JVan_02 wrote:
From Dustin's review and my own experience with the 40 G: SA is not perfectly corrected on the 40 at least. Not a lens designer, but feel like the secret sauce allowing these compact primes to be as performant as they are (aside from the lens design/element choice) are Sony's knowledge of aspheric lens design coupled with their use of dual linear motor focusing groups. Ever since the 24 GM, we've seen a trend for Sony to develop lenses that for their specification are remarkably performant for the size and cost, while simultaneously achieving MFD/reproduction ratios that are uniformly excellent for internally focusing lenses....Show more →
I’ve got it, (the 40) and I must say I see very little RSA, and if anything, a slight tendency to overcorrection (detected by comparing rendering in front of and behind focal plane). But I do like it a lot nonetheless, sharp small and fast focussing.
But I agree it’s likely the dual linear motors with independent control that is the main advantage. I don’t think the aspehres are — given the residual roughness inside the bokeh balls, it looks like these Gs use aspheres no fancier than anyone else, not the super smooth ones Sony uses in GMs (and which apparently Canon can now put into their top RF lenses)
Dave Sanders wrote:
This conversation needs a caveat... At even f/4 or infinity, the 45 is nearly as sharp as my Art, GM or Loxia lenses. The issues with the 45 are at f/2.8 and close focus distances.
It's a great dial purpose lens, at least for my uses.
Of course. It’s been clear from the start that medium and long distance performance is great, and the issue is not having the space for floating elements or the tech for multiple independent focus groups. Which affects performance at different distances, not performance at the optimised distance.
DavidBM wrote:
I’ve got it, (the 40) and I must say I see very little RSA, and if anything, a slight tendency to overcorrection (detected by comparing rendering in front of and behind focal plane). But I do like it a lot nonetheless, sharp small and fast focussing.
But I agree it’s likely the dual linear motors with independent control that is the main advantage. I don’t think the aspehres are — given the residual roughness inside the bokeh balls, it looks like these Gs use aspheres no fancier than anyone else, not the super smooth ones Sony uses in GMs (and which apparently Canon can now put into their top RF lenses)
I mean, definitely little SA in comparison to the 45i—but it can be observed in loss of contrast. At least that seems to be mine (and Dustin's) understanding. Observe in this head to head.
F=2.5
F=4
As for the aspheric elements—I wasn't necessarily talking about the grade of the elements themselves. Sony seems to be more invested in aspheric lens designs than any other lens designer I could think of. What I meant to imply is that perhaps they've found a way of incorporating ashperic elements into a dual focus group design that other manufacturer's haven't thought of or can't do—as AFAIK Sony is the only manufacturer to use multiple focus groups so prolifically throughout their entire range, or even at all. I'm not as well read on super telephotos, but the only other standard lens I can think of with multiple focus groups is the 50 1.2 S.
It's something I'd really like to see spread out to other manufacturers. Not only would Sony be kept on their toes, but it would be great to see what other companies can think up with the technology. Considering Sony's stated 20 year investment in aspheric development—perhaps they have too great a lead to be meaningfully imitated, but I always hope.
JVan_02 wrote:
I mean, definitely little SA in comparison to the 45i—but it can be observed in loss of contrast. At least that seems to be mine (and Dustin's) understanding. Observe in this head to head.
As for the aspheric elements—I wasn't necessarily talking about the grade of the elements themselves. Sony seems to be more invested in aspheric lens designs than any other lens designer I could think of. What I meant to imply is that perhaps they've found a way of incorporating ashperic elements into a dual focus group design that other manufacturer's haven't thought of or can't do—as AFAIK Sony is the only manufacturer to use multiple focus groups so prolifically throughout their entire range, or even at all. I'm not as well read on super telephotos, but the only other standard lens I can think of with multiple focus groups is the 50 1.2 S.
It's something I'd really like to see spread out to other manufacturers. Not only would Sony be kept on their toes, but it would be great to see what other companies can think up with the technology. Considering Sony's stated 20 year investment in aspheric development—perhaps they have too great a lead to be meaningfully imitated, but I always hope. ...Show more →
Loss of contrast can be down to many things. When I said the Sony 40mm was well to over corrected for SA I meant by a test of specular highlights at set distances before and behind the focal plane. A pretty good test for well corrected DA is equal and low mounts of brighter edges. If the edges are brighter in the foreground it’s undercorrected, if brighter in the background it’s probably undercorrected.
I agree that multiple independently driven focus groups seems to be a Sony special at the moment although maybe Canon are doing it. But Canon are slightly ahead in the quest for perfect aspherics - they seem to have combined Sony’s super precision moulding with automated polishing that gets the surfaces ten times more even (though how much that matters I don’t know). But those two are certainly producing smoother aspheric surfaces than anyone else I know of (sorry don’t have reference links)
DavidBM wrote:
Loss of contrast can be down to many things. When I said the Sony 40mm was well to over corrected for SA I meant by a test of specular highlights at set distances before and behind the focal plane. A pretty good test for well corrected DA is equal and low mounts of brighter edges. If the edges are brighter in the foreground it’s undercorrected, if brighter in the background it’s probably undercorrected.
I agree that multiple independently driven focus groups seems to be a Sony special at the moment although maybe Canon are doing it. But Canon are slightly ahead in the quest for perfect aspherics - they seem to have combined Sony’s super precision moulding with automated polishing that gets the surfaces ten times more even (though how much that matters I don’t know). But those two are certainly producing smoother aspheric surfaces than anyone else I know of (sorry don’t have reference links)...Show more →
Well, here's the background specular highlight from lenstip.
As for Canon, saw that in the 50 APO thread by the guy who went through like... 7(?) copies of the CV 50 APO. What wasn't clear to me was if the moulding tech was also co-opted by Canon, or if it was just the polishing they were ahead on. Do you know? (And can some optical engineer educate us peasants on how much any of this stuff matters? )
As for Canon, saw that in the 50 APO thread by the guy who went through like... 7(?) copies of the CV 50 APO. What wasn't clear to me was if the moulding tech was also co-opted by Canon, or if it was just the polishing they were ahead on. Do you know? (And can some optical engineer educate us peasants on how much any of this stuff matters? )
That rear specular looks to my fallible eye probably slightly over corrected, which was my take on the 40G (neutral to very slightly over) but you need to compare with one in front of the focus plane to get a better idea.
So the article I saw suggested that Canon’s moulding (don’t know if it’s the same tech or just similar results) had caught up with Sony but the polishing was ahead. Sorry I can’t quote.
How much does it matter? In terms of smoothness not much. GM lenses have as smooth an internal bokeh area as I care about. I wouldn’t pay for further improvement.
One respect in which it might pay off is in terms of repeatable thickness of elements. Polishing and moulding variation seems to result in variation in thickness of elements which requires careful matching of elements to reduce the resultant sample variation (one of our forumers in Singapore has written about this).
DavidBM wrote:
Loss of contrast can be down to many things. When I said the Sony 40mm was well to over corrected for SA I meant by a test of specular highlights at set distances before and behind the focal plane. A pretty good test for well corrected DA is equal and low mounts of brighter edges. If the edges are brighter in the foreground it’s undercorrected, if brighter in the background it’s probably undercorrected.
I agree that multiple independently driven focus groups seems to be a Sony special at the moment although maybe Canon are doing it. But Canon are slightly ahead in the quest for perfect aspherics - they seem to have combined Sony’s super precision moulding with automated polishing that gets the surfaces ten times more even (though how much that matters I don’t know). But those two are certainly producing smoother aspheric surfaces than anyone else I know of (sorry don’t have reference links)...Show more →
Sometimes it's difficult to say what a certain manufacturer does or does not because we are pretty much depended on how much they communicate about it and how the communication it is tied into marketing. Regarding aspherical elements, I know that Zeiss does something similar as Sony with their XA-elements. They were surprised about the GMaster announcement when Sony published the first GM lens, but not so much about the technology behind it. Zeiss just doesn't believe in technology-orientated marketing, they are more about user experience.
One should also remember that within E-mount Sony has the biggest incentive to push lens design and technology forward - because they also sell the mount (in form of camera bodies). That's why they can invest more into developing two- or four- stage focusing motors, whereas other lens manufacturers for e-mount simply falls off because they can't justify the expenses. It's not that Tamron or some other manufacturer couldn't design something equivalent for example to 135GM regarding sharpness and corrections (focusing motors might be an exception, though Tamron has it own VXD-system), it's just that it might not be profitable enough for them.
The size of the 90 (along with the fact that my other 2 primes on either side are 50 & 135mm) make this an intriguing lens. Casual portraits would be breeze due to the size, and for travel it'd be a good "throw it in just in case I need some reach" lens. Will have to decide how much it'd actually get used.
The 90 OOF rendering does look quite nice. That to me may be worth the price of admission. Have to wait for a few users on here to get one, test it out and post their thoughts
P.S. Kind of surprised Phillip Reeves website hasn't reviewed any Sigma i series. Hopefully he also gets his hands on a few for review.
gocolts wrote:
The size of the 90 (along with the fact that my other 2 primes on either side are 50 & 135mm) make this an intriguing lens. Casual portraits would be breeze due to the size, and for travel it'd be a good "throw it in just in case I need some reach" lens. Will have to decide how much it'd actually get used.
I like the combo, I was on a trip with the 24/85 f1.4's and it worked out very well, love the contrasting style. The only thing was the size and weight, double of what a samyang 24mm f1.8 and sigma 90mm f2.8 would weigh, my current configuration since I've preordered
along with the 28-60, I think I can end my search for a travel kit. Since I had the m43's 85 equivalent, I've always wanted a closer focusing 85, 1:5 may not mean much for most folks, but for me, that's great, considerably better than the sony 85. I'm typically too shy to bring around a longer lens because they are a bit too bulky for my liking, and this will be the change I need.
)
