View Full Version : The official Leica Press release on new lens info for M8
Press Release
June 2006 / No. 17e/06
Release date: June 6th, 2006
New 6-bit coding of Leica M system lenses for image optimization in connection with the planned digital camera model
Leica Camera AG, Solms will be giving the lenses of the Leica rangefinder system a new code on the bayonet ring in future to enable the planned digital Leica M camera to recognize the lens type. The information on the lens that is being used helps the camera to optimize image quality. All lenses leaving the factory from July 1st, 2006 onwards will have the new coding, although they can still be fully used with the current analog cameras LEICA MP and LEICA M7 as well as classic models built after 1954. Lenses in the current range as well as many earlier models can be retrofitted at the cost of the owner to benefit from the image optimization in the camera. The lenses are compatible with the planned digital M camera even without retrofitting, except that the additional features cannot be used.
The lens coding is called ‘6-bit coding‘ because six fields in the bayonet ring are marked in black or white to represent a number from 1 to 64 in binary code. The planned digital M camera reads this information optically and can identify the lens on the basis of this code. Apart from the improvement in image quality, this information is also written into the EXIF image file.
“On account of their legendary quality, nearly all Leica M lenses are ideal for digital use. However, the new 6-bit coding also uses the performance reserves in the image processing of the camera to give our customers the excellent image result they expect from Leica,“ says Rainer Bültert, product manager for the M system at Leica Camera AG.
Lenses bought in the past will be converted at the request of the customer for 95 euros at the Customer Service of Leica Camera AG in Solms or the Leica agencies of other countries. Many of the lenses made from 1963 onwards can be converted. A list of such models is available on the internet (www.leica-camera.com) or from the Leica Info-Service (Tel. 06442/208-111). The only lens in the current range that will not be given a 6-bit coding is the LEICA APO-TELYT-M 135 mm f/3.4. It is not codable later, either, as its extension factor of 1.33 makes it unsuitable for use on the planned digital M camera.
The launch of the digital Leica rangefinder camera is planned for the second half of 2006.
Contacts for your editorial team
Gero Furchheim / Extension –450 (tel.)/ –455 (fax)/ gero.furchheim@leica-camera.com
Sandra Looke / Extension – 404 (tel.)/ – 455 (fax)/ sandra.looke@leica-camera.com
bmattock
06-05-2006, 14:00
Interesting. Apparently, Leica is planning to apply a CPU-based tweak to the images, according to what type of glass is attached. Different lens, different tweak. If you don't get your lens bar-coded, it won't know what you have and won't apply the tweak (whatever that might be).
I suspect a) There will be a small market in third-party peel-n-stick bar codes and b) someone will start to figure out how to 'hack' the bar-coding to get different effects from lenses, not intended by Leica.
Best Regards,
Bill Mattocks
b) someone will start to figure out how to 'hack' the bar-coding to get different effects from lenses, not intended by Leica.
so this would be similar to trying out this film at the EI, with that developer, at this temperature for this time. (no bashing intended)
"I love the look of the 50'cron on TriX, when set to the 35'lux." :D
bmattock
06-05-2006, 14:11
so this would be similar to trying out this film at the EI, with that developer, at this temperature for this time. (no bashing intended)
"I love the look of the 50'cron on TriX, when set to the 35'lux." :D
That's what it sounds like, eh?
Interesting. Apparently, Leica is planning to apply a CPU-based tweak to the images, according to what type of glass is attached. Different lens, different tweak. If you don't get your lens bar-coded, it won't know what you have and won't apply the tweak (whatever that might be).
I suspect a) There will be a small market in third-party peel-n-stick bar codes and b) someone will start to figure out how to 'hack' the bar-coding to get different effects from lenses, not intended by Leica.
Best Regards,
Bill Mattocks
I think Kodak was the first who did that.
A friend has a DCS Pro and a SLR/n and both have some sort of a lens databse built in.
trittium
06-05-2006, 14:31
Looks like Leica is also looking to boost profits through charging an arm and a leg for black and white dots.
Isn't that the exact same text that was on the French site, except a better translation than the Google translation? It doesn't say anywhere that the coding will be required to actuate framelines, that's a huge relief.
erudolph
06-05-2006, 14:51
I can't seem to find the list of lenses that can be converted. Anyone have a link to it?
Ed
John Camp
06-05-2006, 14:52
"Looks like Leica is also looking to boost profits through charging an arm and a leg for black and white dots."
Nah. I understand it, they'll mill shallow recesses to take the paint and then, of course, get the paint right... If it takes half an hour each, to do all of it, plus check, unpack, repack and send, I doubt that there'll be much profit in it. A milling process for this kind of thing won't be cheap.
But let me ask this: I have seven lenses. Do I put them all in one Fed-Ex box and ship it off and then say rosaries until it comes back, or do I put them in (say) three separate boxes so I'd be unlikely to lose them all, should the package get lost? And take three chances of losing something, rather than one?
Also...a la Bill's suggestion...the codes will probably be on the Internet about a week after the first lenses are sold this summer, and I wouldn't be surprised if some carefully applied black and white nail polish wouldn't do the job. Anybody know a Goth?
And, of course, that would probably solve the problem of using Leitz lenses. Leitz might even tell you where to get those peel-off stickers, wink-wink, nudge-nudge...
When are they going to start doing the adaptions on the current lenses? Anybody?
JC
bmattock
06-05-2006, 15:00
And leave us not forget the gear-heads that will claim that once the paint is applied, the lens is no longer pristine glass and is now unworthy junk that should not bear the name Leica. And then those who will swear that when the lens goes back to Leica, they perform some sort of elf magic that no one else can properly do, so that if the lens is not converted by an actual gnome, the glass is horribly damaged and won't take photos anymore.
And everyone will claim that they can tell at ten paces in the dark with sunglasses on if a 4x6 print was made by a converted or unconverted lens.
Yep. I can see it now.
Best Regards,
Bill Mattocks
trittium
06-05-2006, 15:06
And leave us not forget the gear-heads that will claim that once the paint is applied, the lens is no longer pristine glass and is now unworthy junk that should not bear the name Leica. And then those who will swear that when the lens goes back to Leica, they perform some sort of elf magic that no one else can properly do, so that if the lens is not converted by an actual gnome, the glass is horribly damaged and won't take photos anymore.
And everyone will claim that they can tell at ten paces in the dark with sunglasses on if a 4x6 print was made by a converted or unconverted lens.
Yep. I can see it now.
Best Regards,
Bill Mattocks
Bill,
I think that is exactly how it will go
back alley
06-05-2006, 15:13
i wonder if leica has programmed more 'glow' into the image enhancing capabilities of the camera?
does this mean there can be no more claims for leica glass characteristics?
Maybe someone will figure out what dots are suitable for CV, Canon, Zeiss lenses too?
I am already thinking of ideas for an alternative and am just waiting for the specs.
JohnL shares my question -- what about all the great lenses not made by Leica. I am a very dedicated Leica shooter, but I also have lenses from Konica, Zeiss, Voigtlander and Canon...so now what? Perhaps DAG or Sherry Krauter will start offering this as a service for 3rd party lenses?
bmattock
06-05-2006, 15:50
i wonder if leica has programmed more 'glow' into the image enhancing capabilities of the camera?
does this mean there can be no more claims for leica glass characteristics?
Perhaps - but it will be replaced by Leica electron characteristics. Leica CPU characteristics. Fact - Leica electrons just 'flow better' than non-Leica electrons - and Leica photons will glow when properly encouraged by the Wizards of Wetzlar.
Best Regards,
Bill Mattocks
Bertram2
06-05-2006, 15:59
P The information on the lens that is being used helps the camera to optimize image quality.
There is a anti-vignetting variable in the software, for the wides ?
bertram
bmattock
06-05-2006, 16:01
Case in point.
Yes, but assuming there is an anti-vignetting feature, how could it work, given that it varies dramatically with working aperture (which will not be transmitted)?
patrickjames
06-05-2006, 16:20
The 10mp seems to me to mean that they are going to use the same sensor as the R digital back. Does anyone have any more specs on the camera?
I don't believe it will be the same sensor as the DMR, as the DMR sensor crop is 1.37, and the M8's is supposed to be 1.33. Also, even if it is the same, performance is likely to be different as Imacon wrote the firmware for the DMR, and Jenoptik is doing it for the M8.
There is a anti-vignetting variable in the software, for the wides ?
bertram
Yep, it's called Adobe Camera Raw. This M8 thingy is going to produce .DNG RAW files, isn't it?
rs
How about a gentleman's bet that it's a hoax?
H O A X ?
The date is 06/06/06 (666) and it uses a strange word length of 6 bits.
Dunno. It's worth a beer bet.
OK, Pico - I'll take that bet. Loser pays to have the winner's lens of choice fitted with Leica dots?
OK, Pico - I'll take that bet. Loser pays to have the winner's lens of choice fitted with Leica dots?
Heck, I'll dot it for you! :p
We can even make dot-hack decals until we hack the chip.
Bob Ross
06-05-2006, 17:54
There is a anti-vignetting variable in the software, for the wides ?
bertram
Besides correction of fall off, some lenses that are excellent on film, do new trcks on digital such as chromatic aberration, lateral chromatic aberration and because of the reflectivity of the sensor surface, a strange red dot in the center of the image (flare?). This is not to say that our Summicrons will behave strangely on digital, but things like this could be handled by firmware using a look up table using the mount codes or going into the menus (hopefully). I have found that digital can sort a lens collection better than any film could. I suspect Leica and/or their beta testers have encountered most of the anomolies. Let's hope they can identify where we might encounter image quality problems using non-coded lenses. I have gone through this with my E-1 and KM5D DSLRs with a collection of legacy lenses.
Bertram2
06-05-2006, 18:07
. Let's hope they can identify where we might encounter image quality problems using non-coded lenses. I have gone through this with my E-1 and KM5D DSLRs with a collection of legacy lenses.
I bet they do not intend to code afterwards any non-leica lenses. The digital world, a nice opportunity to go back to the happy times of proprietary solutions. :D Will work only if the coding soon turns out to be a real feature and not a marketing gimmick.
bertram
Pico,
This came on an original Leica Letter. I can scan it and post it but it is legit. Here is a link to the PDF
http://www.rangefinderforum.com/image_dir/rff_uploads/leica.pdf
Wow! I owe someone a beer.
This might be the first digital camera I buy.
Thank you, Sir!
Bob Ross
06-05-2006, 19:28
I bet they do not intend to code afterwards any non-leica lenses. The digital world, a nice opportunity to go back to the happy times of proprietary solutions. :D Will work only if the coding soon turns out to be a real feature and not a marketing gimmick.
bertram
I'm sure that you are right about non-Leica lenses. When Oly provided an adapter for using their old OM lenses on the E-1, it came with a list showing which lenses and apertures would produce acceptible results. The list was fairly accurate for my old lenses. At least Leica's approach may be open to hacking and self application.....where did I put my Moto Tool?... :)
I wonder if some of the beta testers, who might do write ups/reviews, like Erwin Puts with the M7, might be asked to refrain from commenting on non-Leica lens compatibility, for a time. This feature of the M8 personality will be fun to watch unfold.
does this mean there can be no more claims for leica glass characteristics?
I absolutely agree ... there'll be many folks (re)discover Leica. :D
Leica is such utter BS IMO. They could have just made it a menu selection without you having to ship your glass off and get it milled. But the release implies that you will not get the software tweaking if you do not get your lenses milled. Maybe I am not reading it correctly. I f I am ....
You know what Leica, you can keep that M8.
...The date is 06/06/06 (666) and it uses a strange word length of 6 bits...
... and the PDF file has been created at 06:59. :rolleyes:
Best,
LCT
Peter Klein
06-05-2006, 21:12
Hey, John Camp, great minds think alike!
I posted this on the LUG yesterday:
"Now, anyone want to take bets on how long it will be before a do-it-yourself lens upgrade method gets posted on the Web? Something involving a rat-tailed file, a micrometer, and nail polish stolen from your Goth-girl daughter or girlfriend . . . "
:angel:
--Peter
Personally, I hope that the 6 bit data just references the focal lenght of the lens and maximum apeture. That will give us all the EXIF data we need. And I much prefer that the milled dots do the communication rather than a menu in the camera. The whole point of an analog expererience with a digital rangefinder is to minimize as much as possible any use of menus and toggle buttons and any other non-intutive camera functions. At least Epson got it right on that score.
I don't see how, with only 6 bits of information, issues of vignetting much less lateral color, can possible be addressed by the lens. Nor would I want the camera/lens to be making these desicions for me.
I turn all the auto stuff off anyway. However, the EXIF information would be nice.
Rex
Mark Norton
06-05-2006, 22:03
I think this is pretty much what we have been talking about. I certainly agree with the point about image optimisation being limited if you do not know the working aperture. So the lenses are going to have ROM chips after all, except this ROM has only 6 bits and the chips are paint chips, not silicon chips.
I am surprised they are saying the new processing cannot be used if there are no dots, they could have put a menu in the camera with the supported lens types listed from which you select the lens you use, plus a "recent lenses" list to allow you to rapidly select the lens you are using.
I'm slightly disappointed this is not being used for frameline selection as well because that indicates the bayonet lug will continue to be used which suggests we won't be moving to a world of single frame lines, new frame lines for lenses such as the 21mm and so on.
As regard the modification cost, I think it's a pretty modest charge but still 700 euros for me to have mine modified. I'm wondering whether they will doctor your own lens mount or simply replace it with a new one. More likely a new one I would have thought to streamline the workflow.
J. Borger
06-05-2006, 22:05
Keep in mind that the M8 will be used and bought by a lot of people who have ZERO experience with digital cameras. Those people will judge the M8 on bases of the in camera processed off camera files ... not the files they can get by spending a lot of time learning how to get the most out of a raw file.
I can already hear the complaints about vignetting and soft files zoomed in on screen at 100%, the dissapointing on-line samples, too high noise levels at higher isos and soft corners shooting a brick-wall .. .... the complaints have already begun and the camera is not released yet.
For those who do not want the auto-stuff ... i bet there will be no reason at all to get the contacts into the lenses. Just waite and see.
I prefer my older lenses .. those not on the upgradable list .......anyway to my latest crons/ ASPH lenses ..... so the performance (resolutionwise) will probably not be up to Leica's modern cq. general digital standards anyway.. ... so what??
The raw converter software that comes with the camera should have the option of selecting the lens and shooting aperture used, and applying the corrections at the raw conversion stage. And also having the option to accept default corrections for a certain lens or reduce/increase the corrections being applied.
I'm a bit puzzled as to the remark about the 135 3.4 apo. Why should an extension factor of 1.33 make the 3.4 unusable and not the older 4.0 for instance? I suppose they are referring to the fact that rangefinders are less than optimal for focussing beyond 135 mm lenses, but the use of a thing like the Okular 1.25 should be a useable workaround.
J. Borger
Good point about the average user wanting good quality right out of the camera. I stand corrected as being a little bit to elitist in my distain for the "auto everything" processing of the in camera image. I am not asking that a popular and necessary feature like this not be included in the camera design, I only ask that it can be turned OFF.
I guess a little tape over the dots could go a long way.
Rex
J. Borger
06-05-2006, 23:59
Rex,
I hear you ... i never shot a jpeg with any digital camera myself ... so i just hope for and count on the raw converter with all the options Nick suggests.
Looks like Leica is also looking to boost profits through charging an arm and a leg for black and white dots.
Compared to what they currently charge for a lonely red dot....
100/150-ish dollars for black AND white dots (six of them!) looks like a real bargain.... :p
Six bits isn't really all that much, I wonder. Especially with lenses such as the Tri-Elmar I don't see how they can serve for image correction in any sensible way, since this is very much lens specific and Leitz have produced more than 64 different lens models. It's probably just for EXIF.
It would have been no problem to include a little bit of electronics in the lens mount. Out of six contacts you can get a lot more than 64 combinations, even if you just use diodes. I wonder why they went this awkward route with optical coding. Probably because of longevity and because of electronics aversion in the Leica user base.
Philipp
I would say that it is probably because adding rom contacts would have cost a lot more than adding the dots.
Mark Norton
06-06-2006, 04:08
J. Borger
Good point about the average user wanting good quality right out of the camera. I stand corrected as being a little bit to elitist in my distain for the "auto everything" processing of the in camera image. I am not asking that a popular and necessary feature like this not be included in the camera design, I only ask that it can be turned OFF.
I guess a little tape over the dots could go a long way.
Rex
A little WHITE tape, that is...
As regards the contacts, the dots are effectively a 6 bit static ROM and using electrical connections would not have worked reliably because they are outside the lens mount and susceptible to damp - there is no room inside the lens mount, it's full of glass in many lenses.
When a Tri-Elmar is mounted, the camera can use the position of the viewfinder frame lever to determine which of the 3 focal lengths is being used so it will be possible for Leica to optimise each of the focal lengths separately.
If you look at the list of supported lenses, it's about 30 going back to the early 60's and ignoring black/chrome differences so there is room for future growth. Seems to me each new lens introduced will now come with a CD ROM to allow you to "teach" the camera about the new lens.
John Camp
06-06-2006, 05:34
I think the paint spots are an elegant solution to a compicated problem. If they made a mistake in this, I think it might be that they didn't go for an eight-bit code right off the top. I think they may eventually find that they have more infromation to transmit than they originally thought.
Apparently they feel that 64 cases are sufficient to cover the lenses they wish to support, and my bet is that they are right. Breaking the 64-case barrier will indicate an entirely new generation of camera/lens combination, and they have time to do that gracefully.
The infinity ramp/cam and frame position gives information without the code. Taking-aperture can be read using led lasers in the camera, turned on for a microsecond; glass won't significantly defract it. Perhaps they can use the same laser idea to further detect what lens is mounted by reading other internal lens information such as lens mounting steps.
So, even an uncoded lens could still have the frame lever and infinity cam, and possibly the laser data to indicate what lens is mounted.
Personally, I look forward to what they do. I want to believe that Leica will pursue a conservative course of compromise that will leave us manual-mavens with the control we wish to have.
Taking-aperture can be read using led lasers in the camera, turned on for a microsecond.
They can skip all the James Bond/Star Trek stuff as far as I'm concerned. All I want is a digital version of the M6, and it sounds like that's what's in store. These coded blobs sound like they're as useful as ROM on an R8 and that's just fine with me.
Mr. Camp mentioned paint spots. We don't know what kind of spots they will use, do we?
By reading dots they can exploit the analog quality of a physical dot.
For example, one could use engraving or lasered printed dots with different orientations of lines in each spot (like half the dot vertical, half horizontal, or even thirds or quarter orientations) thereby adding new dimensions to the word. That's something we can't do with straight binary representations where a bit is just a bit.
Spacing can be changed to shift the dots into alternate readings - say shift them all one dot right to indicate a different class of data.
So, it's possibly not one 6 bit word, and depending on the technique, it could yield far more bits of information than can possibly be used.
I would really like to know if the framelines will be affected by the dots
erudolph
06-06-2006, 08:14
I would really like to know if the framelines will be affected by the dots
Jorge, can you point me to the list of lenses that can be converted? I don't seem to find it on the Leica website.
thanks, Ed
Leica is such utter BS IMO. They could have just made it a menu selection without you having to ship your glass off and get it milled. But the release implies that you will not get the software tweaking if you do not get your lenses milled. Maybe I am not reading it correctly. I f I am ....
You know what Leica, you can keep that M8.
Go into a menu and change the settings each time you change lenses?? Or each time you switch between focal lenghts on the Tri-Elmar?? Now there is a brilliant idea......
phototone
06-06-2006, 08:30
Lens optimization is a valid concept, which is/was used by Kodak on their Full Frame digital cameras, the SLR/n, and 14n. These Nikon mount cameras use the "chip" that is in the lens to tell the camera what it was.
The Leica solution is just as valid as the Kodak solution was.
However, I am sure, just as the Kodak cameras had, the Leica will have a way to manually set the optimization for the lens in use. Probably thru a menu selection on the back, and from a database built into the firmware on the camera. While this will work just fine, it doesn't address the issues from using non-Leica brand lenses. Kodak had a database that included all popular brands of lenses available in Nikon mount, as well as "generic" optimization settings for lenses not listed in the database. The big question will be: Will Leica include lens optimization settings for non-Leica lenses, such as the new Zeiss lenses and the Voigtlander lenses? Perhaps there will be a way to input parameters for lenses not in the database?
Lens optimization is a valid concept, which is/was used by Kodak on their Full Frame digital cameras, the SLR/n, and 14n. These Nikon mount cameras use the "chip" that is in the lens to tell the camera what it was.
The Leica solution is just as valid as the Kodak solution was.
However, I am sure, just as the Kodak cameras had, the Leica will have a way to manually set the optimization for the lens in use. Probably thru a menu selection on the back, and from a database built into the firmware on the camera. While this will work just fine, it doesn't address the issues from using non-Leica brand lenses. Kodak had a database that included all popular brands of lenses available in Nikon mount, as well as "generic" optimization settings for lenses not listed in the database. The big question will be: Will Leica include lens optimization settings for non-Leica lenses, such as the new Zeiss lenses and the Voigtlander lenses? Perhaps there will be a way to input parameters for lenses not in the database?
If not in-camera, I'm sure the post-processing software will have options. That is probably the more practical way.
Mark Norton
06-06-2006, 09:03
Go into a menu and change the settings each time you change lenses?? Or each time you switch between focal lenghts on the Tri-Elmar?? Now there is a brilliant idea......
If there is a menu option to select the lens you are using, the whole point of the dots is to fast track that selection so that you don't have to. And with the Tri_Elmar, you won't have to do it either because the computer program which is handling all this can say, "lens coding is Tri-Elmar, I need to look at the finder lug to determine the focal length".
The question in my mind is whether they've figured out a way of sensing the working aperture...
If there is a menu option to select the lens you are using, the whole point of the dots is to fast track that selection so that you don't have to. And with the Tri_Elmar, you won't have to do it either because the computer program which is handling all this can say, "lens coding is Tri-Elmar, I need to look at the finder lug to determine the focal length".
The question in my mind is whether they've figured out a way of sensing the working aperture...
Sorry -I forgot to hit the light-touch-non-malicious-sarcasm-smiley;) I just meant to say that it seemed a good idea to me to make that automatic instead of saddling the user with menu options, and I took issue at calling common sense BS... Additional in-camera and post-processing corrections are fine with me.
Lens optimization is a valid concept, which is/was used by Kodak on their Full Frame digital cameras, the SLR/n, and 14n. These Nikon mount cameras use the "chip" that is in the lens to tell the camera what it was.
The Leica solution is just as valid as the Kodak solution was.
For the record, Hasselblad is also using lens optimization. The H2D-39 uses "digital APO correction" to enhance the lenses. Personally, I would rather have optical APO correction, but as long as it works....
Jon Graham
06-07-2006, 02:31
1. Even though the aperture is not 'transmitted', it can easily be calculated (to a very close approximation) by the light meter and shutter speed. Thus it could correct for vignetting at various f stops.
2. It seems likely that Leica would replace the bayonet mount for the $125. Re milling is cost prohibitive and applying a 'sticker' is ridiculous.
John: I see how number one could be true for an SLR that uses open aperture metering, but what about for a rangefinder? The camera has no way of knowing at what f/stop the camera is at, so even if it knows that the lens's maximum aperture is f/2, how does it know that the lens is at f/22, versus it just being dark out? I think you would either need a separate ambient exposure meter that was not TTL, or some way to read the actual aperture of the lens...
Stuart, when you're shooting at a metered EV9 with 1/25th you probably have f2, only unknown variable is the use of filters.
I am not sure what you are trying to say Volker. Perhaps I am dense on this matter, but doesn't the meter "see" only the amount of light coming through the lens and the shutter speed? Since the amount of light coming through the lens is dependent on the aperture, I don't see how the meter or camera can tell the difference between dim light and low shutter speed because the lens is at f/16 or f/22 and dim light and low shutter speed when the lens is at f/2 and the building is dark. Probability doesn't really enter into the question -- there are many times when you would be using f/2 if you were handheld, but f/11 on a tripod...
If someone can explain how the camera can differentiate between the two without some auxiliary meter or aperture reading device, I would be interested to hear it.
Ken Ford
06-07-2006, 04:30
I can see no way the aperture can be derived by the body with the current mount - sorry.
I am not sure what you are trying to say Volker. Perhaps I am dense on this matter, but doesn't the meter "see" only the amount of light coming through the lens and the shutter speed? Since the amount of light coming through the lens is dependent on the aperture, I don't see how the meter or camera can tell the difference between dim light and low shutter speed because the lens is at f/16 or f/22 and dim light and low shutter speed when the lens is at f/2 and the building is dark. Probability doesn't really enter into the question -- there are many times when you would be using f/2 if you were handheld, but f/11 on a tripod...
If someone can explain how the camera can differentiate between the two without some auxiliary meter or aperture reading device, I would be interested to hear it.
It does know the sensitivity, too.
With known ISO and known shutterspeed you can look up aperture here:
http://www.fredparker.com/ultexp1.htm#Introduction
Thats what the Contax G Databack did when it recorded an "estimated" f-stop, estimated because it could not know about filters.
humanized_form
06-07-2006, 04:52
i'm really really hoping Leica gets this right because i can't imagine anyone else putting this kind of investment into developing a digital rangefinder camera. not Zeiss. not Epson. nobody.
if they can carry over as much of the design and simplicity of operations from the M7, while minimizing the need for menu diving that would be great.
because i love my M7. i think it's nearly perfect and a digital version of it would make me very happy. so it may be crazy but i'm thinking it's worth spending the $5k to finally have a digital camera that i will really want to use. now i gotta start selling some stuff to get that 5k! hello ebay.
I presume that Leica are doing something quite similiar to Hasselblad's H2D-39 system. This is taken from their website.
"Getting the most from every lens: digital APO correction
The H2D-39 captures an extended set of metadata and then performs an automated correction for color aberration effects with every shot. This means that your digital captures are automatically optimized to provide the finest detail that a given lens can resolve. We have named this feature “digital APO correction” (DAC), signifying the digital, APO-chromatic correction of the images that takes place. Implementation of this feature includes detailed mapping of each H system lens, ensuring that each image represents the best that your equipment can produce. We are confident that the image quality you achieve as a result of the DAC functionality will make you - and your customers - look twice."
With known ISO and known shutterspeed you can look up aperture...Thats what the Contax G Databack did
AFAIK there are 4 interdependent parameters to exposure: ISO, shutter speed, aperture, and ambient light level. Unless I'm mistaken you need to know 3 to divine the 4th. I can get an "ideal" histogram at ISO 200 and 1/125 at any aperture so long as the light is right for that combo. Since the aperture is between the ambient light and the meter cell, the only way the camera could tell what the ambient level is independent of the set aperture would be if there was an external metering cell.
AFAIK the Contax G bodies had such an external metering cell in addition to the TTL meter. Perhaps that's what was used to measure the ambient light for the aperture guess?
It does know the sensitivity, too.
With known ISO and known shutterspeed you can look up aperture here:
http://www.fredparker.com/ultexp1.htm#Introduction
Thats what the Contax G Databack did when it recorded an "estimated" f-stop, estimated because it could not know about filters.
Isn't the lens's aperture functioning just like a filter in this case? In the case of the Contax G, the aperture was transmitted to the camera, if not the data back. In the case of the leica, the aperture on the lens is covering the meter, so it is functioning like a filter would in the example you described above. Think about it, the only light that reaches the meter is that which the lens lets through. ISO doesn't matter in this case. Sure, if you know ISO and shutter speed you can calculate aperture if you are actually IN the light you are shooting, but the meter is not seeing the full light of the scene, only the light that is being transmitted through the lens. So the question remains, how does the camera differentiate between a low light situation caused by small aperture and a low light situation caused by low light?
If I have not explained that well enough, I am sorry. But I do think I am understanding this correctly. If not, please just shoot me in the head so we can all move on.
The Contax G body uses the external meter cell only when a lens blocks the internal one and if the manual is right, the lenses don't pass aperture information to the databack.
So put it the other way around, my Gossen Sixtar measures EV14 and I could use 1/1000 f4, 1/500th f5.6, 125th f11 1/6th f16 for ISO100.
I shot it with a D60 at ISO100 with 1/125th at f11 (got the same reading from the camera set to centerweighted metering)
And I realy need to clean my window and a lenshood for the 24/2.8 would be nice, too :)
So put it the other way around, my Gossen Sixtar measures EV14 and I could use 1/1000 f4, 1/500th f5.6, 125th f11 1/6th f16 for ISO100.
Your Gossen is not behind a lens with stopped-down aperture though. A TTL meter measures light available *in camera*, not outside, and I too fail to see how one can derive working aperture from that.
Your Gossen is not behind a lens with stopped-down aperture though. A TTL meter measures light available *in camera*, not outside, and I too fail to see how one can derive working aperture from that.
Does sunny 16 work? Are shutterspeed and aperture direktly linked to the amount of light reaching the sensor?
If yes, then it will work within the limits of the sunny 16 rule under the assumption that no filtes where used.
Since most digital cameras can indicate under- and overexposed areas, you could even make a good assumption wether the photographer under- or overexposed the picture.
Have a look at the histogram, the handheld meter, the TTL meter and the histogram aren't too far from each other.
This can be fooled, yes, but Sunny 16 is a good indicator of what the aperture should have been, without filters, and the sensor data show what they got, with filters and intentionaly "wrong" exposure.
The histogram is not in any way related to working aperture, I don't see how it is relevant: it is an issue of building distribution of pixel values.
Sunny 16 trick could maybe work if there was an external meter available for reference. I don't see how can it work in case of TTL metering with non-automatic aperture: you don't know the actual EV, the exposure you get is the amount of light you have through the lens (which can be stopped down to unknown amount).
Ok, I must be wrong. I thought exposure is the amount of light hitting the film and the amount of light can be controled by aperture and shutterspeed (and filters).
Come to think of it, has anybody here ever used a Rolleiflex where you could couple aperture and shutterspeed?
That's exactly what I mean.
EV14 on ISO100 can be 1/1000th f4, 1/500th f5.6, 1/250th f8, 1/60th f16 and so on.
Ok, I must be wrong. I thought exposure is the amount of light hitting the film and the amount of light can be controled by aperture and shutterspeed (and filters).
Of course it's true. It's just that I don't see a way to derive aperture from shutter speed, ISO and TTL metered amount of light, unless you know what amount of light is actually on the subject (i.e. how much light was lost in transmission through the lens).
It's just that I don't see a way to derive aperture from shutter speed, ISO and TTL metered amount of light, unless you know what amount of light is actually on the subject (i.e. how much light was lost in transmission through the lens).Transmission losses aren't a problem. On one hand with modern lens coatings there isn't really all that much light lost in transmission, probably not on the order of magnitude of half a stop. On the other hand the camera more or less knows what lens is attached and can compensate by using a lens database, since transmission losses are lens specific.
Philipp
Phillipp, my point was about the loss due to lens effective aperture (and that's what we were trying to figure out), not coating reflections which are indeed neglectible. As others pointed out, from a given luminosity value on a TTL meter we can't really tell if is a subject that dark or just a lens stopped down.
The Digilux2 has an external lightbalance sensor. Such a thing could easily read ambient light levels and then the camera could calculate the f-stop through TTL. I think it very possible the M8 will have an external lightbalance sensor as well, as it works exceedingly well on the Digilux2
As others pointed out, from a given luminosity value on a TTL meter we can't really tell if is a subject that dark or just a lens stopped down.Yes, that way it's true, of course.
Philipp
Phillipp, my point was about the loss due to lens effective aperture (and that's what we were trying to figure out), not coating reflections which are indeed neglectible. As others pointed out, from a given luminosity value on a TTL meter we can't really tell if is a subject that dark or just a lens stopped down.
But I rely on the luminosity beeing the same with a varity of lenses, the Gossen Sixtar I mentioned was bought by my father for his Robot with a Schneider Kreuznach lens, a Rolleiflex with a Tessar and a Linhof Technika with a Rodenstock lens. I used it with a Zorki and a Kiev and somehow it agrees with my Contax SLRs with, hard to admit, Contax, Yashica and even Soligor lenses.
So I just assume by observed evidence that a Contax lens at a given aperture transports as much light as a Yashica lens at the same aperture, same from my two Jupiter-8 and one Jupiter-12. It works.
So I just assume by observed evidence that a Contax lens at a given aperture transports as much light as a Yashica lens at the same aperture, same from my two Jupiter-8 and one Jupiter-12. It works.
And..? I know what aperture is :) I don't see a way how you deduct an aperture from a value of TTL meter, not knowing the ambient light. Up to the point that I started to doubt my sanity and tried to put down it into math, but all I end up is one equation with two variables unknown, aperture and subject brightness.
For example, we know the shutter speed is 1/125, and film speed ISO 100. TTL sensor shows 100 Lux/second.
Now, how exactly you determine aperture?
Varjag, unless my fundamental understanding of the universe is wrong, you, Ben and I are correct in this matter. Jaapv's idea of an external ambient reading would be the only way I can think they might do it, but even that could be tricked, as you can be standing in one light and shooting at something in another (i.e. standing in the shade and metering on the sunny mountain top 2 miles away).
And..? I know what aperture is :) I don't see a way how you deduct an aperture from a value of TTL meter, not knowing the ambient light. Up to the point that I started to doubt my sanity and tried to put down it into math, but all I end up is one equation with two variables unknown, aperture and subject brightness.
Hm, how does the meter do it?
As far as I know it tries to represent the scene as 18% grey no matter what the color or illumination of the metered scene is.
But I may have cheated a bit, I started with a negative exposed to what a meter would read. Think autoexposure.
Thus I reduce it to the subject illumination as metered with a given sensitivity and shutterspeed, then only the aperture is unknown.
At the moment I have EV13, bright light but no shaddows sunset in about 2 hours, I shoot with ISO100 and the built in meter sets my shutter to 1/125th then my lens must be at f8. If it sets the shutterspeed to 1/250th the aperture must be at f5.6 and so on.
"At the moment I have EV13, bright light but no shaddows sunset in about 2 hours, I shoot with ISO100 and the built in meter sets my shutter to 1/125th then my lens must be at f8. If it sets the shutterspeed to 1/250th the aperture must be at f5.6 and so on."
If the camera/processor know the light level is EV13 then it can deduce the aperture. Otherwise, you might well have a lower light level, and a larger aperture. The ambient sensor, as used on many P&S cameria would allow the processor to deduce aperture, in that there will be two meter cells, one behind the diaphragm, one on board the camera, and it would be a reliable sustem unless you obscure the on-camera sensor.
In any case, I would doublt that deduction of the aperture is necessary. If the processor needs to know the lens type, in order to map for vignetting due to oblique light rays, I would think the vignetting is independent of aperture. Although I'd be interested to know if it varies slightly with subject difference.
Baaah! Now I get it! The camera has no way of knowing how much light is out there and if I try to shoot a supernova or the moon.
Damn! Sometimes it takes realy long until the obvious trickles down to my brain.
I would think the vignetting is independent of aperture..
From my own experience, which confirms what I've seen in print multitudes of times, vignetting very much does depent on aperture. Most lenses vignette less and less as they are stopped down, in fact some vignette significantly at maximum aperture and not at all by mid-apertures. My 21mm Elmarit pre-ASPH is a case in point (and one of the codable lenses according to the list). At 2.8 there's maybe 2-2.5 stops of vignetting which is all but gone by 5.6. How could software compensate without knowing the shooting aperture?
Of course we're talking about the vignetting inherent in the lens, which in fact would be lessend due to the cropped sensor. Perhaps there is some vignetting caused by the digital sensor/microlenses which is not aperture-dependent, but more severe with some lenses than others. That might be compensated for by the camera only knowing the specific lens that's mounted.
Good point Ben. It would be interesting to know whether that type of vignetting is aperture dependent as well. Perhaps RD-1 users can speak to that...
And Volker, I am glad we worked that out...I thought I was going crazy there for a second...
Sorry Ben, should have made myself clearer; as you say, vignetting on a lens does vary with aperture.
However, I have read that vignetting due to light rays hitting the microsites (or microlenses) on the CCD sensor at an oblique angle does not vary with aperture to a significant degree. But I've only read this on the 'net somewhere, so you are within your rights to take this statement with a pinch of salt. If aperture did make a difference, I would have thought it's pretty easy to deduce what the aperture is using two light sensors as suggested earlier, and it wouldn't be hugely expensive.
...vignetting due to light rays hitting the microsites (or microlenses) on the CCD sensor at an oblique angle does not vary with aperture to a significant degree...
Yes digital vignetting does not vary much with aperture.
Best,
LCT
herbkell@shaw.c
06-07-2006, 14:28
I see that the 135 is "unsuitable" for the M8. Does this mean it can't be used at all or is is simply too difficult to focus. I would have thought that a smaller field of view was a bonus focusing a rangefinder lens but I am a rookie in the rangefinder world and may have this wrong
Yes digital vignetting does not vary much with aperture.
Best,
LCT
I just looked again into Sean Reids R-D1 lens review, and yes, most of the lenses don't improve much at f8.
Hopefully Leica found something more than a software tweak to improve this.
I see that the 135 is "unsuitable" for the M8. Does this mean it can't be used at all or is is simply too difficult to focus. I would have thought that a smaller field of view was a bonus focusing a rangefinder lens but I am a rookie in the rangefinder world and may have this wrong
The ease of focus would not change, but their argument is that the framelines would be too small to accurately see what was in the picture. I think this is more or less a legal argument so that people won't say: "Hey, I just bought your damn 3000 dollar lens, and I can't get a single good shot!" The 135mm will work fine on the M8, but the framing would probably be very difficult, even with a 1.25 magnifier. Think of it this way, the 135mm frameline WITH the 1.25 magnifer will be smaller than the current 135mm frameline is without it. Just like they left the 135mm framelines out of the .58 Leica, they will probably leave them out of the M8.
I see that the 135 is "unsuitable" for the M8. Does this mean it can't be used at all or is is simply too difficult to focus....
Any 135mm lens should be usable on the M8 but assuming that the base length of the rangefinder and the finder magnification (.72x) will be the same as that of classic Ms, it should be difficult to focus at f/4.5 and wider apertures IMHO.
Best,
LCT
Mark Norton
06-07-2006, 23:03
My hope is there will be a frameline for the 21mm lens - which is no more than providing one for a 28mm on a film M.
vBulletin® v3.6.8, Copyright ©2000-2010, Jelsoft Enterprises Ltd.