Nikon TC-E14III vs. DX-Crop on D800: Nikkor 300mm 2.8 VRII extra reach?
Вставка
- Опубліковано 26 сер 2024
- Getting more reach with a Nikkor 300mm 2.8 VR II on a Nikon D800?
Is it worth buying the new teleconverter TC-14EIII? Or just switch to the DX-mode?
Additional information about comparison:
Apperture (in video): f6.3
Additional tests: f2.8 (DX only), f4
Shutter speed (in video): 1/640
VR: off
ISO: 100
very interesting , thanks for the info
My 1.4tceiii with the vr1 version of this lens is sharp but not tack sharp wide open at F4, stop down to F5.6 and its stupendously sharp.
Excellent video, Mr. Kopp. You present your case quite conclusively. After the TC-14EIII came out, you had one patron of DPReview (initials LB) go to great lengths to "sell" this latest iteration, arguing that it represented a significant improvement over its predecessor, and advancingwhat is in retrospect an unmistakably shoddy argument,by posting some birding/wildlife photos.
Reputable sources, however, are unanimous: the TC-14III offers no tangible benefits compared to the TC-14II; the latter has been well-established to supply no benefits over DX crop mode on the D800, and your results appear to be congruent with this.
The shills at DPR, a site predicated on selling gear, no doubt disagree. I'm glad to have all but severed my ties to its disreputable fora!
Did you AF fine tune the lens + TC combo? For example when I use the 14TCE III on my 200-500mm f/5.6, I have to set AF fine tune to +5, even though without the TC the lens doesn't need any tuning at all. It looks like your TC images are a bit soft compared to the lens alone.
Excellent work on this. I use both DX crop and the 1.4 with my D810. I prefer the simplicity of the DX crop but I then have to shoot inside the DX lines.
I don't understand this comparison, because the pixel to image scale ratio is different with teleconverter compared to without. You cannot change the size of the pixels.
I think a more fair test would be still shots. Video quality can be misleading. I used a doubler (2X) by Kenko I purchased in Japan in 1968 with very good results. On a Nikkor 300mm prime and I got some fantastic sunset shots where minor imperfections do not spoil the shot. In the end, It's the subject and composition that matter.
Best crop
Why not D7100/D7200 + 300mm 2.8 lens?
was the d800/300f2.8/tc14iii combo auto-focus fine-tuned? (we can't tell if you used the viewfinder or the live view focus because of the video editing)
Right now I have D750 + 300mm F4 VR + TC 20iii combo. 600mm full frame reach is great but perhaps too much loss of sharpness.
So I am thinking about changing the teleconverter to TC 14iii. In DX mode wouldn't that combo would give me 630mm?
Since it is 24mp resolution , would it be less effective in crop mode than 36mp D810?
What tripod head were you using? Cable release or no? Just curious. Interesting comparison.
+Photo314159 Sirui M3204X with Gimbal Head, no cable release but shot with release timer
Do you recommend it for a Nikon 200-500 f/5.6, in a D810 or D7100?
Better on a D600 maybe ?
I have used the TC 1.4 ii and 1.4iii with my D800e and now D810 and compare with DX mode as well. 1 I took it to Nikon to ensure it is calibrated with naked lens as well as with all three 1.4iii, 1.4ii, 1.7ii and 2.0iii converters. non of them needed any adjustment. My test and using it, 300 f2.8 vrii with the TC 1.4ii for 90% of my wildlife photography clearly show that the IQ loss is almost nothing between naked lens and the 1.4ii, I saw no improvement with the 1.4iii on the D810. Even the TC 2.0iii is still fantastic. So sorry, I don't think your test here is a very good reflection of the reality. Using a cluttered image like this also reflect little detail. I have shot more than 3500 images with the D810, and the 300 f2.8 vr2 with the TC 1.4 and 2.0. the focus speed slows down with the Tc2.0 as the light becomes low and there is a +/- 10% loss in detail. You cannot do a test like this if your lens calibration is not done. On my D800e I needed a fair amount of adjustment on the naked lens as well as the TC's. After that, the improvement was significant.
You can view them all on my 500px page or on facebook
500px.com/JacobusDeWet
I agree with Jakes. I tried TC1.4-EII and TC20-E3 with 300 f/2.8 VR and D800.
My sample didn't need calibrating. I didn't notice loss of sharp with TC14. With TC20, image quality was a bit less accurate and sometimes erratic according to exposure, although this combo is able to produce very sharp pictures. Anyway, I have never noticed such a loss, even with TC20. In my opinion, AF speed is the main drawback of this combo. Your sample probably needs AF fine-tune, Thomas.
www.flickr.com/photos/40967138@N05
***** I agree. Teleconverters are convenient to reach high focal length, but always degrade image quality and optic performance. So we should always avoid to use teleconverters when it's possible, in particular with stellar optics such 300 f/2.8 VR.
Firstly, I don't think D800 is the best choice for assessment of the sharpness of such a high resolution lens as Nikkor 300mm f/2.8G VR II. The AA filter at D800 sensor is capable of degrading the image sharpness much more readily than any Nikkor 1.4x TC can. Nikkor 300mm f/2.8G VRII much outresolves the D800 sensor constrained by its AA filter. This is evident when comparing DxOmark Perceptual Mpixels one gets with Nikkor 300mm f/2.8G VRII on D810 vs D800: 32 vs 24 P-Mpix. One loses 25% sharpness the lens could deliver just due to the AA filter of the camera. In terms of MTF50 (used for Imatest Score) 25% perceptual Mpixels loss translates into 13.4% loss of MTF50.
Basically this is the reason why photographers use TCs on cameras with high resolutions sensors - the loss of sharpness is more than compensated with Mpixel gain of the longer focal length (higher magnification).
Imagine you have a 300 mm telephoto lens that exactly matches your 36 Mpix sensor of D810 or D800E. When you attaches 1.4x TC, it eats 10% of MTF50 in the center of the frame, so you get 19% P-Mpixels less, i.e., 29 P-Mpix. But 1.4x TC has made the object 1.4 times bigger in your frame, so the 29 P-Mpix of the frame now cover 51% of the area of the initial object. It means that 29 P-Mpix now cover 51% of what was covered with 36 P-Mpix before. How many Mpixels covered 51% of that object without the TC? 36 P-Mpix * 51 % = 18 P-Mpix. 29 P-Mpix vs 18 P-Mpix is 58% gain in sharpness.
The gains of focal length (i.e., magnification) always overcompensate any loss of sharpness with a TC, especially with high resolution sensors. The TC should be extremely pour in order not to do that. That is why even the 2.0x TC, which degrades the sharpness much more than 1.4x TC, is beneficial for wide-aperture telephoto lenses when the reach is important and AF isn't becoming an issue.
You can even calculate how much MTF50 (or Imatest Score) should drop with the TC so that you don't gain using it. The MTF50 should decrease by exactly the magnification factor of the TC in order not to give you any benefit with a high resolution camera sensor. Thus, in the case of 1.4x TC, the MTF50 should decrease by 28.6%. As you can see with many Imatest reviews this is not the case with Nikkor TC-14E III.
In order for the 2.0x to become useless, MTF50 with it should drop by 50%.
***** Thank you for your answer. Your methodology was very good: the same aperture, same ISO, etc. I'd only like to draw attention to one more confounder that could potentially have played a role here - the RAW conversion algorithm in the camera or in the RAW converter depending on what you used. If any sharpening was on at the camera when taking jpeg or at the RAW converter when converting the RAW file manually, this function could apply sharpening differently to the image of the object (e.g., a flower) that took 20 pixels on one photo and 30 pixels on another, because the sharpening algorithm improves micro-contrast on a pixel-to-pixel basis, the fewer pixels the object is depicted with the more chances its micro-contrast will be enhanced by this algorithm.
On the images you demonstrate, the lower overall and micro-contrast with the TC is quite evident even without zooming. So I would repeat this interesting test with the software sharpening completely off, of course if this hadn't been done already.
This is not a fair comparison as F4 on the tele-converter is actually 2.8 on the lens, but some of the light no longer hits the sensor, and the ratio is different, its now a 420mm, so the camera tells you 4 instead now. you need too test 4 vs 5.6 on the tele-converter, or better 5.6 vs 8 on the tele-converter. The lens is probably slightly softer wide open and you cant notice it in crop mode because there are less pixels in the same area.
in fact I tested up to f10 but unfortunately without positive effect on TC images :(
That lens isn't soft wide open not a bit.
Damn. This is very interesting. Do you have any thoughts on the other TCs together with the d800 for maximum reach?
***** Yeah. I'm thinking of trying the TC 20 III for maximum reach in good lighting situations. If I have to boost the iso too much Im sure it's not worth it (not to mention AF issues). If you get to keep the pixels you also have the benefit of downscaling slightly for more sharpness.
Great video, why 76 bad rating ???????
Probably those who spent money for this TC and can't believe that it was not worth it :)
But you can still use DX crop with the TC and get even closer.
+Thomas Kopp I've never did a side by side, but I've always been told/read that cropping yields better results than using a TC. And cropping from 24mp to 16mp is still giving plenty of resolution for prints.
I don't see any reason whatsoever in not using a tc, especially the tc 20e iii with the 300 2.8 vr/vrii? There is enough evidence to show these two are a great combo! It's certainly good enough on my 70-200 f2.8 vrii, so on the prime it will be even better! Some of you folks spend way to much time behind your big screens pixel peeping!
***** well if you're looking to save money and get the extra reach, why didn't you buy a DX body instead in the beginning? D7200 over a D810 for nature would be my choice, and that's a huge saving financially!
***** I have a D7200.
I have also just picked up 300 2.8 VR and use it with a TC 20e iii. So far I'm impressed with the little time I've spent with it. 900 5.6 is not to be sniffed at, especially given the price point too! The lens is nothing short of stunning with the D7200, and does seem to work well with the converter too. Converter also works well with the 70-200 2.8
With good light the results with the 70-200 2.8 and x2 are pretty good! It's nice to be comfortable handheld at 600mm! I know it's not a substitute for the prime 600 f4, but a D7200 with the 300 2.8 and the 70-200 2.8 with TC 20e iii is a force to be reckoned with!
I would love to see the DX body go 36mp in the future, if at all it's possible?
***** yes, 300 2.8 VR (not VRii)
+mokotramp I have the same lens - 300 2.8 VR and a D7000.I am curious how fast is the autofocus with the 20e III and D7200, because I tried mine with tc 17e and it was slooow, not good at all for birds in flight.And probably you get great results with the bare lens on the 24mpx body without AA filter :)
+Dilomski I'm very happy with it, although I don't shoot BIF. I just use it to get me closer to the static/slower stuff. I find the results very good indeed, perfectly acceptable. It's certainly a very cost effective option to get you closer!
This is face palmingly stupid.
If you were to compare the image between FX mode and DX mode, you will notice something rather interesting, there will be no difference in the image when zoomed in.
DX mode just cuts out the middle 16 million pixels of the sensor and stores that as the image. You are not getting extra reach at all as the number of pixels is the same for that portion of the image as you get in FX mode. All you are doing is pre cropping the image, the exact same effect can be achieved in lightroom by just cropping the FX image.
What you really need to compare is a higher resolution dx crop sensor vs the 1.4 teleconverter on d810.
Contrary to some opinions, DX sensors do not actually give the lens a longer reach, they just tend to be of a higher pixel density so that the crop area of the DX has more pixels than its full frame counterpart.
So a 24MP dx sensor has the illusion of 1.5x the reach over a 24MP FF sensor.
Just switching to DX mode does nothing, it's there so you can use DX lenses that would vignette heavily on the FF area.
+magottyk that's exactly the point: DX Mode just crops and with FX + teleconverter you will get about the same "crop" but you will have more pixels in this area. And this comparison just shows that it is not worth having those more pixels, because they won't be sharp enough. So you can just use DX crop mode instead of using your TC. You can try it an you will see!
*****
The only IQ difference noticed in your comparisons was a slight loss of contrast with the TC (more glass), but that was more than made up with the extra detail that 36MP gives over 16MP.
Let's not forget that you are comparing jpegs too, not a very good representation of the actual performance.
There really is no point switching to DX mode unless you want to save on file size.
I'll point to your description.
_"Getting more reach with a Nikkor 300mm 2.8 VR II on a Nikon D800?
Is it worth buying the new teleconverter TC-14EIII? Or just switch to the DX-mode?"_
and as I said, switching to DX mode doesn't give you more reach, but a 1.4 TC does do that and it does it optically.
i did not compare jpeg files, but RAW files. And nobody needs to switch to DX mode, you can shoot in FX mode and crop it yourself. But thats not the key message. I bought such a TC and was disappointed because I did not get more reach. Any results supporting the contrary are welcome. But your arguments don't make the TC any better.
+Thomas Kopp
My mistake on the jpeg, it's the sidecar file. The blocking in the comparisons must be due to mpeg. Either that or you are previewing the sidecar jpeg embedded in the raw file
_"I bought such a TC and was disappointed because I did not get more reach."_
No you got more reach, physics says you did, you just personally were not happy with the results. Your 300mm f2.8 became a 420mm f3.9 physically.
Switching to dx mode did nothing but give you a smaller file size, not extra reach.
If you look at 3:00 in the video, you clearly have a lot more detail with the TC in FX mode, the bit of green beside the pipe has a v shape, but just a blob in the dx crop.
There will always be degradation with a TC, but it's swings and roundabouts, in my opinion the tc image is better as far as detail goes, just a little bit softer and less contrast but not by enough to counter the detail gain.
But then it comes down to the subject and how well you can frame it. If you have to crop, then a TC might come in handy. Otherwise do it in post.
I see what you're doing, a better explaination would have been Teleconverter vs post production cropping in (essentially you are doing the same thing & I realise you know that). I found that with my D750 and 70/200 f2.8 images overall actually looked better with a post crop than with a TC. I can only imagine this would be more so on a D810 with more pixels
in camera crop is digital and lousy, step up to refurbished D810 AND use any TC1.4,1.7,2.0 versions lll and get best images
The music overshoots horribly.
Plot twist - buy a 800mm sigma 5.6. dont use a TC or DX mode. Haha.