Jump to content

Leica FAQ has been posted -- with 1 error


wparsonsgisnet

Recommended Posts

Advertisement (gone after registration)

The Leica FAQ has been posted as a download, here:

Leica Camera AG - Downloads

There are 2 general items covered so far:

1. Downloading/Compatibility, and

2. Use of IF/UV filters

 

There is an error in the FAQ in that the last question in category !,

Why does the Leica M8 store raw data in 8-bit only?

has the answer repeated from the previous question.

 

Leica will have to fix this particular part of the FAQ.

Link to post
Share on other sites

Guest sirvine

This reminded me that I have been using a Heliopan on my 50/2. Does anyone know of a retailer with 39mm Leica filters in stock? (Not B&H, they're on Passover holiday until next week).

Link to post
Share on other sites

 

There is an error in the FAQ in that the last question in category !,

Why does the Leica M8 store raw data in 8-bit only?

has the answer repeated from the previous question.

 

Leica will have to fix this particular part of the FAQ.

 

this faq is correctly answered in the german site. it says (rough translation):

 

"the m8 has a near-lossless compression of data saved in dng format. this doubles the speed of saving and halves the space needed for the files. during testing with prototypes, a 16-bit version was tried, but did not show any visible improvement of the image quality. therefore, leica dispensed with the 16-bit option."

 

hope this helps,

günter

Link to post
Share on other sites

Guest guy_mancuso
this faq is correctly answered in the german site. it says (rough translation):

 

"the m8 has a near-lossless compression of data saved in dng format. this doubles the speed of saving and halves the space needed for the files. during testing with prototypes, a 16-bit version was tried, but did not show any visible improvement of the image quality. therefore, leica dispensed with the 16-bit option."

 

hope this helps,

günter

 

 

This is the same info i got in a e-mail from leica , they saw no visable improvement. although we have requested it in a future firmware

Link to post
Share on other sites

Yeah, and the M8 images are certainly nice, but they also saw no visible need for IR filters either, so I'd be curious to see the difference between 8-bit and the full file.....maybe less posterization in the highlights, which are compressed by their scheme?

Link to post
Share on other sites

LFI is a good magazine, and I am sure that they also saw no difference in their test on the computer. However, the question is if a white subject is sufficient. This is where it loses the most information compared to 16 bits, but not necessarily where we perceive the difference most readily.

Link to post
Share on other sites

Well, the white car was meant to be exactly that. Since a white car and the sky are both well lit, I am not sure that it would make a difference. However, if a shot were underexposed and then the exposure was brought up on the computer, that would give banding with a JPG. I might try this with the M8 DNG to see what happens.

Link to post
Share on other sites

I saw good news and bad news in the FAQ. The good news is that they are working on SDHC support, so we won't be restricted to Transcend for 4GB cards, but will be able to use the faster cards that SanDisk, Lexar and others have put out.

 

The bad news is that they are still perpetuating the confusing "thick IR filter" nonsense that has permeated Leica's marketing communications about their filter choices. Interference filters are a set of layers a few microns (thousandths of a millimeter) thick on a glass substrate that is no thicker than you need it to be for handling, in this case a few tenths of a mm. Absorptive filters need thickness, and making them half a mm thick makes them not very effective with present materials. They do say they are looking for more effective materials, which is good news. I guess marketing still hasn't a clue what the technical issues are on that score, and engineering finds it simplifies their lives not to try to educate marketing further.

 

I think it is up to us the users to find the test cases where in postprocessing you see posterization in the 8bit nonlinear encoded DNGs. I certainly don't expect to see it right out of the camera, but as Carsten suggests, our ability to pull exposures up or down a long way may be limited by the compromises made in resolution.

 

scott

Link to post
Share on other sites

Scott, I don't know why you call the IR filter choice nonsense. Leica had to choose what filter to put on the sensor. The angle of incidence from many lenses, screw and M-mount, is such that an interference filter cannot work in this camera without filtering out visible light in the corners. Hence they had to use an absorptive filter. The thickness of this causes optical problems, as seen on the Epson R-D1, for example, so they had to choose how thin to make it to make their excellent lenses able to pass the information to the sensor. They chose a thickness which gave sharp results and minimised artifacts, and had to accept that an absorptive filter of this thickness cannot filter all of, or even enough of the IR light. Hence the filters on the lenses.

 

This whole train of thought fits together, and although I am not happy about having to use filters on my lenses, I understand that with the technology of today, it has to be this way, if I want the performance of the M8.

 

I imagine that there is no company in the world as interested in improving IR filtering technologies as Leica, at this time. Their clear interest in this echoes their and our expectation of excellence.

 

I hope that when thin, yet effective, IR-absorbing filters have been developed, and I imagine that this is only a question of time, that there will be some upgrade path for the M8.

Link to post
Share on other sites

Carsten, I object to the garbled description in the FAQ which suggested that interference filters were rejected because they would be too thick. That's the part that is nonsense.

 

The rest is a series of decisions made step by step, I think, ending up in a difficult compromise. An interference filter on a clear cover glass less than .5 mm thick would be no less sharp in the corners than the green cover glass which was used in the M8. But it would have caused a potential for internal reflections that might have required further work on baffling at a time when the body design was already frozen. The correction for red vignetting due to an IR filter is smaller when the filter is at the image plane than when it is in front of the lens because the angle variation across the image is less than the full angle of view (modern lenses are are at least somewhat telecentric). But the corrections for red vignetting in an interference filter are stronger than with an absorptive filiter, so Leica opted to go with a fairly approximate set of weaker firmware corrections. Since the cover glass decision was probably made about a year before Photokina 2006, they may have hoped to get along with no color vignetting correction, and only added a little bit when they realized that even the absorptive filter causes vignetting because of its greater thickness when light angles through its edges. There are glasses which absorb more strongly than the material used in the M8. but apparently these were not qualified for manufacturing to Leica's tolerances.

 

Since these problems are being solved after the product has shipped they are costing Leica more than if they had been addressed earlier, but in return we have the M8 to shoot with (and speculate about) now, not later this year sometime. In each decision that had to be made I suspect that schedule was an important factor, and it should be. I like the Steve Jobs quote, "Real artists ship."

 

scott

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...