My observation is that, even outside of the problematic 50-100 mm range, the Tamron has a bit more CA compared to the GM. It's not much by any means, but if you compare them directly, it's easy to see the difference.
Regarding the problematic range with fringing, my theory is that it's mostly excessive LaCA (as it behaves exactly like it), which is too big be corrected in-cam (jpeg). That's why I hope a raw profile from Adobe/DxO/... will alleviate the issue.
aCuria wrote:
I have always suspected it’s the coatings that Tamron uses being yellow… this has been the case even 15-20 years ago with Tamron’s very first 28-75 on EF mount
Regardless the Tamron yellow is a real thing and getting it out of skin tones is annoying to say the least
Maybe blue light is being suppressed somehow
The only brand that I've learned to associate with yellow cast is Samyang...
I'll have to compare this Tamron to 35-150, but I haven't really noticed any of it in the latter.
To be honest, GM's colors are a bit too blue to my eyes in those grayish shots, but I blame the combination of gloomy weather + auto-WB in both cases.
So, all in all the Tamron 50-400 is pretty much on par with the Sony 100-400GM, or at least not far behind? Based on the test results here it seems to be the obvious choice for me as a landscape photographer, considering it has 8x zoom instead of 4, and a much more appealing price tag than the GM. Could it be that the lens was not 100% perpendicular to the initial test on the test chart? I mean, in the irl testing it was sharper than the gm in the corners 🤷
I'm not quite sold yet on the Tamron, at least not at a price over 1000€. The fringing (which also seems to be present in OOC jpegs) and the yellow tint are a bit annoying, especially since I normally only use the jpegs and do not postprocess the images.
Is anyone considering to switch from the GM to the Tamron?
* smaller and lighter, but still a bulky lens (feels heavier/bigger than 35-150). In general, it reminds me of how 100-400DN handles, probably because it's nearly the same weight/size . You can feel the difference vs the GM IMHO.
* 0.5x macro at 50mm is a nice bonus, very sharp but you're practically touching the subject with the lens. I guess it depends on the situation if 0.35x at 400mm is preferable or not.
* black, not flashy. You can actually use the lens hood without the lens starting to look like an RPG.
* It starts to loose its weight advantage when you mount the lens collar (145g), I'll see if I can weight both with and without lens foot/collar. On the other hand GM's hood is also kinda heavy.
* Shaky EVF at longer focal lengths - massive difference vs the GM.
* Looses some sharpness/contrast at closer distances, you can see the difference even at ~2.5m @ 300 / 400 mm. GM has a rather uniform performance across the focal range, especially retaining sharpness close-up.
* Vignetting seem a bit higher than the GM, but I didn't check in detail.
* Some color cast (yellow), but the GM goes towards blue on the other hand (a bit less I think though)... It seems that it mostly affects AWB and it's less noticeable on a chosen preset.
* Purple fringing, or rather LaCA near the borders of the frame on contrasty edges at 50-100mm. Still hoping a raw profile can fix this more than in-cam jpeg corrections. LoCA seems very well controlled, but the GM looks a tiny bit better in that regard.
* AF is super fast (AF-C), hard to tell which is faster, though I think the GM is a bit more accurate.
* Main selling point 8x is the reason (combined with semi-macro and smaller size) why it's hard to let it go
I have a new outdoor test scenes to process (they are not in the review yet). I played with focus bracketing and I have two images at 400mm f/8 that show field curvature is present in long distance focus.
In normal AF-S focus with flexible L spot in center, I was generally getting images that were slightly behind 100-400 GM in sharpness in center but often better on sides and corners. I wonder what makes camera to prefer imperfect focus at focus point (though it improves side sharpness). Does E-mount protocol include focus shift correction tables that could be used thay way?
So, you're saying that shots focused on center were less sharp in the center than those focused elsewhere? If so, I think that's not really intended and simply some variability in how the exact focus point is chosen by AF.
j4nu wrote:
So, you're saying that shots focused on center were less sharp in the center than those focused elsewhere? If so, I think that's not really intended and simply some variability in how the exact focus point is chosen by AF.
Generally, when I use AF-S with center AF point on distant landscape with 50-400 at 400mm, I get image that has slightly worse center sharpness than if I do the same with 100-400 GM (a lot of examples in my review show this already, even old Sigma shows better sharpness in center often as is also described in my current review). On the other hand that same image from 50-400 often has better sharpness on sides than 100-400 GM (the same examples).
But yesterday when I was playing with focus bracketing (using Bluetooth remote), I have got two images: one shows great central sharpness (as good as 100-400 GM image of the same scene, though there was quite a lot of air distortion) where sides were clearly worse than GM. The second image shows very good side sharpness and center defocused somewhat.
So I guess the lens could be better focused for center sharpness all the time, just AF is not doing that. I would prefer better center sharpness when I want to crop a lot, but many times I would prefer what I am already getting as it is: the compromise between center and side sharpness.
I hope I will be able to process and upload the images in the evening/night (CEST timezone).
I think that's not intentional, i.e. the camera tries to focus on the selected point so that it's the sharpest but there's some variance to it. I think Tamron issued firmware updates in the past specifically targetted at long range AF-S accuracy (for 28-200 I think).
On your Riv diffraction kicks in at f/5.7, so having all the test shots at f/8 means everything is probably diffraction limited… may want to repeat at f/5.6 or wide open (6.3?)
On your Riv diffraction kicks in at f/5.7, so having all the test shots at f/8 means everything is probably diffraction limited… may want to repeat at f/5.6 or wide open (6.3?)
I have some shots comparing different apertures.
Day 2 Scene 4 shows 400mm f/6.3 vs f/8. F/8 is better there.
Day 2 Scene 5 - 50mm shows f/8 vs f/11. The f/11 shows less sharpness due to diffraction and not much advantage in corners except for DoF.
Auto seems to vary as expected, depending on the shot corrects one lens more than the other, but I think that when a certain WB preset is used (e.g. Daylight) then the new Tamron is more yellow-ish...
j4nu wrote:
Don't look at shaprness, 50-400 shot is shaken a bit. I was only interested in colors...
Overall you are right though, 50-400 is exteremly sharp at 50mm but the performance deteriorates with longer focal lengths at close distances.
In my comparison with 55/1.8ZA at f/8 (Day 2 - Scene 1 - 50mm) the Tamron was as sharp in center but very close to the sides it was weaker. However when compared with 85/1.8G (Day 2 - Scene 1 - 85mm) the Tamron was almost matching the prime. So my copy (at least) is performing better at 85mm than at 50mm.
. You can feel the difference vs the GM IMHO. 
