Not sure i will fix it.

[Jamie Roberts]

Advertisement (gone after registration)

Well, depending on what Leica says, of course, I'm thinking I'm 90% in the "I will fix it" boat. It's the streaking /blobs that's the only problem for me right now.

 

I honestly don't believe that a higher frequency IR cut internally, coupled with a software tweak or two, will degrade the admittedly mind-blowing (ok, so it blows my mind, anyway) quality of the M8.

 

Leica will not take a step backwards, I don't think.

 

They've capped my favourite digital camera... the DMR. The files are uncannily good. I can't quite believe I'm looking at a 10mp camera.

 

So I don't think they will change the fundmental accuity of the sensor.. I really don't. But it's all just a guess.

 

Hopefully, someone like Guy will have a new one and an old one and let us know. Or my dealer will let me try a new one before mine gets fixed!

[Guest guy_mancuso]

Let's think of it in terms of film first. The colours of the spectrum that film is sensitive to are diffracted by lenses differently. The cover of Pink Floyd's Dark Side of the Moon provides an illustration most of us can bring to mind. Or think of how light is broken up by raindrops to produce a rainbow. This breakup causes unsharpness as light reflecting off a point source goes through the lens and hits the film at slightly different places. Good lens design can reduce this effect.

 

Now let's turn to the situation with M series cameras and the M8 in particular. There's no mirror box and the M lenses sit closer to the film/sensor plane than with SLRs. The range of angles that light hits this plane, from the center of the image to the edges, varies more than with SLRs. The light coming out of the back of an M lens has to fan out more sharply to reach the edges of film or sensor than with an R lens and body. Those well designed lenses (perhaps apochromatically corrected to keep the different colours together by the time they hit the film/sensor plane) have done their job, only to have the light diffracted again when it goes through the glass in front of the sensor! Geez. With an M8 the extent of diffraction varies across the sensor, from center to edges, moreso than with digital SLRs. And, importantly, if the glass is thicker there is more depth for diffraction to spread before it hits the little pixel buckets in the CCD.

 

So keeping the glass thin helps retain imaging sharpness where the lenses sit close to the sensor. Some chromatic aberration still occurs - which may be compensated for by the body knowing which lens is mounted - hence the lens coding (which also helps account for vignetting).

 

Leica have given us a double helping of sharpness retention. The cover over the pixels is thin and, there's no anti-aliasing filter.

 

Turning to skimmel's questions -

 

- Given that the sensor is sensitive to IR which will split from other colours as it passes through the lens elements and the glass cover of the sensor, if the IR is filtered out before passing through the lens there will be an improvement in image sharpness.

 

- Putting an additional or thicker glass cover over the sensor will increase the effects of chromatic aberration, causing light to spill into adjacent pixel buckets and reducing sharpness.

 

It is conceivable that a different sensor cover with greater IR filtering properties could be used though it would introduce other light transmission problems. It may be that very clever algorithms could compensate for this, however the variety needed for different lenses, the variation within lenses of one part number (not to mention third party lenses)and the effect of differing light conditions would surely make this a difficult challenge. Chopping out the IR before it enters the system is most likely the best solution.

 

 

Rick very well said and I have the same thoughts ,you folks just can get it on paper better than me.

[reddawn]

Guy

 

As you know, I'm a committed Leica fan and DMR owner, I'd also like to own a digital M sometime in the future.

 

I've not taken part in the debate so far but I'd like to add this and hope you and your correspondents understand where I'm coming from.

 

1) How much of the added sharpness in the M8 over the DMR is really sensor and how much is down to to the excellent M lenses, which appear to have better sharpness than many R equivalents, anyway.

 

2) It's pretty hard to turn a fault into an advantage, no matter how loyal I want to be - having enhanced IR trasmission is great for a few specialists and astronomers but pretty unhelpful (note we British are masters of understatement) to the rest of the community - unless you invest time and money in work arounds and filters, you have a problem of unreliable colour.

 

3) This is likely to cause the greatest problems to those who really want to use the M8 in its traditional roll in journalism, weddings, reportage of diplomatic and other public events - unless they happen to have paid out for the right filter, they are going to get an unpredicatable result.

 

4) Someone said earlier that the eye is a poor measuring instrument but a good comparator - I genuinely beleive that someone viewing a photograph will probably not notice a few microns lost in sharpness (in reality my subjective feeling is the sharpenss is likely to be excellent by other makers standards, even if a little was lost in coatings / internal filtration / firmware) but will notice his son's tuxedo has changed colour.

 

5) Personally, as someone who would like to go on the M buyers list sooner rather than later, I would rather all the cameras were withdrawn and leica get it right - if the new M8's then came back with good colour and a slight loss in image quality, then so be it - they got quite a lot in the image quality bag to loose if they had to and still be outstanding.

 

I think the reality is the camera buying world can live with a tiny loss in image quality, when it's so good anyway - it can't live with a camera that you have to use with filters and colour profiling as standard to get a reliable result.

 

Robert, I just want to say, I agree completely with you on every count.

[Guest malland]

... if the new M8's then came back with good colour and a slight loss in image quality, then so be it - they got quite a lot in the image quality bag to loose if they had to and still be outstanding.

 

I think the reality is the camera buying world can live with a tiny loss in image quality, when it's so good anyway - it can't live with a camera that you have to use with filters and colour profiling as standard to get a reliable result.

 

Exactly: in my view the medium format-like image quality is not really necessary, particulalrly as I like the "35mm aesthetic", which I wrote about in the following thread:

http://www.leica-camera-user.com/digital-forum/9735-m8-pogo-paradigm.html

 

—Mitch/Bangkok

[osera]

Not a real life possibility, but just a thought:

 

What if in-camera IR filtering were offered as an a-la-carte option-

choice A - "standard" IR filtering level, little or no IR contamination

choice B - "thin" in-camera IR filtering, on-lens filter required for some applications, "better" performance/sharpness.

 

How much of a performance difference would get me to choose B?

[reddawn]

Exactly: in my view the medium format-like image quality is not really necessary, particulalrly as I like the "35mm aesthetic", which I wrote about in the following thread:

http://www.leica-camera-user.com/digital-forum/9735-m8-pogo-paradigm.html

 

—Mitch/Bangkok

 

I think we are all in the minority on this.

 

I always felt the low noise images from Canon files look more medium format like than the gritty look at high ISOs of the kodak sensor. And looking at Sean's samples in his reviews, I feel the 5D has a very slight edge in resolution.

 

But of course, we "love" the Ms more. But I would also accept the tradeoff of a little drop in sharpness or resolution in return for faithful color reproduction ala the 5D....

[Guest guy_mancuso]

Advertisement (gone after registration)

Straight out of the box in C1 with WB 50 lux , fill flash and IR filter

 

Top one normal WB.

 

Bottom lowered the kelvin a little

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[gogopix]

Here is where, although I think I would use the profiles for a 'fix' of WA, you actually get a positive effect from the IR filters. The slight shift to cooler colors as well as the elimination of the 'extra red' (for that is what the IRproduces, more red signal, there is no bayer IR) takes some of the warmer 'smearing' of the IR. The bayer matrix is interpolating and putting some red ALLOVER where it need not be.

 

In these, it would be interesting to A/B vs the non IR; I would bet some of the 3d and sharp edge look would be lost (look at the clean, lifelike edges of the black cloth and the faces against the grass.

 

nice example.!

 

more and more nervous about any excess IR filtering on the sensor that degrades

 

the board fix for the frequency I always thought was the issue (didnt give sensors recovertime from 'overflow'

 

For me, board fix, IR filter 35mm up, and profilers for 28mm and smaller seems the way to go.

 

Victor

[jonoslack]

Straight out of the box in C1 with WB 50 lux , fill flash and IR filter

 

Top one normal WB.

 

Bottom lowered the kelvin a little

 

Hi Guy

You aren't supposed to HIT them with the IR filter - how much is her mom suing you for the loss of that front tooth?

 

 

Lovely shot - I'm one who would rather deal with the IR than lose that splendid resolution.

[Guest guy_mancuso]

I need to play with it a little more but with the filter seems to be the key , those black shirts no doubt would have went magenta

[skimmel]

Straight out of the box in C1 with WB 50 lux , fill flash and IR filter

 

Top one normal WB.

 

Bottom lowered the kelvin a little

 

Yikes, you are good! Really nice picture!

 

2 questions:

 

1) What flash did you use to fill and was this, I assume, manual exposure setting.

 

2) On my (calibrated) monitor, the shirts don't look totally dark black. Is that a correct representation?

[Guest guy_mancuso]

Stephen I used the Metz 54 with the 3502 shoe and a stofen diffuser. Yes there black but there not like a wool black that absorbs light. The left shirt is polyester and the right cotton and the middle cotton. Plus i am not exactly dead on in color , still playing with that

[skimmel]

Stephen I used the Metz 54 with the 3502 shoe and a stofen diffuser.

 

Thanks Guy. Excuse my ignorance, but is the 3502 shoe the one for the M8 (I tried to find at B&H but couldn't on a quick search)?

 

Also, where you using manual exposure? (Curious as to how the flash does in AE mode.)

[Guest guy_mancuso]

Yes sorry i use the 3502 on both the DMR and M8 but still no TTl on the M8 and someone report that version 5 does not help with TTL also. But Ishot these on manual full power at about 5.6. In A modeit woks pretty good but Fill flash may not work as well when on A mode since the flash sees plenty of light

[skimmel]

Yes sorry i use the 3502 on both the DMR and M8 but still no TTl on the M8 and someone report that version 5 does not help with TTL also. But Ishot these on manual full power at about 5.6. In A modeit woks pretty good but Fill flash may not work as well when on A mode since the flash sees plenty of light

 

Thanks Guy.