Lens coding confusion...

[jquimby]

Advertisement (gone after registration)

I seem to be confused with the whole 6 bit lens coding 'enhancement'. If I shoot raw, saving the data in the DNG file format what exactly does coding do besides provide a little more exif data?

 

Or am I missing something?

 

I have two summilux 35 asph. If I get one coded and leave the other as is, uncoded, what is the difference in the raw data when I take the same picture with each one?

 

I was under the impression that any benefit from using a coded lens, with the exception of more exif data, is only realized when saving the photo as a jpeg; as this is when the photo is actually rendered as a photo. Otherwise it is just that, raw info that needs to be processed.

 

I bring this up because all of the quoted solutions for dealing with the M8's IR issues involve an IR filter and a coded lens and possibly a new profile for C1. Leica is aware that not everyone shoots jpegs or uses C1 right? Hope they don't forget abouts us.

 

Thanks for any insight

 

-joe

[ho_co]

Joe--

From tests made by other forum members, coded lenses do bring corrected data to the RAW file.

 

Check for threads started by "robsteve." One of them shows a comparison on the DMR with the 19mm R. His take is, Look: Here's what Leica has already done; it's reasonable to think that's what they'll be doing with the M coding.

 

There are a number of other examples around, scattered thru a lot of threads here.

 

Apparently there is good reason for zebra striping.

 

--HC

[leicapfile]

RE: 6 bit lens coding.

 

I would think it a positive first step (given their current problems) would be to include the information derived from this lens coding, into the menu for setting manually.

 

That would save shooters some money (but not time) and provide a little much needed goodwill with their public.

 

Do any of you who may be software/firmware literate think this could be done with out having to send their M8s in for a fix?

 

Jerry

[marknorton]

It could certainly be done in software, however, I think Leica will be wary of manual selection because it could make image quality worse, not better and it removes one of reasons for going with Leica's lenses insteasd of other peoples. But for the fact that the lens coding impact to date is subtle at best, I think we would have seen early M8 users putting stickers on their lenses to try out the lens coding. Actually, the lens coding has been met so far with a shrug of the shoulders.

[ho_co]

Jerry--

(Oh, drat! While I was drafting this convoluted reply, Mark got in and out succinctly. Well, here's my take anyway:

 

I'm not software literate. Your idea sounds good, but as Bob Ross and others have pointed out:

 

First, what if I forget to change the setting when I change lenses? What if I'm applying a correction designed for a 28 when I use the 75/1.4? Allowing a manual setting would thus make the camera less 'automatic,' more menu-burdened, less Leica-like.

 

Second, the modification would allow use of other manufacturers' lenses, even if not optimally corrected. So allowing manual settings could work to Leica's detriment.

 

And, to your question: I'm pretty sure it could be done simply by downloading a later firmware update. It's just adding a section to the same code module that now has the 'use/don't use lense codings' toggle. (My opinion, unencumbered by education.)

 

Personally, I would not want to see it done. I can't afford to have all my lenses coded, and some of them are of the design (bayonet screws in the wrong place) that won't allow coding. But I don't want to see us adding more levels to a simple menu system.

 

--HC

[jrgeoffrion]

A few 'safeguards' could be put in place should Leica opt to add a manual lens selection in their menu. For example,

 

- the ability to "tell" the camera which lenses are owned so that the menu subsequently only present those lenses.

- lens confirmation upon start-up if the manual lens selection option is selected.

- add a switch that is pressed when the lens is remove (just add to the lens removal button) which would either 'reset' the lens selection or revert to the lens confirmation above.

 

None of this is relatively complicated but would make the camera much more versatile and "powerful". As for not adding the coding for third party lenses, I'm against that. Since the coding is required to get the performance a reasonable person would expect from "the first professional rangefinder system in the digital age" (as per the Leica brochure), it would only be right to add them.

[ho_co]

Advertisement (gone after registration)

JR--

Good arguments. Hadn't thought about your first points, simply because I dislike firmware-intensive cameras. But what you suggest would be pretty natural and unintrusive. I concur: As you define them these would be good, usable design additions.

 

As for your second paragraph, "the first professional rangefinder system in the digital age" would in Leica's mind be a Leica system, not a Leica/Zeiss/Voigtländer/Canon system, wouldn't it?

 

--HC

[jquimby]

HC, thanks for the reply. I found the post you were talking about, here is the link for those playing along at home: http://www.leica-camera-user.com/digital-forum/9457-why-coded-lenses-ir.html

 

I guess I am even more confused now than before. How does Leica apply corrective measures to essentially something that doesn't exist yet?!?! The raw data is just that values pulled off of the sensor, plus a WB value among other info. All of which needs be processed, put into a color space, interpolated and so on... before the correction can be applied.

 

While this is very interesting, I am rather perplexed as to how this is actually happening.

 

Hopefully whatever voodoo Leica is doing fits within the standard DNG spec so any raw converter can use this additional info for correcting photos as this would drastically cut down on post production time. Especially important to those of us that need to move images while still out in the field.

 

Thanks again for reply

 

--joe

 

 

Joe--

From tests made by other forum members, coded lenses do bring corrected data to the RAW file.

 

Check for threads started by "robsteve." One of them shows a comparison on the DMR with the 19mm R. His take is, Look: Here's what Leica has already done; it's reasonable to think that's what they'll be doing with the M coding.

 

There are a number of other examples around, scattered thru a lot of threads here.

 

Apparently there is good reason for zebra striping.

 

--HC

[ho_co]

Joe--

Thanks for finding and posting the thread I mentioned.

 

It's beyond me. They know the specs of their lenses and how much vignetting they generate. Theoretically they could also add to the formula the amount of change an IR-cut filter would generate in the visible range in the corners.

 

That's where the coding comes in: This lens is being used; with our sensor it will show X amount of vignetting and Y amount of colorshift due to use of IR-cut filter, so we will adapt the output of the sensor to compensate for that. The corrections are applied before the RAW data are written. That is, the DNG file will not show the vignetting because it has been balanced out. So any RAW converter will see the file 'properly.' But I've got no idea when in the chain the correction is applied.

 

"guy_mancuso" has done a lot of testing with his M8, including comparisons with lens-recognition both on and off. There are a number of threads, some started by him, in which he has posted his results. There is a visible difference, but how it's obtained is beyond my understanding.

 

As Mark said, having the lens coded may or may not be worthwhile depending on the specific lens, the focal length, and what you shoot.

 

This is one heck of a sophisticated piece of equipment, as its results show. "Voodoo" may be the more general term, but the manufacturer would probably prefer "Leica mystique and know-how." They have definitely broken new ground!

 

--HC

[hammam]

Jerry--

(Oh, drat! While I was drafting this convoluted reply, Mark got in and out succinctly. Well, here's my take anyway:

 

First, what if I forget to change the setting when I change lenses? What if I'm applying a correction designed for a 28 when I use the 75/1.4? Allowing a manual setting would thus make the camera less 'automatic,' more menu-burdened, less Leica-like.

 

--HC

 

Did you ever forget to change ISO settings when changing film in your metered M?

[marknorton]

A few 'safeguards' could be put in place should Leica opt to add a manual lens selection in their menu. For example,

 

- the ability to "tell" the camera which lenses are owned so that the menu subsequently only present those lenses.

- lens confirmation upon start-up if the manual lens selection option is selected.

- add a switch that is pressed when the lens is remove (just add to the lens removal button) which would either 'reset' the lens selection or revert to the lens confirmation above.

 

None of this is relatively complicated but would make the camera much more versatile and "powerful". As for not adding the coding for third party lenses, I'm against that. Since the coding is required to get the performance a reasonable person would expect from "the first professional rangefinder system in the digital age" (as per the Leica brochure), it would only be right to add them.

 

Good ideas...

[ho_co]

Did you ever forget to change ISO settings when changing film in your metered M?

Olivier--

No, that I haven't done. But when I had been away from photography for ten years I did once open the bottom plate to change film without first rewinding.

 

And I don't see the two as analogous. One changes film every 30 to 38 exposures in my experience; lenses might get changed after only one or two. Changing film requires a bit of a break in 'the action'; changing lenses doesn't.

 

And as I said, JR's points are well considered. I simply stated my position, and don't argue with those who hold another. But I have a strong distaste for the menu frenzy of a number of the Japanese cameras.

 

If I had older Nikkors, I would be glad to be able to set their information on current cameras so as to continue using them. But Nikon didn't set out to make a pathway for people to use Zeiss lenses on their bodies.

 

--HC

[leicapfile]

Thanks for the several replies noting the advantages of coded lenses.

 

I drafted my thread regarding the possibility of "menuizing" the same info provided by the 6 bit lens coding mainly because of a flyer I received today from Leica NJ.

 

Therein was a listing of the old and current lenses which they would apply the coding to at $125 a pop. For me that could total $1375 plus shipping handling, etc. were I to choose to have all the lenses accumulated over 45 years done.

 

Now I'm not a candidate for the M8, prospective or otherwise. However, given that I fall into the amatuer class of shooters, I felt that menu selection should be an available choice to me were I to be thinking of purchasing a M8. Were I a professional (and could write off that expenditure) who well could be using several lenses in a session, coding makes good sense.

 

But I'm not, so it would seem to me to have been a good idea to allow for that info to be set by the shooter if they wished. Perhaps it will be in the future. After all, to have the best of both worlds you have to have more than one world. Menu selection and lens coding could certainly co-exist.

 

Jerry

[jrgeoffrion]

I hope that the reason the DMR firmware update is delayed is because Leica is adding the ROM coding to the menu of the DMR... just like they are doing with the M8.