Jump to content


Photo
- - - - -

Rokinon 7.5mm MFT on EPL5 sweet spot and focus


  • Please log in to reply
15 replies to this topic

#1 jhelmuth

jhelmuth

    Triggerfish

  • Member
  • Pip
  • 35 posts

Posted 26 July 2014 - 11:10 AM

Hey guys (and gals),

 

I've got an Olympus EPL5 (housed in the Oly PT-EP10) and have recently added a new port and lens combination (currently shoot mostly the Oly 9-18mm zoom in a Zen 100mm port) in the Athena 100mm dome for the Lumix 8mm fisheye along with the Rokinon 7.5mm fisheye. Since these are very close to the same focal length and physical size, I do believe this combination will work well.

 

So... since the 7.5mm fisheye is a fully manual lens, I want to set this up for nearly any circumstance where I would be shooting say from 1ft out.

 

From reading, I've decided that the aperture sweet spot should be right about f8.0 (I've read that it could be from a low of 5.6 to a high of 11.0, with most agreement on 6.8 to 8.0). So I've experimented just a bit (all on land so far) and found that if I set my focus ring infinity index mark to align with the aperture ring index mark (between 3.5 and 4.0) that I seem to get very good all around focus on shots from about 8" and out (again - this is @ f8.0).

 

My question to you is: [A] does this seem correct to you? [B] any suggestions? [C] How might this change when I place this in the Athena dome?

 

I'll plan to follow-up with a review (including whatever I learn) on this combination...

 

 

Thanks in advance,

 

Jim

 

PS - any tips on dual strobe positioning for fisheye?


Edited by jhelmuth, 26 July 2014 - 11:42 AM.


#2 Interceptor121

Interceptor121

    Great Hammerhead

  • Member
  • PipPipPipPip
  • 842 posts
  • Gender:Male
  • Location:Weybridge, UK

Posted 27 July 2014 - 01:10 AM

It is quite difficult to predict the behavior of a fisheye lens in a dome port.

If this was a rectilinear lens you could think of setting the aperture to achieve the hyperfocal distance of the virtual image in this case you would have something like f/22 and everything from 20cm to infinity would be in focus (one foot won't work as in a dome the virtual image is at 4r in your case 20cm)

However the lens is not rectilinear and therefore it is going to be a bit of trial and error to see what is the right compromise between sharpness of the lens, features of the sensor and how much depth of field you will achieve

For sure your focus point will range between the minimum distance and 20 cm so you won't have too much to play before you find out.


Check my video, pictures and blog

YouTube Channel

Flickr Sets

Blog


#3 FanchGadjo

FanchGadjo

    Moray Eel

  • Member
  • PipPip
  • 96 posts
  • Gender:Male

Posted 28 July 2014 - 04:36 AM

Not M4/3, but APS-C : I use a samyang 8mm on Canon 400D. I set focus to minimum (30cm)

before closing the housing, then I try to close the aperture as much as possible.



#4 jhelmuth

jhelmuth

    Triggerfish

  • Member
  • Pip
  • 35 posts

Posted 31 July 2014 - 05:21 PM

So it would seem to me that the virtual image DOF is "squished" - correct?   That would make the DOF more narrow? (IE - the actual distances the lens "sees" at the virtual focus point out to infinity)



#5 jhelmuth

jhelmuth

    Triggerfish

  • Member
  • Pip
  • 35 posts

Posted 31 July 2014 - 06:25 PM

Not M4/3, but APS-C : I use a samyang 8mm on Canon 400D. I set focus to minimum (30cm)

before closing the housing, then I try to close the aperture as much as possible.

If I am absorbing this correctly, I'm thinking the virtual image is at 10" from the Image Sensor (true focal distance).  I arrived at that by 50mm (radius) * 3.03 (factor) = 151.5mm (6.01") in front of the dome vertex (v) + ~100mm from (v) to Sensor (fp) = ~ 251.5mm (9.99").

 

So I set the focus ring index to 0.8 ft (~10") which also coincides with the infinty index DOF mark.  Took some test shots (bathtub) from ~ 6" out to about 2'+ (from the dome front)  and did the same for 0.5 ft (focus ring setting)...

 

results:    the images set at 0.8ft were all mostly in focus (short, med, and long - the long shot was slightly out-of-focus)

 

Comments?     Suggestions?

 

PS - Was able to use ISO 400, and 1/125th sec. shutter on 2x 2500 LED lights (reflected)


Edited by jhelmuth, 31 July 2014 - 06:34 PM.


#6 Interceptor121

Interceptor121

    Great Hammerhead

  • Member
  • PipPipPipPip
  • 842 posts
  • Gender:Male
  • Location:Weybridge, UK

Posted 31 July 2014 - 10:21 PM

The virtual image at infinity will be at less than 20cm as that depends on the dome size the refraction index and the nodal point of the camera (unknown) so you should be looking at less than that as a starter around 0.6 ft or less

How did the 0.5 ft image come out?


Check my video, pictures and blog

YouTube Channel

Flickr Sets

Blog


#7 jhelmuth

jhelmuth

    Triggerfish

  • Member
  • Pip
  • 35 posts

Posted 01 August 2014 - 05:25 AM

The virtual image at infinity will be at less than 20cm as that depends on the dome size the refraction index and the nodal point of the camera (unknown) so you should be looking at less than that as a starter around 0.6 ft or less

How did the 0.5 ft image come out?

 

Not nearly as sharp as the 0.8ft image (not even close).

 

I'm going to head to the pool sometime within the next few days and take more time to carefully measure exposures at various distances.  I'm not too concerned about having the infinite distance sharp (would be nice to have - but not a deal breaker).  I DO want the image to be sharp from as close as 6" (1.5cm) to about 2-3' (0.7-1.0m) from the front of the dome. These are the anticipated distances where I will likely "shoot" my subjects (at least that is what I am anticipating)  


Edited by jhelmuth, 01 August 2014 - 05:32 AM.


#8 Interceptor121

Interceptor121

    Great Hammerhead

  • Member
  • PipPipPipPip
  • 842 posts
  • Gender:Male
  • Location:Weybridge, UK

Posted 01 August 2014 - 05:26 AM

Makes no sense 10" is more than 4x the dome radius maybe the focus wheel is a bit approximate way to judge this

Check my video, pictures and blog

YouTube Channel

Flickr Sets

Blog


#9 jhelmuth

jhelmuth

    Triggerfish

  • Member
  • Pip
  • 35 posts

Posted 01 August 2014 - 06:17 AM

Hmmmm....   maybe I am not understanding all this, but I do believe that 4x gets me only to the front of the lens (that is where the base of the dome is - exactly at the lens front).  Unless I'm missing something, I also have to add in the distance from the front of the lens to the sensor.  If I add all that, I'm at ~ 10".

 

I'm not saying you're wrong. It's just what I've observed (and tried to learn and absorb) so far.

 

Again, I'll take time to do some more accurate and in-depth testing in the pool. I can lawn lay down a tape measure, and take more detailed tests.


Edited by jhelmuth, 01 August 2014 - 06:18 AM.


#10 Interceptor121

Interceptor121

    Great Hammerhead

  • Member
  • PipPipPipPip
  • 842 posts
  • Gender:Male
  • Location:Weybridge, UK

Posted 01 August 2014 - 07:56 AM

It is correct to add back the sensor distance however the camera should focus from its minimum focal distance to the max determined.l (4r-b)
So the fact the image at 0.5ft is not in focus in any scenario is strange. If you put something close to the dome it should focus at least there.

Check my video, pictures and blog

YouTube Channel

Flickr Sets

Blog


#11 jhelmuth

jhelmuth

    Triggerfish

  • Member
  • Pip
  • 35 posts

Posted 02 August 2014 - 06:20 PM

OK...   I still did not get to the pool (and I might get some time in tomorrow), but I did get some time in with better testing in the tub!  (just and old fiberglass bathtub).

 

Here is what I did...

 

> Used an Inon D2000 strobe as my "subject" since it had several qualities that I thought would help me decide on focus clarity.

 

> Used 2x BigBlue 2500 lumen LED torches as my ambient lighting (which afforded me an ISO of 400 and shutter of 1/100 - 1/125 at various apertures)

 

> Snipped several bit of old measuring tape (the type that is stiff and rolls into the case by itself) so I could measure distance, but also because the fine lines and graduations would give me a better objective to judge focus.

 

> Tested distance of 2", 4", 6", 8", 10", 12", 18", 24" (inches) all at f11 (I've been able to prove to myself that this seems to be the aperture sweet spot), but different focus settings (0.3ft, 0.4ft, 0.5ft, 0.65ft, 0.8ft)

 

> I wanted to test 2" and 4" shots, but I'm weighting them quite low as I do not intend to shoot 4" and in. Still, I took them as a reference to help me figure out what was going on

 

> As a final test, I used several objects set in background and foreground and tried f16 and f22 at the 0.65ft focus, as well as f11 (again) and f8.

 

Here is what I found...

 

* 0.3ft (minimum focus) and 0.4ft were all pretty much rubbish except at the 2" and 4" distances (actually, even the 4" distance was rubbish with the 0.3ft focus)

* 0.8ft looked best only beyond 6"

* The 2 contenders for best DOF seemed to fall to the 0.5ft and 0.65ft shots.

* As one might suspect, the background focus was slightly better at the 0.65ft shots, while the foreground focus was decidedly better with the 0.5ft shots

* f11 was a clear winner in the final test shots (I think this is the sweet spot of the lens overall)

* @ f11 and 4" - 6" - 8" - 10", 0.5ft & 0.65ft were virtually identical - I could not call a winner

* @f11 and 12" (and beyond) 0.65ft was more clearly the "winner" as the distances became longer.

 

Final Results...

 

So now I considered f11 focused at 0.65ft (which was actually as close to exactly between 0.5ft and 0.8ft as I could guesstimate) to be my decided "winner"

Now I wanted to "see" what f22, f16, an f8.0 would produce... and f11 was the clear winner.

 

Conclusion:  In my limited testing, I would choose f11 focused at as close to exactly between 0.5ft and 0.8ft on the focus ring as my correct hyperfocal setting for anything 4" and beyond (excepting 2" and 24"+ - in which case I'd choose 0.5ft for the 2" shot, and 0.8ft for the 24"+ shot)

 

sqm9gx.jpg

ISO 400 1/100th f11  (2" with focus set to 0.5ft)

 

34gpqir.jpg

ISO 400 1/100th f11  (6" with focus set to 0.65ft)

 

ve14yo.jpg

ISO 400 1/100th f11  (12" with focus set to 0.65ft)

 

aemteq.jpg

ISO 400 1/100th f11  composite

 

PS - I do get the tinyest mechanical vignetting with this Dome/lens combination (again - Athena 100mm fisheye dome for Oly PT-EPxx [designed fromLumix 8mm] with a Rokinon 7.5mm fisheye)



#12 Interceptor121

Interceptor121

    Great Hammerhead

  • Member
  • PipPipPipPip
  • 842 posts
  • Gender:Male
  • Location:Weybridge, UK

Posted 03 August 2014 - 12:43 AM

You will probably find once you have more depth and detail to shoot in the distance that as aperture reduce everything is in focus. You probably are looking at f20 to have super sharp corners however that is probably a bit hard on our ISO and may need to take a compromise somewhere

Check my video, pictures and blog

YouTube Channel

Flickr Sets

Blog


#13 jhelmuth

jhelmuth

    Triggerfish

  • Member
  • Pip
  • 35 posts

Posted 05 August 2014 - 08:34 PM

The more I think about the testing (which was all "fresh" water - not salt), the more i'm leaning toward setting the focus ring to 0.5ft, as the change in diffraction due to density ought to cause the focus to shift closer...


FYI...

 

Here is a sample (cropped) shot from the pool test...
1grpi.jpg

...and some pics of the lens/housing...
2q9ia1i.jpg

35ko5du.jpg



#14 Interceptor121

Interceptor121

    Great Hammerhead

  • Member
  • PipPipPipPip
  • 842 posts
  • Gender:Male
  • Location:Weybridge, UK

Posted 05 August 2014 - 11:49 PM

0.5 ft makes more sense since the start but really is a matter of testing out and finding your best setting for your subjects

You won't be shooting a bathtub in the ocean


Check my video, pictures and blog

YouTube Channel

Flickr Sets

Blog


#15 jhelmuth

jhelmuth

    Triggerfish

  • Member
  • Pip
  • 35 posts

Posted 08 August 2014 - 06:17 PM

Here is a sample of today out at Pickles. Bad storms last night had things stirred up a good bit, so water quality wasn't the best...
x6k60n.jpg
ISO 200 f11 1/125th dual YS-350 Pros with H/W Digital Adapter in conjunction with Oly EPL5 in Fill mode and RC on.



#16 jhelmuth

jhelmuth

    Triggerfish

  • Member
  • Pip
  • 35 posts

Posted 08 August 2014 - 06:28 PM

If I look at this at 100%, focus is not that good.

 

I shot this with focus fixed at 0.5ft  (or 0.15m).  Give the DOF scale, I'm now wondering if this wouldn't have been better at something closer to 0.8ft (0.24m)...

 

Strobes are a bit hot too...  so I think I should have maybe shot it at 160th sec. (or have set the camera's EV to maybe -0.3)?

 

PS - really kinda amazing (to me anyway) that the YS-350s work so well with this (thanks to the H/W Digital Adapter).  I only had 2 misfires all day and by setting the Olympus EPL5 to "use" RC (Remote Control ON) and the YS-350s to TTL, I had  a fair bit of success (at least after I set the ISO to 200 down from 400, and increased the Shutter to 1/125th)