Candles reflection (M8)

[arthury]

Advertisement (gone after registration)

May be it's so faint, it is not easily recognizable. I'll re-check it later tonight.

[Advertisement]

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

[Jamie Roberts]

Birthday parties for kids sometimes consists of various cloth materials that may be black and synthethic. It's unpredictable until you're there and so not using the IR filter seems like a moot point to me.

 

I still cannot accept the fact you guys are seeing this phenomenon in other DSLRs produced by other vendors.

 

{snipped}

 

Hey Arthur--

 

Rex is right about this being systematic and unpredictable. He's also right about both the lack of IR light and the relativity of colour balance for candlelight.

 

And it's not faint when it shows up.

 

And just to show you that it's neither wides nor the M8, I dug up an earlier birthday party with my mom... three years ago, so she's only 81 here. Still it's not a pic I'd show on aesthetic grounds, but it's the clearest (of about 10) to show the problem.

 

This the Canon 1dm2 with the wonderful 85 1.2L--shot at 1.2, with a hideously costly BW UV MC pro filter.

 

See the red splotches that match the candle flames!? They go entirely away by f4; they also go away wide open without the filter.

 

This is the only one I could dig up quickly; there are others. Including film, and including Nikon--all taken with fast primes, BTW.

 

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!

 

PS--I left the EXIF intact when I re-sized for you doubters out there

[arthury]

Thanks, Jamie. OK, I believe you guys now before you change my name to Thomas.

 

Here are the differences I saw so far between the Canon's version and the M8's:

• the ones produced on the Canon is situated laterally to the candles whilst the ones from the M8 were all hovering above
  
• the color of the reflections from the Canon's seems to match the candles whilst the ones from the M8 are closer to the blue end of the spectrum

I'll have to try harder with my Nikkor 85mm/1.4 tonight with the same B+W filter.

[Jamie Roberts]

Hey Arthur--

 

Agreed on the differences. I have no idea why they're positionally different, or even what affects that (other than "it's the angle of reflection".

 

But even with the Canon I used to get the artifacts in different locations before I just ditched the filters entirely. I have one shot of my mother where I thought they disappeared, but they were actually in her hair, no less (talk about magenta!).

 

I also think the colour is probably due to the particular filter coating / lens & sensor combination, but we'll see.

 

The only reproducible artifact I've seen on the fixed "v2" M8s so far is the horizontal banding when you've got a really overexposed light source half way up the frame on the very edge of the frame; someone said that's just a sensor readout issue and can be fixed in firmware, too.

[arthury]

OK, after repeated tries, I have finally able to perturb the D2X to show forth this problem but it is very mild compared to the M8. It is reproducible on the 50mm/1.4 Nikkor. My experiments tell me that it is not easy to get this effect with a longer lens: 90mm APO and 85mm/1.4 Nikkor --- I was not successful in doing it.

 

#1: 50mm/1.4 ; @1.4 --- possible to reproduce but hardly noticeable

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!

 

 

 

#2: 85mm/1.4; @1.4 --- not possible to reproduce

[Jamie Roberts]

Arthur--don't try to reproduce this with your Nikon--just take the filters off!

 

In my experience, the variables you're trying to chase down here are too hard to reproduce.

 

When this happened to my Canon, there was a certain distance, angle, sensor amplification (speed), and even ambient light as well, and I think that all plays a part here.

 

So just because your tests show "negligible" I'm pretty sure based on contrasting experience that given just the right conditions, it won't be negligible at all. You may never see them--or you may. Hence my reasoning in always removing filters in that kind of lighting.

[Micha67]

Advertisement (gone after registration)

Hey guys!

 

Did you ever think about the physics of a multilayer antireflective coating? Maybe you should do. The antireflective effect is best when the coating has a certain thickness that scales with the wavelength of the (otherwise) reflected light. Any antireflective coating has to be optimised for a certain range of wavelengths and is, thus, not too far extending into the IR range. The candle light, however, contains a huge moiety of IR and is, thus, more prone to reflections at the UV or protection filter.

 

The good message: if you use the IR-cutoff filter, the amount of IR light to be reflected will be vastly reduced. For this reason, you should not worry too much about refections (seen with UV or protection filters) to occur with the same intensity using the IR-cutoff.

[cocker]

This is nothing to do with filters. I noticed this effect a few weeks ago and the two images show its a function of the lens. The first one is taken with a 35 lux at f4 the second one is with a TriElmar at 35 f4. The 35 lux shows the "reflexions" the 35 TriElmar doesn't. Neither lens was wearing a filter. There is (almost the same amount of wine in the bottle in the second shot as in the first so it's a valid test.!

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!

[arthury]

So, you are saying that newer lenses are more equipped to deal with this issue?

[Jamie Roberts]

Keith, that's interesting, and I've never seen it before with my 35 1.4 lux.

 

Can you please tell us if this is a "fixed" or original version M8?

[rvaubel]

This is nothing to do with filters. I noticed this effect a few weeks ago and the two images show its a function of the lens. The first one is taken with a 35 lux at f4 the second one is with a TriElmar at 35 f4. The 35 lux shows the "reflexions" the 35 TriElmar doesn't. Neither lens was wearing a filter. There is (almost the same amount of wine in the bottle in the second shot as in the first so it's a valid test.!

 

Well your example doesn't have anything to do with filters. But 99% of the time its a filter issue. I must admit though that the causes of the phenomena are more complex than my test of last year would indicate. It might be interesting to do some more testing. But from a practical point of view, I'll just take off the flippin filters when a very bright object is included almost on axis.

 

Rex

[Micha67]

This is nothing to do with filters. I noticed this effect a few weeks ago and the two images show its a function of the lens. The first one is taken with a 35 lux at f4 the second one is with a TriElmar at 35 f4. The 35 lux shows the "reflexions" the 35 TriElmar doesn't. Neither lens was wearing a filter. There is (almost the same amount of wine in the bottle in the second shot as in the first so it's a valid test.!

In this case, the ghost image is scaled down compared to the tue light source. Thus, reflection has occurred not at a planar surface in front of the objective, but somewhere within the lens. The reason for internal reflections is the same than for filter reflections: the multilayer antireflective coating is not capable of suppressing reflection of IR light. Coatings have beeen optimized for those wavelengths that film materials are commonly sensitive to. Now we have an IR-sensitive sensor and must cope with the problem of poorly suppressed internal and external reflections.

 

BTW: the reason why one lens shows these internal reflections and others do not, highly depends on the shape of the reflective surfaces and on their position within the optical path. Possibly most surfaces will reflect some of the red/IR light. But in most cases, these reflected rays won´t be focussed on the sensor again and, thus, impose only as stray light. In unlucky cases, re-focussing of the reflected image may result in these poorly suppressed ghost images. It will be interesting to see, whether the IR-cutoff filter can solve this problem.

[cocker]

Keith, that's interesting, and I've never seen it before with my 35 1.4 lux.

 

Can you please tell us if this is a "fixed" or original version M8?

 

Jamie,

 

Its a fixed version, i.e bought new (Jan 2007) with the 1.9 firmware not original and "fixed" by solms