PDA

View Full Version : Lens Test - Fujinon NW 150mm f5.6 vs Sironar-S



jbmia
11-Oct-2015, 20:06
I'm posting this in the interest of sharing. No need for flaming or negative commentary. Please don't consider this a scientific test or conclusive in any way. I do plan to do another round of testing on this pair... The results very well may be due to sample variance...

With that said, here are the details. Having recently built out a 4x5 kit over the last year, I've read all the usual lens reviews and many forum posts to evaluate lenses for my bag. As many of you know, there are a few sites with well documented lens tests, but other than that, there's a lot of commentary and anecdotal evidence on this and other forums. In my research, I've read the phrase, "you probably won't notice the difference between it and something way more expensive" more than a few times. So much so, that I'd figured it was it basically true. Based on reading this phrase so many times, I wanted to really see for myself if the Fujinon NW 150mm, I picked up on the cheap in fact provided 90% of the performance of the de facto champion, the Rodenstock Sironar-S.

So, with that, I set up my tripod, a Manfrotto 410, Sinar F2 and positioned the combo about 10 feet from the front of the brick oven I built some years back. I leveled the camera and and squared it against the wall as best as possible and set to focusing the Fuji lens. Once I exposed a sheet of HP5+ (it's all I had around today) with the Fuji, I inserted and refocused the Sironar-S and exposed another sheet of film. Both of these were shot at f22 for 21.1 seconds on HP5+ with the Sinar F2 and tripod locked in position. No movement was made between exposures other than double checking focusing on each lens mounting. NO photoshop work (e.g., levels, toning, sharpening, or anything else) has been applied to these images. Other than the jpg conversion and the snipping these 100% crops from the original image, nothing has been done to these images. Have a look the results by clicking on each thumbnail below...

Center - Oven Arch:
140884140885

Center Lower Left - Oven Landing Crap:
140882140883

Honestly, I'm pretty surprised.. So much so that I have myself wondering if I goofed up the Fujinon shot. But, thinking back, I focused down pretty carefully and set the aperture at f22 for each of them. Either way, I do intend to retest the pair with a target a little further off in the distance. If that turns up anything obviously different than what I've found here, I'll retest this scenario again.

I'll post more pics in the next post, but if you're interested in reading up more or seeing higher resolution images, you can check out my write up here: http://wp.me/p2PvoS-3h

jbmia
11-Oct-2015, 20:09
More examples...

Lower Left - The Hatchet:
140886140887

Upper Right - Brick Work
140888140889

Again, take it for what it is.. Just a side by side comparison. Hope someone finds it useful.
As mentioned above, more details here if interested... http://wp.me/p2PvoS-3h

Old-N-Feeble
11-Oct-2015, 20:20
I'm shocked at the performance of the Fujinon. I'm guessing there's something wrong with it.

richardman
11-Oct-2015, 21:00
I have Fujinon 150, 210, Nikon 75 and 300, SSXL 110, Cooke PS945. I have printed some of the 150mm Fujinon shots 24x30, they are as sharp as anything. I agree with O-n-F

Tim Meisburger
11-Oct-2015, 23:20
Could this be focus shift? Are you focusing wide open then stopping down to f22? If it is focus shift you should not see it if you focus at the taking aperture. If it is focus shift that is creating this, I think I remember reading one time that this can be caused (i.e. focus shift) by reversed elements.

If its not focus shift, be happy, as you got a soft focus lens at a bargain price;)

Mick Fagan
12-Oct-2015, 03:19
I have two of the 150 Fujinon lenses in question. One is attached to my Razzle camera, which is a converted Polaroid and focus is done with a rangefinder and a cam ground especially for that lens on that camera, it is brilliantly sharp.

The other 150 is used as my standard lens on my Shen Hao and recently after acquiring a Toyo 45G I did some 1:1 stuff using it and in all instances from infinity on one camera through to 1:1 on the other camera, that 150 Fujinon is also very sharp.

Somehow I do think there may just possibly be an issue with your particular Fujinon 150 lens, maybe something simple like a spacer missing?

Mick.

Edit: Oops, I just re-read your stuff, my lenses are both Fujinon W f6.3/150 lenses, sorry.

Ken Lee
12-Oct-2015, 04:57
Please don't consider this a scientific test or conclusive in any way.

1) Focus shift (already mentioned by others) - it happens and ruins lens tests

2) "both of these were shot at f22 for 21.1 seconds" - Why not try a typical shutter speed, less vulnerable to vibration, like 1/60 ?

3) Why not turn off the software setting which detects blown highlights. We don't need to see that.

4) Enlarger ? Scanner ? You don't mention how these images have been acquired.

jbmia
12-Oct-2015, 06:17
Please don't consider this a scientific test or conclusive in any way.

1) Focus shift (already mentioned by others) - it happens and ruins lens tests

2) "both of these were shot at f22 for 21.1 seconds" - Why not try a typical shutter speed, less vulnerable to vibration, like 1/60 ?

3) Why not turn off the software setting which detects blown highlights. We don't need to see that.

4) Enlarger ? Scanner ? You don't mention how these images have been acquired.

Thanks for the reply Ken. Your work has been inspirational to me... Thanks for keeping your site up. Since you asked:

1. I really have no idea.. Any way to verify this?
2. The brick oven facade is under a patio roof and the exposures were made in mid-afternoon with no direct sun. I wanted to shoot at f22 to ensure both lenses were well stopped down. I intended to use Delta 100 for this test, but only had one sheet left. I've plotted curves according to BTZS methods for both of these films and used Expodev and an incident meter to set the exposure. I suspect the long exposure was due to HP5+'s reciprocity.
3. When someone sends me a check for my effort, I'll starting caring what people think about stuff like this. No offense please. In all seriousness, I quickly made the image captures and penned the post during commercial breaks of The Walking Dead season premier... ;-) Whether that setting was on or off has no bearing on the results, which was the only reason for the post.
4. Didn't want to write a book here about it, so indicated in my original post, you can read more here: http://wp.me/p2PvoS-3h Since you asked specifically though. Scanner is an old Epson 2450. Both at 2400 input 48 bit RGB output to 16bit grayscale tiff. Crappy old scanner (I'm waiting on lens adapter so I can move to an LED light table/D800+macro rig), but clearly enough to demonstrate the difference between the two. I went back and re-scanned the Fuji negative and got the same results (just to make sure it wasn't an issue with that pass).

One thing. These snips are 100% crops of 11147 x 8776 image, so what we are looking at is around 93"x73" on a 120dpi monitor.. somebody correct me if I'm wrong here.. I'm not a pixel peeper (well.. err..?!? ).. At least not intentionally.. Anyway, at that unnatural resolution all kind of flaws should be glaringly obvious. Here's a couple side by side 100% crop snips reduced down to 24 x 30 resolution at 120dpi (typical flat panel monitor resolution). The fuji image is from this mornings re-scan.


Fuji, Rodenstock
140891140892

FWIW, I think I'm going re-shoot the fuji image later today to make sure there wasn't a process error with this test. Could be an error.. Could just be a crappy sample of this lens. If anyone wants to send of another sample, I'd be happy to waste another sheet of film and post. ;-)

Bob Salomon
12-Oct-2015, 07:14
Just to try to simplify a bit.
There really is no difference between a 21.1 second exposure and a 20 second one. And 20 seconds is a hell of a lot easier to time.

jbmia
12-Oct-2015, 07:28
Agreed. Let's not get hung up on little things like that though... That value is just what ExpoDev suggested the film be exposed to, nothing more. The point of the post is not about the method used to calculate exposure or whether I was able to or needed to expose to a 1/10 of second or not . And of course a second or two either way wouldn't really impact the exposure significantly at that point.

Ken Lee
12-Oct-2015, 08:07
Focus shift is very easy to observe, particularly if the subject is brightly illuminated and at moderate to close distance: Focus wide open, then stop down. After changing the aperture, check if the the target is still in focus. Some lenses exhibit it, others don't.

For a clear demonstration of where I blundered in this regard, click here (http://www.kenleegallery.com/html/tech/BokehComparison.html) for a test of several ~210mm lenses. All lenses were carefully focused wide open - on the fine print of the vitamin bottle - then stopped down to f/11 for exposure. The first photo in the series was made with an old Heliar and the focus shift is plain as day.

You can see the same focus shift in this test (http://www.kenleegallery.com/html/tech/BokehComparison150.html) of three 150mm lenses. Again, the old Heliar shifts focus when stopped down... oops :rolleyes:

jbmia
12-Oct-2015, 08:17
Gotcha. I'm going to re-shoot the Fuji and post an update taking that into account. Thanks for the example.

Drew Wiley
12-Oct-2015, 08:28
Most likely neither the lens is NOT the culprit, nor alleged focus shift, unless there is an odd case of mismatched cells. There are plenty of other variables that need to be pegged for an objective test.

Kevin Crisp
12-Oct-2015, 08:54
Focus shift on a modern plasmat? I have never experienced this with any Schneider, Rodenstock, Fuji or Nikkor LF lens. On my protars when used as single elements? Sure.

jbmia
12-Oct-2015, 09:19
Getting ready to re-shoot the Fuji shot Drew. If you list the variables that should be pegged, I'll make my best attempt to account for them. The two variables that I want to re-confirm. 1) Was the Fuji shot in fact taken at F22 2) Was the Fuji shot properly focused. Let me know if there's anything else I'm missing.

Ken Lee
12-Oct-2015, 09:57
Let me know if there's anything else I'm missing.

To evaluate sharpness, a 20 second exposure is best avoided due to potential for vibration while waiting out the time.

With apologies for the over-statement: if I told you that this test of a 610mm APO Nikkor (http://www.kenleegallery.com/html/aponikkor610/Nikkor610Test.php) was made with a 10 minute exposure, wouldn't you be skeptical about the results ? (For the record I used the highest shutter speed available on the Sinar Copal Shutter, 1/60 second.)

Try 1/60 or higher if you can: the higher the better. Change location if required.

jbmia
12-Oct-2015, 10:09
Agreed Ken. But that wasn't an issue with the other lens though and they were both shot under the same conditions. I'm going to reshoot the Fuji now in hopes it was something like that and it is brighter at the moment. Either way I do intend to shoot both lenses with a different subject in conditions where I can take a much shorter exposure. I'll post an update.

Sal Santamaura
12-Oct-2015, 10:23
To evaluate sharpness, a 20 second exposure is best avoided due to potential for vibration while waiting out the time...


Agreed Ken. But that wasn't an issue with the other lens though and they were both shot under the same conditions...In my opinion, given a lack of wind, a stable camera (which yours appears to be) and the fact that your Apo Sironar S result is quite sharp, the 20-second exposure better avoided vibration than one at 1/60 second would have. Any perturbation from the shutter's operation constitutes a much smaller portion of total exposure than it would at the shorter exposure time. Also, your Fujinon result exhibits simple unsharpness, not artifacts of motion blur. By the way, some of the sharpest enlargements I've ever made were from HP5 Plus negatives exposed using the 80mm lens on a Mamiya 7II.

It will be interesting to see what your Fujinon re-test shows. Right now I'm betting that lens is sub-par for some reason. The difference between your initial result from it and the Apo Sironar S is much greater than could reasonably be expected were both samples performing up to at least "nominal" standards for their types.

I predict that, if you test a greater distance, the Apo Sironar S result will be slightly less good, since it's corrected for 1:10. Also, the Fujinon will be just about as bad, since it's probably defective, perhaps as a result of mismatched cells.

Bob Salomon
12-Oct-2015, 10:38
"I predict that, if you test a greater distance, the Apo Sironar S result will be slightly less good, since it's corrected for 1:10. Also, the Fujinon will be just about as bad, since it's probably defective, perhaps as a result of mismatched cells."

Nope. The S will hold its performance at f22 from 1:5 to infinity.

Sal Santamaura
12-Oct-2015, 10:51
...I predict that, if you test a greater distance, the Apo Sironar S result will be slightly less good, since it's corrected for 1:10...


...Nope. The S will hold its performance at f22 from 1:5 to infinity.Please note that there's a whole lot of room for wiggling in these quotes. :)

"Slightly" less good doesn't mean anything close to the difference between the OP's Apo Sironar S and Fujinon results. It means that, under extreme magnification, one would likely be able to see a slight difference.

"Hold its performance" doesn't mean there's no difference in an Apo Sironar S' performance between 1:5 and infinity. It means that, given the 1:10 design optimization, performance is up to Rodenstock's standards throughout the specified usable range of magnifications.

Absolutes don't work in this discussion. Even after Rodenstock's production test/acceptance regime, the most appropriate descriptors are relative ones.

ic-racer
12-Oct-2015, 11:07
The fuji needs to be torn down and inspected. Perhaps it has been tampered or has separation. Everyone should test their equipmnet. I did similar tests between two Horseman lenses. The one with separation was indeed inferior upon side-by-side testing. http://www.largeformatphotography.info/forum/showthread.php?85695-The-BEST-way-to-solve-lens-separatio

Henry Ambrose
12-Oct-2015, 11:17
You can find any focus shift (or just plain operator error in focus) by looking at other areas of the film. The scene set-up is at a slight angle to the film plane so somewhere across the film you will find the precise plane of focus for each lens. It may not be in the same place on each piece of film.

If there is no sharp focus anywhere on the Fuji test film then there may be a lens problem.

jbmia
12-Oct-2015, 13:30
Ok. Here's a new capture of the Fuji at F22 & 1/4'. It's the best I could do with the lighting. Also, below are new snips of the original Rodenstock images without the clipping function on. I'm only posting the images that were in my intended plane of focus this time...

Fuji Retest, Rodenstock Orig - Archway
140918140919

Fuji Retest, Rodenstock Orig - Upper Right Brick Work
140920140921

The 2nd Fuji exposure is under completely different conditions and has higher contrast due to the lighting, so no longer a valid comparison. It does show that there was clearly something wrong with the original Fuji exposure though. Probably just out of focus. At this point, I'm going to consider this one closed and move to re-testing these two lenses under better conditions at a higher shutter speed. I didn't intend to turn this into a science project, but then that's kind of what these end up being I guess. So, take the results for what they are. Thanks to everyone who responded. I found the various responses to be very informative.

Note: Deleted posts above due to issues with image uploads.

Lachlan 717
12-Oct-2015, 14:59
Check the film holder that you used on the first Fujinon shot. It could be shot/bent/worn/stuffed/loaded incorrectly.....

Sal Santamaura
12-Oct-2015, 15:41
Check the film holder that you used on the first Fujinon shot. It could be shot/bent/worn/stuffed/loaded incorrectly.....Absolutely. With 4x5, I've settled on all Toyo holders. They're consistently closer to the ANSI "T" depth than Lisco/Fidelity holders, which can vary wildly, some of those being far outside the specified +/- 0.007" tolerance. Toyos are usually within +/- 0.002" and, fortunately, I bought mine before Toyo moved production off shore, then started delivering reeking holders.

For 5x7, I individually depth checked lots of holders and selected those that were close to the 0.228" "T" specification. When it comes to 8x10, although Toyo was very precise, the holders are much heavier than Lisco/Fidelity. Thus, I went to Calumet back in 1998 and depth checked piles of holders on the shelf, then bought a dozen that were very close to the 0.260" "T" specification.

Note that the smaller the film and shorter the lens focal length, the more critical emulsion position is. The OP's first Fujinon result could easily have been caused by bad film positioning.

Drew Wiley
12-Oct-2015, 16:00
To do anything resembling a true optical bench test you'd need a vacuum filmholder verified flat (measured with appropriate precision tools), equal measurement
of the film plane on the camera itself, verification of both standards for perfect parallel and flatness, the lensboard checked too, a precise target properly aligned.
Also note what Sal just said. This is how it has to be done.

Lachlan 717
12-Oct-2015, 16:11
For this test, I'd simply use the same holder (same side, of course). Have a changing tent next to the camera for a quick change to maintain light as best as possible.

One variable removed.

Lachlan 717
12-Oct-2015, 16:14
One question for the OP:

How did you differentiate which sheet was from which lens? Given the Fujinon came out okay in the second "test", perhaps the Rodenstock is stuffed and you mixed up the negs?

Sal Santamaura
12-Oct-2015, 16:41
For this test, I'd simply use the same holder (same side, of course)...One variable removed.Well, one variable that might result in different results for the two lenses removed, but still a potential problem. That could cause both lenses to look like dogs.

Better to first depth check the holders and either use two sides with "T" very close to 0.197" or, if you can only find one like that, then do the 'change film between lens trials' routine. :)

jbmia
12-Oct-2015, 17:21
No mix up with the lenses.

In the initial test, I noted the the lens used in the each of the exposure files in the ExpoDev software on my iphone. Each film holder's labeled side and it's # was noted there as well. Each label was removed from the film holder and attached to the developing drum as it was loaded. As each sheet of film came out of the drums it was marked accordingly.

In the future though, I will check the depth on the film holder and include the lens id in the image itself.

Appreciate all great details provided here.

Lachlan 717
12-Oct-2015, 17:36
Well, one variable that might result in different results for the two lenses removed, but still a potential problem. That could cause both lenses to look like dogs.

Better to first depth check the holders and either use two sides with "T" very close to 0.197" or, if you can only find one like that, then do the 'change film between lens trials' routine. :)

Given you're comparing the lenses' performance, you should still see a variation if there is one, regardless of T dist. If a difference IS detected, then a more empirical test platform could be conceived. Seems to be spinning wheels doing anything more at the initial stage.

Sal Santamaura
12-Oct-2015, 17:59
Given you're comparing the lenses' performance, you should still see a variation if there is one, regardless of T dist...If the OP's first Fujinon blurred result was caused by bad film positioning, I doubt one would see much variation with the Apo Sironar S exposing film in that same bad position.

Lachlan 717
12-Oct-2015, 18:17
The blur isn't it that bad, Sal... One feels that one would see a difference in one's results if one's subsequent negatives are relative to one's initially unsharp result.

Sal Santamaura
12-Oct-2015, 19:57
...the OP's first Fujinon blurred result...


The blur isn't it that bad, Sal...It looks pretty darn bad to me!