The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

DF+ & CMOS IQ Backs

Dogs857

New member
Curious why you would shoot with an GND if you are bracketing shots??

Expose for the foreground, add your ND filter then shoot the sky exposure. The GND is not worth putting into this. Unless of course you didn't have enough ND filtration to get the effect you were after.

Can't say I have ever experienced this problem bracketing and aligning shots myself. I noticed in the shots you posted here that the vignette in the upper part of your frame is larger in the ND shot as well.

Anyway at least you found something that suits your photography, I can't be sure but I think you are a fan of CMOS??? :D
 

GrahamWelland

Subscriber & Workshop Member
Regarding matching exposure images at the pixel level, I'm actually not totally surprised that you couldn't achieve it between shots given the filter on/off steps between exposures. I can easily understand there being very slight movement of the camera/tripod/lens just but touching the filter assembly, not to mention environmental aspects such as wind and vibrations. Heck even the shutter release.

Now if we were talking multiple pixels it might be a different situation.
 
Curious why you would shoot with an GND if you are bracketing shots??

Expose for the foreground, add your ND filter then shoot the sky exposure. The GND is not worth putting into this. Unless of course you didn't have enough ND filtration to get the effect you were after.

Can't say I have ever experienced this problem bracketing and aligning shots myself. I noticed in the shots you posted here that the vignette in the upper part of your frame is larger in the ND shot as well.

Anyway at least you found something that suits your photography, I can't be sure but I think you are a fan of CMOS??? :D
a) An ND filter itself without an ND grad would cause the same alignment problem as well. I deliberately chose this example to show people why ND grad cannot always workaround the dynamic range limitations of a CCD sensor - see how the ND grad cuts the buildings in this example?

b) Yes I'm a fan of the Sony CMOS sensors, just like many here are fans of the Kodak CCD sensors :D
 

Ken_R

New member
Hi Void,
Took a quick look at your raw files and not sure why they don't line up, but they clearly don't, as you have pointed out. If I didn't know better, I'd say they were shot at slightly different apertures?
In any case, a simple 100% overlay-difference map shows where they do not line up. I could do it using helicon or even warp tools if it were mine, but I'd be curious as to why they don't line up too. ;)
I have shot many multiple exposure-blended images with Credo80 and IQ250 (for masking reasons with fly-away hair on products and models) and have not experienced this problem. But Mine are in studio conditions using a DF+ standard primes, etc...

Bingo. Thx for posting.
 

Paul2660

Well-known member
Actually the two images line up fine. I will try to post some crops later today.

I simply converted them in C1, made sure that any "lens" corrections were off, then opened them in CC 2014, via bridge. Loaded them into photoshop layers via the tool menu, then ran an auto align. They line up 100%, no problems with the bulidings, the Belfast, (I love that ship), etc. The only thing that won't align is one of the cranes on the left side, but that is obviously motion of the crane. You can clearly see that it "moved" considerably in position in the two images. All the other stationary cranes all line up If you are doing this I have found it's "key" to turn off any type of lens corrections in C1 or LR as these tend to screw up the alignment process. It's much simpler to combine. C1 really doesn't have any corrections for most of the tech lenses, however I believe in the latest version they have added some for the rodies. I wasn't sure which lens was used anyway. All the building aline fine, my way to test this is paint back one layer from the other and see if you get any misalignment in the lines of the buildings, which I did not.

There are some very very faint movements on the tall building's edge on the right but no where near enough to matter, I found it very easy to combine the two with a soft edge brush. There are also a few issues on the extreme left where some people are standing on a quay or dock, but I believe that may be movement.

I also did not apply the LCC in my test, but I can't see how that would alter anything, but will try that later.

After I ran auto align layers, I simple painted back to get the composition I wanted.

This is standard proceedure for me with any bracketed images where I am not using the camera to shoot the bracket, and I have hands on the lens, i.e. in this case the shutter speed setting. Considering that several filters were added etc. the alignment is very good IMO.

In the crop, I have brought back the lighter shot of the building, you can see that one of the cranes on top has moved. That is not an alignment issue, nothing would align that. It moved. But the building have no issues at all. There is a slight slight halo on the right edge, that would be very simple to remove, just did not take the time. The other hot spots just show where I painted in the other image so you can see both are aligned around the building edges.


Paul
 
Actually the two images line up fine. I will try to post some crops later today.

After I ran auto align layers, I simple painted back to get the composition I wanted.

But the building have no issues at all. There is a slight slight halo on the right edge, that would be very simple to remove, just did not take the time. The other hot spots just show where I painted in the other image so you can see both are aligned around the building edges.
Could you upload the psd file, or a full sized uncropped jpg? ;)
 
Actually the two images line up fine.
Hi Paul,

Many thanks for spending the time and effort to demonstrate, and also I appreciate that you took the time to upload the psd file to dropbox! :thumbs:

I agree that to some extent CC 2014 does a better job than my CS6 (that you get better alignment than I could), however I can still see misalignment in the psd file you uploaded. For some people this may not be of a problem, but for people who dumped a serious amount of cash into MFDB and obsessed in pixel peeping to justify the investment, it may still be less than ideal / total satisfactory. The problem with such misalignment is the excessive amount of effort in post-processing attempting to remove halo, as well as decreased sharpness in the transition areas of blending (which defeats the purpose of image quality). With a Sony CMOS sensor (e.g. IQ250 / D800E / D5300) instead I could just push the shadow from a single exposure and be done with it and spend my time for other stuff in my life.

Below shows the 100% crop of the two aligned layers on both sides of the psd file you offered:



 

Paul2660

Well-known member
Yes, I agree it was not perfect, but I think that I could have gotten even better if I had just laid the two on top of each other and then faded opacity on one and moved the other around. That takes a bit more time.

However it's pretty close, not exact, but I could have made a print from it. :angel:

The areas you have pointed out may not even matter when blended, I need to look at them again later.

Paul
 
Yes, I agree it was not perfect, but I think that I could have gotten even better if I had just laid the two on top of each other and then faded opacity on one and moved the other around. That takes a bit more time.
I spent 6 hours on that using various things like the warp tool in Photoshop (50% opacity for the layer above), or Helicon Focus, trying to justify my IQ260. In the end I reached a conclusion that I would give up and accept the fact that even a Nikon D5300 could handle that scene better. This forced me into the decision to give up the IQ260 for an IQ250.
 

Paul2660

Well-known member
Looking at the shot again, I still feel I can get it to pass inspection, really with just a bit of tweaking, I know we will just have to agree to disagree on this one.

The vast majority of the stuff lines up with no problem, I don't see the issue on the tall pointed building as it blended perfectly. The egg shaped building is problematic, but there is a lot of green aberration on the right side of the file, (I did not check C1 to have it work on CA, and the top of the egg building, on the right side has a lot of it. Still easy to fix with the clone stamp.

I thought I was picky, but I would have no problem printing this shot up to 30 x 40, as the eye will not see the issues, at least my eye.

One other point, it's really amazing to see just how much detail the lens can pull in as you are getting some really fine details to show up, like the airplane warning lights on building tops, the finer cranes, antennas, a lot of detail on the Belfast, etc. I assume this was with a tech lens, the 32 Rodie?

The only thing that bothered me was the noise in the sky on the left side near the buildings but a round of neat image takes that out easily.

Here again, is a good example of CCD needing light. You have excellent details on the buildings, in the longer exposure shot, probably not as good as shadow detail that a D810 would get, but this is almost 2x the resolution. If you uprez'd the D810 shot to this level of resolution you would lose a good bit of the details, so it's a pretty good trade off.

Just one opinion.

Thanks again for sharing.
Paul
 

GrahamWelland

Subscriber & Workshop Member
Paul,

As you mentioned, the secret really for the IQ260 is simple - lot's of light. One thing I'd never complain about was the highlight recovery potential of the back and when I transitioned to shooting tethered it was immediately obvious that I was often shooting up to 1.5 stops under exposed compared to what C1 Pro would very comfortably handle, especially with C1 v8 and highlight recovery.

What I have already noticed is that the IQ150 works the other way around and shadow recovery is this sensor's forte.
 

Chris Valites

New member
I for one am excited about this. It'll be interesting to see what developments come out of Sony simply for consolidating the manufacturing lines from two different types into one (CCD and CMOS into CMOS only) as well as what R&D budgets would turn out when they combine CCD and CMOS.

This presumes, of course, that they're going to combine them, and it's not just a total shutdown/erasure of the CCD side of things. I feel like the resources would be combined, but that's just a gut instinct on my side of things.
 
Top