• Log in with Facebook Log in with Twitter
Photo

Next PZ lens test report: Sigma AF 85mm f/1.4 DG HSM | A ("Art")


  • Please log in to reply
32 replies to this topic

#21 Brightcolours

Brightcolours

    Advanced Member

  • Members
  • PipPipPip
  • 4,832 posts
  • LocationThe Netherlands

Posted 16 March 2017 - 09:21 AM

Another lenstest-site is less convinced of the 105/1.4 E and more of the Sigma.

 

Since they are not using only a 24MP sensor, but one with 200MP plus optical element between lens and test sensor, I'm wondering how this totally opposite results are possible?

 

i-wgWgRV9.png

 

I wonder how both lenses would behave in direct comparison? All Zeiss Otii are Apo-lenses, and in lenscore's database the 135/2.0 Apo is also better than the 105/1.4 E

 

Meanhwile I really doubt if that dated D3 is still an appropriate reference for these tests.  :huh: ?

That is the 3rd or 4th time I hear about that website to come to odd results, is that not reason enough to doubt their workflow, just like we can doubt DXO lens tets?



#22 JoJu

JoJu

    Advanced Member

  • Members
  • PipPipPip
  • 2,838 posts
  • LocationSwitzerland

Posted 16 March 2017 - 09:40 AM

Well... Nasim Mansourov also was waxing poetry around the Fujinon 16-55/2.8 which Klaus did find "However, given the very high pricing, the Fujinon XF 16-55mm f/2.8 R LM WR should have done better honestly speaking. If you can't live without a lens with these specs, you have no other choice."

 

Different sites, different results. I'm also not sure how I should rate the idea of lenscore's "one sensor to judge them all" and the optical element in between which is not onboard of any of the DSLRs using these lenses. And - they are only testing one sample (like PZ as well...) meaning there's some space to debate about tolerances and reliable production results.

 

However. When I see the fringing of the Nikon and the less fringing of the Sigma I start asking questions about a lens with 5 stars, high end of the rating scale. After that nothing goes above. And seeing the resolution numbers of the Otii against the Nikkor on lenscore I have to say, I would not have given so generously the highest possible verdict. I start asking questions because the reference camera is dated (but the same for the Sigma, so it is comparable)

 

PZ doesn't want to test the Zeiss Apo / manual focus / high end price items, I can understand that, but I notice in myself some doubts about the ratings. I feel like only a lens without any flaws deserves the highest rating.

 

If I wanted to replace the old Nikkor 85/1.4G I would rent both candidates and see which makes me more happy. Both have some downsides like weight and size.



#23 mst

mst

    Advanced Member

  • Moderators
  • 2,095 posts
  • LocationWesterwald, Germany

Posted 16 March 2017 - 10:04 AM

To be honest, I find this somewhat amusing: for years there were complaints in many places that we're way too critical... just because we use the full spectrum of the star rating, so an average lens really just gets an average (3-star) rating. Which also means, we keep the very high ratings, especially the 5-star one, to lenses we think deserve it. This rating hasn't been given too often, yet. Now that we do, for the 2nd time, if I remember correctly, we're too generous ;)

As to the Sigma vs. Nikon debate: the difference in rating is mainly because of the MTF performance wide open, this is where the Sigma lost that half star. The bokeh fringing, as explained already, is not tested in a very formal way, in real life (= in the field) the difference between the two lenses is insignificant, IMO.

And about that single figure star rating: honestly, I don't like it. And to my knowledge, Klaus doesn't either. It wasn't there when PZ tests started more than 10 years ago, there was just the written verdict at the end of each review, but PZ received so many requests to provide a simple/numeric final rating that we just couldn't ignore it. It turned out to be as misleading as we thought, because even though we provide fairly short reviews, that's obviously still way too much for some to read.

The D3x: I know it's dated. Trust me, I know very well, because it's still my main camera, so I (have to) use it quite often :) The issue is, that we hesitate to change a test camera, because each time we do, we lose comparability with already existing reviews. That's why we didn't upgrade to the D800E (or the D810 later) when it was announced, because the fairly moderate increase of resolution per axis just didn't justify to more or less start the FX reviews from scratch. If we had known back then that 36 MP would be the high-end mark for Nikon for that long, we probably should have decided otherwise, yes. However, NOW it makes absolutely no sense to switch to a D810, because it is very likely that something with higher resolution is on the horizon. D810-based reviews would just be a waste of resources, because we could only do a few of them until we switch to whatever is hopefully announced later this year.

I know it's sometimes hard to be patient... I'm actually suffering with you. The plan was to switch to a high-MP Nikon last year already... but unfortunately Nikon did not adjust their roadmap to our plan ;)

-- Markus
Editor
photozone.de

#24 mst

mst

    Advanced Member

  • Moderators
  • 2,095 posts
  • LocationWesterwald, Germany

Posted 16 March 2017 - 10:11 AM

PZ doesn't want to test the Zeiss Apo / manual focus / high end price items, I can understand that, but I notice in myself some doubts about the ratings. I feel like only a lens without any flaws deserves the highest rating.


Well, that's not quite true: show us an affordable way to buy/rent/loan the lenses, and we will test them for sure. However, so far, that simply hasn't happened, yet. I haven't come across a really afordable 2nd hand offer of any Otus lens for example (and I'd still have to think a lot about buying one, because from all I've seen so far they seem be very hard to sell).

With a few exceptions, we don't want to rely (or even ask for) loaners from the manufacturers. That means, we either have to buy the review units, or hope for loaners from the community. You can imagine, that not that many readers out there owning high-end fancy lenses are willing to give their "babies" (which they often need for their work) into unknown hands for a period of two or three weeks. It sometimes happens, yes (see the Noctilux for example), but that's really an exception.

-- Markus
Editor
photozone.de

#25 Brightcolours

Brightcolours

    Advanced Member

  • Members
  • PipPipPip
  • 4,832 posts
  • LocationThe Netherlands

Posted 16 March 2017 - 10:13 AM

I just doubt lenscore resolution result reliability. And with reliability in doubt, one can't say which lens test results to go on and which not (a bad issue with DXO).

 

On the 105mm f1.4, lensrentals was very impressed with it too.

https://www.lensrent...tf-bench-tests/

 

As was lenstip.

http://www.lenstip.c...resolution.html

 

On the Sigma, it scores very high on lenstip too:

http://www.lenstip.c...resolution.html

 

It is not a big secret that the end score from Photozone mostly is about the resolution test, maybe a bit about exorbitant distortion, and the rest (LoCA, CA, flare, veiling, contrast, fore- and background bokeh at different distances, astigmatism and coma, light loss, field curvature at different distances) play only a small role if even that. 

The thing is that different aspects are of different importance to different individuals. So you can't really give a optical lens score that is right.


  • mst likes this

#26 mst

mst

    Advanced Member

  • Moderators
  • 2,095 posts
  • LocationWesterwald, Germany

Posted 16 March 2017 - 10:14 AM

Saw markus comment about apo - i think that apo lenses significnatly reduce the fringing; however I'm not sure he is describing bokeh frinign or sensor bloom (purple on water); I think the thought on sensor bloom is angle of the light.


Definitely bokeh fringing... was able to change it from purple to green with the focus ring :) But yes, you're correct otherwise, APO lenses show a lot less of it, but not none.

However, for real-world use, it's probably safe to say that with most APO lenses and most common light situations, bokeh fringing is usually not an issue.

-- Markus
Editor
photozone.de

#27 mst

mst

    Advanced Member

  • Moderators
  • 2,095 posts
  • LocationWesterwald, Germany

Posted 16 March 2017 - 10:15 AM

The thing is that different aspects are of different importance to different individuals. So you can't really give a optical lens score that is right.


Very true.

-- Markus
Editor
photozone.de

#28 Brightcolours

Brightcolours

    Advanced Member

  • Members
  • PipPipPip
  • 4,832 posts
  • LocationThe Netherlands

Posted 16 March 2017 - 10:21 AM


I know it's sometimes hard to be patient... I'm actually suffering with you. The plan was to switch to a high-MP Nikon last year already... but unfortunately Nikon did not adjust their roadmap to our plan ;)

-- Markus

And other, more important, life things kept you from testing last year, anyway  ;)



#29 mst

mst

    Advanced Member

  • Moderators
  • 2,095 posts
  • LocationWesterwald, Germany

Posted 16 March 2017 - 10:24 AM

And other, more important, life things kept you from testing last year, anyway  ;)


Well, actually for a little longer than last year, but yes, that's also part of the whole story :)

-- Markus
Editor
photozone.de

#30 JoJu

JoJu

    Advanced Member

  • Members
  • PipPipPip
  • 2,838 posts
  • LocationSwitzerland

Posted 16 March 2017 - 10:31 AM

Markus, thanks for the time answering.

 

I once got an Otus for two days free of charge in Switzerland, as a kind of "testing lebs". Zeiss runs some kind of trial occassion. But I have to add, the lens very much looked like a lot of people used the offer and there was no one responsible for cleaning. And the rubber ring is not pleasant to look at after many fingers touched it. Talking about 2-3 weeks and about rental companies which just rent and cash, but don't do the maintenance - I'm not sure if that's leading to reliable test results.

 

I've seen the first Otii 2nd hand in a shop in Basle and I think the market will even out the steep prices.



#31 mst

mst

    Advanced Member

  • Moderators
  • 2,095 posts
  • LocationWesterwald, Germany

Posted 16 March 2017 - 07:27 PM

I'm aware of this try-out service, plus Zeiss (and other's, too, Sigma for example) of course also have copies dedicated to be loaned to reviewers. There's a few issues with this approach though (for us): getting your hands on one of these requires patience (we're probably very low on their priority lists), but once you do, you need to be fast, because others are waiting for it, too, so you can have it for a few days only. The condition of the lenses is probably similar to the "regular" loaners, it has gone through a few hand already and might have been mistreated. If not, you never know if you got a hand-picked lens that might perform better than the average copy. And since we feel we'd have to share the source of the tested copy, we'd be under suspicion to write flattering reviews in return for a free loaner.

Regarding APO lenses again: sorry, I wasn't aware that part of the discussion already happened in the Canon section (just caught up there).

-- Markus
Editor
photozone.de

#32 JoJu

JoJu

    Advanced Member

  • Members
  • PipPipPip
  • 2,838 posts
  • LocationSwitzerland

Posted 30 June 2017 - 12:25 PM

Yesterday I got a rental specimen of this lens. I think it was the same I had troubles to AF when I borrowed it briefly at Sigma's event in Tour de Peilz.

 

Now I know why. Given the AFMA numbers I can get with the dock are roughly the same I can get inside my D810, I and to set up camera and lens this way:

 

0.85 m: +31

1.25 m: +25

2.50 m: +21

∞        : +17 (not completely tested)

 

No wonder I didn't get a single shot with the AF in focus. That is pretty much the poorest performance I ever saw from a new Sigma lens. From a rental lens, that is - I wonder if this importer wants to scare customers away?

 

Next to it, it's also the poorest performance of FoCal. At the end I had to use the fully manual FocusTune and LensAlign. because Reikan FoCal's business model is "throw the user out only three years after he paid for a full Pro license". FoCal told me the D810 is not supported although I have dozens of PDF documents from 2015 focus tests done with the D810...

 

The elder version recognized the body and froze after I wanted to set up the target. I always thought Adobe has a mean business model, but Reikan's is just plain betrayal. 70 quid for sh...t

 

However, I would be very astonished if I end up buying that lens. It's really over the edge of weight and size. I will take it next week to The Netherlands and before and after that I hope to do some comparison shots with the Fuji 56/1.2 APD.

 

Edit: FoCal 2.0.x has problems with Mac OS X 10.12, hence it was not working. The developer/supporter recommended me to go in file mode. Meaning: Taking a lot of pictures at various AFMA adjustments, load it into the software and get a result. That's just what LensAlign does only better.



#33 JoJu

JoJu

    Advanced Member

  • Members
  • PipPipPip
  • 2,838 posts
  • LocationSwitzerland

Posted 24 July 2017 - 11:02 PM

Took me a little to go through all the pictures I brought home, but the effort to adjust the AF values was worth the time. It really leaves the Nikon far behind - in terms of sharpness wide open, contrast and clean of CA wide open with front light not to mention the (for my eyes) superb bokeh.

 

i-Q8DLSSx-XL.jpg

 

i-4WhpMB2-L.jpg

 

Will I buy one? Well, in dark situations I didn't get a lot keepers - but that's the same with the Nikkor, and it's more the camera than the lens, I think. But in case I would do more portraits, then yes, it's worth the costs.

 

Just because it was possible:

 

i-xx952Gt-L.jpg

 

That's f/1.4 and here's the full res






0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users



© by photozone.de