Luminous Landscape Forum
Equipment & Techniques => Medium Format / Film / Digital Backs – and Large Sensor Photography => Topic started by: EricWHiss on May 24, 2011, 12:45:00 am
-
I know I have promised a few some comparison images between the AFi-ii 12, CFii-39MS, Canon 5d2 and P20 and I'll get to that very soon. For now here's a shot I took today hand held with the Rollei 90mm apo macro at f/13 and 1/250 using a twin metz macro flash unit with the 356 module.
First I'll attach the full frame image and then a couple 100% crops. I processed this out in LC 11.5 using basically the defaults for sharpening and noise but added +5 on the saturation and brought in the dark and light points with the curves tool.
Make sure to expand the 100% crops all the way. These jpgs got sharpened on output from LC 11.5 and then again on output from LR3 (screen sharpening set to low) and you could argue its too much but anyhow wow!
-
Wow. Impressive stuff. Makes my D700 & 60mm Nikkor micro's results pale into insignificance
-
nice!
-
have to say its impressive
-
this is insane!
-
I want more detail !!! :P
-
I want more detail !!! :P
Open up to f8 just careful as it may cut your eyes ;)
-
Open up to f8 just careful as it may cut your eyes ;)
So the 5.2u pixel size on your 80MP sensor places the diffraction limit at f8 and lower ? Is the per pixel degrading that obvious between f8 and f11 ?
-
So the 5.2u pixel size on your 80MP sensor places the diffraction limit at f8 and lower ? Is the per pixel degrading that obvious between f8 and f11 ?
The size of the first ring of the diffraction ('Airy') disc is about 1.35 times the f-stop (for green light). People tend to say having 4 sensels capturing an Airy disc is still good.
So - that would result in an allowed f-stop of (5.2*2)/1.35 = 7.7 - voilá !
-
Wow.. that is indeed pretty impressive stuff.
-
So the 5.2u pixel size on your 80MP sensor places the diffraction limit at f8 and lower ? Is the per pixel degrading that obvious between f8 and f11 ?
All I meant was that the 90mm Macro is sharper at f8 than at f13 but as you can see in Eric's images at f13 it's not so bad...
-
All I meant was that the 90mm Macro is sharper at f8 than at f13 but as you can see in Eric's images at f13 it's not so bad...
Actually I get a feeling of craving seeing this - I believe drug addicts feel this at times ...
-
I tested the IQ 180 several times already and seriously these new 80 mpx sensors have a wow factor when it comes to detail. When shooting there is no way in the world your eye is getting what these sensors are picking up, not even a remote chance of it. It's actually kind of frightening. Certainly a lot of horsepower and more than many of us need for sure but when you see it than you want it. I just upgraded but I am more concerned about back speed so I chickened out for a IQ 160 besides the delta jump is not easy for sure. Even 60 mpx is a lot and maybe more than my commercial needs are but I wanted FF this time over the cropped P40 but for those jumping into 80 there is a difference in detail for sure among other things. As more folks are getting there new 80 mpx sensors you will certainly hear more about the detail jump. It's there alright
Nice shots BTW
-
I tested the IQ 180 several times already and seriously these new 80 mpx sensors have a wow factor when it comes to detail. When shooting there is no way in the world your eye is getting what these sensors are picking up, not even a remote chance of it. It's actually kind of frightening. Certainly a lot of horsepower and more than many of us need for sure but when you see it than you want it. I just upgraded but I am more concerned about back speed so I chickened out for a IQ 160 besides the delta jump is not easy for sure. Even 60 mpx is a lot and maybe more than my commercial needs are but I wanted FF this time over the cropped P40 but for those jumping into 80 there is a difference in detail for sure among other things. As more folks are getting there new 80 mpx sensors you will certainly hear more about the detail jump. It's there alright
Nice shots BTW
Have you or anybody else tested the files from IQ180 against the Leaf 12?
Henrik
-
I want more detail !!! :P
me too! me too!
-
The size of the first ring of the diffraction ('Airy') disc is about 1.35 times the f-stop (for green light). People tend to say having 4 sensels capturing an Airy disc is still good.
So - that would result in an allowed f-stop of (5.2*2)/1.35 = 7.7 - voilá !
Thanks, that's a really quick and useful calculation method. I did test through the range of apertures to see how diffraction would impact stopping down since getting the most DOF (without stacking) is sometimes important to me. You can indeed start to see diffraction creep in beyond f/8 but its not making a big impact on sharpness until about f/13. I shot those side by side with the CFii-39MS so maybe I'll post those in a separate comparison thread.
-
Not bad for detail, but I can't quite see the individual molecular bonds holding the flower together--perhaps a little more detail would be in order. :) Looking forward to see the side-by-sides--how much does ~40% more resolution (40MP -> 80MP) make an impact? It will be great to see.
Thanks for taking the time to do this, Eric.
Since the PSF for diffraction is known (Airy disc), does anyone know if the damage can be undone (or partially undone) with R-L deconvolution?
-
All I meant was that the 90mm Macro is sharper at f8 than at f13 but as you can see in Eric's images at f13 it's not so bad...
Got it, thank you :) . I can certainly relate to the craving comment of the previous poster :)))
-
Not bad for detail, but I can't quite see the individual molecular bonds holding the flower together--perhaps a little more detail would be in order. :) Looking forward to see the side-by-sides--how much does ~40% more resolution (40MP -> 80MP) make an impact? It will be great to see.
Thanks for taking the time to do this, Eric.
Since the PSF for diffraction is known (Airy disc), does anyone know if the damage can be undone (or partially undone) with R-L deconvolution?
Get a freaking microscope. ROTFLMAO
Just kidding
Seriously Bradley as much as i love my P40+. Its a fairly good margin difference between the 40 and 80. You realize it kills me to say that too and cost me a boat load of money realizing that but settled on the 60 but if speed of back and money (wife) where not the issue I would get the 80. LOL
Besides the detail and can't speak for the Aptus 12 but I would assume like the IQ 180 the tonal range and smoothness of the files is certainly different to the good not to mention the extra DR that really is there. I'm adding black to this 180 images I have shot just to look more normal. So this new sensor has some mojo going on. Does one need this much horsepower ? That is up to the buyer but it certainly has a wow factor.
-
Since the PSF for diffraction is known (Airy disc), does anyone know if the damage can be undone (or partially undone) with R-L deconvolution?
Hi Bradley,
That should be no problem for R-L deconvolution. With these mild amounts of diffraction, even a Gaussian PSF will work quite well. However, one should be careful to not 'enhance' aliasing at the same time (the risk depends on subject matter and good focus).
Cheers,
Bart
-
Since the PSF for diffraction is known (Airy disc), does anyone know if the damage can be undone (or partially undone) with R-L deconvolution?
Funny... that was my first thought too
-
Hi Bradley,
That should be no problem for R-L deconvolution. With these mild amounts of diffraction, even a Gaussian PSF will work quite well. However, one should be careful to not 'enhance' aliasing at the same time (the risk depends on subject matter and good focus).
Cheers,
Bart
Bart or others,
Any recommendations for R-L deconvolution software? I'm on a Mac if that matters. I'd be curious to try this out, and will share the results.
Eric
-
Try Raw Developer (http://www.iridientdigital.com/products/rawdeveloper.html)
-
Do any of the Schneider Digitar macro view camera lenses compare to the performance of this 90mm?
-
Try Raw Developer (http://www.iridientdigital.com/products/rawdeveloper.html)
Thanks - just downloaded the demo... here's what it looks like with radius of .75 which I narrowed down on quickly as at least near the best value. Obviously there are other differences in the way the image is handled but I don't really know the software well enough yet to really make the rest look the same, however it does appear more detail is being extracted. Why doesn't C1 or LR3 have this method?
-
btw - should also mention that in this image I think LC 11.5 does a better job with the fine detail than does C1.
-
Bart or others,
Any recommendations for R-L deconvolution software? I'm on a Mac if that matters. I'd be curious to try this out, and will share the results.
Hi Eric,
It is possible to process a TIFF with e.g. RawTherapee (http://www.rawtherapee.com/), which is freely available for a number of OS platforms. The Richardson Lucy deconvolution was already implemented in the older V2 (non-alpha V3) version, should a recent build not function. Adobe Lightroom and Photoshop ACR (which can open a TIFF as an ACR file) offers a form of deconvolution under the detail slider when set to maximum, and Photoshop's Smart sharpen filter, in advanced mode also uses a sort of deconvolution. I find the RL implementations to be more effective.
Cheers,
Bart
-
I think Leaf Capture renders finer detail in high contrast areas but it's a pain to work with.
Leaf Capture
(http://copiasxl.com/images/aptus12/LeafCapture.jpg)
C1
(http://copiasxl.com/images/aptus12/CaptureOne.jpg)
-
Hi Eric,
It is possible to process a TIFF with e.g. RawTherapee (http://www.rawtherapee.com/), which is freely available for a number of OS platforms. The Richardson Lucy deconvolution was already implemented in the older V2 (non-alpha V3) version, should a recent build not function. Adobe Lightroom and Photoshop ACR (which can open a TIFF as an ACR file) offers a form of deconvolution under the detail slider when set to maximum, and Photoshop's Smart sharpen filter, in advanced mode also uses a sort of deconvolution. I find the RL implementations to be more effective.
Cheers,
Bart
Bart,
Thank you. I've downloaded the only build I found for Mac and it's certainly got a lot of promise, but I need to figure out how to make the most of the settings. Besides the RL sharpening, the fine detail contrast slider is also bringing out more apparent detail.