More EFCS woes.

Mhoner's post on the previous page along with the link to the non-telecentric lenses explanation seems very plausible to me.   I can take a look at this over the next week or two perhaps. (if the sun comes out ever again!)

The sample image i posted can be almost entirely cured by a 1.5 stop ND grad applied in lightroom with the centre exactly in the middle of the image. That sort of implies their is something geometrical going on here but the maths is beyond me.   

Also, I have never heard the word 'telecentric' before today but from mhoner's page, I think I get the general idea.   

Does someone have a list of telecentric lenses or even a way to tell if a particular design is telecentric or not?

Photosopher wrote: I'm having difficulty accepting that aperture is the cause of this. If it were, would not the image display a vignetting affect all around... circular... rather than just one side bright and one side dark?

No, it wouldn't. The effect would appear with the incorrect exposure showing up across the long dimension of the frame, extreme at the edge and tapering off. But still, the out-of-sync aperture theory has other problems.

Below we see one user's example photos from a recently posted thread in the DPReview forum.





1) Those were shot at 1/2000s (not really all that fast) and f/9 (closed down a lot).

2) The first shot of the sequence is unaffected. Why is that? Did the EFCS not engage for the first shot? If it didn't, is that normal behavior?

3) The underexposure in the next two shots is EXTREME, way more than the first example in the current thread (and, I would think, way more than can be explained by any theories proposed so far, especially at this shutter speed).

4) The shutter blades in recent SLT models move from bottom to top, right? If so...

5) The underexposure of those frames is at the top, meaning that it occurred at the bottom of the shutter gate, at the beginning of the exposure.

6) There is also visible overexposure at the bottom of those frames, meaning that it occurred at top of the shutter gate, at the end of the exposure.

7) Is this just a red herring example of a malfunctioning camera or is it some weird thing about the Tamron lens used?

robertsmx wrote: There is a good chance you're experiencing what I did with my A55 a few months ago.

Who are you talking about... the OP or the person who made the shots directly above (not me - I only referred to them)?

And in any case, the A55 doesn't have an EFCS.

I just tried a few more shots with my sigma 70-200, Sony 50/1.4 and Sony 16-50/2.8.   I set 1/5000 and f3.2 as in my earlier shot a took a few blank wall photos. I had to boost the ISO to 6400-12800 to get any sort of decent exposure due to the gloomy day here again.    In every single EFCS shot, I see a darker horizontal band at the bottom 1/3 of the screen and a rather fainter band across the top 1/3. It is more pronounced at 1/8000 but with the same layout. I tried f4 and the effect was still there but it was getting to dim at that f stop to conclude anything useful. F3.2 vs f2.8 made little difference but then again, it's only 1/3 of an f-stop. Perhaps a more controlled test could conclude a bit more.

I couldn't quite reproduce the more extreme band as seen in my original pic. And perhaps my original pic also had the lighter band too but it is to faint to notice in a real pic.

With efcs off, the bands go away entirely.

Revisiting my hss stripy pics, apart from the main effect of narrow stripes, you can also make out the 1/3 and 2/3 bands but they aren't as pronounced as my quick tests from this morning. At the time, I thought they might be just from light fall off from uneven flash light but I think there are two effects in those pics - HSS narrow banding and general 1/3-2/3 bands.

More than ever, I'd like a menu option to offer "EFCS off if shutter speed faster than 1/4000". If this effect is there every time I go over 1/4000, I'll be getting fed up switching efcsvon and off all summer!   I didn't notice it much because it's been winter and gloomy perhaps.

bartman wrote: I don't want to be a devil's advocate, but is the shutter system a relevant disadvantage of the SLT design?

It has noting to do with SLT (and the function is only available in A77/65 but not in the previous SLTs). The electronic first shutter curtain is also used in Canon DLSRs. It is in fact a significant advantage, because in DSLRs with MLU and in SLTs vibrations caused by the shutter will not appear before the end of the exposure. This feature combined with the non existing mirror slap of a SLT is a great benefit, especially when you are shooting with long focal lengths.

As we see in this thread and also here under rare circumstances the electonic first shutter curtain can also have disadvantages. That's why you can disable this function in the menue.

sybersitizen wrote: robertsmx wrote: There is a good chance you're experiencing what I did with my A55 a few months ago. Who are you talking about... the OP or the person who made the shots directly above (not me - I only referred to them)? And in any case, the A55 doesn't have an EFCS.

EFCS or not, my camera (A55) is with Sony for a fix to address a similar issue (albeit, in my case, as I said, it happens at shutter speeds faster than 1/800s). It may be due to a different component, but ultimately, that is the problem with image I've seen (initially, occasionally, and later, almost always).