LFI Comments

[marknorton]

Advertisement (gone after registration)

I'm pleased to see that LFI are not afraid of telling it as they see it in the 1/2008 issue:

 

First, a comment on "White Balance with the M8"

 

- "Automatic settings: This picture is useless - much too warm"

 

Then, a comment on Four Thirds:

 

- "As for the experiment with the FourThirds format, the company from Solms will have drawn their own conclusions by now"

 

It would be easy for LFI to just become a mouthpiece for Leica corporate speak but instead the message to Leica seems clear - the M8 AWB needs fixing and FourThirds is a dead duck.

[ho_co]

I'm glad you mentioned LFI's editorial integrity. I've previously been impressed by their honesty, though it's sometimes couched in terms like "possibly not the best, but then again, not everyone needs this feature," or some such.

 

I'm miffed at the M8's AWB problems, and the fact that AWB still doesn't work, a year into the product's life, despite the 'frequent though stepwise' firmware revisions we thought we might see.

 

But can a programmer tell me: How would you get a camera to tell what kind of light it's looking at?

 

I can imaging programming an auto-flush toilet not to respond to mere changes of light intensity (and waste water), but only to the change a departing user brings; but how would you get a camera to look at a scene and say for example, "Okay, the scene is primarily red--but that's likely a red backdrop, because the light seems to be daylight"?

 

Are AWB algorithms proprietary? The P&S cameras do it well. Nikon has (uniquely, I think) put a sensor atop the prism of some models to help with their white balance.

 

What should the camera look for? Specific wavelengths? Specific ratios of wavelengths?

 

 

Hmm. Looking back at this post, did I really compare the M8 to an auto-flush toilet?

 

--HC

[sdai]

Nikon ditched the hybrid WB sensor in D3 and computes WB numbers with the data collected on CMOS sensor ... Canon did this first starting from 1D Mark 2, Nikon followed suit ... works quite well.

 

I'm not surprised at the LFI comment at all ... great minds always think alike.

[ho_co]

I'm not surprised at the LFI comment at all ... great minds always think alike.

Simon--it took LFI a year to mention it? Better late than never!

 

Thanks for update on hybrid sensor. From what you say, the AWB issue with other brands is also still in flux, even if they are a tad ahead of Leica's efforts.

 

--HC

[zapp]

If a comment on WB shows independence of LFI, well that is not enough. So far, all real issues with Leica gear have been downscaled by a lot. Remember the focus shift article. 4 pages written around the topic, telling you how great Leica is and half a page giving real numbers to show the problem. You get all the info from LFI, but in different formats.

[sdai]

Thanks for update on hybrid sensor. From what you say, the AWB issue with other brands is also still in flux, even if they are a tad ahead of Leica's efforts.

 

Howard, the AWB results from both C and N's top dogs are very constant actually and they keep improving no matter what changes they've put in place, it's just their algorithm and processing speed have reached the level when a hybrid sensor has become totally redundant. This is no longer an issue in modern DSLR design.

[sdai]

Advertisement (gone after registration)

If a comment on WB shows independence of LFI, well that is not enough. So far, all real issues with Leica gear have been downscaled by a lot. Remember the focus shift article. 4 pages written around the topic, telling you how great Leica is and half a page giving real numbers to show the problem. You get all the info from LFI, but in different formats.

 

Focus shift is really nothing new IMO ... it's just when people shot with film, the errors were masked by film grains, and pixel peeping hasn't gone extreme just like today.