New M8: magenta highlights=magenta cast??

[Jamie Roberts]

Advertisement (gone after registration)

If you don't over expose them then, without an IR filter, the foliage has a brown muddy apperance due to the IR sensitivity and the very well known high IR reflectivity of foliage.{snipped}.

 

Um, Peter, not to put too fine a point on this, but Jono's shots do not suffer, believe me! Are the greens up to his particular snuff? Probably not, but I seriously doubt it's the IR response of the sensor and more to do with the absolutley craptastic job various parties did in making profiles for the sensor (though no doubt IR plays a role, as Sean has shown elsewhere).

 

But under normal lighting conditions, do you have any shots you can post to prove your point about horrible degraded tonal response with foliage and the M8 causing purple fringing?

 

What would you call the purple in this non-foliage water shot? IR sensitivity? I think not...it's the sensor & lens plus however the file was converted (in other words, the optical system) not able to handle the extreme exposure. Oddly enough, it looks precisely the same as the leaf shot posted here... and I say that's probably no coincidence.

 

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!

 

BTW--we can guess what camera and lens this 100% crop came from

[mike prevette]

Jamie that effect doesn't resemble the original post at all. The fringing looks totally different, and to my eye looks to be lens+filter+sensor related. The original shot did not look like that to me. Notice the fringing in your shot effects all the highlights, vs the original shot where the fringing only appears on the overexposed foliage, and not on the equally over exposed rocks on the ground. You would think the rocks, being lighter than the foliage would exhibit the same issues, but they don't because they are not reflecting as much IR.

 

Keep in mind that IR focuses at a different point than visible light. So If you shoot without a filter a slightly out of focus IR image is basically polluting your visible spectrum shot. And if the visible spectrum is maxing out the chip, that means the out of focus IR image is almost as strong in the surrounding pixels. Imagine laying two slides of the same image on top of each other, one slightly out of focus.

 

_mike

[Peter Branch]

I maintain that we will all be surprised with ourselves when the IR problem if finally solved.

 

Today, now that we have got rid of the streaks / bands / stripes, vertical and horizontal together with the green blobs which, for very good reason, dominated the initial thinking of many of us, - though some seemed to escape this horror and concentrated from the start on the IR related problems – we are all concerned about the image quality when the camera is used without IR filtering. I’m now persuaded that the image quality for those with IR filters is very good – but for the rest of us; we are kidding ourselves. You don’t need some exotic test equipment, just use an ordinary Digilux2 and take comparison shots. In many – but not all – conditions the Digilux images are superior in respect of colour and often exhibit better definition. Why is this?

 

The evidence suggests that if IR contamination is not a factor then the M8 is significantly better than the Digilux but if IR is involved then it is the other way round. I maintain that this is not a coincidence and that there are solid technical explanations as to why it is so.

 

IR is insidious, you don’t know its relative strength, unless you have specialist equipment and it is very prevalent. The issue is what proportion of the light entering the camera does the “Far Red / IR” component represent. There are many circumstances which tend to increase the ratio of IR to visible. One is anything to do with water. Water is a strong absorber of IR so any light which enters the water and then gets reflected / refracted out and enters the camera will have a very low IR component. But the surface reflected component will, relatively, have an increased IR component. Foliage is another well documented example. IR is also not focused at the same plane – remember the lenses with an IR focusing mark?

 

As I said we can often “correct” for the false colours and sometimes this is perfectly acceptable but often it adversely affects other parts of the image unless extreme effort is applied selectively in, say, PS CS2. In my experience foliage and flesh tones don’t go well together in an M8 image without IR filtration. I recall some of the early forum posts about the IR problem and people sharing images with and without filters. One particular one involved foliage, fabrics and flesh tones and there was much comment about a 05CC shift in the flesh tones but nobody commented about the massive colour shift on the plants in the background.

 

I strongly suspect that the undoubted problems with AWB are also related to the IR sensitivity.

 

We need the filters, urgently, before we can fairly and properly evaluate the camera under a wide range of normal usage.

[jaapv]

You don’t need some exotic test equipment, just use an ordinary Digilux2 and take comparison shots. In many – but not all – conditions the Digilux images are superior in respect of colour and often exhibit better definition. Why is this?

 

I shoot both camera's side by side, the M8 often unfiltered, and colour I can in some cases confirm. But definition???? Sorry, you lost me there. No way.

Your theoretical argument is valid and I'm surprised we did not see it mentioned more often before, but on the Digilux2 it is more than offset by the smaller sensor size, lower resolution and smaller pixel size. If you want to make the comparison, compare the M8 filtered with the M8 unfiltered and use the same lens. And yes, then the difference in definition can be quite dramatic, also in B&W.

[Jamie Roberts]

@ Mike--I respectfully disagree with your assessment of the M8 vs a Digilux. Any Digilux, and the difference you're seeing is more likely sharpening and relative pixel size than anything else. BTW--the fringing should look different; it's fringing from a Canon 85 1.8 on a 5D, which isn't supposed to fringe at all To me, the original post looks just like that, except the Canon is a better overall color lens, and there's no question of IR contamination on the 5D.

 

@ Peter--I'm not sure what to say. Yes, there is IR contamination on the M8. I've seen the results in foliage, especially using artificial light. But I've never seen fringing like the post that is related to IR. But your hand-wringing doesn't reduce the beauty or effectiveness of the M8 in the hands of someone like Jono...

 

@ Jaap--exactly right. Compare the M8 with the M8--filtered and unfiltered, but first lets put a lens that was made for colour photography on it

[Peter Branch]

@ Peter--I'm not sure what to say. Yes, there is IR contamination on the M8. I've seen the results in foliage, especially using artificial light. But I've never seen fringing like the post that is related to IR. But your hand-wringing doesn't reduce the beauty or effectiveness of the M8 in the hands of someone like Jono...

 

Seems like I can't win!

 

I don't for one moment dispute that under certain conditions the M8 produces outstanding images. Best I’ve ever seen from a “35mm” digital camera. I don't dispute that where images, as taken, show colour defects that these can often be corrected using software. Often this correction has no adverse consequences – but not always. I accept that I don't fully understand the precise mechanisms which result in the particular defects being complained about but I know that the root cause of many of them is probably IR sensitivity. I spent my working life at Kodak including some time in the Colour Group of the Research Laboratories. I've seen all these defects before in their analogue form on experimental films etc. I was trying to help by providing an explanation. If I have failed then I apologise.

 

My point is that until we all have the IR filters on our lenses, which everyone seems to accept - reluctantly - as being inevitable then we will have to learn to put up with a range of "defects". The camera is not yet as capable or as versatile as previous M cameras loaded with film or let it be said many digital cameras like the Digilux2. It shows great promise and is already a delight to use – but we are not by any means out of the woods yet.