Steve's Digicams Forums

Steve's Digicams Forums (https://forums.steves-digicams.com/)
-   Panasonic / Leica (https://forums.steves-digicams.com/panasonic-leica-29/)
-   -   [Recovered Thread: 82624] (https://forums.steves-digicams.com/panasonic-leica-29/%5Brecovered-thread-82624%5D-80616/)

shene Feb 10, 2006 11:23 PM

Started to think about my wide angle converter comparison recently. The first thing to do is to find scenes that can easily reveal the "problems" of each and every lens. This scene has to be somewhat weather independent so that I could reshoot some results at will. Yes, I could use lab equipment such as optical bench; but, the results may not be very reader friendly. As a result, shooting actual scenes is my favorite way. Please keep in mind that these scenes are not for artistic expression, they are chosen because they can reflect real shooting and at the same time can easily reveal problems of the lenses being tested. This is the first test run and may not represent what will follow on my comparison pages.



This scene is a section in the library of my university. I choose this scene because the subjects are virtually on a flat surface and full of little features. I did a white balance present and then set up my camera about 2 meters away. Since library illumination is not strong as usual, the aperture and shutter speed are f/5.6 (the middle one) and 1/8 sec. This is a slightly underexposed to make sure features on white surfaces can be seen for comparison. Unfortunately, I can only set the noise reduction to low rather than turning it off. Consequently, I do not know what the impact of NR on image sharpness is. Additionally, as usual, all shooting parameters are set to the lowest settings.



Converter lenses used include: Minolta ACW-100, Nikon WC-E80, Nikon WM-E80 and Olympus WCON-08B (the 0.8X group), and Olympus WCON-07, Panasonic LW55, Raynox HD6600pro-55 and Sony VCL DH0758 (the 0.7X group). The Nikon WM-E80, which is not discussed on my FZ-30 user guide page, is a single element and very light design (for Coolpix 8800). It does not have a usable thread for the FZ-30, and, hence it is hung over the FZ-30 lens. The WCON-07 uses a 55-62 step-up ring and a 62-55 thread converter to remove the center blur spot as described in my FZ-30 user guide.



The following shows the scene at 0.8X (by Minolta ACW-100) and 0.7X (by Panasonic LW55). In this post, I only looked at the center and lower-left corner marked by yellow rectangles.



http://www.cs.mtu.edu/~shene/FORUM/F...1/1-sample.jpg





*** 0.8X GROUP ***



Let us look at the 0.8X group first. The following has the four center crops from the 0.8X group. From these four crops, I would say they are very similar to each other with the WCON-08B slightly ahead of the Nikon WM-E80 and Minolta ACW-100. Surprisingly enough, the once well-known Nikon WC-E80 did not produce an image as crisp as the other three. However, I'd like to point out one important fact: like the TC-E15ED, the WC-E80 has a deeply recessed rear element. Whether this is an important factor is not known.



http://www.cs.mtu.edu/~shene/FORUM/F...-CENTER-08.jpg





The following is the lower-left crops. It should be clear that the Minolta ACW-100 and WCON-08B take the lead. IMO, the ACW-100 is slightly better; however, it also produced a very slight touch of purple fringing. Blow up the crop to some degree and the purple fringes can be seen along the lower left corners of high contrast areas. The Nikon WM-E80 seems slightly better than the Nikon WC-E80, and the former has some greenish fringes while the latter has some purple fringes. These fringes are so thin that may not appear in prints unless the print size is unusually large.





http://www.cs.mtu.edu/~shene/FORUM/F...ER-LEFT-08.jpg



*** 0.7X GROUP ***



The following shows the center crops of the four 0.7X wide angle lenses. None of these four is as good as the 0.8X lenses. The Raynox HD6600pro-55 seems to have lower contrast than the other three, and, as result, its result does not look as sharp as those of the other three. Of the other three, the Panasonic LW55 and Olympus WCON-07 are equally good, and both are very slightly better than the Sony DH0758.





http://www.cs.mtu.edu/~shene/FORUM/F...-CENTER-07.jpg



The following is the lower-left portion. Clearly, the Panasonic LW55 is better than the other three. The LW55 did not show chromatic aberration even in 150% enlarged size. The Olympus WCON-07 is also sharp, in fact, maybe as good as the Panasonic LW55; however, some purple fringes can be seen along the left edges of bright areas. The Sony VCL DH0758 is not as sharp and has clear greenish fringes. The Raynox HD6600pro-55 has a soft edge which is worse than that of the Sony DH0758. This softness also amplifies its greenish fringes.





http://www.cs.mtu.edu/~shene/FORUM/F...ER-LEFT-07.jpg



*** BARREL DISTORTION ***



The following image shows the barrel distortion of the four 0.8X lenses (Minolta ACW-100, Nikon WC-E80, Nikon WM-E80 and Olympus WCON-08B). It is clear that the WCON-08B is slightly better than the Minolta ACW-100, and both are better than the two Nikon's. Of the two Nikon's, the WC-E80 is better than the WM-E80. The latter has a very bad barrel distortion.



http://www.cs.mtu.edu/~shene/FORUM/F...-BARREL-08.jpg





The following one shows the 0.7X group (i.e., Olympus WCON-07, Panasonic LW55, Raynox HD6600pro-55, and Sony VCL DH0758). In this group, the Raynox is the winner. This is followed by the Olympus WCON-07 and Panasonic LW55, with the Panasonic very slightly worse. The Sony does not look good here.



http://www.cs.mtu.edu/~shene/FORUM/F...-BARREL-07.jpg





*** CHROMATIC ABERRATION (CA) ***



While in the first post CA was discussed, I intend to add one more piece of information here. The following is a 100% crop of the upper left corner of the 0.8X images. The focus is the high contrast area between the bookend and background. The contrast is not very high actually; however, the reflection makes the situation bad. Of these four, only the WCON-08B did not show fringing, and the Minolta ACW-100 is just slightly behind. Both Nikon's show purple and greenish fringes with the WM-E80 worse than the WC-E80. Note also that the Minolta ACW-100 is sharper than the WCON-08B! In fact, both the Minolta ACW-100 0.8X and ACT-100 1.5X perform respectfully at corners.



http://www.cs.mtu.edu/~shene/FORUM/F...ER-LEFT-CA.jpg





The following shows the upper-right corner of the 0.7X lenses. The Panasonic LW55 is very clean. The WCON-07 is the worst with strong purple fringes between high contrast areas. Although the Sony and Raynox both show some purple and greenish fringes, IMO the Sony is slightly better.



http://www.cs.mtu.edu/~shene/FORUM/F...IGHT-07-CA.jpg



*** LIGHT LOSS AT CENTER ***



By definition, adding a converter lens does not affect exposure value (i.e., EV) because the added converter lens changes the focal length and magnifies the entrance pupil by the same ratio. However, all glass elements will absorb and reflect some incoming light. High transmitting ratio glass elements and better multi-coating technology can reduce absorption and reflection, respectively. However, this light loss is still there.






To measure this light loss, I used the following setup. I open a large white window on a flat and good LCD monitor, and have a 1-degree spot meter pointing at the center of this large white area. Take five direct measures, mount a lens in front of the spot meter, and take five measures again. The difference between direct measure and the measure with a lens mounted is the light loss in EV. Since my spot meter is only accurate to 0.1 EV, the resolution of this difference is 0.1 EV.



Here are my measured results. All numbers are measured differences averaged and rounded/truncated to one digit after the decimal point.



*** 0.7X GROUP ***



Olympus WCON-07: 0.2 EV

Panasonic LW55: 0.1 EV

Raynox HD6600pro-55: 0.1 EV

Sony VCL DH0758: 0.1 EV



*** 0.8X GROUP ***



Minolta ACW-100: 0.1 EV

Nikon WC-E80: 0.1 EV

Nikon WM-E80: 0.1 EV

Olympus WCON-08B: 0.1 EV



Except for the Olympus WCON-07 that has a light loss of about 0.2 EV, all others have a light loss of around 0.1 EV. Since camera meters have 1/3 EV step, they perhaps could not recognize this 0.1 EV or 0.2 EV difference unless the original measure is at the border of 1/3 EV.



Note that I only measured the center portion. The difference between center and corner can vary from less than 1 EV to 2 EV depending on the lens. This can be visualized by shooting a uniform blue sky as shown on my FZ-30 user guide pages



*** SHORT SUMMARY ***



In this indoor short distance comparison, I believe the Panasonic LW55 is the winner of the four 0.7X lenses, and the Olympus WCON-08B and Minolta ACW-100 are the best two of the 0.8X group. However, the ACW-100 does show some very slight CA although not very significant; however, its off-center sharpness is the best.



Lastly, I have to point out that we should not use one set of images to judge the quality of a lens. This is the major reason that I always use multiple scenes each of which is designed to reveal one set of problems. The above scene works for lower contrast and short distance shooting. In the near future, I will add other scenes for other purposes (e.g., higher contrast, long shooting distance, moderate shooting distance, etc). It is very naïve to show a high contrast scene in which strong CA can be seen as a counter-example because shooting parameters are completely different. It is also very naïve to say one's image is more "real life-like" than the others in terms of lens testing. As long as a scene can reveal the problems (YES, lens testing is to tell the problems as well as good thingrather than hiding them), it is a good scene to use, and this is not a "real" vs. "not real" issue. If a "real" scene does not reveal the problem in a test/comparison, it is not fair because that test simply hides the weakness of a lens.



CK

http://www.cs.mtu.edu/~shene/DigiCam

Nikon Coolpix 950/990/995/2500/4500/5700, Panasonic FZ-10/FZ-30, and Canon A95 User Guides



msantos Feb 11, 2006 12:00 AM

Thanks for sharing this useful information:cool:

PeHa Feb 11, 2006 2:39 PM

Yes, thanks a lot for the evaluation, Shene! Though it confirms that I own a soft image lens... (the HD6600). :(

polar905 Feb 11, 2006 5:36 PM

Good shots and information, as I am in the market for a WA lens for the FZ30. I really want to go with a 0.7x. I am probably going to still hold off a bit till I see some test pics of the Kodak Xenar 0.7x lens on the FZ30, since after reading about it on the FZ4/5 I was really impressed.



Zake Feb 18, 2006 6:49 PM

Thank You Very Much Shene for your hard workand precise findings. I am about to purchace a WC. Your work has made my NEWBIE experience a whole lot easier towards my purchase.

HIGH 5!!!!!!

xedrox Feb 19, 2006 4:39 AM

Great, i just had made my mind up to by a raynox 6600, i'm confused now.

Great review:-)

Cwood Feb 19, 2006 7:47 AM

wadeatljust did some tests on the Kodak Xenar 0.7x lens and his finding show that at times the Kodak Xenar 0.7x lens does the same thing the Wcon-07 does, the center blur.:(

polar905 Feb 19, 2006 10:30 AM

The Kodak WA does do the centre blur on the one photo I saw, but the others (if they still were takes with lens on) did not show any "soft focus". Maybe I am blind or need a new monitor! :-)

There are fixes to the issue with the blur, using rings seems to be the most popular.

shene Feb 19, 2006 5:17 PM

polar905 wrote:
Quote:

There are fixes to the issue with the blur, using rings seems to be the most popular.
Yes, please see my FZ-30 user guide for possible solutions. They are discussed on the WCON-07 warning page.

CK

http://www.cs.mtu.edu/~shene/DigiCam

Nikon Coolpix 950/990/995/2500/4500/5700, Panasonic FZ-10/FZ-30, and Canon A95 User Guides





stnkline Feb 20, 2006 7:26 AM

I am hoping someone will post the off-set of the two rings, 55/62 step-up and 62/55 thread converter for the application of the WCON 07 to the FZ30!

TIA!!!!..................

Stan

fmoore Feb 20, 2006 7:40 AM

stnkline wrote:
Quote:

I am hoping someone will post the off-set of the two rings, 55/62 step-up and 62/55 thread converter for the application of the WCON 07 to the FZ30!
Why not just use a 55mm uv filter? Remove the glass if need be.

stnkline Feb 20, 2006 7:57 AM

fmoore wrote:
Quote:

stnkline wrote:
Quote:

I am hoping someone will post the off-set of the two rings, 55/62 step-up and 62/55 thread converter for the application of the WCON 07 to the FZ30!
Why not just use a 55mm uv filter? Remove the glass if need be.
I understand from LoveLife and CK that there is a finite distance the WCON 07 has to be offset from the FZ30 to reduce the center softness of this lens. A glassless filter may well have the correct offset. But till one knows what is correct one will never know without a lot of experimenting. Easier to know the correct offset. I have in my notes 0.186 inches but did not record where I saw that offset. I am trying to verify the 0.186 is correct.

Thanks

Stan


shene Feb 20, 2006 6:19 PM

stnkline wrote:
Quote:

But till one knows what is correct one will never know without a lot of experimenting. Easier to know the correct offset. I have in my notes 0.186 inches but did not record where I saw that offset. I am trying to verify the 0.186 is correct.
Youhave been fooled by this meaninglessnumber 0.186 inches (or approximately 4.72mm). As a scientist and engineer for many years, if someone quotes this number tome saying thatis the "best" gap, I would justask that person to go back and retake the Physics 101 or Engineering 101 courses. Here is a list of reasons.

First,tools for making consumer productsin general do nothave a 0.001 inch accuracy. Yes, there are computer controlled machines that can produce a surface of thickness accurate to 0.001 inch for sure; but, these tools are for advanced engineering purposes mostly forhigh tech applications rather than for step ring making. One certainly can use those tools to make rings of thickness of 0.186 inch, but just think how much you are going to pay for the machining and material that can provide a constant 0.186 inch thickness.

Second, even though you have a ring that was manufactured with the 0.001 inch accuracy, can a human being mount that ring on a camera lens with a 0.001 inch accuracy? What if the ring is screwed onto the lens cannot maintain that 0.001 inch accuracy? I don't thinkthe hands of a human being can manipulate a ring that requires 0.186 inch accuracy onto a camera lens. So, the "optimal" 0.186 inch is not reachable.

Third, the making of the camera lens and converter lens also subject to manufacturing error. It may be and very likely belarger than 0.001 inch. This means even though you have a 0.186 inch ring, the camera lens and converter lens may not be up to the expectation of using that space age accurate ring. Therefore, the 0.186 inch ring does not make sense.

Fourth, I have never seen a careful and correct derivation of the 0.186 inch number. Nobody published the reasoning behind it. Some one kept saying the same thing here and elsewhere; but, that person was not able to provide any convincing argument to justify his/her result. As a result, I don't even consider this 0.186 inch number a valid one.

Fifth, due to the optical theory of converter lenses, this 0.186 does not make any sense. The theory is too complex to be discussed here; but, if you like, the why 1.5X is 1.5X page of my FZ-30 user guide may offer you some clue.

Given the above reasons, I would suggest just forget this meaninglessnumber. Use whatever ring that can remove the center blur spot. The WCON-07 warning page of my FZ-30 user guide has the details.

CK

http://www.cs.mtu.edu/~shene/DigiCam

Nikon Coolpix 950/990/995/2500/4500/5700, Panasonic FZ-10/FZ-30, and Canon A95 User Guides





stnkline Feb 20, 2006 8:12 PM

shene wrote:
Quote:

Given the above reasons, I would suggest just forget this meaninglessnumber. Use whatever ring that can remove the center blur spot. The WCON-07 warning page of my FZ-30 user guide has the details.

CK
OK. If the .186 is wrong what is the offset you used to mount the WCON 07 when you made your tests ofyour lens on your camera.That is if you measured it.The optimum offset may be different with my camera/lens stack.

Granted there will be a tolerance build up of the camera threads plus that of the step up ring and the thread converter one would get .xxx +/- .yyy but I have to start somewhere. . To me that is very simple.No 101 required!

If I had given the offset to one of my machinists in the past it would be a measurement plus or minus some specified tolerance. I could not tell them I did not know the dimension/ tolerance and get a part machined. Precision gearing and mounting in mechanical air data computers in the 50's were in tolerances of +/- 0.0001 inches. Re J4 air data platform for the B58 A shaft size might have been 0.0625+0.0000 -0.0002" Yes: Not a consumer product.The parts would be segregated by measuring to 0.000050" Measuring equipment to 1/10 of that. The bearings were manufactured and inspected and sized in the same manner. Run-out too! But I would get the part machined and inspected to Mil standards and hand assembled in a white room. Granted this is not possible with the consumer camera as you mentioned.

From reading and printing out your reviews you did mention a step up ring and a thread converter. Then: there are nominal dimensions and their tolerances. I.E. the real world. I was trying not to reinvent the wheel by buying various rings and converters and attendant assemblies in order to reduce the soft spot..If this information is propriety informationI can and will live with that.

I have read and printed out all of your reviews thatrelated tomy interests in photography. In this instance I was hoping to benefit from your experience from mounting the WCON 07 on the FZ30. Extremely informative. I salute you.

Thanks

Stan



All times are GMT -5. The time now is 12:34 AM.