de-mosaic will, I think, increase the spatial resolution.

 

Dead on. The 'too sharp?' critics not withstanding this was a perceptive goal by Leica. Lenses are made for years; not so much cameras. If the 50 APO can not out-resolve the best sensor Leica can produce in 2013 it would have to be considered a failure. Your 50 Summicron would be good enough. In a few years the sensor game will have changed, a lot, and no matter what other improvements there are increasing pixel density is a given. Don't need them? Who cares, that's where the technology is going and Leica would be crazy not to go there, even with its milder timeline.

 

Against my better judgement I'm envious of those who get to play with such a superlative device, but I'm close to pulling the trigger on an MP and Tri-X is not going to give the APO much of a run for its money. :D

Link to post
Share on other sites

Dead on. The 'too sharp?' critics not withstanding this was a perceptive goal by Leica. Lenses are made for years; not so much cameras. If the 50 APO can not out-resolve the best sensor Leica can produce in 2013 it would have to be considered a failure.

 

I totaly agree, the M9 or M Monocrom is no match for this lens, it has been designed by Leica for 60+ Mpix sensors.

 

From what I am seeing from the APO50, it will be a nice match with the 70 Mpix M241 Monocrom that we will see in 2015 :)

 

This sensor is already a standard product from CMOSIS:

 

CHR70M - CMOSIS

 

I am very happy that Leica has chosen the most innovative sensor design company on the planet today!

 

BR

 

Trond

Link to post
Share on other sites

Against my better judgement I'm envious of those who get to play with such a superlative device, but I'm close to pulling the trigger on an MP and Tri-X is not going to give the APO much of a run for its money. :D

 

At least you won't get any moire from it...

Link to post
Share on other sites

I totaly agree, the M9 or M Monocrom is no match for this lens, it has been designed by Leica for 60+ Mpix sensors.

 

From what I am seeing from the APO50, it will be a nice match with the 70 Mpix M241 Monocrom that we will see in 2015 :)

 

This sensor is already a standard product from CMOSIS:

 

CHR70M - CMOSIS

 

I am very happy that Leica has chosen the most innovative sensor design company on the planet today!

 

BR

 

Trond

 

oh wow, thanks for sharing this. very interesting indeed! i wonder why they would show it so early...perhaps the M240 Monochrom version is not that far off afterall

Link to post
Share on other sites

Advertisement (gone after registration)

At least you won't get any moire from it...

 

Ha! I suspect moire will eventually be auto-correctable in post and eventually in-camera. Technology advances will make it easy. Just like the 'new' technology of the waterhouse stop solved the pesky problem of too much light. :D

Link to post
Share on other sites

I will give it a try another day.

 

It is snowing heavily today, so the images comes out very "soft" no moire whatsoever :D

 

BR

 

Trond

 

It's snowing here too, but orange. First APO50 shots. No flare earlier with sun directly into the lens.

 

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

Link to post
Share on other sites

Hej Trond, interesting photo. Thanks for posting.

 

Was aperture did you use? It's a pretty large crop. Would you mind posting one of the yellow house at the extreme left of the image or, alternatively, of the red and blue (people?) at the right-hand side of the image?

 

On my monitor here at work (admittedly a rubbish monitor as monitors go) there is also moiré on the side of the yellowish barn in the crop.

 

Cheers

Philip

Link to post
Share on other sites

sorry, couldn't view your DNG attachment. something about my mac not being able to load quicktime 7.7.1 ? v. 10 is installed.

 

but thanks for starting this discussion of the new 50 APO.

 

greetings from hamburg

 

rick

Link to post
Share on other sites

Full of moire and "false color" pixels. The only conclusion: You got to use film with apo 50 until M with 70 Mpix comes. ;)

 

Seriously, if anyone is asking whether something is too sharp the answer is in fact always "sensor is sub par". As long as there is even a hint or moire, sensor is not good enough for the lens, not the other way round. I wonder how long it takes before "megapixel race accusers" get the idea.

 

Of course more resolution is not easy to get, and especially getting the sensor to accept steep ray angles at the same time is something we still have to wait for. So probably the new M is a good compromise anyhow.

Link to post
Share on other sites

As long as there is even a hint or moire, sensor is not good enough for the lens, not the other way round.

 

This happens not to be the case. Projecting a regular pattern onto another regular pattern will result in interference patterns.

 

You can avoid moiré either by applying a band pass filter to the projected image or by using a sensor which does not form a regular grid.

Link to post
Share on other sites

Philip, I do disagree, and I do know the causes or moire and all kinds of aliasing in pictures and in audio and the ways to avoid them. The most logical way to me for any lens+sensor-combination to avoid moire is that sensor outresolves lens, and we are far from it at the moment. Low pass is a serious compromise, not an ultimate solution.

 

Ultimately the limit is diffraction, beyond that there is no need to increase sensor resolution. Micro 3/4 sensors have 4 times the pixel density compared to Leica, and when technology permits Leica will get to 70Mpix minimum for 36*24. Apo 50 needs it, deserves it and so do many other lenses too.

 

Surely many people don't need that kind of resolution, I don't, but Leica is about being great, not mediocre. Should be at least.

Link to post
Share on other sites

The Foveon does not help against moiré, either, as its cells form a regular grid, too.

 

Yes, moiré would be greatly reduced if the lens could not project details which have the same magnitude as the sensor's grid. However, without doing all the math I think we'd be speaking about an order of magnitude that might be needed. The resulting file sizes would be something to behold. Also, since the filters would presumable need some thickess, the cells might become quite deep. That might cause a lot of further problems with the incident angles.

 

Arranging the sensor cells in an irregular fashion instead of a regular grid might work, too. However, I'd like to see the software to handle that sort of thing.

Link to post
Share on other sites

 

Arranging the sensor cells in an irregular fashion instead of a regular grid might work, too. However, I'd like to see the software to handle that sort of thing.

 

That would be the best way. Fab the sensor in the usual, cheap, rectilinear grid then let the camera's logic pick out a pseudo-random selection of pixels to use in the image. You'd probably need 200+ MP but, trust me, you'll see that sooner than you think. For all the high tech glitter, in the commercial arena fab plants are little better than the slaves at the oars in Ben Hur. :(

Link to post
Share on other sites

Hej Trond, interesting photo. Thanks for posting.

 

Was aperture did you use? It's a pretty large crop. Would you mind posting one of the yellow house at the extreme left of the image or, alternatively, of the red and blue (people?) at the right-hand side of the image?

 

On my monitor here at work (admittedly a rubbish monitor as monitors go) there is also moiré on the side of the yellowish barn in the crop.

 

Cheers

Philip

 

Hei Philip!

 

The aperture on this shot is f4.0, which also seem to be the aperture with peak performance.

 

At f5.6 the moire pattern is a little less severe, which tells me that diffraction starts to come into play.

 

Moire is still visible at f11, but completely gone at f16.

 

You can play with the file yourself by downloading it from my server using the link on the top of this thread.

 

BR

 

Trond

Link to post
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...