Nikon 50/1.4g vs. Tamron 60/2?

I have been in a shop again today. In the other one with reliable lighting.

I tried Tamron 60mm f/2 only, on my D90. It was underexposing wide open too. Compared to my Sigma 30/1.4 it was very noticeable.

I took some pictures in manual mode so metering was not involved, only optics and aperture lever. At f/2.2 and f/2.8, focus was fixed to 1m, but there was a table at closer distance. I framed shots so that whole frame was filled with single color and evenly lit. ISO 200, ADL off.

At f/2.2 1/100s average of 300x300 pixels in center has 77% light value (V part of HSV mode of color picker in GIMP). In other photo with same parameters it was 78%V.
At f/2.8 1/100s it was 75%V, in other photo 78%V.
At f/2.8 1/80s it was 83%V, in other photo 83%V.
At f/2.2 1/80s it was 85%V.

A lens should let pass 1.62x more light (+0.7EV) at f/2.2 than relatively to f/2.8.

Comparing 1/100s to 1/80s, the slower speed should let pass 1.25x more light (+0.32EV). This 1.25x more light caused light value in the photos at f/2.8 to go from about 76%V to 83%V. While what should be the 1.62x light change due to aperture change from f/2.8 to f/2.2 caused barely measurable change from 83%V to 85%V at 1/80s, or from 76%V to 78%V at 1/100s. (numbers are rounded to favor a bigger change in light for apertures)

Approximating EV change for aperture change:

from shutter speed: 0.83V/0.76V = 9.2% change for +0.32EV; for aperture change: 0.78V/0.76V = 2.6% change 2.6/9.2*0.32EV = +0.09EV (1.065x light).
This corresponds to step from f/2.8 to f/2.7.

So what is the point of using f/2.2 aperture on this lens that is actually f/2.7 in terms of light coming in relatively to f/2.8, has no effect on background blur and underexposes.
Also it underexposes at f/2.8 but less so, at f/4 it is ok.

Perhaps there are good copies like alffastar's bought in US. But all copies I have tried (4 so far) in Czech Republic and Photozone's copy (I assume it was obtained in Germany) were not.

I also guess the Photozone's resolution figure for f/2 is not correct due to this defect.

In Chinese review done on Canon 50D http://review.fengniao.com/143/1433849.html there was a noticeable drop in resolution for f/2 compared to f/2.8 especially on borders, while there is almost none in the Photozone's test.

Of course, my weird computations are not relevant to taking good photos at all. But at least they are relevant for buying decisions like Tamron 60/2 vs. Nikon AF-S 60/2.8 or something else.

I just recently purchased a 60mm Tamron F/2. Basically, I really didn't like the purple halo effect my Nikon 50mm F/1.4G had at F/1.4 on bright highlights. So I sold it and my Nikon 105mm F/2.8D Macro and bought the Tamron. The hope was to consolidate lenses (1 lens to do the work of 2), and to get a fast prime that didn't have purple halos under F/2.8.

The first Tamron had three issues. First, it underexposed (as all the 60mm appear to do on D90's, from what I can tell). Second, the iris when partly closed, had a nonuniform aperture (part of one of the leafs would protrude more than the others). Third, and the reason I returned it, was that it failed to come to focus on my D90 at F/2 (F/2.8 and smaller were fine). I've used 10 other lenses on this camera, so I'm doubtful it's the camera.

So I exchanged it for another 60mm Tamron. This one focuses perfectly at F/2. But it also underexposes to the same degree as the other, and it's iris is even more uneven than the first. (keep in mind the unevenness is minor and unlikely to affect the image) But that's fine. I bought it as a macro and as a fast prime, and as long as the exposure can be corrected, it'll do.

Optically, both 60mm were very impressive. Razor sharp at F/2, and best of all, none of that horrible purple halo effect I saw with the 50mm F/1.4G. When I researched the 50mm's problem, I found that the purple halo is "standard" for very fast primes. I'm happy to say that's not a problem with the 60mm Tamron. Too bad the first 60mm backfocused, as manually focusing it showed it was razor-sharp.

So that leaves the exposure problem. It turns out it isn't a simple problem to solve. Different metering modes (matrix, center-weighted, spot), yielded different levels of underexposure. At first I thought it was the rather significant vignetting that the Tamron has at F/2. But I compared the vignetting to my Nikon 35mm F/1.8G and it's about the same. So I have no idea what it is. Fortunately, it's consistent. My correction is about +1.3EV for Matrix mode and +0.7EV for spot/center-weighted modes. This was independant of the aperture used. Note that this value was determined by comparing the 35mm F/1.8 shutter speed to the Tamron's (the 35mm produces correct exposure every time from what I can tell, and is consistent with the different metering modes). Also note that I left Active-D on as that's how I shoot, the EV corrections might be different with that off.

Speaking of aperture, I checked to see if the aperture might be misbehaving. I photographed myself in a mirror at F/2.2 thru F/8 (can't get F/2 close enough to see the iris, unfortunately). F/2.2 was noticably wider than F/2.8. Given the nearly identical exposure in the middle, the fact that the lens vignettes at F/2 and not at F/2.8 (that I could tell), I'd say it's genuinely opening up fully at the F/2 setting.

So where does that leave me? Happy and sad. Happy that I have a very sharp F/2 prime that can do macro, which autofocuses great (except for the first one I got), it's lightweight, and the flash isn't shadowed when used in 1:1 mode. Sad in that I have to manually overexpose the image. This doesn't bother me, but I know I'll forget one day to turn that off after using the 60mm and blow highlights with another one of my lenses. A hassle, but for my needs, it's worth it. Now if Nikon ever makes an F/2 60mm Macro lens, I'll probably jump ship

Does anyone know if it's just the D90 that has the exposure problem? Maybe it's the D90's metering and that other Nikon DX cameras would be ok?

Hey, I am really not sure my copy is 'ok' in terms of your computations I do not think I will take the time to make them for my copy - have a newborn at home, which is the reason I bought this lens. I tried yesterday again, and judging by looking at the pictures and the histograms, I thought that my lens more or less equally underexposed for 2.2, 2.8, 4. But it still underexposes, so it is not so ok. I just dial + 1.0 exposure compensation and it works fine for me.

Jtra, what is your conclusion? That the lens opens its aperture only to 2.7 instead of 2.0? This is so by design for this specific model or for some copies, what do you think? I am not sure this is the case in my lens, but I am not sure.

I guess I will take it to a photo camera service to hear their opinion.

One more thing - I tried quickly my lens on a friend's D300 body and it underexposed much more than on my D90 but I did not have time for detailed testing...

Everything you say about underexposure I have the same observations - around +1.0 for matrix, +0.7(0.5) for partial and spot. And it is consistent as far as I could test it and I tested it may be 3 or 4 times com[aring the histograms and exposure for different apertures.

I also have the same feelings about this lens - it is just very good, I just need time to get used to compensating the exposure when I use it. Otherwise it si great. Only thing I do not like very much is that autofocus sometimes hunts when the focus point is not the center one and the light is dim. Other lenses also hunt, but tamron just takes too long to go through its focus length. But this is minor. I am keeping this tamron in favor of my nikon 85 1.8 and 50 1.8 - I also replace two lenses

I tested my lens with D300 and it underexposed much more than with my D90 - did not have time to do tests to detirmine by how much, though.

I love the bokeh of this lens, also it is very sharp. The adventages over Nikon's 60 micro are: aperture 2.0 (although it is 2.0 for portrait distances); 1:1 focus is 10 santimeteres from the lens (on nikon it is 4,5 santimeteres); consequently, when shooting 1:1 I can use D90's flash (of course, without using the hood).

I took it to shoot portraits of our baby and it dose that just fine. I am still not sure that sigma 50 1.4 was not a better option, but for really dimly lit occasions I can use my nikon 35 1.8 for better results. Anyway, in relation to sigma 50 1.4, Tamron has this big advantage for portraits (especially of a baby) the close focus ability )))

Look also at my reply to 'hoof' - this is what I think about my lens. I do not think I need to take my lens to a photo camera specialist, may be only to understand what is the reason for underexposure, but I believe my lens CONSISTENTLY underexposes and it isn't due to problem with the aperture at wide apertures.

alffastar said:

Jtra, what is your conclusion? That the lens opens its aperture only to 2.7 instead of 2.0? This is so by design for this specific model or for some copies, what do you think? I am not sure this is the case in my lens, but I am not sure.

Click to expand...

My conclusion so far is:

that my D90 (and the other new D90 which was tried with the Tamron too in one of the shops) were not pressing aperture lever as much as they should or lens aperture lever has some different tolerances. So maximum aperture was about f/2.7 in pictures. Reason for underexposure could be the metering is done with wide open lens up to f/2.2 (at 1m), but for actual shooting it is not as open (timing incompatibility perhaps).

There are other possible conclusions from the facts, but I see this one as most likely.

From what you report, I think your lens is ok, except for metering. But it could be similar issue that the lens is not opening as much as it should - unlike the ones I tried, it does not do this at widest apertures only, but at all apertures. You can find out by comparing with other lenses you have, but beware, lenses differ in t-stops. http://en.wikipedia.org/wiki/F-number#T-stops

The protocol for lens-camera communication is Nikon secret I assume so there might be some issues like Tamron not following it correctly. Perhaps Nikon has planted some obstacles for competitors in the protocol causing such underexposure

Interesting thing is that lens is probably ok in Canon mount. At least the reviews in Canon mount do not tell any underexposure stories.

jtra said:

Hi all,

did anybody tried to compare Nikon 50/1.4g vs. Tamron 60/2 on DX camera. I want a portrait lens about 60mm long. The Nikon 50/1.4 is known for not being very sharp at widest aperture on DX and macros are known to be quite sharp so

Click to expand...

...you need to demo these lens yourself to see whats up. I was given the 50g as a gift and havent used it on my D3 yet but thought its very good on my kick around D50 (Dx). Actually this pic below was its first try out in Nov09, last picture of mom before she lost to cancer..Ill be keeping this lens because it aced the best picture I ever took.

F1.4 1/640 D50



-------------------

Best
Shaun
http://shaun.zenfolio.com

Shaun_Nyc said:

...you need to demo these lens yourself to see whats up. I was given the 50g as a gift and havent used it on my D3 yet but thought its very good on my kick around D50 (Dx). Actually this pic below was its first try out in Nov09, last picture of mom before she lost to cancer..Ill be keeping this lens because it aced the best picture I ever took.

Click to expand...

I have tried them side by side, see my others posts in this thread.

Shaun_Nyc said:

F1.4 1/640 D50

Click to expand...

Only think I disliked on 50g (aside from being a bit shorter in focal length than I would like) is green colored circle sides in background blur. This can be seen in the picture you shown, in the top left corner. The Tamron 60mm was much better in this even at same apertures.

jtra said:

Shaun_Nyc said:

...you need to demo these lens yourself to see whats up. I was given the 50g as a gift and havent used it on my D3 yet but thought its very good on my kick around D50 (Dx). Actually this pic below was its first try out in Nov09, last picture of mom before she lost to cancer..Ill be keeping this lens because it aced the best picture I ever took.

Click to expand...

I have tried them side by side, see my others posts in this thread.

Click to expand...

Shooting bar codes and reading MTF charts has nothing to do with photography. My nikkor micro 60 is clinically razor sharp wide open but not a portrait lens imho..

jtra said:

jtra said:

F1.4 1/640 D50

Click to expand...

Only think I disliked on 50g (aside from being a bit shorter in focal length than I would like) is green colored circle sides in background blur. This can be seen in the picture you shown, in the top left corner. The Tamron 60mm was much better in this even at same apertures.

Click to expand...

F2 has nada a thing in common with F1.4.. The backround in that area was green, hence green oof highlights..lol

F1.4 D3 green circles ? lol Good luck



-------------------

Best
Shaun
http://shaun.zenfolio.com

Shaun_Nyc said:

F2 has nada a thing in common with F1.4.. The backround in that area was green, hence green oof highlights..lol

Click to expand...

The AF-S 50 of course shows pronounced LoCAs, which can lead to greenish borders on background highlights (or any other high contrast subject in the background). In fact, with only very few exceptions, any fast prime behaves this way.

It may be no issue to you, but it is to some (me included).

-- Markus
-- Nikon lens review, photozone.de

I got our lens back from Tamron.

They say there is no issue, but I repectfully disagree.

I'll keep you updated when I have done further testing.

-- Markus
-- Nikon lens reviews, photozone.de

Shaun

We've all heard the one about a picture being worth a thousand words.

Some are worth a book.

Nice capture.

Rick

jtra said:

Interesting thing is that lens is probably ok in Canon mount. At least the reviews in Canon mount do not tell any underexposure stories.

Click to expand...

I have checked this. The Canon EF and EF-S mount uses an electronic aperture control. The Nikon F mount uses a mechanic lever except for TS-E lenses which use an electronic control but they do not work with entry level cameras (even D90).

So if there are no bad reports on the Tamron 60mm underexposure in the Canon mount, the issue I had is most likely related to the mechanical lever.

The issue with constant underexposure at any aperture is more likely a lens-camera protocol incompatibility.

Freedolin said:

The AF-S 50 of course shows pronounced LoCAs, which can lead to greenish borders on background highlights (or any other high contrast subject in the background). In fact, with only very few exceptions, any fast prime behaves this way.

Click to expand...

I agree. The 50/1.4 has a fair amount of CA at f/1.4 but it is extremely easy to clean. On the other had, at f/2 it is practically all gone and at f/2.8 there is just nothing.

Freedolin said:

It may be no issue to you, but it is to some (me included).

Click to expand...

CA is indeed an issue. It reduces detail and creates an ugly bokeh in some situations. It is not disturbing in every image, like in images with many coloured OOF lights, but there are situations when it is clearly disturbing. Luckily lens makers recognize this as an issue as well, and do the best they can to reduce CA as much as possible and economically reasonable. After all, we would not be paying 10x more for the 50/1.4 if we knew it had zero CA in any situation since the rest is relatively easily handled in CNX2.
--
http://www.olyflyer.blogspot.com/

olyflyer said:

Freedolin said:

The AF-S 50 of course shows pronounced LoCAs, which can lead to greenish borders on background highlights (or any other high contrast subject in the background). In fact, with only very few exceptions, any fast prime behaves this way.

Click to expand...

I agree. The 50/1.4 has a fair amount of CA at f/1.4 but it is extremely easy to clean. On the other had, at f/2 it is practically all gone and at f/2.8 there is just nothing.

Click to expand...

That wasn't my experience, and why I ended up selling my 50mm f/1.4G. Maybe the D model is better.

Basically, the purple haloing at F/1.4 wasn't the easily correctable type. I was photographing some guys in armor at a Medieval Times show, and the glints were getting really nasty purple halos on them. I tried both of Capture NX2's CA removal modes, and it simply couldn't remove it all, there was always a purple tint to the remaining glint flare. And this was one example of many during that photo shoot. IMO, this made the 1.4G very undesirable to me as it added much purple to a scene that shouldn't have any. In addition, my D90 doesn't have an easy way to tell it to not use an F ratio under a certain amount ("A" mode sets it completely, it doesn't give it a lower limit allowing it to go smaller). If it did, I'd tell it to never go below F/2.2 or so and be happier, except I spent 2x the cost of the F/1.8 version to get that extra bit of aperture that turned out to be useless (for me). That was an educational experience in fast lenses that I wish hadn't cost me so much money.

My conclusion was to look for a lens that might be slower, but didn't have this problem. And with the Tamron 60mm, I found that lens (although I now have to deal with the exposure issue, but that's manageable).

The Nikon 50mm is no better with CA's, you're comparing different backgrounds, looking at the previous shot, looks like there was either mist or dew or rain remaining on the forrest trees, any lens will do horrible with water droplets in the defocused area, try it one time doing wide open shots with dew on the background of leaves, see what you get

Shaun_Nyc said:

F2 has nada a thing in common with F1.4.. The backround in that area was green, hence green oof highlights..lol

F1.4 D3 green circles ? lol Good luck



-------------------

Best
Shaun
http://shaun.zenfolio.com

Click to expand...

jtra,

Wow, your image is so sharp at f1.4! Would you tell me any sharpening done with it? If not, then I'm very impressive . I'm interesting in this 50 1.4G .
Thanks

Shaun_Nyc said:

...you need to demo these lens yourself to see whats up. I was given the 50g as a gift and havent used it on my D3 yet but thought its very good on my kick around D50 (Dx). Actually this pic below was its first try out in Nov09, last picture of mom before she lost to cancer..Ill be keeping this lens because it aced the best picture I ever took.

F1.4 1/640 D50



-------------------

Best
Shaun
http://shaun.zenfolio.com

Click to expand...

enjoynikon said:

jtra,

Wow, your image is so sharp at f1.4! Would you tell me any sharpening done with it? If not, then I'm very impressive . I'm interesting in this 50 1.4G .

Click to expand...

That was not my image, it was from Shaun_Nyc who contributed the image to this thread.

I think at this "for web" image sizes, the differences between lenses sharpness are rarely visible. If they do, it is more likely a photographer set the camera or lens badly. Sharpening can affect the perception. But the bokeh is clearly visible even at such sizes. I disqualified the Nikon 50g from my choice just for the bokeh.

The background bokeh comparison for Nikon 50g, Tamron 60/2 and Voigtlander 58/1.4 is here:



I prefer the background bokeh of Tamron, it is creamy enough and without noticeable colored bokeh circle borders (compared to the other two).

It has a bit harsh foreground bokeh as can be seen in Photozone pictures, but background is excellent. See the last picture here: http://www.photozone.de/nikon--nikkor-aps-c-lens-tests/481-tamron_60_20_nikon?start=2 It clearly shows both background and foreground bokeh.

Often a lens with a good foreground bokeh has a bad background bokeh and conversely a lens with a good background bokeh has a bad foreground bokeh. It has something to do with spherical aberration and its correction - at least from what I have read.

In the Shaun's picture, there are some areas where background bokeh is not optimal, it shows a bit of what the Tamron shows in foreground in the picture on Photozone site I referred to above. It is not very pronounced to affect initial perception, but it is there.

jtra said:

Often a lens with a good foreground bokeh has a bad background bokeh and conversely a lens with a good background bokeh has a bad foreground bokeh. It has something to do with spherical aberration and its correction - at least from what I have read.

In the Shaun's picture, there are some areas where background bokeh is not optimal, it shows a bit of what the Tamron shows in foreground in the picture on Photozone site I referred to above. It is not very pronounced to affect initial perception, but it is there.

Click to expand...

It can be seen in the bottom picture here:
http://www.photozone.de/nikon--nikkor-aps-c-lens-tests/413-nikkor_50_14g?start=2
that 50g has a bad background bokeh and an excellent foreground bokeh.

So what are the other fast prime lenses in range 50-85mm with an excellent background bokeh? (like the Tamron 60mm) Or at least neutral (also called ideal) bokeh where both background and foreground are equal.

85/1.4 AF-D does not count - it performs great, but it is too pricey.

response deleted by hoof because I can't find a delete post option, and thought the previous post was referring to my post, not someone else's. Sorry if you read my post before I changed it!