Canadian Railroads > The Big Lens

I used to own a Canon x2 converter that would work on a couple of my long lenses.  The usual caveats about converters, or "cheaters", are that they cost you some sharpness and speed of the lens, and the autofocus will likely slow down as well.  On my 200/2.8 lens these factors could be dealt with by stopping way down (this particular image was shot at f11 effective aperture) and pre-focusing where you think the train is going to be (easier said than done).  On the 70-200/2.8 zoom it always seemed noticeably soft, no matter what I did.  What nobody seems to tell you about, is that these nasty little ghost images will show up in unpredictable places.

I thought about retouching this shot of 701 (what the gypsum turn was called before it became 511) at Elmsdale, NS on December 22, 2004, to get rid of those phantoms, but then decided it wasn't worth it since I had cut it so close at the top of the image anyway.  The loaded gypsum train is on a long tangent between Lantz and Elmsdale and is likely rolling along at 50 mph, having entered the main line from the mine a few miles back. Some film types might be better than others, I never really formed an opinion about that.  This one is a Fuji Provia F scan.

The ghosting with the converter got even worse when I went digital.  Apparently the camera's sensor reflects almost all the light that hits it back through the lens, whereas film absorbed more light making the reflections less objectionable.  I sold it  to a woman who was shooting the eagles at Sheffield Mills in the Annapolis Valley.  At least the birds don't come equipped with headlights!

GTD

---

In my experience the ghosting of headlights is mainly down to filters/skylight filters on the front of a lens.., remove those and the ghosting vanishes...

I've read about reflections off the sensor and had some thoughts about it. I did find a convincing article that, basically, says it's both reflections off the sensor and in the lens but we only see it because of the improved range of the newer sensors. Here's a link, this guy really gets into it, see the last paragraph at the bottom, which is included in a discussion about Sony Mirrorless cameras:

http://www.sansmirror.com/cameras/camera-database/sony-mirrorless-cameras-2/sony-mirrorless-issues.html

To add to the confusion, on TO we've seen a similar effect from the slide to digital process, perhaps something optical inside the box but also the digital noise reduction removing what it thought was noise of repetitive images, a type of ghosting.

I was looking around for a tele-converter and am finding the newer products cost almost as much as a telephoto lens of the same resulting size. Big money. My cost vs results equipment expense budget has to justify gas and hotels, so I would prefer the high road or expect to be really lucky. We know how that works out.

Great subjects and photos, a slice of of Canada we haven't been seeing on TO.



Edited 2 time(s). Last edit at 03/31/20 15:24 by TCnR.

Interesting read about the Sony "problems".  Almost all of them can be worked around if you're aware of them, as is the case for many other cameras and devices.

He did mention aperture being a factor which I hadn't considered before.  Although in the case of a teleconverter, the aperture is in front of the converter, so I wouldn't think it would change the reflection from sensor to rear element.  I was looking tonight at another slide (scan) taken with that same lens combination, and it's free of any ghosts.  But it's also at f8 instead of f11, a slightly broader angle to the train, and taken with Astia instead of Provia F.  Not exactly a scientific comparison.

I didn't mention it in the first post, but the image was taken without any extra filters on the front of the lens.  I've been a big believer in letting the lens hood protect the front element rather than putting a relatively cheap piece of glass there instead. At least I have been after being upsold on UV filters for some early lenses that produced ghosting like crazy.  I recall reading somewhere that the front element on the Canon 70-200/2.8L cost about $150 to replace if it was damaged, which is about what you'd have to pay for a high end 77mm UV filter anyway.

GTD

I wonder if the fast lenses are worse for this... I use a 70-300mm (non L) on my canon dslr so the max aperture is 4 but i'm usually up around 8 or so. I haven't had a single ghost light since i stopped using skylight filters on this lens many thousands of shots ago. 
I never saw the point of a f2.8 lens for me as I like to shoot in good light pretty much all the time. Maybe those faster lenses have more of a ghosting issue?

exhaustED Wrote:
-------------------------------------------------------

> I never saw the point of a f2.8 lens for me as I
> like to shoot in good light pretty much all the
> time. Maybe those faster lenses have more of a
> ghosting issue?

The article linked to above seemed to imply that shooting at a larger aperture made the ghosting less of a problem, but I've never tried that under controlled conditions.  You're right about f2.8 not being that useful for railfan shooting.  I generally shoot at f8 or f11 in order to get some depth of field and to cover some sometimes questionable focus, since you don't get a second take for most of these.  The larger aperture gives you a brighter image in the viewfinder though, which I find makes composition easier.  I also do concert photography from time to time, so low light and shallow depth of field is an advantage there.  However, the focus recently quit working (both auto and manual) on my 20-year old 70-200/2.8.  If it can't be repaired economically, I will seriously consider one of the f4 versions to replace it.  For now, I'm shooting all primes again.  It's like 1999 all over again, except there are GEVOs everywhere. :-)

GTD

The article is written by a Photographer who really knows the business, from experience. There is some info out there on generating lots of data on a lens as an OEM research Lab would do, I don't see conclusive evidence of what effects ghosting. But it certainly suggests there's a way to imperically differentiate between lenses, or settings, that have ghosting and those that don't.

Most of my rail photos simply stay away from tight, long lens compression compositions, so I'm of no use on that.