My Framing Test Photos - Please help me decipher them

[Larry Kellogg]

Hello,
Please visit:

http://homepage.mac.com/mac.hive/PhotoAlbum6.html

to see some shots I took in order to test the framing. I think I have a
problem but I want to make sure.

The top 3 shots were taken with the white framing lines on the top shelf and the
left side. The framing lines don't look square to my eye when I look through the viewfinder.

The last three shots were taken with a bubble level in place, again, trying to line up the top shelf with the top framing line and the left side of the shelf with the left framing line.

What do people think?

Thanks for the help!

Larry

 

[driggett]

Larry,
Is the bookself selves level? Whe I use photoshop I have to adjust .75 degrees ccw to get a straight line on the first bubble shot. I would hope that the sensor is level.
Cheers,
Chris

 

[driggett]

Larry,
On the third photo of framing with frame lines I have to do a correction of .75 cw to get the bookcase selves level.
Cheers,
Chris

 

[Larry Kellogg]

Larry,
Is the bookself selves level? Whe I use photoshop I have to adjust .75 degrees ccw to get a straight line on the first bubble shot. I would hope that the sensor is level.
Cheers,
Chris

Chris,
Thanks for the response. I just threw a torpedo level on the top shelf and it is almost perfectly level, perhaps lower a tiny bit on the right side. I don't know if it is off enough to cause the error, it might be. I'll pick up the bookshelf a little bit on the right side and shoot some more shots tomorrow.

Is this error similar, in terms of magnitude, to what you've seen in your camera? Did you bring yours back for adjustment or exchange?

Regards,

Larry

 

[driggett]

Larry,
I just had a thought. Assuming your sensor is level and the bookcase support is not then your framelines is 1.5 degrees ccw off (.75 + .75) which is what my frame lines are when using a 50mm lens.
What serial number is your camera? If it is close to mine hmmm....
Cheers,
Chris

 

[driggett]

Larry,
I have not had it adjusted yet. I will call Epson in the morning to see what they say. Does anybody know how to contact this DAG person to see what he would charge. He was mentioned in earlier posts about helping another person with a focusing problem.
Thanks,
Chris

 

[pfogle]

Larry... the first thing that strikes me is that the lines of the bookshelf converge towards the right. This means that the sensor plane isn't parallel to the wall, which makes it very hard to interpret your pix in terms of the lining up of the framelines.

What I noticed in my case was a rotation of the picture, but all the parallel lines were still parallel. There are some samples in my original post (sorry, I don't know how to link to them from this reply)

cheers
Phil

 

[Larry Kellogg]

OK,
Here are another set of shots to test the framing:

http://homepage.mac.com/mac.hive/PhotoAlbum7.html

The first set of shots are are lined up with the top shelf and left side. The sensor does not seem to be correct. By the way, I picked up the bookshelf a little to make it perfectly level and I used a laser level on the edge of that top shelf to check it.

The second set of shots use a two way bubble level.

What do people think?

My serial number is 18xx...

Larry

 

[jlw]

To get a valid test, the imaging plane of the camera has to be absolutely parallel with your test surface, and square to the center of your test target.

Any deviations will result in 'skewing' of the image, which may get misinterpreted as frame misalignment. You're looking for errors in fractions of a degree here, which makes alignment extremely critical.

Also note that if the lens you use for the test photos has any rectilinear distortion (barrel, pincushion, or skew) it may also distort the image in a way that could be mistaken for frame misalignment... so, repeat your test with several lenses if you can.

 

[Larry Kellogg]

Ok,
I guess I'll have to do some measurements from the shelf to both sides of the camera to figure out if the image plane is absolutely parallel to the bookshelf when the camera is on a tripod. Any other suggestions?

So, in real world shooting situations, say, on other rangefinders, do people expect their framing lines to be parallel to an object they're shooting? I don't know what to expect here but it does seems to me that all my handheld shots where I put the top white line on the bookshelf come out with the bookshelf rotated clockwise a small bit.

I saw the other test shots of the wood slats and the venetian blinds and it looks like it is the same type of misalignment.

Larry

 

[jlw]

I'll be honest and say that in real-world shooting I don't worry about it.

For the type of shooting I do with a rangefinder camera, I seldom manage to hold the camera exactly level within a fraction of a degree and exactly plane-parallel to a flat subject anyway, so it wouldn't make any difference whether my framelines were perfectly level or not.

If I've shot a subject in which it would drive me crazy if a particular straight line were not perfectly level, I just level it in software.

(Note that if you shoot a straight-line subject at even a slight non-parallel angle, perspective will cause all straight lines to converge toward a vanishing point -- so if you level one that doesn't pass through the exact center of perspective, the others will still be non-level. This is why it's important to be square to your target in test shots. It's very difficult to reproduce this condition in non-controlled shooting.)

 

[Larry Kellogg]

Yeah, I understand what you're saying. By the way, I'm shooting with the Ultron 35mm lens. Another thing, why do my pictures include so much more than has been framed by the white lines?

Larry

 

[jlw]

The extra area is a safety factor. It's necessary because the effective focal length of a lens actually increases as you focus closer.

If the framelines were designed to so that the actual picture area was exactly the same as the frameline area at infinity, then the actual picture area would be less than what's in the framelines at close distances -- in other words, objects positioned at the edges of the framelines would be cropped out of the final image.

This is considered undesirable: If the customer looks through the finder and sees that Aunt Marge is just barely squeezed in at the edge of his family group shot, Aunt Marge had darn well better be in the actual picture!


So, most RF manufacturers design the finder frames so that the actual picture area is slightly larger than the finder frame area at infinity. This provides the needed safety margin, and it gets larger as you focus closer (because of the increase of the lens' effective focal length.) That way, if you position Aunt Marge at the edge of the frame, you can be sure you'll get ALL of Aunt Marge... plus a little extra, which easily can be cropped out.

[Note: A few manufacturers -- notably Konica, in some of the Auto-S cameras and the Pearl IV rollfilm folder -- incorporated "field size correction" in their RFs' bright frame mechanisms. This caused the framelines to shrink as you focused closer, compensating for the lens' increase in effective focal length. Designing this feature into an RF camera with interchangeable lenses would be extremely hairy, and most photographers don't expect super-accurate framing from an RF camera anyway, so most manufacturers opted not to follow Konica's lead.]

 

[FrankS]

What jlw said. It is perfectly obvious from the photos of your bookcase shelves that you are not directly in front of it. I can see the inside of the right side of the shelf but not he inside of the left side. You are standing too far to the left when taking this photo. This completely invalidates any conclusions you want to draw about framing and tilting of the framelines.

Aside from that, I don't think in real world picture taking situations (unless it's archetectural photography) a slight discrepancy in framing like this matters one iota.

 

[pfogle]

Just to clarify...
I agree with Frank's last post, that the sort of discrepancy I'm seeing in these tests (Larrry's) is not enough to be overly concerned about.

In the case of my camera, I've shot hundreds (maybe thousands) of frames and never been bothered about things being out of whack UNTIL I tried this particular body. The last R-D1 I tried was perfectly fine. The tilt with this current one (which is being replaced today) was obvious from the start, so this isn't some tiny nit-picking thing I'm talking about, but a glaringly obvious fault.

Also, I do care about things lining up - especially on digital where you want to do as little cropping as possible. One of the reasons I like the RF is that there's less barrel distortion to worry about, so you can line things up nicely.

Anyway, to recap... from what I can tell, your machine looks fine to me! Also, I started this thread to alert people to a possible QC issue - I apologise if I have made people anxious about slight variations in their own cameras. This wasn't my intention!

cheers
Phil

 

[David Kieltyka]

I've never bothered testing frameline alignment 'cuz I figure if they were off enough to matter I'd notice this with actual photos. My approach is: shoot, shoot, shoot...and if I notice any technical issues then I test. Otherwise not.

-Dave-

 

[Larry Kellogg]

Frank, you're right, I guess I wasn't standing exactly in front of the bookcase when I took those shots. I think I have to stop taking pictures of the stupid bookcase and go shoot something else. ;-)

I think my framing problems is not as pronounced as what I saw in the shots that Phil showed so maybe I should be happy. There is a no guarantee that another camera would be better in this regard, and like you said, it really doesn't matter when shooting in the real world.

I just had a conversation with a friend of mine with regards to this issue and his take was that if I want all of the "organic" feel of this analog camera, and controls, that I should be willing to deal with small issues like this one. He called it "listening to the music" and not worrying about the measurements. If I want something with more perfection then I should shoot the SLR.

Larry

 

[pfogle]

Larry, I've had two R-D1's that I've sent back, the first because of hot pixels that I wasn't overly concerned about, but the supplier wanted to exchange it as it was, in his words, 'a bespoke product' - ie premium price, premium service.

the second had this unfortunate frame issue (and incidentally, just as many hot pixels!), so I had to send it back.

BUT I still think this camera's great, and just added another lens to the swag bag ( a new CV 21mm f4), so I'm still prepared to invest in it. I really think it has something unique to offer... a combination of great glass and RF sponteneity, with the digital output. For me, it's the bee's knees 🙂

Minor issues- I'm not a collector, so I've never had a system I didn't have to hack so some degree!

Phil

 

[jlw]

Frank, you're right, I guess I wasn't standing exactly in front of the bookcase when I took those shots. I think I have to stop taking pictures of the stupid bookcase and go shoot something else. ;-)

I think my framing problems is not as pronounced as what I saw in the shots that Phil showed so maybe I should be happy. There is a no guarantee that another camera would be better in this regard, and like you said, it really doesn't matter when shooting in the real world.

I just had a conversation with a friend of mine with regards to this issue and his take was that if I want all of the "organic" feel of this analog camera, and controls, that I should be willing to deal with small issues like this one. He called it "listening to the music" and not worrying about the measurements. If I want something with more perfection then I should shoot the SLR.

Larry

While that's basically my take as well (at least as long as we're talking about fractions of a degree) I'd concur that if it really is consistently "off" -- significantly enough to constitute a manufacturing defect -- you're better off documenting it clearly and then exchanging the body while it's still under warranty.

My point was that unless you make your test shots pretty carefully, you won't be able to distinguish whether the problem is really CCD/frameline misalignment, or something else.

A few rigorous test shots -- carefully lined up on a tripod so you're square with a definitely rectilinear subject such as a brick wall -- should not only put your mind at rest as to whether or not you've got a problem, but provide irrefutable evidence of the problem if you decide it needs to go back to the seller.