ReidReviews has started an analysis of IR-cut filters on the M8

[wparsonsgisnet]

Advertisement (gone after registration)

Sean has started a (rolling) review of the IR-cut filters on the M8. He's analyzing the Leica and B-W 486 versions first.

 

The results are disturbing to me as the corner effects change with aperture and the filters seems not to work linearly. That is, sometimes one type appears better to me and sometimes the other.

 

When he's done, we'll have some very authoritative data to evaluate.

 

I suspect I will hang on to my preference for fixing this in post-production (that, of course, eliminates the need to code any lenses).

 

If the post-processing can be automated, then we would not have to

1. Do without the lenses while they're away being coded,

2. Wonder why the corection has a different effect at different apertures, and

3. Spend money for coding,

 

I am not currently interested in an "alien" lens, so Leica's decision not to open up the lens awareness part of the camera with a menu entry doesn't bother me.

 

I can see, however, that it is inconvenient to do this in two ways:

1. If we can't tell the camera what lens -- of any manufacture -- is on the camera and have the camera do its correction dance, then both users of "alien" lenses and those who use older, uncodable Leica lenses are inconvienced, but

2. More to the point, the whole thing is a drag. I would rarther work a little to keep track of the lens sequence during a shoot (or work it out if I don't keep track) and automate post-processing correction for the color and vignetting aberations.

 

The reason for my last remark is that I don't think Leica's going to get it right. Changing the aperture of the lens makes it perform differently with regard to vignetting and probably with regard to color shift. From the point of view of precision in this matter, I would rather run some tests myself, at different apertures for each lens (this is really only onerous for Guy, who has 10,000 lenses), and build some automated processes that take care of the problem.

 

Most of the time I only shoot at 2 apertures, anyway: open and f8. The task of calibrating is not so difficult, after all.

 

I also expect that the problem of cyan-drift and vignetting is going to change with subsequent releases of firmware. As Monty Python would say: "My head hurts."

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[Guest guy_mancuso]

We have to remember leica based all there calculations based on the specs. of the Leica filters. Not saying B+W or Heli's would not work in most cases just that things maybe slightly different or change. Seans report will help understand some of that change with the B+W. We pretty much figured under 21mm you will want the Leica filters.

[Guest tummydoc]

The results are disturbing to me as the corner effects change with aperture and the filters seems not to work linearly.

 

The effects change with the quality of light as well, as I noted the first day with 1.102. The in-camera correction is fixed at some kind of average value. Therefore regardless of whether you use Leica, B+W, 486 or any other IR-cut filter the cyan correction will have times when it is either less effective than needed, or over-corrects and get a bit of red in the corners. For my personal preference I will stick with Heliopan filters, code or no code, because I have found that they have much less propensity to ghost and flare than the 486 (and extrapolating from Erwin's findings, the Leica filters too). I can better deal with slight re-correction of cyan corners, or overall colour balance, in post-processing than try to clone out ghost images and compensate for overall contrast-killing veiling flare (if that is even possible).

[sean_reid]

Hi Bill and all,

 

One of the consequences of the split forums is that I just now saw this thread. Thank you very much Bill for mentioning the article. I'm still wondering what an "alien lens" is...<G> It must be slang for "excellent lens that is also inexpensive"...<G><G> Just think of all the alien electronics in the M8 itself....

 

The external metering cell allows the M8 to estimate aperture and I believe that this information factors into the cyan drift correction calculations. That said, tummydoc (gastroenterologist?) is right that there will necessarily be some variation in the effectiveness of the correction, depending on various factors (including the type of subject lighting). Its not so much that Leica "hasn't gotten it right" (in fact it seems pretty well dialed-in so far) but rather that they're shooting at a moving target with what must be a mostly-fixed gun.

 

Perfect correction for all the coded lenses, at all apertures and in all lighting conditions is likely impossible. That said, I think that owners of Leica coded lenses who use them with Leica filters will rarely, if ever, have problems with cyan drift in "real world" photography. My test board pictures are done for a purpose but they're far more ruthless (with respect to revealing color drifts) than is normal photography.

 

As Guy said (and he and I have been discussing this), the Leica filters seem to be the best way to go for 21 mm and wider lenses. The 486 filters are also working quite well with lenses longer than that.

 

Correcting in post for cyan drift may work for some but it would be an awful solution for me. I make a lot of pictures during a shoot, switch lenses, change apertures, etc. As Guy and I agreed some time ago, correcting for this manually in post would be a nightmare for us. YMMV So far, I'm quite happy with the "real world" results I'm seeing from the filters and new firmware together.

 

Cheers,

 

Sean

[Guest guy_mancuso]

I agree correcting this stuff after the fact is just going to be too difficult for heavy shooters with too many images to correct manually and frankly IF this was going to be the case than the m8 would hit the sale block. Luckily that is not the case and we can in real world shooting work as normal but i still would like to see C1, lightroom and other programs have the facility to work our images with corrections for the cyan drift in post production . this would give EVERYONE no matter what a chance to correct this after the fact . so i think it is important that leica has it dialed in for in camera work and look forward to third party vendors to be able to do this in post also. Than if we do or don't get a menu option we still have a way to get stuff corrected . so this would help us on all fronts and for everyone. Sean and i talk a lot about this stuff and we both are on the same page as too what maybe great for us maybe not work for others so all of our comments and suggestion will and always have been for the overall M8 shooters and not just our needs. we both want a working camera that is for everyone and there needs too.