645 negatives with vertical vignetting

[mfogiel]

I have three 645 cameras: Bronica RF, Contax 645 and Pentax 645N, all fantastic photographic tools, but...
All of them to some extent, and particularly Bronica RF create vertical vignetting on my images whenever I expose a tad more generously. This is particularly evident with low latitude films, like Acros, but it even happens with Tri X. The vignetting runs along the long side of the negative ( which means across the film) so it cannot be from light fogging the film from the outside, and in any case you would expect stray light to create a denser negative, thus a "white vignetting". Here, the effect is opposite: a darker edge looking like dodging on the final image, but it feels like "fogged" in the sense, that all details are uniformly darker. I do not recall anything similar happening with any other format camera I've ever used.
I attach a frame on Tri X exposed for the shadows, where this effect is quite striking. Anybody has a good explanation?

 

[DavidKKHansen]

Interesting, never seen that before. Since it runs along the vertical part of the frame, we can rule out a filter or the hood... wild guess: lack of film flatness?

Wait, that wouldn't make sense, as the film is pulled left to right... hmm...

 

[Dwig]

'Tis a puzzelment.

Main points:
It is a darker area along the edges of a positive
It is ruler straight so it is not a lens issue
It occurs in multiple cameras
It runs across the negative rather than parallel to the edges of the film

Things it is not:
not fogging which would cause light areas not dark
not a camera issue as it occurs with multiple cameras
not caused by developing reels as it runs across the film.

Questions:
can it be seen on the negative?

The posted image is a positive so it is NOT what the camera produced. Some other agent was involved in producing the positive image posted. I suspect that agent is the evil one. Could this be a scanner or enlarger carrier problem?

 

[sevo]

All of them? Then I'd suspect a scanner or enlarger issue. Sure you are using the right mask in the right place?

Aged medium format cameras are somewhat more subject to mirror or shutter capping issues than 35mm cameras with their smaller, less heavy mechanics - but capping would affect one side different than the other, not both the same way...

 

[charjohncarter]

These things drive me crazy, I have no suggestions other than what Dwig said. I'm really marking this thread so I can see what it ends up being.

 

[Bill Clark]

How is the film processed?

Sent out, try a different lab.

Tank processing, then developing may not be the issue.

Doesn't seem to me to be camera related as you said it appears from a variety of cameras.

Does this appear on the negative? If not, perhaps a scanner issue?

It almost looks like a wee bit smaller mask than the negative being used. But then again it probably wouldn't look like vignetting.

My guess, it's something other than the camera.

Do you use the same lens on each camera? Adaptors etc.

 

[Spanik]

I'd rule out any camera related issues. If all three have it, then it isn't the shutter as the Bronica has a leaf shutter and at least the Pentax a focal plane one (don't know about the Contax). And because the Bronica having an electric controlled leaf shutter these lenses won't work on any other camera so we can rule out using the same lens on different camera's.

No, I'd search it outside the camera. Got a Bronica RF myself and never seen that.

If it is scanned, how is the film with respect to the scanning direction? Does it looks the same if you put a 6x4.5 in the 6x6 holder?

 

[mfogiel]

Thank you all for suggestions. Following your intuitions I have concluded, that it has to be the scanner. I scan on Nikon CS 9000 in fine mode. When a negative is dense, the scanner slows down considerably, and these vertical signs are most likely an effect of some internal reflected light. It is usually a bit difficult to notice small tonal differences on a dense negative, but I've checked this neg now with a diapo 6x6 loupe, and there is no discernible density difference. The solution then could be to scan a 6x6 field and crop later. I wonder if the problem is native, or if it has been caused by some scanner defect/deterioration. I've been using mine for about 7 years without maintenance.

 

[sevo]

TI wonder if the problem is native, or if it has been caused by some scanner defect/deterioration.

Native. It looks like the usual result you get when scanning negatives on a CS8000/9000 without the proper masks.

 

[brbo]

Yes, a "feature" of CS8000. CS9000 was said to be much better in this regard so you should probably clean your scanner's mirror and lens. It's easy and there are instructions for this on the net.

One other thing to try is to not let scanner software set automatic exposure. What software do you scan with? It's very easy in Vuescan to set exposure just under clipping value (on the clear part of negative). This way you have the best shadows and Nikon CS9000 should have reasonably high DMax to get also good detail in highlights even on a dense negative.

(I still can't believe how good Microtek 120tf scanner is in this regard. I could put any film negative/slide on a glass with no masking and never got ANY light creeping from unmasked areas into the scanning area.)

 

[Fred R.]

Run a negative through the scanner both ends first and see if the dark edge changes sides.

 

[mfogiel]

I use Vuescan too. I'm thinking of widening the selected field to embrace the black space as well, but before I will fix the black and white points.

 

[Dan Daniel]

I use Vuescan too. I'm thinking of widening the selected field to embrace the black space as well, but before I will fix the black and white points.

In Vuescan, under the CROP tab, there is an option called 'Buffer.' What this does is reduce the outer edge area used for calculating exposure. So it is easy to over-scan and include the black (clear film) area. With other scanning programs I have used, including the border will throw off exposure. But not with Vuescan using the Buffer setting.

I have mine set for 5%, sometimes 10%.

 

[menos]

In Vuescan, under the CROP tab, there is an option called 'Buffer.' What this does is reduce the outer edge area used for calculating exposure. So it is easy to over-scan and include the black (clear film) area. With other scanning programs I have used, including the border will throw off exposure. But not with Vuescan using the Buffer setting.

I have mine set for 5%, sometimes 10%.

Thanks a lot for this explanation!
I have seen similar issues when scanning 6x6 with a Minolta Multi-Pro and VueScan and will experiment with the scanning.


Marek, this is very likely related to the Scan - looks exactly a what I see sometimes when scanning 120 film!

 

[brbo]

'Buffer' setting won't help much. It will lower the analog gain if you have it set to auto exposure (not locked), but this won't help much.

I've now experimented with my CS8000 and even lowering exposure way down won't completely prevent light bleeding from clear areas of the negative or in the transitions from very bright to very dark areas in the film.

Cleaning scanner optical path (mirror, lens) should reduce the problem, but some people suggested the glass on the ccd sensor housing (on most scanners this glass is a cheap glass that causes reflections) was the main source of those internal reflections. Read this thread. So, lets remove the glass...

Example of a scan with ccd sensor glass still installed (scanner is Minolta Multi Pro):

Glass removed:

Wow, no flare/ghosting!


Getting to the ccd sensor in CS8000/9000 is very easy, the problem is removing the glass. The person that provided the above examples had the glass removed by a professional service that uses special equipment that sloooowly (several hours) grinds the glass off the ccd sensor housing. So far I haven't been able to locate a service in EU that would remove the glass from my sensor (sent pictures of the sensor to a few firms around EU).

Nikon CS8000 ccd sensor:

Yesterday I dug out an old CanonScan flatbed and took out the sensor (looks pretty similar to my Nikon CS8000 sensor). I had the glass off of it in under a minute:

Somebody please talk me out of trying this on Nikon's sensor 😉

 

[Peter_S]

I have a 8000ED and get the same thing occasionally, using 6x6 and 6x7 from a Bessa III. I scan with the Nikon software.