Thursday 25 February 2010

Penta 110 24mm: flare and softness

Hi

I've seen some pentax 110 lenses now (bought a few, looked at a few) and noticed that many are dirty (back and front) and need a little love.

With a clean-up they aren't very prone to flare (as has been mentioned in some forums). I took these today with my 24mm in what can only be called provocative circumstances

parkBridge

now, orienting the camera into portrait you can see that direct sun is falling on the lens ... look at the wood under the bridge which the handrail is attached to for detail and contrast changes.

fullFlare

but if I cast a shadow over the lens I get just a small reduction in contrast (and a colour change cos I had AWB set ...

shieldedFlare

exposure was 1/4000th for all

so if you're seeing way more flare than this then your lens may be grubby.

You can also see that the edges don't look really soft on this lens. Well I've found that it depends on what you're photographing. For instance here:

winterBridgeSoftEdges

softness is quite apparent but here:

snowHandRail

doesn't really intrude much

Personally its better than C mount stuff and if you have the 24 and an adaptor then just go get the 50mm as its a great portrait lens. This is about as close as you can go

tiger110

naturally you should be careful with a tiger in your bed

autofocus adaptor for micro4/3 - rumors are afoot

Hi

a rumor site 43rumors has just put an article out discussing a potential adaptor from Olympus for micro 4/3 cameras to adapt manual lenses and importantly add at least one (and perhaps two) features:
  • auto focus (for sure)
  • aperture (one certainly hopes so)


adaptorThe rumor suggests that the adaptor will reduce the focal length by 2 meaning that your 200mm lens will have the same view as a 200mm lens on a full frame camera.

Now for those of us who have been used to 4/3 giving us x2 advantage in focal length (making a 200mm into a 400mm) this sounds a bit poor, but it brings with it the advantage that light gathered will be essentially doubled so that could mean an extra stop of brightness.

Suggestions are that this could actually work well. A poster on one forum has suggested that this technology works well in Telescopes:
Focal reducers are common in astronomy, you take a 2000mm f10 SCT lens and drop it down to 1000mm f5.0. Wider field, faster, etc. But you don't see them for camera lenses. You can get a well-corrected f5-f6.3 focal reducer for telescopes for about $150.00.
Also the adaptor is rumored to be equipped with a SWD focusing mechanism, which sounds great as the SWD technology of Olympus is similar to the Canon USM.

While all this sounds very positive, in communication with a friend I was given this information relative to a previous historical precedent in Nikon's stable:
However, because of the small diameter of the entrance pupil of the relay system due to the limited space in the lens mount, any f-stop wider than f4.8 could not be used. Consiering that the space in the 4/3 or m4/3 adapter is also very limited (if you are to incorporate AF mechanism, the space should be even more limited)

which is less than ideal if you ask me. Time will tell.

For those who are interested, here are some links:

interesting times

a quirk of nature

Well in a strange quirk of nature Kouvola (where I live) has more snow than any other part of Finland.

This is pretty unusual, and as far as I can tell this is the first time in recorded weather history.

Normally there is more snow in Lapland or on the Eastern Border near Ilomantsi.

Either way I'm not complaining (although the people who do the snow clearing sure are) as the local forest around my home is filled with winter wonderland scenery like this poor little Spurce saping which is totally smothered in snow.

My ski track is in better condition that it ever has been

and the whole place looks great


wow

Wednesday 24 February 2010

things bogans hate

a friend of mine has a cute blog called "Things Bogans Like"

Its rather a laugh, pop over and have a giggle if you have time.

Anyway, this morning I was reminded of things bogans hate ... being dopes in public when they should be showin off to the missus an the mates. This dope forgot to engage his front hubs, and his back quietly sank into the sunrise.. despite the valiant effort of one lone fisherman (who was sick of laughing) they had to pull out the shovels.
naturally we all offered advice.

Tuesday 23 February 2010

micro four thirds: four fifties flare

For no apparent reason I decided I would test my lenses for flare. I'm not sure if this method is the most ideal, but I lined up the lenses against my kitchen wall under a fluro tube. I put a black circle with a hole cut in it on the left and a lens bottom cap (with the depth of the bucket facing me) on the right.

Lenses are:
  • Canon FD 50mm f1.4 (top left)
  • Canon FD 50mm f1.8 (top right
  • Pentax 110 50mm f2.8 (bottom left)
  • Olympus OM 50mm f1.8 (bottom right)

Method


all lenses were tested wide open and the camera (Panasonic G1) set to AWB and, camera on a tripod unmoving between each shot. (Note the Pentax 110 lens does not actually have any aperture control, so wide open is all you can pick from)

results

firstly lets look at the overview of each image, then zoom in for details:

overview

I find it interesting the colour renditions are quite different between the lenses. The lens absent from the group confirms which lens is on the camera at the time (to help me remember

the focus point
I used the following lens to confirm focus as best as I could using the magnifier to focus.

what is interesting here is at screen sizes the Pentax 110 looks to have the best contrast and sharpness, but when pixel peeping (click any image to load a full screen snapshot) I would have picked the FD 50 f1.4 ... just goes to show you need to evaluate things at something other than 100% magnification to see other factors than total resolution.

the rubber ring



The lower row seems to show better contrast to me than the top row, but the Pentax 110 lens (lower left) has significant CA purple fringing on the high contrast area, but mainly towards the outer side (away from the lens center). Makes sense.


the base cap

again purple fringing CA on the 110 lens shows up, but with better contrast than the two FD's

outside of the extreme contrast of being against the fluro the CA stops being obvious and you can get some detail in the blacks inside the caps.

Lastly, it was pointed out to me that perhaps my exposure was too short for reliable colour rendition on artificial light. Since its snowing outside and there is plenty of ambient light spill from all the street lights I would try one last thing.

On this I set the colour balance manually to be "cloudy" and manually focused the lenses to infinity. I used an exposure of 2 seconds for all of them and set all to f2.8 This puts the Pentax to a disadvantage as f2.8 is wide open. I then picked up my 3 led ultra-bright headlight so that it shone straight into the lens for the entire 2 seconds of the exposure. I tried to hold the torch in exactly the same spot.

Below are the results


Its interesting looking through the halo of light trying to see my face and how big the halo is. Then there is the secondary flare ... man the FD1.4 is really flaring. Interestingly there is a really large though less obvious purple flare on the 110 lens (lower left)

There are still colour differences, although its really hard to see colours in the outside in that sort of light.

conclusions

Well personally I like the OM and the itty bitty better and the FD's least. The FD 1.4 and 1.8 are almost dead ringers with the 1.8 showing better colour rendition in this situation to my eyes. I know from previous tests however that the FD 1.4 cleans up immensely when stopped down to 1.8 or 2 and shows itself as being much better than the 1.8 in a typical outdoor sunny day.


See my other tests of some of these lenses:


So, hope this was useful to someone (as it was to me). If you have a better suggestion for contrast and flare testing please put a comment in and I'll try to do that if I get time in the future (like I have now)

I think I'll have to try this again on a sunny day ...



Post scriptum (today is sunny) so I thought I would repeat this with better "natural" light.

Please note that the order is:
Top row FD 50 1.4 left and 1.8 right
Bottom row OM50 f1.8 left and Pentax 110 50f2.8 right
(this is also visible on the frame in the names)

some 100% segments from the middle of the frame


and some 50% segments up closer to where the flare is stronger

Again I find myself liking the looks and contrasts of the Pentax 110 lens wide open better than the others.

Next I picked a strongly backlight scene and put the tree trunk to block the direct sun from hitting my lens:

overview


zooming in on the segment on which I focused ...

its clear that the Pentax (lower right) has better contrast (the FD1.4 is softer at 1.4) and the DoF is less (because its a f2.8 lens).

Throwing the Kit zoom into the mix:


which (being 45mm) has a smaller feature size and looks harsher and of course at 5.6 has different DoF ...

Lastly (for what its worth) there are complete image overviews of each of these images linked to a flickr account

P1070418.fd.1.4

P1070420.fd.1.8

P1070421.om.1.8

P1070422.p.2.8

P1070423.fd.1.4

P1070424.fd.1.8

P1070425.om.1.8

P1070426.p.2.8

Monday 22 February 2010

Panasonic 45-200mm

I wanted to try the 45-200 because I have had good results with my previous Canon EF100-300 USM and miss having a zoom at this length. I didn't want to pay the money for this lens when it first came out, so I picked up a FD 200mm f4 for not much and have been using that for about 10 months now, I like it its compact and optically great.

I will naturally compare the 45-200 to this lens so I thought I'd state that up front.

However its important to put perspective on the fact that the FD 200f4 was bought just because I wanted to try it and it was €30, yes that's right peanuts.

I did however want the 45-200 because I think that AF is a benefit. However, since I had not heard much good about it, I delayed purchasing. I recently decided to hell with the expense (and the prices have dropped) and have bought myself a Panasonic 45-200mm zoom for the G1.

There are a few articles on the WWW about it, including some lens tests on Photozone.de which is a good read if you want a more through examination of the optics. I don't want to go deeply into the sorts of testing they did, but I will mention a few observations, quite likely I'll also do a followup in future with some more testing.

The first thing I noticed about it was just how fat and stubby it is:


its quite a little barrel on the G1 and I can see now why there is so much space just below the flash overhang where LUMIX is written. In fact the lens sits flat on the table exactly even with the camera base. So it sort of makes the G1 look as tiny as it really is.

When you zoom to 200mm the lens extends out of the main housing:


which starts to make it quite comparable to Canon lenses like a EF 70-200 USM in size and operation

As I've been talking about the FD 200 f4 lens I thought I would compare them to give you something to go by, the lens on the camera:


which you will notice is essentially just as long as the 45-200 is when its being used at 200. Side by side however reveals that as a traveling companion it is shorter (and 100g lighter) but noticeably fatter.

Interestingly they both share 52mm lens caps, the zoom shows what gets added to allow controls and interior sliding mechanisms. I've left the FD on its adaptor so as to make the comparison more fair in some ways. Off the adaptor they are quite similar in length. Its a bit like the 45-200 has had a poor diet.

In use the 45-200 zoom operation is smooth and the controls are well placed. I like the location of the OIS switch as its well placed (better in my view than that of the EF lenses I've used) and certainly nicer to move. The manual focus ring is well located and seems to operate well.

This is however a focus by wire arrangement, meaning that while there is gentle damping to the movement (turning to rather a solid struggle on a -10C cold day) there is little correlation between turning and effecting the focus ...

AF is reasonable but not as fast as my Canon EF100-300 USM which has a ring motor type operation and full time manual. (Note: that lens is not an expensive L but a reasonable consumer grade lens check it out on the Canon website, but its now discontinued).

I would put AF as being faster than my Sigma 75-300 micro motor lens was but not as fast as the ring motor USM (careful, they have micro motor lenses which have USM on them but are not the same as the ring motors)

An interesting point emerged when I started to compare the lenses. I have noticed before using legacy lenses that Panasonic applies some funny curves to what it does to the recorded data, exposing right is tricky to implement on this camera. I'm now just thinking of this has something to do with the camera knowledge of the lenses its using.

I put the camera in Av mode put on the 45-200 set f5.6 and took a shot, swapped out to my FD200 lens and set that to 5.6 mechanically and took another shot.

God dam it the Lumix was darker! Looked at the data, and sure enough it had changed the shutter from 1/6th to 1/25th ... Hmmm ... why did it do that?

Right, I thought, I'll normalize things by using Manual, and setting each at f5.6 and choosing the same shutter (1/15th of a sec). I was not expecting such differences.



Quite a bit darker on the Panasonic isn't it (Note: there is some nasty JPG artifacting in the red up there, but that's just the thumbnail blogger creates, click the window into a new window and you get rid of that). My only view on this is that the aperture is one thing, but the lens transmission is another. This lens clearly drops more light though than the aperture values suggest. In case you are wondering, T-stops are used in film industry where the actual brightness levels recorded on film are critical. In motion picture the post production crew do not want part of a spool of film to be darker or lighter than another just because a lens was changed. From Wikipedia:
modern cinematographic lenses now usually tend to be factory-calibrated in T-stops. T-stops measure the amount of light transmitted through the lens in practice, and are equivalent in light transmission to the f-stop of an ideal lens with 100% transmission

anyway, to really look at this I thought I'd use RAW and employ dcraw to convert the image, because by carefully selecting parameters I can get consistent results which I can not reliably get using tools like ACR or other plugins (designed for photographers). Now, unlike the JPG, I can set it to do no colour corrections and do a fixed standard demosiac. The idea here is to remove anything which the camera may do differently to the JPG. This should sort out if the sensor is getting the same exposure or if it was just curves ... first the FD lens (following the order of above image):

FD-tiff-manual

then the Panasonic 45-200

pana-tiff-manual

Firstly the colour looks odd because unlike the JPG's made by the camera (using auto white balance) and the light was not a standard colour temperature such as daylight. However even though there is a known standard curve applied to these images, the RAW images are more related to what the sensor captures than the JPGs.

They are still darker, but not as dark. Also you can see the levels difference in the histograms. There is more at work on the rendition of the images than just darkening and lightening. This means that contrast and colour rendition of the two lenses is quite different.

What does this mean


This means (among other things) that your sensor will be getting less light to use when you pop on the 45-200 lens than using the FD200 mm lens. So even though you might have the same f-stop set on the lens, it will be darker and thus choose a different shutter to compensate.

You might notice at this point that I have done everything on the FD200 at f5.6, there are two reasons for this; firstly the 45-200 is a f5.6 yet the FD lens is f4. This would introduce differencs making comparison harder and so it wouldn't really be as simple to compare them at different apertures. As it happens the FD200 f4 is actually a little soft at f4 but works out nicer at f5.6

For the mega-pixel peepers:

FD on the left and Panasonic zoom on the right. You may notice that there is some chromatic aberration around the edge of the garlic stem (against the black of the bottle so it stands out), however close examination of the Panasonic image shows a similar blur halo around the edge of the garlic (contrasted against the black). Since this raw conversion does not do any lens corrections it means that the Panasonic lens is actually nicely corrected compared to the FD lens. (The really astute observers will perhaps have already noticed that the FD200 is actually just that wee bit more in tight than the Panasonic is...)

To go some ways to explain why it is darker, consider this: the FD lens is constructed of 7 elements, while the Panasonic is 16 elements. Think about this for a moment, it means that the light has to go through twice as much glass to get to the sensor.

Looking at the above RAW results, I estimate its about 1 stop is lost. But the image is complex, so I thought I'd look at the results obtained by photographing a more simple subject which will also make clear the levels of vignetting occurring.

FD200

FD200-whiteWall-tiff


and the Panasonic 45-200

45-200WhiteWall-tiff

the vignetting becomes clear here (where it wasn't as obvious in the normal scene). I've also selected a small section of the central portion of the image to try to minimize the effects which are clearly happening on the edges. Open them up and check them out, it actually isn't as much as a stop in the middle, but it is more in the edges.

Because the camera corrects for this and evens it out I'm feel that this goes some way towards explaining why the exposures are different.

So while there is loss (although uneven) it will mean that the 45-200 will appear darker than the same f-stop on the FD lens.

Perhaps is really going to need that OIS even at the same aperture ... which brings me to the topic of ...

OIS

To check this out I decided that I'd wind up the ISO (to 1600 in the lighting conditions) so that I could pull the camera off the tripod and use about 125th of a sec. Given that its a 200mm lens I thought that it would indicate if the OIS makes much difference.


The left hand side is the FD200 and the right hand side the 45-200. These are camera sourced JPG files, and I have tweaked the exposure on them to try to get black and white levels similar with just a simple small curve adjustment. The 45-200 image is a little sharper which could be because of stabilization and could be due to focus being better.

I actually took quite a few images more than this, but presented this pair because I feel its representative of what I found; that the 45-200 is a little bit sharper at 125th sec hand held with OIS and AF than I could get the FD200.

outtakes

Personally I feel that this means that while there are some optical advantages to the FD that the combination of AF (to get better focus) and OIS (to help smooth out the shake a little bit) puts the Panasonic 45-200 further ahead.

Added to this I also think that now the price has come down to something reasonable (was € 399 I paid 200 for this one) that it is well worth buying.

Next I'd like to look at the AF performance and how it can make use of some of the snazzy features that the Panasonic provides.

PS I was out yesterday in the snow with actual daylight (gone again today) and took these two with the FD and the 45-200. I think this shows how significantly the vignetting can intrude on images. The FD was at f4 and the Lumix was at f5.6 (NB both wide open)

P1070485 P1070484

those images link to my Flickr photostream, so you can inspect them there a little larger if you wish.

Sunday 21 February 2010

micro 4/3: where are the compact lenses

I've been waiting to see if Olympus actually make something great with their announcement of a micro 4/3 version of their 9-18mm zoom.

Sadly, looking on DPReview this morning I don't see that they have:

9-189-18

Ok, it looks compact while its folded in and unable to be used, but to me the amount of extension happening makes it look like a flimsy mechanism and I prefer the Oly 9-18mm for 4/3.

I've been using that on my G1, which you can see shown here on the adaptor beside the very compact 14-45mm

bunchOfLenses

Really this new lens is quite a joke, as my regular 4/3 Olympus 9-18mm even with its adaptor looks to be about the same size. So now to use the lens you'll have to unlock it turn the collar to wind it out and turn on your camera.

The old one via adaptor was at least faster and seems to be the same optical design looking at the lengths of this. Reading the specifications on this saying that:

57 x 50mm and 155g - new lens
79,5 x 73mm and 280g - old lens

reminds me a bit of the amplifier specification wars. While a professional level Yamaha claims 100W RMS per channel some chintzy home theater amp for 1/3 the price claims 200W but don't tell you proudly that this is the sum of all the channels and at Peak (a more attractive number not as meaningful as RMS).

I'm sure looking at the extension in those pictures above that with the lens ready to use that the lens will be at least 79,5mm as it is only 40% shorter in spec and the pictures seem to show it doubling its length.

Perhaps the reviews will tell a different story, but until I get my hands on one I'm thinking its just poor.

Saturday 20 February 2010

are The Australian a lost cause?

I read this article and just shook my head. Titled:
"Ill-fated scheme goes out with a whimper"
it just seems to be written by some student of propaganda with a totally out of control or any sense of proportion .

Stuff like:

Then, too, there was the Melbourne woman, who just yesterday thought she'd hired an insulator, but found a masturbator.
what, are you some sort of dunny wall failed limerick writer?

and then:

but for the ghosts of the four men dead, including a young man with a learning disability.

right, again we see unethical contractors at work, and you blame the Government? Take a look else where for the blame on this one ... unless you reckon that's ok and you'd do the same to some poor bloke ...

to further erode any sense of decency or social comprehension...
There was the story of a violent criminal going around to private homes to lay batts. He'd served time in prison for all the things you'd want in a home insulation guy: false imprisonment, sexual assault, conspiracy to murder.

so, your making an oblique statement that our "rehabilitation system" is a total failure as someone who has been rehabilitated should be treated as a criminal ... perhaps you should raise that as a real agenda rather than persecute individuals?

Just three days ago, Garrett was telling reporters the program hadn't failed; in fact, it was a big success. He then went out and had a look at an angle-headed gecko.


well, he is the ENVIRONMENT MINISTER, what else should he be looking at? You wanking in public with your prose? Perhaps you thought he was the roof insulation minister?

Basically your piece is saying that without a touch nanny state that Australians will just rip other Australians off, that no contractor in the building industry has any ethics... hmm, I'm not touching that one. Either way ... is that the misters fault?

I would feel personally ashamed of this piece if I was the editor.

Unless of course the real purpose of The Australian has become a shonky rag to sensationalise stories to the unthinking and try to peddle a change of government ... now that looks like the real thing to me.

Feedin the chooks again mate?

Wednesday 17 February 2010

Google Notebook lovers rejoyce

I have been avoiding upgrading my Firefox for some time because the plugin for Google Notebook won't seem to work.

However I have for other reasons been forced to upgrade and of course at 3.6 there is no support for the old notebook xpi file

However we seem to have a temporary fix. From the comments in this blog:

My FF3.6 stop to work, so i find (here) a link to the "Chinese Post" and follow the tips:
1. Execute about:config
2. Add this key(note: final "3.6" is my version): extensions.checkCompatibility.3.6
3. Value "False"
4. Restart my FF and Notebook works fine, again! I just don't know for while
Good Luck and waiting a new "xpi" version.

I did just this and it works!

There was much rejoicing!


focal length, angle of view, aperture and DoF

Hi

I often hear people confused about lenses and formats:
will a 50mm on a 4/3 look like a 100mm on a full frame camera?
what effect does the multiplier have on full frame lenses when used on 4/3

surely just cropping the middle is not what it looks like...

Since this topic goes round and round with people endlessly getting confused I thought I would try to answer it here.

Any lens at a constant aperture diameter always looks the same as long as you are seeing the same angle of view. So yes, a 50mm on a 4/3 will look exactly like a 100mm on a 35mm taken from the same spot with the same sized aperture.

Some of the feed back on my blog indicates that a graph just confuses people and more complex stuff like maths (formula which boils down to 4 divided by 2 equals 2) just sends people into a spin. So rather than reiterate that the division x2 essentially is accurate enough (gosh I think I just did) I'll show it graphically, since a picture is worth a thousand words.

So to answer this question I thought I would provide six pictures (equivalent to a verbal length of six thousand words) to try to clear this up for those how are still trying to push the wrong ideas down others peoples throats.

method


I put the camera on a tripod and did not move anything, I took 3 images only replacing the lens and focusing each on the same point and set the aperture to the same size on all shots. I used the following lenses, all Olympus OM series:
  • 21mm
  • 50mm
  • 100mm
these make a nice progression to explore the effects of visuals.

No matter how wide a 21mm lens is on a full frame camera ... if you take a wide picture and crop out the middle then its not wide anymore is it?

This is exactly what the effect of putting a 21mm wide angle on a 4/3 sensor camera is.

Ohh ... it can't be that simple can it?

well, yes actually.

First the overviews:

21mm (normal angle of view, equal to about a 50mm on a full frame)
P1070285-20@3.5

50mm (which on 4/3 is like a 100mm on full frame, and looking exactly like a crop of the above image from the 21mm lens)
P1070284-50@8

100mm and this also looks like a crop of the above or a x4 crop out of the 21mm above
P1070283-100@16

If that's not convincing enough, I thought I'd actually crop the guts out of them and present that here too:

21mm (this one has to be cropped the most ... of course this is more or less 100% as the 4000 pixels became only 744 pixels by the time I've done the crop to essentially x4 the focal lenght {21 x 4 is nearly 100})
P1070285-20-seg



50mm (cropped essentially to half the frame width / height of the 100)
P1070284-50-seg


100mm (and the image from the 100 simply scaled down to be the same number of pixels as the crops)
P1070283-100-full

do you see any differences?

So you can expect that any 35mm lens used on a 4/3 camera will look exactly like cropping it down to half the width and half the height.

Even if you start talking about focal length multipliers (god knows why), its worth pointing out that DoF is a function of the diameter of the hole (the pupil diameter) and essentially nothing at all to do with the focal length.

I've put a page comparing this on my 20D compared to a compact camera here. But as you can spot in the above all the DoF are the same. Well the reason for that is that I used different f-stops but the same diameter of aperture on all the above.

F-Stop is quite simply the focal length divided by the diameter of the hole. So if you have a 21mm lens then the hole size at f3.5 is 6 mm (21 divided by 3.5 = 6 sorry about the complex maths here). It works out that for the other lenses its:

50mm f8 6.25
100mm f16 6.25

which isn't quite perfect, but close enough.

So again looking at the above, for the various focal lengths you can see that the DoF in the image is actually the same when you just crop them to be the same.

I hope these images (and the extra words) help you to understand this point better.

Folks also ask another question which is "is f1.7 on a 20mm on a 4/3 camera then equal to f1.7 on a full frame?"

the answer to that is no, its acutally more like f4 on a full frame in terms of depth of field. Why? well that's another blog post :-)