Sharpness - 28/1.9 vs. 50/1.5??

[gdi]

I just got these lenses and am using them on the R-D1. The 50 is very sharp wideopen, but the 28 is significantly softer.

Does that sound right? I am not sure whether I need to have the 28 adjusted.


Thanks!

 

[jlw]

I agree -- I own both of these lenses and use them on an R-D 1, and my impression has always been that their performance is very similar (and quite good.)

You might check your screw-to-bayonet adapters. An adapter that's slightly too thin or too thick won't have any effect on the focusing accuracy of a 50mm lens (since the rangefinder is based on the focus travel of a 50; the RF thinks the different thickness is just different focus travel.)

However, an out-of-spec adapter will affect the focusing accuracy of any lens other than a 50, and the amount of error is proportional to the degree that the lens differs from 50mm.

As we've discussed in previous threads, genuine Leitz-brand adapters (beware of eBay fakes!) are almost always the correct thickness, which is 0.98-0.99mm; I've never had a bad one. On the other hand, the vast majority of third-party adapters I've bought have been at least slightly too thick, and some were grossly off. You can thin an "off" one slightly by rubbing it against a piece of fine abrasive paper on a flat surface, but I wouldn't try to remove more than 0.01-0.02mm this way.

 

[Adam-T]

I have them both also and have used Three 28 Ultrons and there`s no way that it`s sharp as the 50 Nokton regardless of matching aperture below F5.6 . the only CV lenses which I`ve used which come up to Nokton sharpness are the 35 F2.5 Skopar, the slow 50mm Skopar and 90 3.5 Lanthar , the 75 Heliar and 35 Ultron I tried were very sharp indeed but still not quite up there with the 50 Nokton for critical sharpness..

In Canon and Nikon glass, the 28mm primes don`t even get remotely close to the £80 50mm F1.8s let alone the Nokton (which beats both makers 1.4 models) and cost more than the Ultron does , the Ultron is the second best 28mm Prime I`ve used (only beaten by a Cron) , the Canon 28mm F1.8 is a total joke in comparison .

 

[gdi]

Actually, I had expected it to be slightly softer, due to Canon and other experiences similar to Adam's. It just seems to be softer by more than I expected.

I'll try to post a comparison today for review. If anyone has a close focused (less than 1 meter) example at 1.9, I would appreciate it!

Thanks for the input!

 

[Toby]

Wide angles are pretty much always softer than 50's. At www.photodo.com the sharpest lenses are all telephotos or standards in their MTF tests.

 

[jlw]

Ultron vs Nokton 1 of 2

Ultron vs Nokton 1 of 2

If anyone has a close focused (less than 1 meter) example at 1.9, I would appreciate it!

Goodness gracious, the things I have to do for people on a Sunday afternoon! ;-)

Here are some shots of my test scene. Test conditions: Epson R-D 1; tripod and cable release; focusing via rangefinder with Megaperls 1.3x eyepiece magnifier; EI 200; raw format converted with Adobe Camera Raw at default settings; saved as JPEG at max quality setting. Since in my last comparo people raised questions about possible focusing inaccuracies, this time I focused three times and then picked the sharpest image.

In the first shot of the series, I used the 50/1.5 Nokton at f/1.9, at a focusing distance of 0.9 meter. For the second shot, I used the 28/1.9 Ultron at f/1.9, also at a focusing distance of 0.9 meter.

Alone, these two shots would cause the Nokton photo to look sharper, simply because its fine details are larger. So for the third shot, I used the Nokton but moved the camera back to a distance of about 1.5m, so the details will be similar in size to those of the Ultron shot.

First, here are full-frame shots of the scene, so you can orient yourself:

 

[jlw]

Ultron vs Nokton 2 of 2

Ultron vs Nokton 2 of 2

Now here are 100%-pixel-size crops of a detail area out of the scene. The text of the "Play" button was the point on which I had focused.

First is the 50/1.5 Nokton at f/1.9 and a distance of 0.9m. Second is the Ultron at the same distance; third is the Nokton moved back to 1.5m, so details are more comparable in size to the Ultron shot.

 

[jlw]

Ultron vs Nokton - my take

Ultron vs Nokton - my take

In the detail shots, the first Nokton photo definitely looks sharper. But I suspect that's mostly because its details are larger -- there's a strong psychological tendency to evaluate detail sharpness by comparing how the image looks to how we think it should look, based on our familiarity with everyday objects.

For example, if I look at a portrait photo of a person, I want to see how sharp the eyelashes are. I know what eyelashes should look like, and I base my evaluation on that expectation.

So, a closeup headshot taken with a mediocre lens may seem to me to be sharper than a full-length shot taken with an excellent lens -- just because in the closeup, the eyelashes are larger and I can see more detail in them, so my brain wants to tell me that picture is sharper.

If you look at the views with the Nokton moved back to 1.5 meters, you see the details at a more comparable size. Maybe I'm just not critical enough, but I really don't see much difference between the two lenses here. The Nokton shot looks contrastier, but I'm not positive it's significantly sharper. (Also, most non-macro lenses are optimized to perform better at longer distances, so just moving back to 1.5m could account for some of the difference.)

Okay, this is just one pair of pictures, and the detail area is only slightly off-axis. If you want test shots of a calibrated chart, at a whole range of apertures, with center, edge and corner crops of everything... well, then YOU do it!

 

[jvr]

I have them both also and have used Three 28 Ultrons and there`s no way that it`s sharp as the 50 Nokton regardless of matching aperture below F5.6 . the only CV lenses which I`ve used which come up to Nokton sharpness are the 35 F2.5 Skopar, the slow 50mm Skopar and 90 3.5 Lanthar , the 75 Heliar and 35 Ultron I tried were very sharp indeed but still not quite up there with the 50 Nokton for critical sharpness..

In Canon and Nikon glass, the 28mm primes don`t even get remotely close to the £80 50mm F1.8s let alone the Nokton (which beats both makers 1.4 models) and cost more than the Ultron does , the Ultron is the second best 28mm Prime I`ve used (only beaten by a Cron) , the Canon 28mm F1.8 is a total joke in comparison .

I have to agree with Adam, I own both lenses and the 50 Nokton seems sharper at the same apertures. At least with my samples, it's not a matter of different focal lengths and detail size.

The Ultron (again, I have only one sample) has a very different rendering from all my other RF lenses when mounted on the RD-1s, with less contrast. Histograms tend to be more "compact" than with any other RF lens I own (even old Leicas, unless foggy). Moreover, it flares more than the Nokton. And it shows (sometimes) a kind of "motion blur" on busy OOF areas that can be distracting.

My Ultron is a bit of a "chameleon" regarding aperture and lighting conditions. All photos with flash at f/5.6 - f8 are razor-sharp and I can't tell the difference from the 50 Nokton. But low-light pictures (and f/2-f/2.8...) show a bit of "haze" that could be bothering (especially in colour) that it's totally absent in the Nokton. This "haze" lowers contrast and saturation and I always thought it was a mix of the three characteristics I describe above (less contrast, flare and "moving" OOF).

The adaptor problem could be real and adding to the problem, I'll check it out!!!

As I posted before in RFF, my 50/1.5 Nokton (and also the 35/2.5 Skopar up to a certain point) is a _very_ impressive lens. I've sold my 50/2 Summicron (last gen) because I really love the Nokton and the (almost) extra stop does make a difference, even with the Epson RF being not as accurate as my M3's.

I just can't feel the same way about my Ultron, as much as I like it and use it(and I do). I have to confess I've been thinking about a 28 'cron, lately... 😀

If the 35/1.2 Nokton is on the same league as the 50/1.5 (from all I've read, it's not), it should be one hell of a lens... Or the 40/1.4, that fits the Epson framelines beter than a 35...

 

[jvr]

Update: I went and test focus on both the 50/1.5 Nokton and 28/1.9 Ultron wid-open, at infinity and close range (<1m) and they focus perfectly well, even with the adapters.

So, the slight "veil" or "haze" I talked about on the Ultron on large apertures doesn't have to do with bad focus...

 

[Adam-T]

I have to agree totally with your findings JVR - the 28 Ultron is a pretty low contrast lens, espeically compared to the 35 Skopar (a VERY contrasty lens) - Ultron pics need quite a bit of Punching up to kill that Flat look (even Sean reid pointed this out and it`s why he likes the lens as it extends Dynamic range) ,.....

the Main diff you`ll see with a Cron is Contrast / punch , less hazy but in reality it`s not massively sharper.. If I was going to spend big money on a lens in this area, it`d be a 21 or 25mm, I got the Ultron because I need the F1.9 aperture and it IS usable wide open if not face meltingly sharp like the Nokton 50 is. I`d look into the ZM glass, especialy the 25mm though you`re losing over a stop of light.

 

[jvr]

I have to agree totally with your findings JVR - the 28 Ultron is a pretty low contrast lens, espeically compared to the 35 Skopar (a VERY contrasty lens) - Ultron pics need quite a bit of Punching up to kill that Flat look (even Sean reid pointed this out and it`s why he likes the lens as it extends Dynamic range) ,.....

the Main diff you`ll see with a Cron is Contrast / punch , less hazy but in reality it`s not massively sharper.. If I was going to spend big money on a lens in this area, it`d be a 21 or 25mm, I got the Ultron because I need the F1.9 aperture and it IS usable wide open if not face meltingly sharp like the Nokton 50 is. I`d look into the ZM glass, especialy the 25mm though you`re losing over a stop of light.

It's always nice to feel you're not alone... 🙂

Thanx, that's good input on the 28 'cron.

I have the ZM Biogon 21/2.8 and I can only say good things about it, so the 25/28 should be great. I prefer a 28 to a 25, because it matches the widest framelines on the Epson (useful with flash!) and I already own a 21.

I'm familiar with the rendering of some of the last generation 'crons (I even owned and sold the 35 asph and the 50, both bought new) and I would expect the 28 'cron to be way better than the Ultron. But you don't seem to find that, nor does Sean Reid in his latest review on 28mm RF lenses. I didn't quite "click" with the 35 asph (kept the 35/3.5 summaron and the VC 35/2.5, still controlling lust for a 35/1.4 'lux asph...) but I liked the 50 'cron very much (it's just that between the old 'cron 50/2 collapsible and the Nokton, the new 'cron was redundant and had to go...). Both were very sharp, very constrasty lenses, I would expect the 28 'cron to be like them...

So, I have been doing some research for a replacement/complement to my Ultron, and I narrowed down to the Biogon, the 'Cron and the new Elmarit asph. I even bought (and sent back) the CV 28/3.5 Skopar, as vignetting on the Epson was more than I could stand (it was traded by a CV 12/5.6 Super Wide and boy! am I happy! - just take a look at the staircase photo attached, Rd-1s + C1 BW).

I was _almost_ convinced to go with the 28 'cron but with your input, I think I'll give the Ultron another opportunity...

 

[gdi]

Ok, I was out today and tried again in pretty good light and the results are below- both wide open. Ultron 28 then Nokton 50 full shot reduced:

(I can't figure out how to post an image with the clickable thumbnail - if anyone could give me a tip, I would appreciate it)

100 % Crop (Near) where I focused, Ultron then Nokton:

Processing was with C1 and identical settings. The 50 is definately a little sharper, but I think the 28 is doing pretty well. I focused on the worm gear, but note that it appears the R-D1 is back focusing - it is more evident on the 50 shot. That is another issue for me - getting the R-D1 RF sorted out!

Does it look like my Ultron is up to spec?

 

[jvr]

Yep, your Epson is really backfocusing: from my side, it looks like the big horizontal dented wheel is the best focus point... So:

1) As I don't know how backfocusing affects both lenses, I don't know if this is a good test... 🙂

2) Moreover, you should be comparing them at the same aperture and I would suggest f/2 (for my Ultron, f/2 and f/1.9 are hard to distinguish and f1.5 is much faster to f/1.9 than f/1.9 to f/2). And my Nokton really improves from f/1.5 to f/2.

3) Your crops seem to have the same number of pixels and, to be fair, the Ultron shot should be magnified to the same visual size as the Nokton crop (just like the test with the remote on the same thread). That means the Ultron shot would seem fuzzier when upsampled (or the Nokton sharper if downsampled). Apart from digital artifacts (unavoidable), the increase in magnification would always have this effect, even if shot on film and optically magnified.

4) I would bet the shots were not taken using the same exact exposure: the Ultron shot probably had less exposure, as the FOV is bigger and includes lighter zones (the river, for instance). This makes darker zones more dark and our eys are easily deceived...

5) Finally, the angle of view is different and cast shadows are more explicit on the Ultron shot (seen more from the left).

With all this, I would say they look quite similar on first sight, if you look at the big dented wheel (the one in sharper focus).

However...

If you look really close (especially to resampled copies - so that physical sizes are comparable - that I attach), you'll notice (ok, I notice...😛 ) that all the edges on the wheel seem somehow "doubled", or moving, on the Ultron shot, something that does not happen with the Nokton, where the image is "quieter", even if it's a bit softer.

Although you can easily see the "jaggies" that come from a 200% upsample in CS2 (a bit more for the Ultron), and they have an influence in picture quality, the same "feeling" is present on the original images (and always is in my images when the Ultron is used wide-open).

And this scene is not that bad for the Ultron: although the wheel is not in the center and that "doubling" happens a lot more on the borders, it can become worse. I'll try to come up with a "worst case" shot so that I can demonstrate what bothers me on the Ultron rendering (I can live well with less contrast and saturation and even flare is ok, if you take some precautions).

BTW, I saw some comparison shots from the Ultron 35/1.7 and the Nokton 35/1.2 in some Japanese website and the Ultron 35 showed the same behaviour. I'll try to find them and post a few crops.

Anyway, your Ultron seems better than mine (or your Nokton worse... 🙂 ) regarding contrast and saturation. Even accounting for the different exposure conditions, the Ultron shot is well saturated and contrasty, at least on par with the Nokton. I'll post a few from mine to compare...


Finally: to create thumbnails, you just have to choose "Go advanced" when posting and then click the button "Manage attachments" that should appear downscreen... 🙂

 

[gdi]

Thanks JVR, I really need to get the R-d1 adjusted - I did try after these shots, but I think it will need a top cover off adjustment, unfortunately.

Even so, I think these were enough for me to feel the Ultron is in fine trim. I wasn't really trying to do a perfect head to head test, just get a comparison.

Thanks again!

 

[summilux]

voigtlander 28 color skopar

 

[ronnie_retro]

Toby's got it right. As a nerd I do MTF graphs, and the 28 focal length is noticeably worse compared to the 50, regardless of the manufacturer. It's as true for Leica as it is for Zeiss or Voigtlander, even within their own product lines.

As a practical matter, this is a close focus test where things are different than when photographing objects further away, as is usual with rangefinders. I've used the 28/1.9 and consider it one of the sharpest lenses I own by f 2.8. Transparencies shot with it have that 3-D look under a well corrected 10X loop.