EMWL Critical Sharpness

Hi all!

I've been working with my new EMWL and I love it generally. I’ve been having sharpness issues with the EMWL, though, even when I am shooting with just strobe light, 1/200 at f20 and the 100 degree objective. I haven’t been able to figure out what the cause is. It seems to happen even in good viz (40'+), so I don’t think it’s a problem with muck in between the wet lenses- but i've been diving it in viz from 40' to 15' so it varies.

One thing that might  be important is that I am using a Saga port adapter specifically for the Nauticam EMWL on my Aquatica housing (yes yes I know). Saga has been producing a number of these and they haven't had any complaints, so probably not the issue. It focuses just fine with this combination so it seems like this shouldn’t be the problem?

It's possible it's a focusing problem in general, although even when I can't see focus is spot-on, the critical sharpness just isn't there.

I'm including 2 EMWL images, shot at 1/100, f16 and 2 Fisheye images shot at f/5.6. The fisheye is even on a 1.5x teleconverter!

Anyone have any ideas? Cheers, Kiliii

I’ve been using the EMWL quite a bit recently, though my experience is only with the 130 objective and without the middle relay lens. And I’m shooting on a Sony A7R II with 90mm macro.

First of all, even at its best I don’t think the EMWL can deliver as sharp images as more normal setups. After all, the image is going through a lot of extra glass and getting stretched quite a bit. I am having to apply aggressive sharpening in post to get somewhat acceptable sharpness.

Also, with my setup at least there is a significant tendency for backfocus error with the EMWL that I’m always fighting against. I’m having to be very careful with my focus point positioning so that it isn’t close to overlapping onto the background as well as using continuous focus and giving the camera a second or two to really lock in on the subject. Even so I don’t have a great success rate, so I have to take multiple shots to increase my odds of getting an in-focus shot. But I'm guessing other camera/lens combinations will perform better in this regard. 

Lastly, you might be losing sharpness due to diffraction by shooting at such a small aperture. I did some bathtub testing, and there seemed to be a slight drop in sharpness already at f/11, so I have done all my recent shooting at f/8 (though I need to do a more thorough test on this soon). I suggest you take some controlled test shots of the same subject throughout the aperture range so you know what setting gives you the sharpest results.

I would add that there also seems to be low contrast in the EMWL images - they look rather flat and could benefit from black/white point adjustments which helps them look punchier for want of a better term.  That could be due to particles in the water or the shear number of elements plus the the 4 water/glass interfaces.  How much of a crop are the images?

I would agree that you are well into diffraction with f16-20 - believe you're on a Z7 and the pixels are quite small so it's into diffraction on the pixel level at about f4.5 - but that's only pixel peeping at 100%. I would think the f11-13 range should be fine as those apertures are quite commonly used in UW imaging.

On diffraction it's important to remember that it's only one factor and getting enough depth of field is probably more important and a some of the softness from diffraction can be offset by judicious sharpening.  Doing some capture sharpening like an edge sharpening routine may help as well.

Hello Gentlemen,

We have received some feedback regarding 'focus error' or 'lack of sharpness' from the field around 2 months ago, and have since tested and analyzed this issue extensively (repeatedly with all cameras and lenses that we listed to be compatible with it).

First of all, these problems only happen when taking still pictures.  We have overlooked this issue because we had mostly been taking videos with the system during the limited field testing, and unfortunately the camera and lens we picked for taking still pictures exhibited less issues.  Each every camera and lens (from Canon, Nikon, Sony, Olympus, Panasonic and Sigma) we tested performed very differently.  And due to the pandemic we didn't have actual field tests ourselves for 2+ years.

The EMWL makes use of the macro lens behind it to focus, and the problem relates to the 'focus shift with aperture' nature of the lenses (google 'focus shift' and you can find huge amount of explanation about it and realize how much people suffers from it especially those participate discussions on the dpreview forums).  As the EMWL expands the angle of view to a great extent, it only requires a very short focus movement of the macro lens to focus the EMWL from infinity to extreme close.  As a result, a slight focus shift from the lens results in a huge focus shift of the resultant scene/image.  (in one case the focus shifted backward 2 cm from a target 8 cm from the lens; another camera/lens had the focus shifted forward)

The actual issue of focus shift happens when the camera opens the aperture up for focus, and stops down for exposure.

We were puzzled for a total of 5 weeks testing daily on the bench and haven't been able to conclude a certain answer due to the 2 following phenomena using normal technics that were recommended to overcome focus shift:

1. On DSLRs, the issue remained after we switched to live view and focused manually.  (because the aperture was fully open during live view, and stopped down shortly before exposure.
2. On Sony mirrorless which has depth of field preview, same issue when a target aperture was chosen and AF used.  (because the camera opened up the aperture momentary for AF and immediately closed it for exposure)

So the solution for achieving accurate focus when using the EMWL with any camera (for taking still pictures) is:  1. Use live view (which is default for mirrorless cameras);  2. Choose the target aperture and keep it stopped down all the time;  3. Use MF maybe with the help of peaking.

**When you take videos the camera already use live view and closed aperture so this 'focus shift' issue doesn't exist.**

As the depth of field of both the 100° and 130° objectives are actually quite extensive, you don't need small apertures and 100% accurate focus to get sharp and good contrast images.

Sorry that it took me that long and numerous sleepless nights until we finally conclude the issue.  Enjoy shooting !!

Edward 

1 hour ago, Edward Lai said:

Hello Gentlemen,

We have received some feedback regarding 'focus error' or 'lack of sharpness' from the field around 2 months ago, and have since tested and analyzed this issue extensively (repeatedly with all cameras and lenses that we listed to be compatible with it).

First of all, these problems only happen when taking still pictures.  We have overlooked this issue because we had mostly been taking videos with the system during the limited field testing, and unfortunately the camera and lens we picked for taking still pictures exhibited less issues.  Each every camera and lens (from Canon, Nikon, Sony, Olympus, Panasonic and Sigma) we tested performed very differently.  And due to the pandemic we didn't have actual field tests ourselves for 2+ years.

The EMWL makes use of the macro lens behind it to focus, and the problem relates to the 'focus shift with aperture' nature of the lenses (google 'focus shift' and you can find huge amount of explanation about it and realize how much people suffers from it especially those participate discussions on the dpreview forums).  As the EMWL expands the angle of view to a great extent, it only requires a very short focus movement of the macro lens to focus the EMWL from infinity to extreme close.  As a result, a slight focus shift from the lens results in a huge focus shift of the resultant scene/image.  (in one case the focus shifted backward 2 cm from a target 8 cm from the lens; another camera/lens had the focus shifted forward)

The actual issue of focus shift happens when the camera opens the aperture up for focus, and stops down for exposure.

We were puzzled for a total of 5 weeks testing daily on the bench and haven't been able to conclude a certain answer due to the 2 following phenomena using normal technics that were recommended to overcome focus shift:

1. On DSLRs, the issue remained after we switched to live view and focused manually.  (because the aperture was fully open during live view, and stopped down shortly before exposure.
2. On Sony mirrorless which has depth of field preview, same issue when a target aperture was chosen and AF used.  (because the camera opened up the aperture momentary for AF and immediately closed it for exposure)

So the solution for achieving accurate focus when using the EMWL with any camera (for taking still pictures) is:  1. Use live view (which is default for mirrorless cameras);  2. Choose the target aperture and keep it stopped down all the time;  3. Use MF maybe with the help of peaking.

**When you take videos the camera already use live view and closed aperture so this 'focus shift' issue doesn't exist.**

As the depth of field of both the 100° and 130° objectives are actually quite extensive, you don't need small apertures and 100% accurate focus to get sharp and good contrast images.

Sorry that it took me that long and numerous sleepless nights until we finally conclude the issue.  Enjoy shooting !!

Edward 

I found this information with a little searching, it seems that the Z7 stops down to f5.6 when focusing which might help avoid this issue?

https://www.dpreview.com/forums/thread/4331230

There may be ways with various mirrorless cameras to confirm focus when stopped down which would be worthwhile investigating to get around the focus shift.

@Edward Lai Thanks for the insight. Is there any chance you might produce a wider objective for it in the future (such as 180)?

Interesting discussion! The focus shift you describe Edward seems a bit different from the usual focus shift due to spherical aberration that is quite common with fast lenses like those with an f/1.4 maximum aperture.

I have been using mine with the 105VR Nikkor with decent results shooting a Nikon D4S camera using automatic AF point selection in AF-C mode. I am shooting using a remote control plus intervalometer with a pole cam rig. I am about 2 meters away from the subject on more or less dry ground so only have a marginal view of what I am shooting. As well I am using a SAGA port with 67mm threads like the OP and like Isaac I am shooting without the middle section so just the 100 degree objective and focusing unit.

The pix that are attached are from one of my more successful shoots all done at f/20. Even at f/20 the DOF is very very shallow. These fish are Sockeye Salmon fry about 3 to 4 cm long. In particular the last shot with the salmon at the right edge of the frame - the middle of the fish is in focus but both the head and tail are slightly OOF. The camera was on the streambed so the framing remains constant.

Note quite satisfied with what I just posted I jumped back into LR and did this blow-up of the fish mentioned at the end. Not too bad since the fin rays can be seen and even counted though a bit challenging with this background (the anal fin ray count is an important diagnostic characteristic for identifying salmonids).

Hi everyone, thank you for the excellent and informative responses.

Re: Isaac, I have definitely been experiencing back-focusing issues and I am sure it's what Edward has just informed us about as well. It's interesting that it affects the whole variety of setups with focus shift due to aperture. My macro lens, the new Nikon Z 105mm is supposed to have virtually no focus shift due to aperture changes, but then again maybe the EMWL magnifies even minor errors into significant ones. 

I don't think what I'm seeing is problems due to diffraction really. Almost all of the problem is due to missed focus, after reviewing a hundred of my EMWL images. Very very occasionally I hit focus dead on, and I can see now that it was by accident, as when the lens back-focused and then I drifted into better position by accident.

Re: Thomas, nice clean images! I think you're getting good results but I do think that it's because you're helped out by the f22 DOF. But then again perhaps your particular setup is less prone to this focus shift issue? I would think it would be affected even more!

I've been hunting all day for possible solutions for my Nikon Z setup, or perhaps with a Canon setup RF setup (I like the EMWL enough in theory to consider particular camera/lens combos), but at the moment the only mirrorless camera that seems to provide the option of autofocusing at whatever aperture you set, especially small apertures (f8+) is the Fuji GFX. My Nikons actually all stop down to f5.6 and focus there, so in theory I could use the EMWL at f5.6 and it would focus correctly, but really f5.6 is quite limiting in terms of depth of field, especially considering what I use the EMWL for- macro distances. f16 would be more like it.

Re: Edward's 3 workarounds, only one is actually a workaround- that's using manual focus and focus peaking. And of course it's doable, just a huge bummer given that the promise of using the EMWL to a large degree was the ability to use autofocus. If the focus shift is as sensitive to small changes as it sounds, manual focus might be very difficult as well. 

The first workaround, (use live view) doesn't really help the problem on mirrorless cameras, and the second (ask the camera to focus at the final aperture) is not actually possible on Nikon, Canon or Sony at the moment.

I can really only think of one additional possible workaround- with the lens set at the aperture you would use, and focusing at around the distance you would use, you can preset AF fine tune to correct for the focus shift problem of a particular lens/camera combo, through controlled testing topside (with the EMWL immersed in water through the neoprene sock trick). Even if that works, it would be very restrictive, as it means that your aperture will be locked in at whatever you set the AF fine tune for (although maybe focus shift isn't that much from f11 to f16?), and also your focus distance has to be the same as test conditions as well. That seems like a huge drawback.

Any more thoughts? I sure wish we could find a way to make this thing do what it was supposed to. As a professional I think unless I can find a way to make AF more reliable I can't actually use it for work.

I just happened some more interesting information about focus shift due to aperture- apparently it is a phenomenon that happens only with phase detect autofocus. Nikon Z's autofocus in AF-C mode (all I've used so far) is Phase detect, but in AF-S mode it uses a hybrid mode ending with contrast detect. It also has a contrast detect only mode.

I'll give the contrast modes a test and see what I can come up with. AF-S is not ideal in a moving underwater situation but it is vastly preferable to manual AF! Hope it works!

32 minutes ago, kiliii said:

I just happened some more interesting information about focus shift due to aperture- apparently it is a phenomenon that happens only with phase detect autofocus. Nikon Z's autofocus in AF-C mode (all I've used so far) is Phase detect, but in AF-S mode it uses a hybrid mode ending with contrast detect. It also has a contrast detect only mode.

I'll give the contrast modes a test and see what I can come up with. AF-S is not ideal in a moving underwater situation but it is vastly preferable to manual AF! Hope it works!

 

The contrast detect would only help if the lens stopped down to shooting aperture I would think.  All the descriptions say it is due to the change in focus point when the lens stops down from wide open to shooting aperture and whether focus is detected by the phase detect system or by contrast detect, with the Z7 they are both on the sensor so it should make no difference  unless one method stops the lens down. 

You mention the Nikon lenses stop down to f5.6 which should be better than focusing wide open but the link I found above mentions it only occurs when D8 (Apply settings to live view) is turned OFF.  You could test for focus shift in a pool or a tub once you confirm the camera is stopping the lens down when it attempts to AF.   So shoot at f5.6 -8 -11- 16 without changing focus.

I wonder if AF fine tine may help you get closer to correct focus?  It's normally only a very small adjustment though, but may be worth trying as well.

1 hour ago, kiliii said:

I just happened some more interesting information about focus shift due to aperture- apparently it is a phenomenon that happens only with phase detect autofocus. Nikon Z's autofocus in AF-C mode (all I've used so far) is Phase detect, but in AF-S mode it uses a hybrid mode ending with contrast detect. It also has a contrast detect only mode.

I'll give the contrast modes a test and see what I can come up with. AF-S is not ideal in a moving underwater situation but it is vastly preferable to manual AF! Hope it works!

I hope that method works for you. However, as Chris said, I think the only way it will is if the camera focuses with the aperture stopped down in contrast detect mode. You can figure out whether or not the camera is focusing with the aperture stopped down by looking into the front of the lens and watching the aperture blades during focusing and during the exposure. 

After doing some testing with my A7R II and 90mm I figured out why the method I stumbled upon of using AF-C and letting it focus for a second or two has made the EMWL usable for me. In AF-S mode the A7R II focuses with the aperture wide open and only stops down to the shooting aperture at the moment of the shot, which as Edward has explained results in a focus shift and backfocus error. Likewise in AF-C mode the A7R II initially focuses with the aperture wide open, however once it acquires initial focus it stops down to the shooting aperture and continues to adjust focus until the shot is taken. This means it will focus at the shooting aperture and correct for the focus shift as long as I use AF-C and give it a little extra time to keep focusing after the initial focus acquisition. While I had already figured out that this method gives me better results through trial and error, it’s great to now understand why it works. It’s not perfect though since the A7R II with 90mm constantly hunts back and forth in AF-C mode when stopped down past f/8 (even without the EMWL), which means my setup cannot accurately focus the EMWL past f/8 (unless I have time to focus at f/8 and then stop down further and take the shot without focusing again). I’m very curious if newer Sony bodies also have this f/8 limitation with the 90mm in AF-C mode? Anyway, the above information will be helpful for other Sony EMWL users but obviously not directly applicable for Kiliii and Tom. But it is more confirmation that if you’re experiencing consistent focus error with the EMWL, figuring out a way to get the camera to focus at the shooting aperture will likely give much better results.

10 minutes ago, Isaac Szabo said:

I hope that method works for you. However, as Chris said, I think the only way it will is if the camera focuses with the aperture stopped down in contrast detect mode. You can figure out whether or not the camera is focusing with the aperture stopped down by looking into the front of the lens and watching the aperture blades during focusing and during the exposure. 

After doing some testing with my A7R II and 90mm I figured out why the method I stumbled upon of using AF-C and letting it focus for a second or two has made the EMWL usable for me. In AF-S mode the A7R II focuses with the aperture wide open and only stops down to the shooting aperture at the moment of the shot, which as Edward has explained results in a focus shift and backfocus error. Likewise in AF-C mode the A7R II initially focuses with the aperture wide open, however once it acquires initial focus it stops down to the shooting aperture and continues to adjust focus until the shot is taken. This means it will focus at the shooting aperture and correct for the focus shift as long as I use AF-C and give it a little extra time to keep focusing after the initial focus acquisition. While I had already figured out that this method gives me better results through trial and error, it’s great to now understand why it works. It’s not perfect though since the A7R II with 90mm constantly hunts back and forth in AF-C mode when stopped down past f/8 (even without the EMWL), which means my setup cannot accurately focus the EMWL past f/8 (unless I have time to focus at f/8 and then stop down further and take the shot without focusing again). I’m very curious if newer Sony bodies also have this f/8 limitation with the 90mm in AF-C mode? Anyway, the above information will be helpful for other Sony EMWL users but obviously not directly applicable for Kiliii and Tom. But it is more confirmation that if you’re experiencing consistent focus error with the EMWL, figuring out a way to get the camera to focus at the shooting aperture will likely give much better results.

What would be most interesting I think would be measure how much focus shifts from f8 to smaller apertures.

12 minutes ago, ChrisRoss said:

What would be most interesting I think would be measure how much focus shifts from f8 to smaller apertures.

Yeah you'd expect focusing at f/8 might work OK for shooting at f/11 but maybe not for say f/22. After a little research I think some of the newer Sony bodies like the A7R IV can keep focusing in AF-C at f/11 and the very latest ones like the A1 can keep focusing in AF-C at f/22. If so the A1 might be a good match for the EMWL. 

To follow up on this issue, while I’m very thankful to Edward for sharing this information with us, if you evaluate the workarounds he listed, none of them will actually be of any help for some cameras/users. So unfortunately we have a very expensive product that is nonfunctional for some users and only partially functional for others. A year ago when I first purchased and tried out the EMWL and discovered that all my photos were significantly backfocused, I tried to return it to the store I bought it from, but they refused to take it back. Later on I tried to sell it here on Wetpixel at a discounted price, but I didn’t have any luck (partially because I privately alerted prospective buyers to the issues I experienced). Since I seemed to be stuck with it, I decided to try it again this year and stumbled upon the workaround that gives me partial functionality. However, there are still significant limitations, and I’m not sure whether or not I want to keep it long term. But especially for customers like Kiliii for whom there may be no workaround, it only makes sense to me that there should be a way to return it for a full refund.

1 hour ago, Isaac Szabo said:

To follow up on this issue, while I’m very thankful to Edward for sharing this information with us, if you evaluate the workarounds he listed, none of them will actually be of any help for some cameras/users. So unfortunately we have a very expensive product that is nonfunctional for some users and only partially functional for others. A year ago when I first purchased and tried out the EMWL and discovered that all my photos were significantly backfocused, I tried to return it to the store I bought it from, but they refused to take it back. Later on I tried to sell it here on Wetpixel at a discounted price, but I didn’t have any luck (partially because I privately alerted prospective buyers to the issues I experienced). Since I seemed to be stuck with it, I decided to try it again this year and stumbled upon the workaround that gives me partial functionality. However, there are still significant limitations, and I’m not sure whether or not I want to keep it long term. But especially for customers like Kiliii for which there may be no workaround, it only makes sense to me that there should be a way to return it for a full refund.

Yes, unless you can find someway to compensate for the shift somewhat automatically it does seem like it is going to be difficult to use.  I still think it would be worthwhile trying out to see if AF micro adjust could shift focus enough to bring it into critical focus. 

53 minutes ago, ChrisRoss said:

Yes, unless you can find someway to compensate for the shift somewhat automatically it does seem like it is going to be difficult to use.  I still think it would be worthwhile trying out to see if AF micro adjust could shift focus enough to bring it into critical focus. 

With Sony mirrorless cameras AF micro adjust is not available except when using Sony's A mount lens adapter, so that's not an option for me or other Sony users. I don't know whether or not other mirrorless cameras from Nikon, Canon, etc have it for native lenses. 

3 hours ago, Isaac Szabo said:

With Sony mirrorless cameras AF micro adjust is not available except when using Sony's A mount lens adapter, so that's not an option for me or other Sony users. I don't know whether or not other mirrorless cameras from Nikon, Canon, etc have it for native lenses. 

I know Nikon and Olympus does not sure about others.

1 hour ago, ChrisRoss said:

I know Nikon and Olympus does not sure about others.

Well then I agree it would be a worthwhile thing for Kiliii to try. 

I have some thoughts to add to this (apologies for the long post).

First I think that everyone would agree that the EMWL lens does not have the ultimate image quality of a setup like the WACP-1. But I find the image quality is completely acceptable for my needs - especially with the unusual perspectives possible make the images so compelling. I’ve not done much EMWL suitable diving recently, but found it very productive on the three dives in the Maldives I took it in for last week. Here is a photo of young hawksbill feeding, with crop on the eye (from last week, not finally processed yet - but I find the detail and image quality totally useable - as do my clients). Taken with Nikon D5, 1/50th f/16, ISO 500

I think people’s reaction to the lenses performance depends a bit on where you are coming from. If, like me (and Adam, for example) you have used other probe/relay/bugeye lenses in the past underwater, then the EMWL represents a massive leap forward in terms of image quality. Previous lenses were very borderline for publishing, this is totally publishable. Whereas if you are coming to the EMWL-1 from a WACP-1, for example, it is a step down in IQ and an expensive one at that.

For me, WAM images need to have as much depth of field as possible. The whole point of shooting wide angle it to have the scene be part of the image and this means that the lens needs to be stopped down as much as possible. Two consequences: I try and shoot the lens in bright conditions so that I am not fighting too slow a shutter speed (having a macro lens with a stabiliser is a bonus) and, yes, shooting at f/18-f/22 all the time must be degrading fine detail also because of diffraction. 

I feel that the EMWL would be enhanced by a more extreme optic too. I agree with Isaac about something in the 150-180˚ range (I’d vote for the middle of that range). I passed this on to Nauticam pre-Covid when testing the lens, but of course it hasn’t been a normal time since then for anyone, so I don’t know if they have looked at the idea. I think the conundrum for Nauticam is that a lens in the range 150-180˚ would have more optical compromises, especially if it has a small front element, which is what we’d want. And if you release a lens with optical compromises some people moan. While others of us moan about not having an extreme optic. It is a loose/loose for them in some ways.

In my opinion the Nauticam MWL lens got caught in this space. It grew bigger to give it better image quality and this limits its usability as a WAM tool. Fortunately it has found a great niche for blackwater diving. 

Regarding focus - I have only used this lens with Nikon SLRs. It is definitely a lens where focus is critical and you need to consider this when shooting. When I shoot normal wide angle I just use AUTO Area AF on my Nikons - basically just trusting the camera. But not so with the EMWL, I have found the Grouped AF (diamond) is the best mode with the lens for me. I still discard a few frames for focus, but they are the exceptions, not the rule. For the turtle I moved this to the left of the frame as I shot it feeding. This mode is like a large single point AF - and works very well. I don’t have experience with other cameras and other focus modes with this lens, I am afraid.

I do find that my lens has a strong yellow colour cast. I guess from all the glass and the images need re-white balancing in post. 

I dive the lens with three lens caddies on my strobe arms - so I can easily change optics during the dive. In all my dives with this lens I have only once taken the whole lens off and shot a standard macro shot. As it is so much of a pain totally changing all the camera settings and the strobe positions. But I do like having the flexibility for that once in a lifetime subject. 

Alex