j4nu wrote:
I think a better example of the Contemporary line is Sigma 100-400. It has a very fast AF, but most reviews agree it's a bit worse than the GM.
Sigma 45mm is really one of its kind, both in rendering and AF, so I'll really be surprised if any of the three new lenses exhibits the same character...
Do the other three lenses have the same SA levels (FL / close focusing / aperture) in their design as the 45/2.8?
Not sure I followed about the 100-400 Are you saying that it performs well with AF-C (i.e. better than the 45/2.8)? And, if so ... is that a byproduct of the FL / aperture variance for producing SA?
httivals wrote:
When Sigma designs it's lenses, for example, it's typically to work on different mounts.
+1
And then, the question here is if they are designing the series for the Alliance (Panny / Leica / Sigma) or Sony independently, or just letting the dog wag the tail vs. the tail wagging the dog?
Would be interesting to know how much difference there is between the L mount version on Sigma / Panny / Leica (CDAF) vs. Sony (PDAF) versions?
Would be interesting also to see how the Batis performs on CDAF, but since that's not an offering ...
RustyBug wrote:
Do the other three lenses have the same SA levels (FL / close focusing / aperture) in their design as the 45/2.8?
Not sure I followed about the 100-400 Are you saying that it performs well with AF-C (i.e. better than the 45/2.8)? And, if so ... is that a byproduct of the FL / aperture variance for producing SA?
I think the only one of the new trio that still has a chance to retain that special look is 24mm, but I doubt it personally.
Sigma 100-400 performs great in AF-C, but from what I've read it's not as good as the GM when tracking moving targets (e.g. a dog running towards the camera)... It's "normally" corrected, so no abundance of SA .
All my lenses perform better than 45mm in AF-C because the difference between AF-C and AF-S is minor (which is not the case with 45mm).
j4nu wrote:
I think the only one of the new trio that still has a chance to retain that special look is 24mm, but I doubt it personally.
Sigma 100-400 performs great in AF-C, but from what I've read it's not as good as the GM when tracking moving targets... It's "normally" corrected, so no abundance of SA .
All my lenses perform better than 45mm in AF-C because the difference between AF-C and AF-S is minor (which is not the case with 45mm).
The sample images of the 24mm ... I was kinda "meh" good. Not, "wow" special. But, here again, expectations from the Contemporary Series. My .02 on the sample pics were appropriate bang / buck (which includes size / weight) kinda thing. Adequate, just not uber-inspiring. They reminded that I should be looking at OEM options (and saving my pennies to do so)
But, that could also be that I'm more inclined to 28mm than 24mm (have my TS-E for that FL).
JVan_02 wrote:
Since there's some discussion of 40 CF and I think by and large I have a good amount of experience on the issue I'll say that it's got a lesser version of the 45 C's problems, excacerbated by Zeiss' software meddling. I use an α7 III.
To clarify: in good lighting (bright daylight, sufficient indoor lighting) I have practically no problems using the lens exactly how I wish too until the reproduction ratio/magnification becomes sufficiently high that DoF is under a cm... maybe 35cm, and closer. At this point, focus is only reliable on stationary objects. At a certain point (with regards to lighting getting worse), AF becomes less reliable in the 0.65-.24m range due to two factors. The first is uncorrected SA showing up in that range (almost certainly why Zeiss implemented their forced aperture behavior). The second is Zeiss' aperture behavior essentially interrupting the AF process. This becomes especially problematic in worse lighting, due to both lower contrast exacerbating SA based focus shift and the fact that lower light stresses mirrorless PDAF more than DSLR PDAF (I don't know why this is the issue, just something I've noticed).
This can be somewhat alleviated by stopping down to 2.5 as lighting gets worse. This aperture is where there are only the faintest amounts of SA and it stops Zeiss' software from interrupting the AF process. At this point, your biggest barriers become the same as any lens at 2.5: the -EV AF rating of your camera. At some point, contrast and lighting in the image are simply too poor to focus properly for your setup. The focus limiter actually comes in handy here: it's very unusual for a standard lens, and remembering to engage with it gives the opportunity to discard some AF calculations and make AF more efficient. In the same vein, spot focusing can eliminate a lot of AF calculations and further improve performance.
It's definitely not the almost completely brainless approach achieved by newer Sony lenses like the 20 G (and I imagine the 35 1.8 FE) and admittedly, it's probably not as hands off as say the Sigma 35mm f2. But the 40 gets a bit worse reputation than it deserves, imo. As a last point of info, I'd say the AF motor is exceptionally speedy: cameralabs clocked the average focusing time at 0.35 seconds. For reference, the 24 GM in the same test clocked 0.5 seconds. ...Show more →
I think you have some very interesting observations here.
But I'm not sure I understand your remark "Zeiss' aperture behavior essentially interrupting the AF process".
Could you explain that? Thank you.
RustyBug wrote:
The sample images of the 24mm ... I was kinda "meh" good. Not, "wow" special. But, here again, expectations from the Contemporary Series. My .02 on the sample pics were appropriate bang / buck (which includes size / weight) kinda thing. Adequate, just not uber-inspiring. They reminded that I should be looking at OEM options (and saving my pennies to do so)
Yes, I agree. I personally like the IQ I saw in the samples from the new trio, especially for compactish walk-around lenses.
What I wanted to say is that AF-C performance of Sigma 45mm is representative neither of the Contemporary line, nor of the Art line. 45mm is really unique ...
j4nu wrote:
Yes, I agree. I personally like the IQ I saw in the samples from the new trio, especially for compactish walk-around lenses.
What I wanted to say is that AF-C performance of Sigma 45mm is representative neither of the Contemporary line, nor of the Art line. 45mm is really unique ...
Given the form factor, certainly a consideration. The 24mm with its CF and .5X magnification is something I might have to look into a bit more. Always some torture aspects to a non-macro performing such CF duties, so a grain of salt might need to be included on expectations. Unless I find a different CF lens for the L mount, I suspect that either the 24 or the 45 will be on my radar. I've got a 15mm (30mm equiv) with CF capability right now (m43) that I like, and the 45 Tammy CF (EF) that I like ... so, now I'm looking for something suitable to replicate them in the L mount for CF duties.
If I had my choice, it'd be either a 28/2.8 or 40/2 with CF ... but, nobody's making either of those.
Maximilian wrote:
I think you have some very interesting observations here.
But I'm not sure I understand your remark "Zeiss' aperture behavior essentially interrupting the AF process".
Could you explain that? Thank you.
There's no way of knowing exactly how this works, per say, unless I was able to get a peek under the hood of their code. But essentially, in most lenses you have AF operation wherein the lens does an AF check, calculates the distance to a subject from measurement of the AF check, moves, and repeats the process until the distance is below a certain threshold.
What Zeiss is doing is adding an extra step: at a specified distance (0.65-.24m) the aperture closes. However, there's a lot of potential for this to mess with the AF process—and this is where the uncertainty comes in. I'm pretty sure I'm at least somewhat right, but likely I'll never know. We know that an additional process is included with this lens that is not included with most, so that is for certain slowing AF calculations down by some amount (it'll slow the lens down less with faster processors). Additionally, I suspect the aperture must close before AF continues or people would have figured out a way of tricking the system into actually shooting at f2. So, if that is true now we have a flat delay in the speed of the aperture motor. Additionally, during the closing we now have to re-meter our scene—I know that in AF-C metering is also continuous, but this might force a step in addition to the basic metering schedule. Finally, I don't know how elastic AF response is once you get high enough into the ISO range—or in other terms, I don't know how sudden 1-2/3 stop changes in ISO introduces possibility for error as you get high enough in ISO.
Stopping down to 2.5 in these scenarios just stops the need of the lens from having to run this check except at absolute MFD... which, let's be honest, isn't going to be in focus on any platform without steadying yourself somehow.
It be fun to shoot with but could get old fast the rendering of it.
We certainly have a lot of choices today. Not so much my cup of tea as is this 45 2.8 which is too slow and limited AF abilities but you folks can argue over that all day long.
Dec 14, 2020 at 09:09 AM
imagesfromobjects Offline • • • • Upload & Sell: Off
If you're on the fence about the Sigma 45, I'd say just rent one for a week. It's almost impossible via writing to describe how the AF works, or if it will bother you. You really need to just try it for yourself. Personally I think it's not as bad as folks make it out to be, but I have very little experience with AF lenses, being 99% MF.
Edit- use code FREEWEEK20 to get a week added to your rental at Lensrentals (no affiliation, just happened to get this in an email)
RustyBug wrote:
So, is that to simply suggest that for both the Sigma and the Zeiss ... the design choice for the close focusing capability comes with the attributes of uncorrected SA, thus the root of problems for AF-C on PDAF? It does sound like a familiar story with my Tammy 45/1.8 (also CF). And, is that what is to be expected from such CF / SA designs for PDAF (vs. CDAF)?
Foresska wrote:
Earlier today I tested this lens at 2 meters to compare focusing at f/2.8 and f/5.0 (more details in earlier post). I couldn't tell any difference in tracking in a dimly lit scene. I would expect it should've done better at f/5.0. However, I didn't take any pictures, so accuracy may have differed.
I could be interpreting it wrong, but it looks to me like the Sony FE 85mm f/1.8 also has some, perhaps less, spherical aberration. However, that lens is considered to autofocus well.
If SA is the issue, these new lenses should have issues closer to MFD. They also have significant SA closer to MFD, just less so than 45mm....Show more →
question: if spherical aberrations generally result in more pleasing background bokeh, shouldn't the circles with the harsh rings be in front of the plane of focus and not behind it?
Teo Rey wrote:
question: if spherical aberrations generally result in more pleasing background bokeh, shouldn't the circles with the harsh rings be in front of the plane of focus and not behind it?
Others may know more, but I think it depends if the lens is undercorrected or overcorrected for SA. Over corrected results in brighter outlined bokeh balls, whereas undercorrected is brighter in the middle than the edge. I'd think it would be opposite on each side of the plane of focus but I'm not sure.
I don't really see any similarity in the output of FE85 and Sigma45 so I'm not sure if looking for shared aspects in their optical design can add to the discussion. FE85 is rather clinical while Sigma45, if I were to put it in one word, puts a "veil" (SA I guess) on the background, making it smoother...
Based on that paragraph from lenstip tests, Batis 40 should have the smoothest rendering of them all ...
One thing I don't understand in his 35mm tests is that he claims the Sony has no CA... (on the other hand, I won't be able to get that fringing picture on the Sigma out of my head ).
Also, the portrait crop of the eye looks sharper to me on the Sigma, due to higher contrast I guess, than on the Samyang.
Sadly, not even the Sigma 35mm f/2 DN can resolve the whole sensor of the Sony a7R IV when stopped down. Still waiting patiently for a compact 35mm to be able to resolve the whole sensor at "landscape" apertures. I'm sure Voigtländer got this with a future 35mm f/2 APO-Lanthar. Might even prefer a 28mm APO if they make one. The 65mm DN seems to be outstanding all around, though.
Frederik0711 wrote:
Sadly, not even the Sigma 35mm f/2 DN can resolve the whole sensor of the Sony a7R IV when stopped down. Still waiting patiently for a compact 35mm to be able to resolve the whole sensor at "landscape" apertures. I'm sure Voigtländer got this with a future 35mm f/2 APO-Lanthar. Might even prefer a 28mm APO if they make one. The 65mm DN seems to be outstanding all around, though.
The new Sigma 35/2 seems to capture fine detail wide open at mid and long distance, especially at f/4. Although not quite as good, it performed neck and neck against the Sigma 35/1.2 Art in a review I saw. I think it was from Cameralabs.
Are you saying that it performs well with AF-C (i.e. better than the 45/2.8)? And, if so ... is that a byproduct of the FL / aperture variance for producing SA?
.
