Press release: S2 Technical Specs

[dfarkas]

I never understand why so many cameras come so close yet fail to reach 100%. Why is that? Does it save a huge amount on the mirror or prism size? I agree it's bizarre that a camera striving for such perfection, at a price that should ensure perfection, would do this.

I had a discussion with Stephan Shulz, the project manager for the S2 a few days ago to get answers to some of the lingering questions raised after the specs were announced.

Firstly, Leica measures the coverage of the viewfinder in percent of image area. Canon, for example, measures by linear coverage. The reality is that at 96%, the viewfinder is "cutting off" only about 0.2mm on each side. Measured linearly (same as Canon's method), the S2 viewfinder is 99%.

The reason for lack of 100% is that the sensor requires shimming and a little bit of movement in the calibration process druing assembly. In order for the viewfinder image to correspond to what the sensor records, there needs to be this 0.2mm spill to cover adjustmnets in assembly.

Be wary of any manufacturer stating 100%. They usually mean 99.x% on a linear basis and round up to the next whole integer. I'd be curious to see if anyone is willing to run a test on a camera with a "100% viewfinder" with some grid paper and share the results. I believe this is another case of Leica erring on the side of conservatism and it has nothing to do with physical size of the housing.

David

 

[Lars]

Be wary of any manufacturer stating 100%. They usually mean 99.x% on a linear basis and round up to the next whole integer. I'd be curious to see if anyone is willing to run a test on a camera with a "100% viewfinder" with some grid paper and share the results. I believe this is another case of Leica erring on the side of conservatism and it has nothing to do with physical size of the housing.

David

David,
This sounds like classical FUD. Do you know of any manufacturer making false claims or not?

 

[georgl]

So it's not really 96% coverage in the center but 96% to fit within the actual framing due to sensor adjustment!?
How precise are SLR-viewfinders anyway, the mirror isn't on the film plane therefore coverage changes with focal length?

I'm sure they have a good reason, cutting off 0,2mm on each side of the prism doesn't make sensor from an economic perspective anyway.
That would at least fit into typical German behaviour: taking the unconvenient, difficult and expensive way, point out to weaknesses which others don't and hope that consumers will appreciate it... Consumers stupidity than killed most of these technologies/companies...

 

[dfarkas]

David,
This sounds like classical FUD. Do you know of any manufacturer making false claims or not?

Not false claims, per se. Just a different way of measuring and/or stating on specs sheets. You wont get FUD from me (or Leica). :angel:

This is from Canon's website for the 1Ds mkIII:

Coverage -- Approx. 100% horizontally and vertically​

Approx. means they are rounding up. Horizontally and vertically means they are measuring linearly. In actual area coverage, you'd find that this would work out to about 96-97%, same as the S2.

The 5D mkII (again from Canon USA):

Coverage -- Vertical/Horizontal approx. 98%​

98% linear is about 94-95% by area.


Here is Nikon's take on the D3x:

Viewfinder Frame Coverage -- Approx. 100%​

Again, "Approx." is used. No definition of measurement is given.


So, you can see that there really aren't any "true" 100% viewfinders on SLRs. This is not a jab against C/N, just a clarification of why Leica's spec for the S2 seems off. They (Leica) could have also used the same method of measurement and said that the S2 viewfinder is "approx. 100% coverage" as well.

David

 

[Guy Mancuso]

But David are you not assuming at what they mean by approximate. Obviously it's not flatly stated in those quotes.

 

[dfarkas]

But David are you not assuming at what they mean by approximate. Obviously it's not flatly stated in those quotes.

I think I'll be getting the exact numbers from N or C about the same time as I get measured MTFs from Phase/Mamiya....

My point was that while most manufacturers tend to err on the side of "rounding up", Leica tends to be a bit more conservative/precise. I think in practical terms, you will see very little, if any difference, between a "true" 96% vs. 98% viewfinder.

David

 

[Bob]

Actually, My measured MTFs on the mamiya 150 2.8D well exceed the manufacturers claims, it looks better than any Leica lens I have measured so far.
So we will see.
-bob

 

[LJL]

David,
Not to put any more of a fine point on this....."approximately 100%" can mean just about anything, maybe even 102% or so. Assuming that it should get knocked back to the ranges you suggest is just that, an assumption. Canon and Nikon do not exaggerate wildly on those sorts of specs, as they are pretty easy to measure, and folks would be all over them if they were very far off. I know that I can rely on my 1-seires Canons delivering pretty darn close to what I see in the viewfinder. In fact, I have to be careful not to cut too closely at times. So, if Leica is saying 96% coverage, I am going to tend to believe that number from them, and not think it is a "conservative 100%"....it is probably somewhere between 95.2% and 96.8%, but still notably less than 100%. Why try to make it something it may not be? If Leica came up short on that one, that is just something folks will have to adjust to and live with, much like the frame lines on the M8, which are not even close to what the actual frame is for most lenses. Folks are not happy, but they live with it.

LJ

 

[georgl]

@Bob
Where have you got MTF-data form Mamiya?

 

[dougpeterson]

I don't know why this matters to me, but...

By claiming "approx." 100% on their top camera and 97% on their 2nd tier body they must be rounding to the nearest whole number (or else they are flat out misleading).

So assuming the worst case, that they are rounding both linear directions the minimum coverage would be 99.5% (H) by 99.5% (V). So at worst the 1Ds3 and D3x are 99% viewfinder area coverage. At absolute worst.

The P65+ on a Phase One body is also spec'd at "approx" 100% viewfinder coverage. I don't know if that means 99.5% or 99.9% but regardless anyone who has shot a 1DsIII or D3X or a P65+ on a Phase One body knows that what you see is what you get.

If you don't mind cropping (especially since so many will have to crop the S2 to a 4:3 frame anyway) then this point is moot. If you're an OCD "I never crop" style shooting (different strokes for different folks) then any difference between your composition and the final frame will bug the crap out of you.

Doug Peterson (e-mail Me)
__________________
Head of Technical Services, Capture Integration
Phase One, Leaf, Canon, Apple, Profoto, Eizo & More
National: 877.217.9870 | Cell: 740.707.2183
Newsletter: Read Latest or Sign Up
RSS Feed: Subscribe

 

[dougpeterson]

Actually, My measured MTFs on the mamiya 150 2.8D well exceed the manufacturers claims, it looks better than any Leica lens I have measured so far.
So we will see.
-bob

That 150mm f/2.8 is an absolutely stunning lens.

Doug Peterson (e-mail Me)
__________________
Head of Technical Services, Capture Integration
Phase One, Leaf, Canon, Apple, Profoto, Eizo & More
National: 877.217.9870 | Cell: 740.707.2183
Newsletter: Read Latest or Sign Up
RSS Feed: Subscribe

 

[Guy Mancuso]

NO question about it and frankly MTF charts are worthless to me. I care how it looks and this lens in particular is killer and I don't say that very often about any brand or lens

 

[carstenw]

By claiming "approx." 100% on their top camera and 97% on their 2nd tier body they must be rounding to the nearest whole number (or else they are flat out misleading).

So assuming the worst case, that they are rounding both linear directions the minimum coverage would be 99.5% (H) by 99.5% (V). So at worst the 1Ds3 and D3x are 99% viewfinder area coverage. At absolute worst.

If your assumption is correct, yes, but Canon might also be rounding 99 to 100, so who knows. Since you are a Canon dealer, why don't you try the graph paper test?

 

[tashley]

<SNIP>

If you don't mind cropping (especially since so many will have to crop the S2 to a 4:3 frame anyway) then this point is moot.
<END SNIP>

Hmm, interesting - because of course having spent most of my life shooting 35mm I prefer the 2:3 ratio so I end up cropping a lot of my Phase shots to that. Which means that in effect I will be getting more, not fewer, MP if I buy an S2... 37.5 versus 34.7 to be more or less exact!
:grin::grin:

tim

 

[ptomsu]

NO question about it and frankly MTF charts are worthless to me. I care how it looks and this lens in particular is killer and I don't say that very often about any brand or lens

Guy,

I am really behind you here, for me the same thing - MTF charts are worthless to me, what really counts is the look and feel of the image.

But just for curiosity - have you ever tried to say this in the LUF? I did and was almost killed :ROTFL:

 

[Guy Mancuso]

LOL people need to understand their are some very very good lenses besides Leica. I know hard to say on LUF without arrows being tossed at you but people need to wake up too. Zeiss, Rollie , Hassy and Mamiya make some really great glass

 

[Jan Brittenson]

The reason for lack of 100% is that the sensor requires shimming and a little bit of movement in the calibration process druing assembly. In order for the viewfinder image to correspond to what the sensor records, there needs to be this 0.2mm spill to cover adjustmnets in assembly.

This sounds reasonable. Another problem is that a 100% screen requires a > 100% mirror; the image will vignette the closer it gets to the size of the mirror. Mirror size is a much bigger design and engineering obstacle than prism size - it's part of a complex mechanical assembly, all of which has to be sized up along with the mirror.

 

[Bob]

@Bob
Where have you got MTF-data form Mamiya?

I measured it myself using Imatest.
It was a system mtf since it involved camera and sensor, but it does provide an interesting comparison since the frequencies reported on vendor provided mtf graphs are well within the resolution capabilities of of most sensors.
In comparison of the 150 2.8D Mamiya to the leica 90 apo cron there was very little difference, with the Mamiya winning pretty much across the board.
Also, many of the most prized older Leica glass with the Leica "glow" actually are technically worse, it it is the result of the aberrations that give the lenses their special character.
MTF is an objective measurement, however, vendors use different techniques for their measurement, particularly different frequencies and various apertures often with inadequate documentation. I have also measured significant shifts in mtfs when measured at different subject distances. With all lens makers, Leica included, there are significant sample variations, so your mileage may vary.
So, the thing to do is to see if the lens draws the way you like, is sharp for the purpose and so forth.
There is no mystery or magic in "Leica glass" although some of their optics are amongst the best made. Their "signature look" is not really the best or most accurate, they have, like Bose speakers a particular collection of characteristics that although might not be as "accurate" as Schneider Digitars is generally pleasing.
-bob