Erwin Puts has Posted Part3 of his M8 Report!

[Asher Kelman]

Ahem in the Leica camera user forum let us know Erwins Puts has posted his latest report on the M8.

It is well written and worth reading! Erwin has been a longtime user and tester of photgraphic equipment and Leica in particular. He is therefore experienced. Still he, like a lot of photographers, can have strong opinions. However, in this aspect, he write in his recent M8 review, recognizing how users can differ in reporting the same equipment. I like to follow the reports of Sean Reid, Guy Mancuso, Jimmie Roberts and others. Here, Erwin's report has more congruence with the positive reviews the others. The camera system comes out rather well!

Leica M8 (November, 26, 2006)
Part 3: field test and monochrome imagery

I'm cross posting my critique here,

This part of Irwin Puts' review of the M8 is perhaps the best of his 3 sections. Here his work is done with RAW files and he leaves behind issues of artifacts to examine the capability of the M8 system to deliver the "Leica Look" in B&W, the hall mark of Leica from the beginning.

The description of the dynamic range of the Leica M8 and the 3D shading is, itself, worth reading.

I do have problems however with the rest of the report.

I have issues with how the 5D was compared with the M8 in pictures of a model to demonstate the 3D effect and also to show that the Leica M8 delivered sharper eyes.

When one takes pictures one cannot judge a camera by totally excluding what one would do in real photography. What's the point?

Also in photography, the result is not mathematically added up factors. The probability of getting your vision written to a file requires multiplication of the factors, so for the Canon 5D

THE EYE PICTURES

The general verdict from the previous article, that the M8 is the equal of the best performing digital cameras in the filed is substantiated by these comparison pictures. To eliminate as many variables as possible, I used studio light, a tripod and the Apo-Summicron 2/75 at medium apertures. The Canon had the very good 24-105 at focal length 75 and also at medium apertures. In this comparison the Canon cannot capitalize on its larger sensor area as I wanted to get the same pixel area. Results are almost indistinguishable, but on critical inspection of the Raw files on the monitor, the Leica image showed more three dimensionality, especially in the reproduction of the eye-ball, which has more transparency and depth.

First The Leica M8 Image, cropped to show the eye.

Below the corresponding Canon 5D Image

Focus= .7 (look at the edges of the pupil and the reflection of the soft box)

Sharpening = .8 of the Leica image

This has already degraded your probability of rendering intent to 0.8 x 0.7 = 0.56 or 56% compared to the Leica system.

Add to that a better Lens for the Leica, obviously the Canon system has not been treated, as a professional would do in a portrait session.

To compare the Leica with an Apo-Summicron 2/75 to a 5D with a 24-105mm L lens is already perhaps an issue.

1. I'd rather use on the Canon the EF 50mm Macro or and 85mm 1.2 L.

2. I'd manually focus the Canon lens as manual focus when accurate is better than the autofocus within the errors inherent.

3. The DOF must be the same. Ideally several apertures would be used to get the DOF needed by the portrait one has in mind.

4. Resolution is going to vary in these two systems, of course. The Leica has no doubt more lp/mm to deliver but the Canon has more pixels.

5. Why handicap the Canon by not using the whole sensor. One would not do this in practice. The job is to realize one's vision to get the very best print, not to pixel peep per se.

6. Tests can be done for theoretical purposes but we need tests useful to the photographers intent to make a picture. Therefore the system must be optimized for each camera to show that tool at it's best under practical conditions. The former esoteric tests are interesting and give insight; the latter practical tests are, however, essential.

THE 3D ILLUSTRATION PICTURES

The lighting is different! Also the position of the head is so different, although most likely the conclusion would be about the same! I don't ever see such a "Leica effect" with my Canon systems.

Still, none of this takes away from Erwin's approach to try looking at the M8 in a reasonable way while others fight to solve the artifact issues. That, in itself, is welcome.

Asher

 

[Bart_van_der_Wolf]

Sharpening was wrong on both 'eye' samples

Too bad, since it detracts from the inherent quality.

The M8 shot was oversharpened, I cannot undo the damage on that one, but the 5D shot can be improved reasonably well, for a JPEG source image taken with a 'lesser' lens:

5D sharpened a bit

Bart

 

[Asher Kelman]

Great! Thanks Bart for doing that. I worked with it but I can't say whether mine looks similar as it's on my other computer! i'll check it later.

I think it needs deconvolution as well.

Asher

 

[Will_Perlis]

Yes, IMO the M8 image was sharpened by a high-speed grinder.

Here's the Canon image sharpened by Focus Magic with a setting of 1.

 

[Asher Kelman]

Hi Will,

Do you use Focus Magic as your main sharpener?

That's an improvement. Anyone try deconvolving?

Asher

 

[Will_Perlis]

Asher,

Yes, I like and use FM. I think it's just a little less heavy-handed than USM and "Smart Sharpen" in CS2 and at least comparable to the other after-market plug-ins I've tried.

(Tho' I tend not to have the patience to learn each to the point of getting optimal results for every program I try. I'm thinking that way lies madness. If I can't see a clear difference almost immediately I'm not going to bother exploring in depth.)

 

[nicolas claris]

Here's my take on the sharpening:

 

[Herman Teeuwen]

And my take:

 

[Don Lashier]

The Leica still has better micro-contrast IMO. Also, where did Erwin's reflection go in the Canon version?

- DL

 

[Will_Perlis]

How's this, Don? FM w/ 2pixel shifting. In PS I can see skin texture all the way up to her eyebrow. In Firefox some of that washes out.

 

[John Sheehy]

The Leica still has better micro-contrast IMO.

At my desk, that's called "aliasing".

I really don't understand the public's love of aliasing. I never understood the appeal of Sigma SD images - they look extremely distorted, spatially, like someone took a perfectly good higher-res image and downsized it to 50% with "nearest neighbor". The result is snap-to-grid mosaicing with false, misplaced edges.

There really should never, ever be 100% contrast at the neighboring pixel level, IMO. If you want higher resolution, have more pixels - fortunately, unlike the Sigmas, the Leica has lots of pixels, so the lack of an AA filter is not as disturbing. However, since Leica lenses are so sharp, you may still see the aliasing if you crop or print very large.

No AA-less cameras for me, unless they have so many pixels that aliasing is not possible with the sharpest optics available.

 

[Petter Stahre]

I couldn't resist to have a go at the 5D image. For 1:1 viewing it's oversharpened, but I wanted to bring it up to a somewhat compareable level with the M8 image. (But for print in 300 dpi I think the M8-image and this one works very well.)

Sharpened via Smart sharpen 500/0,3/Gaussian (and 100%/40%/1 in tab 2 & 3) and then Smart sharpen again but approx 200/0,3/G (tabs untouched) - I think, this was two days ago.

While this has increased the JPEG artifacts (and I normally never sharpen twice) I like the details of the fine hair in this one.

I also increased contrast a little (S-curve in Curves).

// Petter

 

[Asher Kelman]

Petter and everyone else,

What does this mean. Here we have the 5D image taken with, 70% of its pixels, essentially a prosumer L lens, the capable 24-105mm zoom presumable with merely autofocus!

This shows to me at least how Canon images MUST be carefully processed to get all the reserve quality they have.

Leica images have more out of the box. Now where is the great reflection of Erwin in the Leica eye as Don asked?

Asher