TC-20E III testing with 300/2.8 VR II

[khollister]

Since there seems to have been some spread in the opinions of how well the TC-20 works with the 300/2.8 and I just purchased both, I thought I would undertake some testing. I would like this thread to be a discussion of the test methodology and the results I'm getting. If we can settle on a test methodology, it would then be interesting to have folks that own these duplicate the tests and post the results just to see if there is, in fact, a unit-to-unit variation in the TC-20 as has been alluded to back when it was released.

I am envisioning two tests ...

1. Shooting a subject with and without the TC at the same camera-to-subject distance and cropping the same approx FOV (Field of View) with the 300mm shot up-res'ed to the same pixel dimensions as the 600mm shot. This will confirm the TC is actually better than just cropping further into the "naked" image.
2. Shooting with and without the TC at different distances so that the subject is the same approx size in the uncropped frame. This is more indicative of the performance ina real-world use case where you want a longer focal length because the subject is too far away to achieve the desired reproduction ratio with the shorter lens.

I am completely open to discussion/criticism of the methodology and want people to suggest alternatives and why they feel it would be a better test.

What I would like to avoid in this thread is posting of anecdotal examples such as "I shot this with the TC-20 and it's great" or "I shot this with the TC-20 and it sucks".

To get this started, I am posting the shots of my test number 1 (same distance, near min focusing distance - about 2.5M). D800, on tripod with indirect flash.

First shot is the uncropped 300mm shot for perspective so you can see how far in I'm cropping in the subsequent images.

Second shot is a 100% crop at the focus point of the lens without the TC (300mm) up-res'ed to the same approx pixel dimensions as the same FOV with the TC

Third shot is the 100% crop of the lens + TC-20E III (600mm). This has no resizing.

All shots are NEF's into LR4 with Camera Standard and default LR sharpening (25/1.0/25/0). Unfortunately, I am not certain exactly what Picture Control sharpening setting this corresponds to for those using NX2 - future experiment I guess.

As you can see, the use of the TC clearly results in a significant improvement compared to just "digitally zooming" in the image. This is a fairly extreme crop, so the whether you would be better off using the TC at full frame/res or doing a 50% crop is left to the final two frames - the first is a roughly 50% crop of the 300mm shot and the last is the un-cropped 600mm (TC) shot. This is a lot closer, but I still think the TC uncropped shot is slightly better.

I plan on doing part 2 of the test later this week and I'll post those as well.

So, what do you guys think?

Native 300mm FOV (uncropped)

Subscribe to see EXIF info for this image (if available)

300mm cropped (50%)

Subscribe to see EXIF info for this image (if available)

600mm - uncropped

Subscribe to see EXIF info for this image (if available)

300mm

Subscribe to see EXIF info for this image (if available)

600mm

Subscribe to see EXIF info for this image (if available)

 

[JPS]

Great idea you had !

As a -rather new- 300 f2.8 VR II and TC20-E iII, I found myself using the lens with the 2x converter, at full aperture very often, and not finding any noticeable loss of IQ ! The only -slight-loss is in the AF, but nothibg that bothers me at all !

Keep testing !

:smile:
J-P.

 

[khollister]

Great idea you had !

As a -rather new- 300 f2.8 VR II and TC20-E iII, I found myself using the lens with the 2x converter, at full aperture very often, and not finding any noticeable loss of IQ ! The only -slight-loss is in the AF, but nothibg that bothers me at all !

Keep testing !

:smile:
J-P.

I have read such opposing opinions, I was really curious to see how I would fare with mine. Being an engineer by training and profession doesn't help either :wink:

I was going to wait on the TC-20 (I already had the TC-14 and TC-17 I used on my 300/4), but figured the cost was minimal compared to the lens Need to move outside for round 2, so it will have to wait until TS/hurricane Isaac moves past - rain for the next few days it looks like.

Thanks for the encouragement!

 

[JPS]

Good luck with ISAAC ! I just read it hit Haiti.... What a drag !

I'll bee looking forward for you testings !

Ohhhh and BTW, the TC20-E seem to have been thought especially for the 300 f/2.8 !!! Check some of my latest threads, there are many shots done with the combo !

.....and now that I just received my D800, I'm impatient to use it with the 300 + 2x converter !

:smile:
J-P.

 

[genera]

Methodology suggestions:

Use a fast enough shutter speed or a short enough flash duration to completely eliminate any possibility of motion blur. I think this, by itself, would eliminate most of the reported variation. Your second shot (labeled 300mm cropped 50%) seems to show either motion blur or missed focus.

Use a target that will give some information on corner performance.

Use a target with focus points that won't confuse the AF system.

Use the same aperture for both shots. Your third shot has more DOF (or better focus/less motion blur). I'm guessing you shot at 300/2.8 and 600/5.6 or some similar ratio.

If you have the room shoot 300mm at say 10' and 600mm at 20' instead of resizing in software. Go somewhat beyond minimum focusing distance in case your lens and TC don't quite meet the published spec.

Also, could you clarify what 50% crop means to you? I think of it as 1/2 the width and 1/2 the height but your second photo only seems to be about 10-15% of the width of the uncropped photo above it.

 

[khollister]

Methodology suggestions:

Use a fast enough shutter speed or a short enough flash duration to completely eliminate any possibility of motion blur. I think this, by itself, would eliminate most of the reported variation. Your second shot (labeled 300mm cropped 50%) seems to show either motion blur or missed focus.

Use a target that will give some information on corner performance.

Use a target with focus points that won't confuse the AF system.

Use the same aperture for both shots. Your third shot has more DOF (or better focus/less motion blur). I'm guessing you shot at 300/2.8 and 600/5.6 or some similar ratio.

If you have the room shoot 300mm at say 10' and 600mm at 20' instead of resizing in software. Go somewhat beyond minimum focusing distance in case your lens and TC don't quite meet the published spec.

Also, could you clarify what 50% crop means to you? I think of it as 1/2 the width and 1/2 the height but your second photo only seems to be about 10-15% of the width of the uncropped photo above it.

All of these were with flash (and not remotely close to full power, so I think we have that covered.

The f stop question is a valid one, but I think it "unfair" to compare the lens 2 stops down from wide open to wide open with the TC. Point taken about DOF, though. More on that in a sec.

Your suggestion for a test at 10'/20' is exactly what I described as test #2 - just haven't gotten there yet.

The second photo is a 100% crop from an up-res'ed 50% crop of the original (first) image. Not the clearest, I know.

Which brings me to your missed focus comment. While there is no motion blur (tripod and flash), after starring at the 300mm 100% crop (photo #2), I think I see where it may have grabbed the knurled part of the knob in front of the lettered dial as a focus target, so you may very well be right. In hindsight it was a rookie mistake and I'm old enough to know better.

Thanks for pointing that out. I'll give this another go with a traditional resolution target. It was a stupid mistake on my part

As far as corner performance, that is of secondary importance to me based on how I use longer lenses.