What will the S2 actually DO?

[tashley]

We've all argued about the price and timing of this beast but I am curious to know what it actually DOES. Leica's website is rather delphic on the question.

For example, does it have Live view? How many AF points, how distributed and of what sort? Does it have continuous focus? What's the highest ISO? Fastest shutter speed? Longest exposure?

I'm sure everyone but me has worked out where this info is so please someone, put me discreetly out of my misery!
:salute:

 

[Christopher]

just my guess, but I really hope Leica will proof me wrong:

- No live view not even on a computer. Perhaps later on a computer. In 6 months or so
- One AF Point Cross
- No continuous focus
- ISO 800/1600 the later probably just a unusable
- 1/4000 and 1/500 with CS lenses
- I really hope something around 5-10 minutes or even longer but my guess goes to: 1 minute

 

[thomas]

http://dfarkas.blogspot.com/search/label/Leica S2

 

[Guy Mancuso]

Trying to get the specs now.

 

[dfarkas]

The final specs are being released by Leica within about a week or so.

What we do know is:

No Live View on LCD- unknown at this point if supported in tethering s/w. Capture to screen time was estimated at 1 second or less, but I've yet to see the tethering s/w in action.

One AF point - supposedly very accurate and tuned precisely with each and every lens for every aperture (to account for aperture-related focus-shift)

Not final, but I would assume that the S2 has single and continuous focus modes. Manual is menu-selectable, but AF can always be overridden with clutchless focus ring on lens. AF can also be dedicated to the rear thumb AF button and seperated from the shutter release.

Not final either, but hints at good 1600 ISO in full res and 3200 ISO in binning mode. We will have to wait a few days for this one.

As was already stated, 1/4000th of a sec is the min shutter speed on the focal plane shutter and 1/500th of a sec is the min speed with central shutter (CS) lenses. Either can be selected with a switch on the upper left corner on the back of the camera.

According to my latest info, there is no hard limit on long exposure. Similiar to the M8, you can use as long an exposure as you want in Bulb. The camera's top OLED screen counts up in seconds. At PMA in March, we tested it to 126 seconds to make sure it would go past 2 minutes. Could have gone longer, but Stephan's finger was getting tired and we were both satisfied with the result. Of course, the camera requires a 1:1 dark frame exposure, so a 2 min exposure would require a subsequent 2 min dark frame exposure for NR.

I'm anxiously awaiting all the final specs and will provide them as soon as I get all the info.

David

 

[tashley]

Thanks David - I knew I had seen somewhere some info beyond what is on the official S2 mini-site but I couldn't remember where or how 'guessy' it was.

That's really helpful.

I might be the only person on planet earth but I am slightly thinking of getting one... I just saw the LL link to an amazing video that show how hard it is to tell G1H video from RED and 5DII and it struck me that the kit to replace ALL my gear (P45+ on Phamiya/Cambo WRS, 5DII, K20D, M8.2, Pen) would be an S2 and a G1H with the Pen as its lens-sharing more social sister. I would be giving up comparatively little in terms of what I generally shoot, and getting far far fewer bags of stuff and things to remember when I switch systems.

Hmmm....

 

[dfarkas]

I might be the only person on planet earth but I am slightly thinking of getting one...

Hmmm....

You are not the only person thinking of getting the S2. I know quite a few myself ....they just don't tend to write as such on the forums.

David

 

[Bob]

Oh, I might be convinced, but It will take a little while...
-bob

 

[Guy Mancuso]

You are not the only person thinking of getting the S2. I know quite a few myself ....they just don't tend to write as such on the forums.

David

Let's hope they actually write the check when the time comes. :thumbs:

 

[LJL]

David,
In your blog posting you mention that Leica says each lens will be calibrated with a profile of corrections and stuff that will feed to any S2 body for use in processing within the camera. Does that mean ROM chipped type of lenses? Does the body and Maestro processing then read these profiles, or whatever they will be, so that the camera will make the on the fly corrections? That sort of information is not mentioned anywhere else that I have seen, and it seems important to the discussion about S-system lenses being tied even more tightly to the S2 body. In other words, the lens may be very good to start, but may have all sorts of variances that can be corrected in software processing, but only if the software knows how to read that information. Meaning the S glass may work or not work so good on anything else folks try to adapt it to that does not use the processing algorithms for that glass. Not looking to make this a big issue, and it will be valuable to see the specs and what Leica has to say about all aspects of the new camera and system shortly.

It will be interesting to see how many folks will write the checks when this camera comes to market.

LJ

 

[robmac]

I love this.

Leica does un-spoken-of in-camera lens-specific tweaking, which as LJ mentions ties the end IQ, especially lacking any custom PP software, indirectly to the lens+stored lens profiles & camera processing HP. It makes some in-camera processing mandatory on RAWs (and raises the question of ability to turn-off, save processing power and tweak outside the camera).

This, Leica thinks is fine/good engineering in Leica's mind, BUT when, Hassy does similar BUT advertises it as an advantage, a way to get more affordable glass out, does the tweaking (IIRC), for the most part, outside the camera at the photogs discretion and Leica does a "harumph" and points to how their glass won't need such pedestrian methods. Give me a break.

While I think the concept is perfectly logical - after all Max IQ is always the end goal (and lenses you can actually pay for in this lifetime), the more details of the S2's modus operandi that come out, the closer Solms seems to be borrowing from the Hassy playbook (minus the custom PP software) than they would like to admit. It also certainly isn't making S2 glass any more affordable than HC/D units.

 

[carstenw]

Rob, I have a feeling that you are deliberately misunderstanding the real point: Leica does lens recognition, and focus tuning. Hasselblad deliberately designs lenses which contain certain optical flaws which are known to be easy to correct in software, such as distortion, to keep size and cost down.

There is a huge difference here. Hard to miss, really. Unless one tries to, of course.

 

[LJL]

Carsten,
To be fair, it is unclear exactly what Leica is or is not correcting for, hence the questions I have asked. The post by David suggests things like focus tuning, as you mention, but there is little other information to support or refute anything else at this point. So we do not know that the Leica lenses are or are not designed with with various in-camera post-shot corrections in mind. The suggestion from Leica is that they are optically superior and do not need any corrections, yet they also say that they are calibrating each lens for corrections....at least in focus tuning and aperture, from what was posted.

LJ

 

[robmac]

Carsten - Not deliberately at all.

Hassy DOESN'T design flaws in (Canon, I'm not so sure.. ;>), what they do is not go the the Nth degree to design 'perfect glass' when the differences can be tweaked in firmware/software. Logical.

Same as Leica is doing (again logical). Does perfect glass, by definition, included lenses with focus shift? For 5K Euros, I would bloody well hope not. Given this is the first we've heard of these in-camera S2 lens-specific tweaks, I suspect there is also more going along the same vein than Leica would like to admit. Again, I see NOTHING WRONG AT ALL with what Solms is doing technically. Makes perfect sense.

They, LIKE HASSY, are correcting for a less than perfect lens-specific performance under certain conditions (aperture, focus distance, etc) by doing it in-camera corrections. Correcting for shift in firmware is NO different than correcting for vignetting, or any other less than perfect short-coming of a lens.

However since you're already reading the lens ROM for aperture, distance, etc to correct for shift, what about CA, vignetting, distortion....? I have no issue with what Leica or Hassy are doing, but:
------------------

1) I wish Leica would get off their high horse - they (on paper for now) like, Hassy have a strong system, but call a spade a spade. The "we don't need that" comments, in hindsight, just make them look childish and make me ask (as it will others) what else is going on when you trip the shutter that isn't talked about? How will the S2 react IF you ever get to chance to bolt dumb-old V glass to it (like that will happen)?

2) I'd be VERY surprised if they were not correcting for more than just shift. Once you have the existing ROM data (lens, aperture selected, distance, etc) and the right-sized lookup table, correction factors and processing speed, there is a LOT you can start to play with -- and a GREAT temptation to play with it. That said, if you do it right and have the processing HP - you can work wonders.

3) It means out-of camera comparison testing will have to ensure that you are comparing like-like.

4) Assuming the tweaking goes beyond shift (or will later go beyond shift), what does that means for the fact that they have no tailored PP software?

5) If they had put the childish BS aside and had advertised what they were doing ; "Our Lenses, are, after all, Leica's, (hint., hint) but the laws of physics are the laws of physics (even in Solms) and since max shooter image quality is always our end goal....", etc it would sell well. After all, no one but Leica seems to object to the Hassy (now Solms) method. That said, they now look foolish - and beg the question of if they are doing more than just correcting for shift - even if they aren't/won't. Not a way to instill potential fence-sitter confidence.

Time to get the system in some impartial user hands - and see if the wallets open like Leica needs them to in fiscal 2010.

 

[ddk]

Funny thing about all this perfection is that one gets bored with it very quickly, G1 and GH1's appeal to a large degree seems be the possibility to use old, imperfect glass! IMO perfection is way overrated, I'm hoping that the S2 lenses would have some of the strong Leica charm and characteristics of their older lenses that many of us love, rather than so called more perfect but sterile looking glass that we're seeing from everyone these days.

 

[robmac]

David - good point. I also don't want my camera to be the worlds most perfect Xerox machine - regardless of who makes it.

 

[carstenw]

Hassy DOESN'T design flaws in

Well, I suppose it depends on your definitions of "flaw" and "designs in". My understanding is that they do indeed do that, for most reasonable definitions of both terms. Note that I have not said it was a bad thing, it is just that deliberately leaving in distortion and correcting it in software does match my description.

Does perfect glass, by definition, included lenses with focus shift?

Leica does not, like Hasselblad, speak of perfect glass. They may speak of the "best glass", and with some right. There is no such thing as perfection, of course. There are always design compromises required. Still, the pertinent difference here is that correcting distortion costs resolution in the corners, measurable in the lab, even if not visible under normal circumstances. Correcting focus shift does not.

Correcting for shift is NO different than correcting for vignetting, or any other less than perfect short-coming of a lens.

Well, yes, there is a difference, as explained above. If you were to shoot a complicated texture with the Hasselblad 28mm lens, for example, you would see that after correction, the corners would suffer a little. Naturally so, and in general I applaud Hasselblad for their approach, since it saves size and cost. Leica doesn't do these types of compromises, however, and reserves in-camera corrections for a lesser class of problems, focus shift and perhaps one or two others, like natural vignetting. One problem they cannot solve in this way is wallet-ache...

The "we don't need that" comments, in hindsight, just make them look childish.

I don't necessarily disagree on this point, but it may make marketing sense. Hasselblad's approach, while in my mind sound, did raise some eyebrows at the time, and I can see that Leica wants to drive a wedge into the crack. After all, Hasselblad is playing "budget" manufacturer in order to get the upper hand over Phase, and possibly Leica, and is thus forcing everyone's hand and making the game more deadly. In turn, Phase has made fun of Hasselblad's definition of full-frame of one year ago. It is war, and all's fair.

 

[robmac]

Carsten,

We KNOW what Hassy does. They're open about it and we both appear to applaud it.
We NOW KNOW Leica corrects for shift - after poo-pooing Hassy for doing lens-specific corrections.
We DO NOT know what Leica does or does not correct for outside of shift...
But since we NOW KNOW they correct for shift, it begs the question....and make undermines one of their own marketing thrusts.

Correcting for ANY lens shortcoming, is correcting for a shortcoming. Shift is as much a shortcoming and anything else and can ruin an image just as much as CA, curvature, etc if not corrected for. To swear you don't need to correct for any, belittle the comp for doing it, and then have it known you correct for one (that is known of) is just not kosher, puts a marketing target on your chest -- and just begs for people to dig further. Digging you may not want.

We DON"T KNOW how well the S2 does it because no real sample images are out there as yet.

We'll also NEVER KNOW how much effect it has if images can't seen in before/after mode - which is unlikely. Leica lenses are overall, superb, but just how much better will the S2 units be than HCD or Phamiya equivalents if the camera firmware is a KEY but HIDDEN component in performance? If I were bellying up that much $$$, I'd like to know.

We come at this from different mindsets. As for if one correction is the same as another or what defines a 'perfect' vs 'best' lens or and what is a 'flaws" and or what amounts to 'designed-in', etc., etc is just debating semantics and perceptions.

Lets just say we agree on some points and disagree on others ;>

What ONLY matters for Solms is if sufficient people vote with their wallets. Sign-up lists are nice, but as the man said, until the cheque is cashed...

 

[georgl]

I've talked to some Leica-people several months ago, so when I got something wrong, David needs to correct me:

- Yes, the camera knows exactly which lens is used and every lens is factory-calibrated and measured, the generated data is written on the ROM-chip (AF, aperture, focus-shift...)

- This data is not used for any work in the RAW-converter, not being able to read it (when adapting the S-lenses) won't be a problem, but....

- The cover-glass of the S2 has different optical entities (thickness, integrated filtering) and these entities are taken into account by the optical design. So using them with a different cover-glass (-> other MFDB) could result in certain problems as well. The Schneider/Rodenstock-digital-lenses are also taking the cover-glass into account, but only the "traditional" type used in MFDB.

 

[carstenw]

I went looking for the famous quote about software corrections, and all I could find was this:

"The optic calculation incorporates all the factors and requirements of digital technology, meaning that software to correct image artifacts becomes superfluous."

Clearly, the accuracy of the statement hinges on the definition of "image artifact" here. So far, they are on track, given that focus shift is not an image artifact. Their implied criticism of the chosen Hasselblad direction is also covered, since distortion is an image artifact. Let's see if they can keep it up when the system goes into production...