Saturday, January 31, 2015

Sony A6000/Nikon 300mm f/4.5 H pre-Ai to Canon 100-400mm L f/4.5-5.6 comparison ~ le deuxieme part

When I took a look at the Sony A6000 - Nikon 300mm f/4.5 H pre-Ai and tried to compare it against the Canon 7D - 100-400mm f/4.5-5.6 L combination, I relied on the Canon's contrast detect AF system.  I also relied on the 7D's smaller sensor and stronger anti-aliasing filter.

I realized later that the comparison likely proved nothing.  If I was interested in seeing how the ancient (c.1972-ish) Nikkor really compared against the much more modern-flourite-element Canon100-400L, then I'd have to normalize my comparison conditions.  For this, I would use the Sony A6000 and the appropriate lens adapters.  I would also need to carefully manually focus both lenses, which would be my only option on the A6000 for the adapters and lenses I have.

Scene setup ~ Nikon 300mm f/4.5 H at f/8

With these things in mind, here is another look at the Nikon 300mm f/4.5 H pre-Ai and the Canon 100-400mm f/4.5-5.6 L.
  • Sony A6000 camera, ISO100, "A" mode, "standard" image style settings, shot in RAW format
  • Big sturdy Manfrotto tripod
  • Nikon 300mm f/4.5 H shot at f/4.5, f/5.6, and f/8
  • Canon 100-400mm f/4.5-5.6L shot at f/5.6 and 400mm (to match the scene dimensions of the Nikkor)
  • RAW images converted to jpg at 100percent quality using Sony's software - no image adjustments were made at the time of conversion
  • 600x600pixel segments were taken out of each file - no adjustments to the image were made during the cut/paste process
This image will take you to the Flickr host site.  Look at "All Sizes" and select the largest file size to view at 100 percent.

Sony A6000 - Nikon 300mm H f/4.5 vs Canon 100-400L

My observations remain similar to the ones I made in the previous post.

One thing that I've noticed about internal focusing lenses is that the image magnification is much less than the old rack-focus optics.  In this comparison the Canon needed to be set to 400mm to match the scene size of the Nikon 300mm focused to around 14feet.  I encountered a similar situation when I using a Canon 24-105L and Nikon 85mm on a near-distance subject.

The Nikon 300mm delivers a little less contrast to the sensor than the Canon 100-400L.  This can be easily accounted for during image processing.  As for resolution, it's very difficult for me to find any meaningful difference between the two lenses.  The Nikon might be slightly softer in the corners at f/4.5 (because of spherical aberrations?) than the Canon at f/5.6.  A light smart sharpening would bring the Nikkor image resolution right in line with the Canon's.

I find this interesting in light of the fact the Nikkor is 4 decades old and more than 1500USD cheaper than the Canon.

Sony A6000/Nikon 300mm f/4.5 H vs Canon 7D/100-400L

Sunday, January 25, 2015

Sony A6000/Nikon 300mm f/4.5 H pre-Ai to Canon 7D/100-400mm L f/4.5-5.6 comparison

The Madness has set firmly on the brain as the Camera Flu continues un-abated.

I was curious how a Nikon Nikkor 300mm f/4.5 H might compare against the much more recent and optically sophisticated Canon 100-400mm L f/4.5-5.6 super zoom.

Sony A6000/Nikon 300mm f/4.5 H vs Canon 7D/100-400L
 Comparison scene setup
[shot with Sony NEX5, Nikon 50mm f/1.4, Lens Turbo II]

Why the curiosity?  Well, because, on a whim I picked up one of the aforementioned Nikon 300mm non-ED lenses for 100Euro.  It's in tres bon etat and came with the proper caps for both ends and the original carrying case.  I'm a sucker and an easy mark for old manual focus lenses in excellent condition.

Checking the serial number of the lens I see that I was made during the early 1970's.  The H model optic preceded Nikon's introduction of their extra low dispersion glass example by at least 3 years.  Commenters across the 'net seem to rave about the 300mm ED, but not much enthusiasm is expended over the H.  Well, as I said, I'm a sucker for a good looking lens.

The old Nikkor is nearly as heavy as the Canon L, though this is nicely offset by the smaller diameter lens barrels.  I've found that a smaller lens size makes a lens easier for me to work with, even if the weight is similar to something with a large diameter barrel.

Being a fixed focal length lens, the Nikon is of course not as flexible the Canon zoom.   But, since the Sony A6000 is APS-C and since I have a Lens Turbo II focal length reducer and a standard straight-thru Nikon to E-mount adapter I have two focal lengths to choose from when using the Nikkor - 300mm f/4 and 216mm f/3.2.

Sony A6000/Nikon 300mm f/4.5 H vs Canon 7D/100-400L
Size comparison
The Nikkor lens was set to it's closest focus point,
making the lens barrels extend to it's longest.
 [shot with Sony NEX5, Sigma 30mm EX DN E]

Assuming for a moment that Sony introduces an in-body IS APS-C camera that solves the no-IS problem in using old lenses, the biggest thing I in the Sony/Nikkor kit give up is AF.  I've come to love AF as it's most of the time more accurate than I am.  The question is how accurate can I be at focusing the Nikkor?  Would I be happy with the results?

The comparison setup was as follows -
  • A very stout tripod (a Manfrotto somethingorother)
  • Sony A6000 at 100ISO in "A" mode (aperture preferred)
  • Nikon Nikkor 300mm f/4.5 H c.1972-ish
  • Lens manually focused at 14x
  • Nikon shot in two ways - with a straight-thru adapter and with a Lens Turbo II focal length reducer
  • Canon 7D 18mpixel camera at 100ISO in "A" mode (aperture preferred)
  • Canon 100-400mm f/4.5-5.6 L super zoom set to a field of view that nearly matched those of the Nikkor/Sony setup
  • Canon contrast AF focusing in Live View
  • Both lenses shot wide open and at f/8 (to see if the IQ improved by stopping down)
  • 600x600pixel 100percent resolution image sections taken from the original scene (though I did not account for the file resolution differences between the 7D and A6000).
As we will see, Canon's AF wasn't as accurate as manual focusing at a high magnification.

The results are...

[The following image links to my Flickr page.  Look at this a full resolution to properly compare the various image sections.]

Sony A6000/Nikon 300mm H f/4.5 vs Canon 7D/100-400L

My observations include -

Nikon single-coated lens images are slightly lower contrast than the Canon zoom.  However, it was very easy to match the contrast of the Canon lens in processing.

Canon's Live View AF system was a rather hit-or-miss affair.  This is easily seen in the image comparisons.  To think I've trusted the Canon AF system in the over 5 years I've owned the 7D. Having said that, using a (mostly) accurate very high speed AF system tied to image stabilization and focal-length flexibility makes for a very powerful kit. 

If I can accurately focus my manual focus lenses the Sony A6000 offers 11fps burst rate whereas the 7D waltzes along at 8fps.  Am I good enough to track and accurately focus moving targets?  Time will tell.

The higher resolution Sony sensor (24mpixel vs Canon's 18mpixel) gives me a to be expected increase in the amount of information that can be used in the image file.  The Sony file details are gorgeous and easy to work with.  The sensor is the resolution limiting fact.  Lenses typically out-resolve a sensor from wide open down through f/11.

From wide open, the Nikon 300mm f/4.5 H pre-Ai single-coated ancient as the hills ray-trace designed and mathematically calculated by hand lens is every bit the resolution match of the computer designed fluorite element Canon super-zoom. 

Sony A6000/Nikon 300mm f/4.5 H vs Canon 7D/100-400L
Sony A6000/Nikon 300mm f/4.5 H vs Canon 7D/100-400L

Size comparison
The Nikkor lens was set to it's closest focus point,
making the lens barrels extend to it's longest.
 [shot with Sony NEX5, Sigma 30mm EX DN E]

Friday, January 23, 2015

Helios 44M-4 and Zhongyi Lens Turbo II comparisons

I recently caught a Camera Flu and acquired a couple new pieces of equipment.  To justify the acquisitions (guilty as charged) I wanted to see how things looked.

The first new piece of gear is a Helios 44M-4 that I traded a Takumar 50mm macro for.  I'd read where someone thought that altering the space between the first element and second lens group in a Gauss design lens could lead to more pronounced Petzval portrait lens-like effects.
[Read Jim Galli's comment here.]

The second new piece of gear is a new Zhongyi Lens Turbo II.  This takes full frame SLR lenses and acts as a reverse tele-converter.  It takes a full 35mm frame field and resizes it to fit APS-C or micro 4/3rd's, depending on the adapter.  Mine takes Nikon F-mount lenses and adapts them to the APS-C sensor'd Sony mirrorless series cameras.

This isn't really a test.  Nothing is being measured and any results are purely subjective.

Here is the image comparison setup.  It was made using a Sony A6000, Zhongyi Lens Turbo II, and a Nikon Nikkor 50mm f/1.4 pre-Ai shot wide open.

Sony A6000 Nikon Lens Turbo test setup

The first image comparison is of the Helios 58mm f/2 in various stages of lens element positions.  Starting with the lens properly and fully assembled, I then moved the entire front element group forward by unscrewing the group to the extent of the thread range (without the group falling out of the barrel body).  Next I moved the front element forward in the front group with the forward cell mount fully seated.  Finally, I tried moving both the front and rear elements in the front lens group apart and moved the cell mount to the front of the threaded range.

If you click on the image, it'll take you to my Flickr hosted image.  From there select full image size to look at this at 100% resolution.  Any differences between the various segments can be easily reviewed.

All comparison images were made with the Helios shot wide open at f/2. 

Sony A6000 Helios Disassembly Bokeh Test

My subjective observations are that there are indeed differences in how the image is rendered as the various elements are moved.  Interestingly, the out of focus rendition becomes smoother as the front element groups are separated and the front group mount is moved forward in their threads.  But, and this is to be expected, resolution suffers to varying degrees.

It appears a "dreamy" 1800's portrait lens effect is possible, though I'm not sure it's worth the effort.

Which led me to a second set of image comparisons.  I'd read that the Zhongyi Lens Turbo II full frame to APS-C field reducer rendered the out of focus regions more softly than the native lens used with a straight-thru adapter.  Here's my look at the question.

If you click on the image, it'll take you to my Flickr hosted image.  From there select full image size to look at this at 100% resolution.  Any differences between the various segments can be easily reviewed.

All comparison lenses were shot wide open.  

Sony A6000 Nikon Sigma Lens Turbo Bokeh Test

I did my best to keep the primary scene composition similar between the various lenses and focal lengths.  The Lens Turbo II knocks 0.72x off the focal length and increases the aperture by the same amount.  More on this in a moment.

Considering resolution it appears the Lens Turbo adapter does not degrade image quality in the in focus areas for the lenses I used.  I once again see how difficult it is to get something in focus with wide aperture lenses.  This is partly due to the amount of spherical aberration I see in many old 35mm film-era lenses.  I'm not always certain where the best focus is.

The Sigma 60mm DN Art f/2.8 is obviously sharp.  It's a modern design.  Manually locating the focus point was easy and simple.  But it's maximum aperture is a stop or two under most of the lenses I tried.  How the old Nikkors would perform at f/2.8 against the Sigma has been left to a future comparison.

Looking at the out of focus regions I have to agree with whomever noted the smoother image areas when using the Lens Turbo II focal length reducer.  In every case I feel the out of focus rendition is "creamier" and "smoother" when compared with a native lens used without an adapter.  I rather like what it does to the 50mm and 85mm lenses.

So how does the Zhongyi Lens Turbo II work?  It's pretty simple, actually.  It optically reduces the image size from full frame 35mm to APS-C or micro 4/3rd's dimension.  Interestingly it seems to do this at no cost to resolution.

The Lens Turbo also _increases_ the effective aperture by approximately one stop.  In the case of the Nikon 50mm f/1.4, when used with the adapter it becomes a 35mm f/1.0 lens.

How is this possible?  Remembering optical physics, a lens' focal length divided by the front element diameter will give you the aperture.  When you take a 50mm lens and reduce it's field of view to, say, 35mm _without_ changing the front element diameter, you effectively increase the aperture.  In the case of the Lens Turbo II the aperture is increased by 0.72.

If you don't follow me, tell me and I'll have another "go" at explaining what's going on here.

What's important to note is that a 50mm f/1.4 lens on a full frame 35mm system as a 35mm f/1.0 does on APS-C and there is no change in the effective depth of field.

Again, if you don't follow me on this, let me know and I'll try to explain things a little better.

So... where does this leave me?

I see it's possible to carry a two lens, two adapter kit and cover four focal lengths.  For instance, I could carry a 24mm f/2 and a 50mm f/1.4 Nikkor.  Used on the APS-C Sony A6000 I would have the effective full frame focal lengths of 24mm f/1.4 (focal length reduced 24mm f/2), 35mm f/2 (effective focal length of a 24mm lens on APS-C), 50mm f/1.0 (focal length reduced 50mm f/1.4), and 85mm f/1.4 (effective focal length of a 50mm lens on APS-C) at my disposal.

The combinations and capabilities are now seemingly endless.