I'm gonna ROCK the adapter world....

Longitudinal and lateral chromatic aberration.....

I'm gonna have to battle these factors of the DVX stock lens, then battle the adapter and its lens... yikes. I've got some work ahead of me.

I'm looking into post production ways to correct for this when using just the stock lens, but man, having a second lens can make this really hard to avoid, all the more reason to strap a killer lens onto the adapter itself!
 
Last edited:
Did some more thinking...

I'm gonna have to find some way to keep the DVX zoom constant. This whole chromatic aberration with the DVX stock lens changes at various focal lengths, at least thats what I've observed so far in my limiting studies...

I guess what I'm getting at is this:

For a true equal playing field, I need to keep the DVX100 lens's chromatic aberration consistent during each adapter test, and the best way for me to do that is maintaining the same focal length, and f/stop.

This could be hard to do given all the designs and such.

I've noticed that this issue of chromatic aberration also affects apparent resolution too, which is why I'm diving into each individual channel, and using only the same Nikon Lenses.
 
Which adapter requires the greatest focal length on the stock lens?

I'll probably operate all the tests at that focal length.
 
OK,

Just an update, I'm working on the chromatic aberration issues with the stock DVX lens.
 
Last edited:
Update:

I've been figuring out ways to eliminate the stock DVX Lens's lateral chromatic aberration in post.

As of right now, I'm about 95% sure I have a solid solution.

In a nutshell, the reason I'm doing this is because I want to keep the testing playing field as even as possible. Every lens has its own CA issues. The trick with adapters is that you have two lenses, each with their own CA issues, and when you use them at the same time, you combine the issues.... If I can correct for the DVX lens properly in each adapter test, I will have a far better view of how well they perform, as well as how well the 35mm lenses perform. I've found that the stock DVX lens's lateral CA varies depending on your focal length, and I suspect the f/stop might also play a role in it.

What I'm trying to do is figure out a focal length that will work for all of the adapters screens. Once I figure out what focal length is the "sweet spot" across all of the adapters, I will then shoot a res chart using the stock lens at that focal length, and then create a custom CA correction template for it.

When I post process all of the adapter tests, I will now have a DVX chromatic aberration free view of how they all perform.

What does everyone think of this approach?
 
Hey Guys,

Here are some screen grabs that show what I'm up to...

This is a res chart shot at 48mm set to f/5.6, threw two magenta filters (60 ccM total) under 3200 Kelvin light.

Before custom lateral chromatic aberration correction:

http://www.chrisnuzzaco.com/special_uploads/no_ca_correct.jpg

After:

http://www.chrisnuzzaco.com/special_uploads/custom_ca_correct.jpg

Look carefully at the edges of the lines to see the effect.

Different focal lengths exhibit varying amounts of lateral CA.

So far, after more testing, it appears that the f/stop does not affect lateral CA (it does affect longitudinal CA), nor does your focus setting, though I'm going to conduct more research on that variable.

So far the custom lateral CA correction templates are working very nicely.

I anticipate this will be the fairest adapter evaluation yet. I haven't seen anyone else correct for the camera lenses CA, which certainly skews the accuracy of those tests.

More updates to come...
 
Thanks Matt!

I'm certainly interested in testing the Brevis again.

As you can see, I've switched my CA correction approach from photoshop's lens correction plugin to a far more powerful custom approach in After Effects.

Can you get you give me the lens settings needed to set it up properly with a DVX100? I can make a custom template ahead of time that way...

Thanks!
 
cardmaverick said:
Hey Guys,

Here are some screen grabs that show what I'm up to...

This is a res chart shot at 48mm set to f/5.6, threw two magenta filters (60 ccM total) under 3200 Kelvin light.

Before custom lateral chromatic aberration correction:

http://www.chrisnuzzaco.com/special_uploads/no_ca_correct.jpg

After:

http://www.chrisnuzzaco.com/special_uploads/custom_ca_correct.jpg

Look carefully at the edges of the lines to see the effect.

Different focal lengths exhibit varying amounts of lateral CA.

So far, after more testing, it appears that the f/stop does not affect lateral CA (it does affect longitudinal CA), nor does your focus setting, though I'm going to conduct more research on that variable.

So far the custom lateral CA correction templates are working very nicely.

I anticipate this will be the fairest adapter evaluation yet. I haven't seen anyone else correct for the camera lenses CA, which certainly skews the accuracy of those tests.

More updates to come...
Click to expand...

OK I found your thread and see you have posted a couple pics ( in reference to the questions i had for you on the Easom 645 adapter thread ) Now I'm curious if this is only due to compression artifacting? It's hard for me to tell from a jpeg.
Steve
 
Steve, I'll post some tiff examples soon...

The reason I'm using the Andromeda as a test bed is that it has no compression whatsoever during capture, just raw sensor data, and I only render my files out to 16 bit tiff image sequences (I capture at 10 bit by the way...).

Hope that helps!
 
Just an update, I finally visited my printer, the silver halide prints I can make there might not be large enough, only 10 by 15 inches..

They do have an ink jet printer though, and it uses a sprayer to print the image, no dots, so its technically continuous tone. In fact, if you had to compare it to a DPI, its on the order of 185 or so DPI, any higher a DPI, and you don't see any difference, so I guess thats how they figured out the DPI equivalent.

They don't offer glossy paper, but an inbetween type of paper called luster, its basically a shiny matte surface.

Any opinions on getting the ink jet blow up? I can print a 3 by 4 foot chart for about $85 USD.... so I wanna make sure I'm spending money on something good enough for the job.

Thanks!
 
Chris, kudos on your thorough examination of the stock lens CA. It's the single largest issue for us in testing, and the one commonly ignored in adapter reviews. Rather than attempting to compensate, I'd just baseline with the bare DVX first. In other words, display the warts of the bare DVX lens along with the adapter footage. That's typically what we do in testing so we know what's there to begin with.
 
Dennis Wood said:
Chris, kudos on your thorough examination of the stock lens CA. It's the single largest issue for us in testing, and the one commonly ignored in adapter reviews. Rather than attempting to compensate, I'd just baseline with the bare DVX first. In other words, display the warts of the bare DVX lens along with the adapter footage. That's typically what we do in testing so we know what's there to begin with.
Click to expand...

Dennis,

I've been having that request quite a bit, my lateral CA correction is purely a post correction, so yes, I do plan on releasing two sets of results, one that shows each adapter without any stock lens CA correction, and one with it turned on.

:)

I can kill two birds with one stone, its pretty cool really, one set will show how well a particular adapter works with a stock DVX (no CA correction), and the other will be more fare in evaluating just the adapters themselves.

I'll be sure to mention what focal length each adapter requires from the stock lens as well.

Any thoughts on the ink jet 1956 res chart?

Thanks!
 
I'm having a hard time using a home printer size chart with the longer lenses... I can't fill the frame properly and then focus on the chart :(

So I'm looking into blowing it up some... 3 by 4 feet might be a bit too large.

What is the standard size anyways? I need to look into that. So far my photo paper home printed chart has been sufficient for looking at CA, but I wanna make sure I have an optimal print for checking resolution. I also plan to re-check my CA templates using the newer chart.
 
I just checked my PDF file of the chart... about 10 by 7 inches.. unless that file is wrong, it seems way to small for the longer lenses.
 
Thanks for the link, I've seen that chart, actually, I have that PDF file.

Is this a better chart than EIA1956? My sensors are actually 4:3, but I see that this chart has marks for various aspect ratios, so maybe it doesn't matter.

Thanks!
 
You must log in or register to reply here.
Back
Top