85mm F1.2 DS versus non DS comparison

usern4cr

R5
CR Pro
Sep 2, 2018
1,376
2,308
Kentucky, USA
Thanks for the post, and start of a new thread.
I do plan on getting this lens, and in the DS version.

The difference in the DS, as I see it is this:
* The lens right before and right after the iris (where the aperture blades are) have a vapor deposited light absorption that starts at the outside edge as heavy (but not total) and diminishes to no difference as you get maybe half way to the center and to the center.
* When wide open (where you really want to leave this lens if possible) this reduces the light from the outer edges of the round bokeh ball so that you do not see any edge from it.
* Towards the edge & corners of the frame the wide open bokeh ball can be mechanically clipped on one side towards a football shape (as in the non-DS version) and this particular edge is not smoothed as much as that edge is getting closer to the center of the bokeh ball.
* Since you have less light wide open, you increase your exposure (or ISO) to compensate. This added exposure increases the light in the center to middle of the bokeh ball. You can actually see this when comparing a DS version with a non-DS version. While the in-focus image is the same brightness, look at a big bokeh ball. The DS version center will be brighter (2x because you doubled the center light exposure) than the non-DS version. This means that your in-focus image sharpness and resolution will actually *increase* slightly with the DS version and it will slightly *increase* the in-focus DOF so that in the (non-DS) times when one eye is in focus and the other eye just out of focus, the DS version might have both eyes in focus. Also, it may also slightly *increase* the AF speed (as the video did notice).
* As you stop down the lens a stop or two, all the light to the sensor is coming in near the center of the iris where there is no DS darkening and thus the image is identical in all ways to the non-DS version.
* I would also like to mention that if you print out the resolution charts (available at https://www.opticallimits.com/canon_eos_ff/1095-canonrf85f12) of the non-DS lens you will find it is possibly the sharpest lens (resolution-wise) in the entire RF lineup. You can still use either version in a stopped down manner when you want more DOF for landscapes (for example) and it will give you sharper images than any other lens on your R... camera - a very good thing! What that means is that if they had printed the chart for the DS version then the resolution numbers would actually *increase* at f1.2 and close to it where the DS has an effect.
* Yes, wide open you need more exposure (or higher ISO) but I still find it quite fast enough. If you don't like it then get the non-DS version.
* Yes, it's $300 or so more expensive, but at that level you have to be very comfortable in your finances to afford the non-DS version so another $300 isn't that much to worry about (IMHO). After all, if $300 matters on a $2700 lens, then I suggest you buy the RF 85mm f2 and save a whole lot more money than $300.

Enjoy!
 
  • Like
Reactions: 1 user
Upvote 0
Not a huge fan of the DS version personally. The bokeh looks like a gausian blur filter has been applied and as such looks a little computer generated. I don't mind the way a lens 'naturally' behaves with bokeh rendering as that's what I and many others are already used to seeing. It just seems less artificial for some reason. I also want all of the light gathering potential of the f1.2 aperture.
 
  • Like
Reactions: 3 users
Upvote 0

usern4cr

R5
CR Pro
Sep 2, 2018
1,376
2,308
Kentucky, USA
Since the 85 f1.2 DS *does* sell some lenses (not near as many as the non-DS one, though) I think it would make sense for Canon to come out with a 135mm f1.9(call it f1.8 or f2 if you must) with non-DS and DS version. This is perfect for the often used "portrait" 135mm view while keeping the filter size to be the same (if f1.9) or close as that for the 85mm f1.2.

I think I'd be willing to buy a 135mm f1.9 DS lens, with more buying the 135mm f1.9 non-DS one. So we both win! :D
 
Upvote 0

YuengLinger

Print the ones you love.
CR Pro
Dec 20, 2012
3,751
2,269
USA
Based on the samples in the review, which is an excellent one, I prefer the original, non-DS bokeh. What can I say? I like a bit of cat's eye effect, and slightly more distinct out-of-focus points of light. Otherwise, the non-DS just produces brilliantly smooth backgrounds and superb separation. I strongly disagree with a few who say that it is "too sharp" or "clinical" at all.
 
Upvote 0

usern4cr

R5
CR Pro
Sep 2, 2018
1,376
2,308
Kentucky, USA
For a long time I've thought that the DS version would be best since you don't see any edges of the bokeh balls (excluding those edges caused by the edge of the front element causing a cats eye on one side).

We do know that the DS reduces the light by around 2x and therefore requires about 2x more exposure (which I don't mind that much). But that results in the center of the bokeh balls (which aren't darkened) being 2x brighter. And since the bright bokeh balls are almost always the brightest parts of the image inherently, they often turn "white-ish" due to overexposure and thus lose their color, as well as being 2x brighter in the center as well as having a smaller diameter. That bothers me enough to change my mind, and thus I would now prefer the original non-DS version.

I've been waiting to see if Canon comes out with a 85mm f1.4 L which I think I'd prefer over the f1.2 L only because it would be smaller & lighter. But I doubt that they will do so, and certainly not within the next year. So I'm leaning on getting the 85mm f1.2 L after all.
 
Upvote 0

koenkooi

CR Pro
Feb 25, 2015
3,574
4,110
The Netherlands
For a long time I've thought that the DS version would be best since you don't see any edges of the bokeh balls (excluding those edges caused by the edge of the front element causing a cats eye on one side).

We do know that the DS reduces the light by around 2x and therefore requires about 2x more exposure (which I don't mind that much). But that results in the center of the bokeh balls (which aren't darkened) being 2x brighter. And since the bright bokeh balls are almost always the brightest parts of the image inherently, they often turn "white-ish" due to overexposure and thus lose their color, as well as being 2x brighter in the center as well as having a smaller diameter. That bothers me enough to change my mind, and thus I would now prefer the original non-DS version.[..]

The DS coating also acts like aperture, so the DoF is slightly larger compared to the non-DS at the same f-stop displayed on the screen.
 
  • Like
Reactions: 1 user
Upvote 0

usern4cr

R5
CR Pro
Sep 2, 2018
1,376
2,308
Kentucky, USA
The DS coating also acts like aperture, so the DoF is slightly larger compared to the non-DS at the same f-stop displayed on the screen.
Yes, that is true, and is the reason that the bokeh balls are smaller when wide open. It's a benefit if you want to get both eyes in focus compared to the non-DS. To make the best decision, I'd have to try both at the same time for a while, and I don't think that's going to be possible. I do wish that they had made the DS version allow more light in at the edges so that it only softened the edges somewhat so that they weren't so noticeable while only removing only 20-ish% of the light instead of 50-ish%.
 
Upvote 0

BeenThere

CR Pro
Sep 4, 2012
1,242
672
Eastern Shore
Yes, that is true, and is the reason that the bokeh balls are smaller when wide open. It's a benefit if you want to get both eyes in focus compared to the non-DS. To make the best decision, I'd have to try both at the same time for a while, and I don't think that's going to be possible. I do wish that they had made the DS version allow more light in at the edges so that it only softened the edges somewhat so that they weren't so noticeable while only removing only 20-ish% of the light instead of 50-ish%.
If you prefer more DoF with the non-DS version, it’s a simple matter to reduce the aperture.
 
Upvote 0

usern4cr

R5
CR Pro
Sep 2, 2018
1,376
2,308
Kentucky, USA
If you prefer more DoF with the non-DS version, it’s a simple matter to reduce the aperture.
Well, of course that's true, but that's like saying that if you wanted both eyes in focus, it's a simple matter to reduce the aperture and just buy the RF 85mm f2 instead.

The whole point of the DS version is if you want to have smooth bokeh edges, while still having the option to have both eyes a bit more in focus.

As I've mentioned just above, I'd love to have the smooth edges of the DS version, but the penalty of the white-ish bokeh ball centers and reduced bokeh diameter are too much of a penalty to me, and so I think that I'd prefer the non-DS version.
 
Upvote 0

usern4cr

R5
CR Pro
Sep 2, 2018
1,376
2,308
Kentucky, USA
Well, I bit the bullet and just got the RF 85mm f1.2L delivered! :)
I thought I'd post a single test photo I took today with my R5, and show how I adjusted it and what resolution the 85 1.2L can get out of the shadows.


This is the original 14.7MB (compressed) CRaw photo at f1.2, iso 100, 1/8000", mechanical 1st & 2nd shutter, with no post modifications and no de-noising, resized to 2.5K pixels tall::

A01_2378_M_99%.jpg



I then adjusted and cropped it to taste in DXO Photolab4, with "deep prime" de-noising:

A01_2378_1_99%.jpg



I then took a square 1:1 640x640 pixel crop of the untouched original around the eye, which is here:

A01_2378_2_eye99%.jpg



I then took the same 1:1 640x640 crop of the eye from the version I adjusted in PL4, which is here:

A01_2378_3_eye99%.jpg



If you look at the above 640x640, you can see a pretty clear reflection in the eye of my house on the right, my neighbors' house on the left, and between them is me, hunched over holding the camera in portrait mode just above the ground to take the photo. Since this was with a compressed raw file, I wonder if a regular raw file would get even more out of the shadows?
 
Upvote 0
Well, I bit the bullet and just got the RF 85mm f1.2L delivered! :)
I thought I'd post a single test photo I took today with my R5, and show how I adjusted it and what resolution the 85 1.2L can get out of the shadows.


This is the original 14.7MB (compressed) CRaw photo at f1.2, iso 100, 1/8000", mechanical 1st & 2nd shutter, with no post modifications and no de-noising, resized to 2.5K pixels tall::

A01_2378_M_99%.jpg



I then adjusted and cropped it to taste in DXO Photolab4, with "deep prime" de-noising:

A01_2378_1_99%.jpg



I then took a square 1:1 640x640 pixel crop of the untouched original around the eye, which is here:

A01_2378_2_eye99%.jpg



I then took the same 1:1 640x640 crop of the eye from the version I adjusted in PL4, which is here:

A01_2378_3_eye99%.jpg



If you look at the above 640x640, you can see a pretty clear reflection in the eye of my house on the right, my neighbors' house on the left, and between them is me, hunched over holding the camera in portrait mode just above the ground to take the photo. Since this was with a compressed raw file, I wonder if a regular raw file would get even more out of the shadows?
Nice lighting and composition there. I would say you aren’t really pushing those shadows enough to notice a difference between cRAW and RAW. Looks like the cRAW responded very well to the adjustments
 
  • Like
Reactions: 1 user
Upvote 0

usern4cr

R5
CR Pro
Sep 2, 2018
1,376
2,308
Kentucky, USA
Nice lighting and composition there. I would say you aren’t really pushing those shadows enough to notice a difference between cRAW and RAW. Looks like the cRAW responded very well to the adjustments
Thanks, Chris.

My goal was to get the best looking photo I could, so I didn't push the shadows any more or they would have looked strange. Only later when I zoomed into the face to check the fine details did I notice my own reflection in the eye. I could have raised the shadows in the eye crop further as a separate virtual image to better show the reflection, but I thought it was better to show everyone exactly how it looks in the finished version.

I started taking R5 photos in RAW & using DXO PL4, but after taking so many photos at ~50MB each you really start to eat up your free memory. So eventually I tried cRAW and I'm becoming very happy with how DXO PL4 can make it look, as it's hard for me to tell the difference between the two.
 
Upvote 0