Photoclubalpha

Here are some pictures of ghosting issues with Sony A55.
In the pictures its a LED deck light.
I Took pictures with Sigma 17-70 2.8-4 HSM and Sony DT 55-200 lenses

1. Sony 55-200, 55mm

DSC05169 by Aram Kostanyan, on Flickr

2. Sigma 17-70, 70mm

DSC05166 by Aram Kostanyan, on Flickr

3. Sigma 17-70, 70mm

DSC05164 by Aram Kostanyan, on Flickr

4. Sigma 17-70, 70mm

DSC05163 by Aram Kostanyan, on Flickr

5. Sigma 17-70, 17mm

DSC05159 by Aram Kostanyan, on Flickr

6. Sigma 17-70, 17mm

DSC05158 by Aram Kostanyan, on Flickr

Yes, that's exactly what happens. But in practice I've had no issues with normal subjects including sunsets.

David

Seems that this effect only can occur with strong, direct lightsources that hit the mirror at a specific angle.

Have you seen it on real life photo's yet Aram?

I aggree with David.
In real life shots I have never seen ghosting unless you to shoot a candle in the darkness a a bulb light on a dark background.
Even in the above photos in some of them I gave +1.7 exposure compensation to see the effect.

I've seen ghosting with my a700 + Sony 50/2.8 macro
So could you tell me if it's a reflection of the "translucent" mirror, the A filter or the sensor plane itself?

it would have been good if you could have taken the same photos with "normal" SLR, has a non-expert seems like a lot of the situations that cause ghosting in the A33/55 could also cause ghosting in normal DSLR/lenses, how do the photos look with out the +1.7 exp?

That looks like standard DSLR ghosting to me, I can get the same effect on my a700's sensor.

As I recall the translucent mirror ghosting was always (when visible) a few pixels (don't recall how many) below the bright object. I copied one of the dpreview test shots that shows this to my dpreview galley here:

http://forums.dpreview.com/galleries/59 ... 71/008-007

If you look at the white cross you will see a band of light "shadow" just below the horizontal arm of the cross. There is no band at the top of the horizontal arm or on either side of the vertical arm. There were other bright/dark contrasts in their test shots that showed the same thing.

I checked other cameras in their comparison and none showed any band. There was also a shot on dpreview where the camera was held in the portrait position. In that case the band was to the right of the bright area.

I haven't seen any ghosting on the various shots posted by users (and I have been looking) except in certain astronomical shots where there were some extra stars in the right position to be ghosts. However, when I have looked at user shots of the moon, I haven't seen any ghosting.

tom

There are two specific kinds of mirror-related 'ghosting' with the current SLTs. First one is a sharp ghost a few pixels below a very bright highlight edge happening by the internal reflection between the SLT mirror surfaces. The second one is a hazy veil ghost happening at the part of the frame quite remote from the bright part of the image, and that's caused by the reflection off the sensor being reflected back by the slanted SLT mirror.
Neither of them is present in the examples above.

The 'normal' digital camera ghosts are caused by the secondary reflections between the sensor and the lens back element (or maybe other lens elements' surfaces). And exactly that type of ghost is there in the examples above.
Then there also may be some ghosts induced by internal reflections in the sensor toppings.

I believe there are some design approaches possible that may actually make the SLTs far less prone to any kind of ghosting than a typical DSLR is.
But we shall see; soon, I hope...

UrsaMajor wrote:

Dr. Harout wrote:So could you tell me if it's a reflection of the "translucent" mirror, the A filter or the sensor plane itself?

No question - it is clearly due to the translucent mirror !

With best wishes,
- Tom -

Exactly, that same translucent mirror which doesn't exist in the camera with which I've taken the shots...

If it is because of the mirror, than why is it next to the main light. As, in my opinion, if it is from the mirror, the light that reflected from the sensor reflected back to from the mirror, than it need to always be below (or upper) the main one.

It's the light reflecting from the surface of the sensor back to the lens and from there a second reflection from the surface of the lens back to the sensor, but this time with diminished brightness.
So my question: whenever they say "ghosting" due to the "translucent" mirror, how can we know it's not a reflection from the sensor?