Fred Miranda wrote:
Hi Carlo,
This is a great test with enough subject distance so blur is not so strong.
Did you crop the images in-post to match their FLs or matched them during the shooting by moving the Sony a bit closer?
Also, did you take more shots for the same scene? It does not look like the Sony was properly focused at the same location (sunflower I guess) since it's not critically sharp there. We all know the Sony is a very sharp lens wide open. That could have skewed the rendering test.
I moved the camera closer to try to match the FOV....which was harder than it sounds due to Bigma distortion and GM focus breathing.
It's focused in the same area but I can try the test again. Sun was pretty intense so not sure if the AF was thrown off or not.
Another from yesterday. Looking at all these shots, I'm seeing that its basically impossible to get an exact FOV match between the two lenses because the focal lengths are ever so slightly different...but also because the focus breathing on the GM really comes into play too...if you switch to MF, you can visibly see that the FOV gets wider the as you move the focal plane to infinity.
What this means is that if you focus on a subject that is further away vs nearer, the FOV is a bit wider, and as such, the DOF gets wider and the degree of blur falls too.
Basically with the same subject framing and aperture, the degree of blur on the Bigma is noticeably higher, but the difference in blur increases as you frame more loosely.
Ya, I also had that issue when I wanted to align both lenses the same to check CA on a test char...
Anyways, you can mitigate distortion effect by enabling in-camera distortion correction. Then, what you see on the screen/EVF when framing is already corrected.
Mystik wrote:
I moved the camera closer to try to match the FOV....which was harder than it sounds due to Bigma distortion and GM focus breathing.
It's focused in the same area but I can try the test again. Sun was pretty intense so not sure if the AF was thrown off or not.
I think this scene proves too much even for the Bigma, which loses on the bokeh fringing front but still remains a bit smoother. Twigs/branches (e.g. right side, middle of the frame) again make it a real torture for 35s .
Mystik wrote:
Another from yesterday. Looking at all these shots, I'm seeing that its basically impossible to get an exact FOV match between the two lenses because the focal lengths are ever so slightly different...but also because the focus breathing on the GM really comes into play too...if you switch to MF, you can visibly see that the FOV gets wider the as you move the focal plane to infinity.
What this means is that if you focus on a subject that is further away vs nearer, the FOV is a bit wider, and as such, the DOF gets wider and the degree of blur falls too.
Basically with the same subject framing and aperture, the degree of blur on the Bigma is noticeably higher, but the difference in blur increases as you frame more loosely.
j4nu wrote:
I think this scene proves too much even for the Bigma, which loses on the bokeh fringing front but still remains a bit smoother. Twigs/branches (e.g. right side, middle of the frame) again make it a real torture for 35s .
I agree. Once FOV discrepancies are normalized (I know it's tricky to do, especially with the 35/2 DG DN), the GM's rendering is actually close to the big f/1.2 and small f/2 Sigma lenses. The Bigma will win on the amount of blur being a f/1.2 lens so if that's one's priority, the choice is easy. However, there is more axial CA (color fringing) and optical vignetting (cats-eye) with the Bigma wide open. (no free lunch)
IMO, at f/2, the GM's rendering is actually more pleasant than the Sigma 35/2 DG DN's, where the OOF background remains soft even at the corners. (With the Sigma 35/2 DG DN, DOF deepens due to moderate optical vignetting)
The thing is, the GM is a super contrasty lens. If reducing contrast slightly in post, rendering looks even better.
Fred Miranda wrote:
I agree. Once FOV discrepancies are normalized (I know it's tricky to do, especially with the 35/2 DG DN), the GM's rendering is actually close to the big f/1.2 and small f/2 Sigma lenses. The Bigma will win on the amount of blur being a f/1.2 lens so if that's one's priority, the choice is easy. However, there is more axial CA (color fringing) and optical vignetting (cats-eye) with the Bigma wide open. (no free lunch)
IMO, at f/2, the GM's rendering is actually more pleasant than the Sigma 35/2 DG DN's, where the OOF background remains soft even at the corners. (With the Sigma 35/2 DG DN, DOF deepens due to moderate optical vignetting)
The thing is, the GM is a super contrasty lens. If reducing contrast contrast slightly in post, rendering looks even better....Show more →
I think it has to come from you, Fred, to settle this matter. We all know you care about the rendering too. You're very meticulous with your testing so I know you wouldn't pass your judgement too lightly. The GM 35 did get a lot of flak with the early round of heavily processed pictures and people tend to remember mostly the best quality of their lenses. Without doing a direct comparison, it's hard to conclude otherwise. I actually bought the GM expecting a harsh rendering; the size, weight, IQ and latest AF tech won me over. Now I'm glad it's not as bad as the initial perception . Can't thank you enough for the work you have done so far for the Sony Alpha community.
Mystik wrote:
Another from yesterday. Looking at all these shots, I'm seeing that its basically impossible to get an exact FOV match between the two lenses because the focal lengths are ever so slightly different...but also because the focus breathing on the GM really comes into play too...if you switch to MF, you can visibly see that the FOV gets wider the as you move the focal plane to infinity.
What this means is that if you focus on a subject that is further away vs nearer, the FOV is a bit wider, and as such, the DOF gets wider and the degree of blur falls too.
Basically with the same subject framing and aperture, the degree of blur on the Bigma is noticeably higher, but the difference in blur increases as you frame more loosely.
Given the focus breathing issue, I imagine that it would be easier if you frame the shot with the GM first then try to match that with the Sigma. Thank you for taking your time doing this.
I encountered some harsh rendering with my ZA 35, accentuated due to CA. It typically happens with foliage during bright daylight, you will have highlight circles with magenta inside and slight green dot in the middle. It's really what prompted me to sell my perfectly centered copy. After I saw it, I couldn't unsee it . And even though ZA is quite decent with AF speed and accuracy, I say the GM is much better in this respect as well. Not quite as responsive as 4 linear motor GM, but still a joy to shoot with.
hiepphotog wrote:
I think it has to come from you, Fred, to settle this matter. We all know you care about the rendering too. You're very meticulous with your testing so I know you wouldn't pass your judgement too lightly. The GM 35 did get a lot of flak with the early round of heavily processed pictures and people tend to remember mostly the best quality of their lenses. Without doing a direct comparison, it's hard to conclude otherwise. I actually bought the GM expecting a harsh rendering; the size, weight, IQ and latest AF tech won me over. Now I'm glad it's not as bad as the initial perception . Can't thank you enough for the work you have done so far for the Sony Alpha community....Show more →
I had the same impression when I started seeing sample images from this Sony. There was definitely too much processing and we were left in the dark when analyzing rendering. The reality is that the lens is super sharp and contrasty and does not need extra help in post, so I would advise not touching contrast, clarity, sliders or curves. Editing white/black points should also be kept to a minimum. I would actually lower contrast and increase the blacks a bit in order for rendering to remain natural. Of course it all depends on the needs and overall effect the photographer is conveying.
All in all, the FE 35mm f/1.4 GM has a modern look (well-corrected), while still delivering a pleasant neutral rendering. It's not creamy like in 'under-corrected for SA' lenses, but it has a great balance between background and foreground rendering. It's an improvement from the FE 35mm f/1.4 ZA in terms of resolution/contrast as well as rendering and CA correction. (and size!)
Many were preferring the 35mm f/2 DG DN's rendering but after so many side-by-side tests from close to far, I've concluded their rendering is very close around the center area while the GM has actually the advantage off-axis with softer corners.
I cancelled my pre order at first. All I saw where over processed images and could not get a good feel for the images. To me it felt cold and just another 35 1.8 type. Than I decided I gotta try it. I ordered the Sigma first and after testing them side by side I actually just sold the Sigma yesterday. To me outside some rendering differences and size , weight, bulk decisions it is not enough difference to keep both. I am not afraid of weight and size as any primary decision.even with a kit that’s Laowa 14mm, Sony 20, Sony 35 1.4 , sigma 65 and a Batis 135. I got a really nice 5 lens kit that’s still small but optically very very good. Obviously my main lenses are 20,35,65
Imho, it's the matter of spherical aberrations or whatever else it is that causes more distant objects to be kinda thinly covered in a veil on Bigma shots, while the GM provides more distinct outlines. This is of course normally hard to see, but when there is the right distance, sufficient and direct light and our favorite branches/twigs it becomes apparent.
I also think the bokeh balls are of course bigger at f1.4 but also a bit rougher around the edges compared to f2 (GM).
After my tests I liked the Bokeh balls better on the Sony more rounder less onion ring looking and the real push was less distortion on the Sony. I may shoot thousands of images in a shoot and I have to always think about processing too. Size ,weight and price are always a secondary thought and they should be. Optically comes first always. That’s Guys rule
Not necessarily a rendering characteristic, but given an equivalent subject framing, the degree of blur will be higher on the Bigma due to f1.2, but also because of more compression from a narrower FOV. This is exaggerated for loosely framed shots where the focus distance is further due to the focus shift on the GM. If the goal is the most pleasing bokeh for wide shots at fast aperture, it is a factor to consider.
GM contrast is on the hard side and that factors into the bokeh too, but it can be adjusted to a degree in post. I also find the skintone rendering to not necessarily be harsh, but a bit too revealing at times. Dynamic skin softener in Nik cleans this up nicely without taking global contrast down, but it is an extra step in post.
In every side by side comparison, I prefer the rendering and transition of Bigma better than the GM. Sony is always slightly sharper and Bigma is always more smooth and "Art" like, pun intended.
GM background is more vivid and well defined and distracts from the subject, even when both lenses are at f/1.4. GM is a great lens and has quite a few things in its favor (AF, size, weight, price, sharpness), but Bigma still remains the top dog in rendeing and bokeh department IMO.
.
