2015/07/28

Small apple and big orange: Panasonic 45-150mm f/4-5.6 vs Olympus 40-150mm f/2.8

Traditional wisdom suggests to compare apples to apples, and so there are many comparisons, and many more opinions, about which of the two telephoto zoom lenses of both MicroFourThirds (MFT) manufacturers is better: Panasonic 45-150 mm f/4-5.6 or Olympus 40-150 mm f/4-5.6
I happen to use Panasonic lens on Olympus body (E-PM2), and that makes the great mid-range zoom combo, possibly the best in image quality per size/price. I do not complain neither handling of the camera (with this lens it is acceptable) nor the lens itself (heavy vigneting, red-ish color cast), as the range between 100mm - 200mm focal lengths (in 35mm terms) seems to be the least popular for me, and for the wide angle and supertele I have more appropriate tools.
Nevertheless, this lack of popularity might be coming from subconcious feeling not to compose it that range due to inferior output. Because of this lack of importance to improve in telephoto department as well as a bit prohibitive price and most of all the size of Olympus 40-150 mm f/2.8,  I've always been lukewarm towards that piece of glass. However when I learnt about Olympus TEST & WOW offer, I was ready for the challenge.



Usually reviews are focused on picking the one from a bunch of comparable products. Challenging results produced by two so different lenses is a different matter: to figure out if an upgrade is worthy. That aspect seems to be missing, so I hope to help folks in similar situation.

Let's start with the most important difference: Oly offers 1 stop more at the short end, and 2EV more light when at the maximum reach. That is a lot of a difference: in case of achievable shallow depth of field and most of all in terms of light collecting ability. That may make a difference between acceptable image and no image. That is the reason it bears 'PRO' badge - it gives greater confidence in succeeding with an assignment. On top of that it is splash, dust and freezeproof, comes with tripod foot and retractable hood. The hood is lovely - it takes no time to extend it, folding for transport is immediate too. The foot is not so lovely: it is built like a tank, but with the lens being front heavy it painfully reminds (if the load of the neck wasn't enough) on the chest what optics is currently attached. It is secured with a nut, and that is located exactly where a palm of left hand supporting the lens is - rubbish (Olympus seems to have devious tradition of locating elements - like neck strap eyelets - so they poke your hand). When loosened, the foot can be spun around the lens, but apart from painted dots helping to align it, there are no detents setting the position at the quadrants of the collar. The foot can be completely removed (turn it to the 'upside down' position and pull towards the lens mount), but then it exposes four small screws - I suspect they are for foot guiding only, but in case they do set optics element I'd rather not touch them. Also without the foot the barrel looks a bit naked - some masking plastic ring would be a nice addition to the kit. Removing saves however quite substantial 121 g of total weight.
 

Let's finish with the most obvious difference: it is large (in MFT terms). I used to have Panasonic 100-300 mm, but 40-150 mm f/2.8 is in another league. Saying that, its metal build is of very good quality, with hefty feel reassuring that this lens is going last. It does not extend when changing focal length. That for sure helps with sealing agains elements, and explains its length.

A bunch of numbers:

Panasonic
45-150 mm f/4-5.6
Olympus
40-150 mm f/2.8
Length without hood
73 mm
158mm
Length with hood
105 mm
220 mm
Length with hood @150mm
143 mm
220 mm
diameter (hood)
70 mm
95 mm (105 top to foot distance)
filter diameter
52 mm
72 mm
weight (lens only)
198 g
877 g
weight (with front cap)
210 g
902 g
weight (with hood, no cap)
220 g
1004 g


Test procedure

Tripod loaded with a dumbell for extra stability, located indoors, test camera (Olympus OMD E-M1 firmware v3.0) set to ISO200, IS off, single shot with 0s anti-shock, external cable remote, AF single. Raw images imported to Lightroom 5.7.1 with no settings applied (other than under-the-bonnet lens profiles and default 25/1/25 sharpening. Since all the images come from the same camera, whatever happens here does not matter too much anyway, as the lenses are the only variables). The best of 3 images made at each focal/aperture setting taken for analysis. Pictures exported to jpeg at 90% quality, sharpening to screen - normal. Crops for side-by-side comparison saved in IrfanView as jpeg at 90% (different percents than LR) quality. I have no intention nor tools to produce absolute numbers, just naked eye to subjectively judge which from a pair of images looks sharper. Taking advantage of the setup, some other lenses occasionally get added to the mix, to see how they cope in such noble party.


40mm
With guest appearance of Olympus 12-40 f/2.8 . Distance to target about 20m, focus (centre point) on the sign. For edge sharpness, focused with rightmost point. That will show the best case scenario, as when focused in the centre, the side will likely be offset (field curvature), although at these depths of field the effect should be negligeble. Click on the images - unless you can already see differences :)


 Centre                                    Side         
 

Olympus 12-40 f/2.8 is a bit wider at the long end that 40-150 f/2.8 at its wide setting, so there is some overlap in the system. Both lenses are compromised, being at the end of their ranges, but 12-40 seem to be affected more, showing clearly worse macro- and microcontrast in the centre part. This lens shows some colour fringing evan at f/4, whereas 40-150 seem to be free from chromatic aberration, even wide open. When looking at the side, situation is different: 12-40 is pretty much equally sharp as in the centre (except for soft f/2.8), but 40-150 now loses at each setting, with delicate purple edges at f/2.8 and f/4. So having both lenses Olymous gives a choice: moderate, but uniform sharpness of 12-40 or excellent centre paid for by worse sides in case of 40-150. Light falloff in 12-40 disappears gradually all the way till f/8, in 40-150 is gone by f/4


45mm
Here Olympus 45 f/1.8 joins the party and the real battle between Olympus and Panasonic begins.
 Centre                                                          Side    
 

Let's start with the lonely prime: at the centre it starts very soft (and at f/1.8 it is much worse!), but at f/5.6 the macrocontrast seem to be in the region of 40-150. Microcontrast is worse though. Panasonic 45-150 starts at f/4 like 45 f/1.8 stopped down to f/2.8 (bar purple fringing) and stays there. 40-150 f/2.8 nice and crisp, until diffraction starts affecting at f/8.
At the side, the story of 45/1.8 repeats: very soft at f/2.8, acceptable (sharper than 45-150 in the cente) when closed more. Since it was designed to be an affordable portrait lens, where out of center area being of poor contrast is actually an advantage (adds to diffraction induced backround blur), that is even better than expected. It is actually as good as 40-150/2.8, this one winning only at f/2.8, but also showing some traces of chromatic aberration. Surprising winner here - even if not a lot - is Panasonic! It also means that at this focal lenth its sharpness profile has got a 'mustache' shape, with peaks towards the sides, what explains worse centre.

75mm

 

At this setting centre only: side for 40-150 mm f/2.8 got missed. Olympus wins hands down, except at f/8, or rather at exposure time of about 1/125s, where the images are clearly blurred, and it seems only in vertical direction. It looks like this is the poltergeist of modern cameras: shutter shock. Leading shutter curtain induces vibrations when it stops at the end of travel. Always! Only for short exposure times the effect is negligeble, as exposure is short enough to avoid moment of significant  movement, and for long exposure, the time of shake is small comparing to the total exposed time. So there is a specific range of exposure times, where the whole magnitude is captured. Specific, as it depends on the lens attached: its weight (and its distribution) affects resonance frequency, focal length translates the offset proportionally (like in normal camera shake). If you're interested in the topic, read the excellent piece by Jim Kasson. To avoid that, new cameras are equipped with electronic first curtain (no blade travel, no shock created) in case of OMD E-M1 described as 0s delay. I swear I enabled it, but clearly that was not the case - thus it is crossed over in the description above. Anyway, 1/125s is barely the recommended exposure time for 300mm (equivalent) focal length (although IBIS or tripod do not render it totally useless)

100mm
           Centre                                          Side             


Olympus again much better than Panasonic. At this setting the latter starts at f/5.5, so there are already full 2 stops of relative aperture difference. In both cases the side is as good as the centre. Olympus is well shaken at f/8 and 1/125 s in the centre, but even with this handicap it manages to makes it pretty equal as Panasonic without such issue - that is the amount of 'natural' blur in the cheaper lens. The side of Olympus at f/8 is even more affected, exposure time was 1/100 s.

150mm
           Centre                                          Side             

Olympus leads significantly unless limited by vibrations, Panasonic again seems to be slightly sharper towards the edge than in the centre

Let's see what happens when the distance increases:


At 45 mm setting Olympus leads visibly at f/4, and the closer to centre the better. At f/5.6 Panasonic pulls forward, with centre matching Oly, but edges being better.
At 100 mm setting pictures from Panasonic look like those made with bigger lens through a delicate fog: contrasty edges look similar, but fine detail is blurred.
At full stretch there is not much of a difference: fine detail is lost due to air quality (heatwaves). That is one of the limitations of far distance photography which one needs to bear in mind before splashing $$$ for the greatest kit, which will be handicapped anyway.

On the second day I repeated some of the tests, this time making sure that anti-shock setting is on. Also included other cameras from my zoo to see what tricks they can make in presence of the Oly beast. Because of that reason images were downsampled to match Nikon 1 V1 long edge, which coincidentally also equals dimension of 4K display width - likely medium of presentation of the pictures in the forseeable future. This time focal length is quoted in full frame equivalent values!

150mm (equivalent: 75 mm for 4/3" sensor, 58 mm for 1")


At f/4 FZ1000 not far from Olympus 40-150 mm f/2.8, and at f5.6 I'd call it a draw: brute force of 20 Mpix does the trick. Setting of f/8 is the best for Panasonic 40-150 mm, Olympus suffers shock, and FZ1000 feels the fate of diffraction. FZ1000 metering system seemed to have protected highlights, thus shorter exposure resulting in darker images.

200mm (equivalent: 100 mm for 4/3" sensor, 74mm for 1")




Not a lot of difference in setting and no difference in results. The 4th set: Nikon V1 with 70-300 CX started off very well, even though handicapped by the lowest number of photosites on the sensor.

300mm (equivalent: 150 mm for 4/3" sensor, 112mm for 1")

FZ1000 starts to soften and lags behind, all 3 other setups head to head. Interestingly, 40-150 mm f/2.8 gets soft quickly when stopping from f/5.6 to f/8, even though this time there is no visible shake, and 0s delay was definitely on! Can there be another source of vibration at around 1/100 s exposure time? The other curiosity is Panasonic 45-150: this time it is not that much worse, unlike when tested the day before. Is it after all also affected by vibrations, just of different frequency/magnitude?

420mm (equivalent: 210 mm for 4/3" sensor, 155mm for 1")
FZ1000 just about stretches (its range finishes at 400mm), but the winner this time is Nikon 1 duo. How did Olympus lens lose, and how did it even reach that far? The answer is in its teleconverter. By adding 15mm of thickness it multiplies focal lenght by factor 1.4x. No free lunch though: maximal aperture is reduced to f/4, image clearly suffers and it is quite chunky piece of glass and metal, adding another 118g (with caps). Because of its construction it cannot be used with any other MFT lens (it engages into the back of 40-150 mm f/2.8)


Focusing
The sharpest lens will be less useful if cannot acquire focus fast enough. That is what traps Panasonic 100-300 for example. Because of good light conditions, single focus was instant in both cases. I did some unscientific comparison of continuous AF as well, aiming fot the passing cas on a motorway. in both cases the result was identical: after about 1s of hesitation, from the moment the camera locked on the targer the whole series was in focus. It was actually a bit easier to track with Panasonic lens attached, because blackout of EVF felt shorter. I suspect it was due to slightly lower framerate achievable with that lens (= longer preview time).


Closing thoughts
At the end of two days with Olympus PRO 40-150 mm f/2.8 I have got mixed feelings. The lens is quite remarkable and thus it is not easy to judge it. It seems like it was designed as a non-compromise optics offering fast aperture and possibly uniform sharpness across the frame. The question is though why would one need both at the same time? Short end covers typical portrait focal lenth, but if one is interested in that, then there is a selection of dedicated ~ 45mm primes by both Olympus and Panasonic (Leica), offering even shallower depth of field, much lighter and cheaper. At this end 40-150 mm f/2.8 isn't uniformity champion neither. The long end is likely to be futher cropped, so the corners aren't that critical. All purposes I can imagine then are some sport activities (or moving subjects in general), but not sure if E-M1 is the best choice when it comes to continuous autofocus. Taking into account price and weight, 40-150 mm f/2.8 seems to be in the same league as 70-200 mm f/4 lenses on APS-C cameras, which in many cases may offer better performance per size/weight/price ratio. Subject separation is also easier to achieve in larger sensor systems to begin with, so the lens brightness in respect to shutter speed seem to be the only consideraton. Within MFT camp, Panasonic offers its 35-100 mm f/2.8 at two thirds of price and less than half of weight. Especially if one is already equipped in something covering focal length above 100 mm (like Panasonic 100-300 mm f/4-5.6, or recently announced Panasonic 100-400 mm), the range of up to 150 mm offered by Olympus is less of the advantage (unless, of course, f/2.8 aperture is a must). In that case included teleconverter also becomes completely obsolete (not that is of much use anyway). The other option to fill the gap in the 45-100 mm range, is to take primes route: to complement 45 mm f/1.8 (or 2 available Panasonic lenses) with Sigma 60 mm f/2.8 (also available Olympus Macro 60 mm f/2.8) and Olympus 75 mm f/1.8. Lower total purchase price and weight, versus flexibility of zoom lens.

Overall the unique selling point of Olympus PRO 40-150 mm f/2.8 seems to be short achievable exposure times in wide focal range (4x zoom) lens with very good image quality, consistent across the frame. That makes the neck heavy and the wallet light though.

How did the Panasonic do? You get what you pay for: at £130 (what I paid, split kit) I have nearly pocketable solution, which with E-PM2 does not take too much space and weight in my bag, yet provides always-at-hand telephoto tool and a backup body. Olympus PRO glass would cost me about 10 times more and put an extra kilogram into the bag. With results comparable to Panasonic at short end, challenged by Nikon 1 from 200 mm (35mm equivalent) onwards and not many reasons to use f/2.8 it definitely wasn't a keeper.

2015/07/27

2015.07.25 Dorchester - Hardy Monument bicycle trip

Our exploration of southern coastline of Britain took us to Weymouth. Since spending all day there seemed too challenging, we started off with a bicycle trip from Dorchester to Hardy Monument and back making a loop. The trip was inspired by this route, we modified it by following National Route 2 all the way from Dorchester and doing last two climbs on foot. We used an excellent map found here.
Bicycles were rented from Dorchester Cycles.
The climbs were challenging, but the views were fantastic and the descent from Portesham just lovely - virtually no pedalling needed. See the tracklog of our journey.