Web Analytics RiceHigh's Pentax Blog: October 2008

Sunday, October 26, 2008

K-m Sample Photos Carnival (as at Oct. 26)

Here is a summary of the K-m sample photo galleries or links that I could find so far:-
Item No.
URL
Source
Full Size?
EXIF?
Language
1.
http://www.pentax.jp/japan/imaging/digital/slr/k-m/ex.html
Official
Yes
Yes
Japanese
2.
http://www.focus-numerique.com/test-73/Pentax-K-m-test-exemples-photos-15.html
Reviewer
Yes
Yes
French
3.
http://pentaxplus.jp/k-m/gallery/index.html
Official
No
Yes
Japanese
4.
http://pentaxplus.jp/k-m/gallery/01.html
Official
No
No
Japanese
5.
http://pentaxplus.jp/k-m/gallery/02.html
Official
No
No
Japanese
6.
http://www.dchome.net/viewthread.php?tid=584867&extra=page%3D1
User
No
Yes
Chinese (Traditional)

My observations and comments? Yes, as follows:-

1. Colour reproduction looks good and accurate for "Natural" and "Portrait" image tone settings, but not so for the "Bright" mode or anything else (Look at URLs #1 to #5);

2. Focusing accuracy seems to be fairly accurate (and they did use AF mostly) (Look at URLs #1 and #2);

3. Exposure accuracy with ambient light, for those pictures of which the EXIF data have been shown, *seems* to be improved as most of the photos with normal scenes are *now* without (the need of) exposure compensation (and they mostly used the "Pattern" metering) (Look at URLs #1, 2, 3 and 6);

4. Noise control and level at high ISO are very good in my opinion, with good amount of texture and details retained. The most favourable thing is that the noise really looks film grains alike, with more visible luminance noise rather than chroma (colour) noise, which just mean very usable images and high image quality up to ISO 1600 (Look at URLs #2);

5. Flash exposure accuracy (in this case with the AF540FGZ flashgun, in P-TTL mode), is yet unsatisfactory and there is a great tendency of underexposure. Just look at the flash photos in URL #6, all photos required a +0.5 or +1 EV exposure compensation so as to make the photos to look properly exposed as by the flash. Besides, the image quality and skin tone produced by the K-m, 540 flash and the DA*50-135 do both just look somehow mediocre and indeed it is just nothing is to be excited about. (Despite that the photographer is an experienced one who has very good skills in taking portraits for girls). So, does the P-TTL underexposure tendency and exposure inaccuracy just a ghost that Pentax could never get rid of??

6. The only sample made by the new DA*55/1.4 lens seems to re-produce very favourable colours and skin tone (Look at URL #3, and direct link for the pic here). In fact, I have missed this favourable genuine Pentax tasted colour response which I like for long since the birth of those DA "Pentax" digtial lenses (most of the time, look here for why and what's different).

Well, there will surely be more full-sized, un-retouched and original K-m sample photos to be come (no matter by end-users or by reviewers), I shall report back here later when I find something new, interesting or special, which are worth my further reporting, if any. And, I do hope I can confirm more firmly most of my findings above when more un-retouched/straight-out-of-the camera sample photos are posted.

Tuesday, October 21, 2008

Who were Telling the Truth?

In April, it was reported that a Samsung high-ranking official disclosed that they were "actively pursuing development of a full-frame CMOS sensor. But now just about six month later, yet another more senior official of Samsung played down "rumours" that Samsung was developing a full-frame sensor in another interview. So, after all, who was telling the truth in this case?

"Coincidently", same thing happened at the Pentax side. In June, the Pentax China Head disclosed that a full-frame Pentax DSLR is under planning. But then just in September (3 months later), an even higher ranked Pentax official said that they had no plan for a full frame DSLR, at Photokina! Yet again, who was telling the truth?

Do you think that the policy, project plans etc. could be changed so quickly and drastically within a company (or just an organisation, anyway)? It is very unlikely, frankly. So, who were telling the truth, then? Well, my guess is that it can be either two or all four! If two, either the former Pentax and Samsung officials or the latter two were telling the truth. But, why it can be all four??

Well, if no one was actually lying (who dared? why needed?) then it could be very likely that someone were intentionally playing with the words so that the full-frame bubble was burst *in the meantime* just because *in the meantime* Pentax and Samsung are just unable to deliver such a full-frame body and DSLR system. Remember what Nikon were telling the world just less than two years right before the sudden arrival of the full-frame D3? Still remember? It seems that history could be re-happening and if we look back in what Nikon said and what Samsung/Pentax have now told us, anything sound so similar and too familiar? (Of course I have my own wishful thinking and imagination here so that some people should just choose to "ignore" me! ;-))

Wednesday, October 15, 2008

Yet Another New K-m Preview

A new K-m Preview is out today at the DC page of the Japanese Impress Watch, here it is:-

http://dc.watch.impress.co.jp/cda/review/2008/10/15/9420.html

The article shows us something new and interesting, say:-

1. The bright green power LED indicator does actually indicate also the power level, namely, strong, weak and empty. With the deletion of the top monochrome LCD status indication panel, I think this design is just clever and brilliant as it is simple but yet provides very clear and easy-to-see indication to the user/operator;

2. Battery type can now be entered into the camera or the user can still let the camera to "Auto Detect" the battery type in use. The supported battery types (of the 4 AA batteries) are namely: NiMH, Alkaline or Lithium. It seems that Pentax have finally done something to improve the power stability when they had to insist on the AA battery solution in their lower level DSLRs but which has been proven to be ultra problematic for almost half a decade. Well, my MZ-S battery grip did use to have a battery type selector as it is told in the user manual that it must be selected for the proper operation of the camera. Now, Pentax could only re-think and re-introduce such an essential feature after 7.5 years. Huh? But Late is better than None, anyway! And also let's wait and see if the problem is really completely resolved!

3. A few more screen shots for camera's recording functions, playback options and digital filter / color setting options etc.

Furthermore, there is a shutter sound wave (.wav) file which is available for download in the page. The first time I downloaded it and heard its sound did make me feel really excited for its light touch and smoothness, which just doesn't look like any previous Pentax DSLRs and it would be an improvement over its predecessors. But however, I noticed very shortly about the file name of "5D2_01" for the wav file which is just highly suspected that it is the sound of the 5D MkII instead and the editor has very probably been mistaken.

So, I just browsed through the front page and then go to their 5D MkII Preview page and it is confirmed that my suspection is correct, i.e., the "K-m sound file" is just the 5DMkII one! :-( Now, I just hope that the editor can make the correction and put a true K-m sound file for my measurbation, like I did myself for those, very soon! Anyway, I have already measured the overall system lag time of the 5D MkII to be 105 ms which is indeed 27 ms longer than the original 5D but nonetheless is still faster/shorter than that of any of Pentax DSLRs ever tested and published for the timings. My 5D MkI produces clunky large shutter/mirror sound, though, although it is yet faster.

Update (Oct. 16):

The editor has noticed the error and made the correction already. There are now two sound file links in which the wave files should be the K-m's ones this time: One for the Single frame mode and the other for Continuous shooting mode, the direct links are as follows:-

http://dc.watch.impress.co.jp/static/2008/10/14/tansha.wav
(Single frame mode: ~2.1MB)

http://dc.watch.impress.co.jp/static/2008/10/14/rensha.wav
(Continuous shooting mode: ~3.3MB)

From the two sound files, I have measurebated the following actual performance figures of the K-m:-

1. Actual shooting rate/frequency is only at 2.87 frames per second (well below and much worse than the claimed Spec. of 3.5 fps);

2. A maximum of 10 frames can be shot up to in a chain (doubled and better than the official Spec., which is 5 only);

3. The system time lag (shutter lag) is measured to be 114 ms or 0.114 second (the best figure I have ever seen for an entry level Pentax DSLR - not bad for an entry level DSLR indeed).

I would comment that it just sounds to be smoother for both mirror up and down actions which are now having a lighter touch or simply gentler (for what it was recorded and as it seems to be - the actual noise level cannot be judged exactly, though). Well, well done, Pentax! (I still would say for that part, but the lower fps is yet something which I do not wish to see..)

Tuesday, October 14, 2008

DA 18-55 on Full Frame

I have taken some test shots with the Pentax DSLR kit lens DA 18-55 on my (K)5D. Indeed, quite some Pentaxians have been wondering and questioning about how those DA lenses could produce if they are used on Pentax film SLRs. However, only just several Pentaxians had actually tried to do some tests and shot through the viewfinder or to scan exposed films and presented the effect(s) publicly before. Now, I think this series of my shots will show exactly how this Pentax standard kit lens will behave when it is put on a full frame (digital) camera. Look at the projected images below:-

@ 18mm (at f/4.5)


@ 20mm


@ 24mm


@ 28mm


@ 35mm


@ 45mm


@ 55mm (at f/8)


Remarks:
1. All images were taken with one f-stop closed down from wide opened;

2. All images were taken in M exposure mode with 1/8" Tv at ISO 800. As such, the tele images came out to be (at most 1.5 stops) darker, so as to exaggerate the vignetting effect at the tele side, for easy inspection.

Well, now you can see the contracted image circle(s) at the wide side of this digital kit zoom which is designed for the APS-C sensor format. As a side note, the infamous (de-)centering defect of (the DA/DA*) Pentax lenses has once again been verified. Just look at the images produced at the wider angles, it can be more easily seen that the projected image circles have been shifted to the right, that is, the optical center of my lens at those focals is actually yet mis-aligned to the left (when I am facing the back of my camera). But the even more strange thing is that when the lens is zoomed to the tele side, the de-centering is reversely shifted to the left (for the images produced, just look at the (positional differences of the) corner vignettings). In fact, the only symmetric image I could get is the one shot at 35mm. Huh? What? How come? Why?? :-0

So, some people may question if the sensor of my 5D and/or the mount/adaptor is/are just mis-aligned. But what I can tell is Not. It is simply because my Pentax F Fisheye Zoom produces symmetric images when at 17mm where a little bit of the built-in lens hood can be shot (for the technical reason for why it has to be, look here for the details). Just see this test picture:-

http://img.photobucket.com/albums/v313/RiceHigh/FF%20Fisheye%20and%2043%20Limited%20on%205D/IMG_6954s.jpg

After all, the QC or simply quality difference(s?) between Pentax lenses that were made in Japan against those are / were made in Vietnam (or just nowadays) has yet been verified.

Thursday, October 09, 2008

Sample Photos of Full Frame Fisheye and 43 Limited on 5D

Further to my last blog entry, here is the sample photo gallery for some test photos taken with my film lenses, the Full Frame (not with a dark circle) Fisheye Zoom F 17-28/3.5-4.5 and the FA 43/1.9 Limited on my new K-mount 135 Full Frame DSLR, the "Cantax" "K5D":-

http://smg.photobucket.com/albums/v313/RiceHigh/FF%20Fisheye%20and%2043%20Limited%20on%205D/?start=all

A few more footnotes about the shooting conditions/settings as well as my observations and comments on the final image quality (for the last listed point):-

1. All pictures were taken in Large Fine Jpeg file mode at a full resolution of 12.8 MPs (4368 x 2912);

2. A -0.5EV exposure compensation was applied for most pictures (as both lenses on 5D overexposed by a half-stop most of the time - all the exposures were accurate and consistent thereafter);

3. Most samples were taken at aperture values ranged from f/4 to f/8 with the most usual used aperture at f/5.6 (and around f/5.6 for the Fisheye Zoom);

4. The K5D was able to focus at apertures not smaller than f/8 on bright objects but with best sensitivity at f/4 or faster. The focusing accuracy and hit rate are both considered to be very high (as you can actually see from the samples);

5. All uploaded pictures are down-sampled to 2048 x 1365 and saved at Jpegs at 90% quality level with Irfanview. No anything else is processed and edited, the EXIF is retained also;

6. The totally dark corners as seen in a few samples of the Fisheye Zoom on the K5D is caused by the rotated built-in lens hood for what the adapted Pentax lenses are all now rotated to the left in about 20 degrees (owing to the consideration to avoid the aperture coupler of K-mount lenses in conflict with the bottom part of the EF-S mount of Canon APS-C bodies so that the K-couplers need not to be removed but can still be adapted). As this Pentax Fisheye Zoom is of 180 degrees in the Angle of View, the original built-in lens hood is designed to be in a pedal / flower shaped rectangular form and thus the upright position is critical. I have already found out a way to adapt in the fisheye zoom with less rotation but slight modification to the adaptor is required and I shall buy another adaptor to do the modiffication shortly. The widest setting for what the full vision of the Fisheye Zoom is not blocked by its lens hood is at around 19mm (With modification to the adaptor, it can be further improved down to 18mm and thus close to the ideal 17mm with full 180 degree AOV);

7. The optical quality of the Fisheye Zoom and the 43 Limited on the 5D are both really really *Amazing*, as I surely knew that they are, for years I used them in the Pentax film age. To list a few:-
  • The flare control is superb, which I have never been able to see any Canon lenses (L or not) can attain that level of flare resistance and retain image clarity with strong light source / contrast in frame. Just look at the sample on board of the boat of which the camera was pointed directly to the morning Sun (which has been burnt in the picture for that part, of course);
  • The resolution is amazing and with all that Canon crispness and sharpness, accurate colours (in "Neutral" Picture Style). Clipping of colour channel(s) is minimum - just look at the red boat shots and those for the flowers;
  • The corner sharpnesses are yet amazing for both lenses on 5D. CAs and Purple Fringings are minimal even for the Fisheye Zoom. There is little corner blur nor obvious vignetting neither (far better than anything produced by the DA Fisheye Zoom on any Pentax APS-C DSLR bodies (which have far too more all the above problems mentioned which are actually not found with my 5D/Pentax F Fisheye Zoom combo);
  • The resolution of the FA 43 Limited on 5D, for just shooting at just f/5.6 or f/6.7 is stunning! The super ever large image file size says it all (I have never been able to see such high resolution neither with my EF 50/1.8 nor my EF 24-105/4 L)! Just inspect this sample yourself! (Of course, to get this super high resolution and image fidelity, a very accurate "AF" system is crucial);
Well, these are just the results of the magical combination and fusion of the best optics with the best DSLR body on Earth! But of course, it needed to be Full Frame at the first place! (so as to get beyond the diffraction limit of Physics first - even putting aside all those other compromised design limitations of the APS-C DSLRs)

Wednesday, October 08, 2008

Preview: Cantax K5D Full Frame! ;-)

Since Pentax have just told the world in the Photokina that they will not have a 135 Full Frame DSLR forever, I have opted not to wait them anymore (as they told us to do so, clearly). I have managed to build my own Full Frame K-mount DSLR model just within an hour, which can adopt many of my old excellent Pentax film glass. Here it is!



More to come! (Modifying procedures, things to note and cautions to take, detailed comments on system accuracies (so far so good!), sample photos with various Pentax lenses and etc.!)

P.S. My special thanks to Steven Lin, who has kindly offered me a helping hand and updated his already excellent blog article on the topic (in Traditional Chinese), for the sake of even more clear instructions on the modifications.


Update:
Compatibility List of Pentax Full Frame AF Lenses on Canon 5D Body

Chinese Dragon Boat Race

Last Sunday, I went to a local Chinese dragon boat race event, brought along just with my Pentax K100D DSLR and a F* 300/4.5 and shot some photos, here is the album:-

http://smg.photobucket.com/albums/v313/RiceHigh/Dragon%20Boat%20Competition/?start=all

(Or playing the Slideshow here)

Whilst the K100D is surely not the fastest DSLR for its AF in the Pentax land, it does let me down each time (and actually made me annoyed) when I got it to do somewhat "action" photography with just a little bit more movements and changes of the subjects. I shot in the early afternoon and at that time it was slightly cloudy and is actually just slightly bright of the ambient environment.

As I have been used to the responsive, fast and accurate Continuous AF (and Single AF as well) of my Canon 5D and the EOS system now, the AF of the Pentax DSLR system is just a pain to great pain to use, including the latest K20D which frankly I have also coincidently used indoor in the evening of the same day in a dinner (and it hunted much).

The most annoying things for that Pentax AF is: 1. First of all: Hunting, for whatever AF mode it is in, and even the environment is somewhat bright. I used the same F* lens with my MZ-S for years - no problem!; 2. Second, the "Continuous" AF is not usable and is just a joke. It is even slower than the Single AF and actually it is just a looping Single AF but not a true continuous focusing action as taken by the camera (yes, it is NOT). The worst thing is that this "Continous AF" (or I would give it a new name - Pentax's unique "Continuous *looping* Single AF") will never give the photographer to release the shutter unless it is to complete one Single AF cycle!(What?) (and UNTIL it thinks that it has achieved a focus and then re-start another cycle again!). So, I was forced to use the Single AF, which is actually faster and with a higher "allowable" shutter release rate but then of course I got quite some out of focus photos then, for "improperly" using the Single AF for moving objects.

People may agrue that higher grade and/or latest Pentax "flagship" DSLRs should have better AF. However, just believe it or not, what I can tell is that it could be better but not much better. My opinion is simply that the old flagship K10D had similar ridiculous AF behaviours and the K20D was somehow debugged. But the problem is that even the K20D is NOT *much* better. Its AF just struggles for achieving focus even for Single AF in just a little bit dimmer environment indoor and the worst is it *still* front focuses for tungsten light source.

Also considering the longer system time lag of the Pentax DSLRs, they are not ideal for any action or sport photography. In fact, I found that I was the *only* Pentaxian in the dragon boat race event and all other shooters I saw were Canonians and Nikonians.

I tried the Canon 1D and 1DS at Canon's showroom last month when I took my Canon flash gun for repair. Indeed, they are really *amazingly* faaast, especially for the nearly non-existent system/shutter time lag, which is surely noticeably faster than my 5D, which is then again noticeably faster than all Pentax DSLRs I have used (including the K20D of course)! In fact, when those differences can be noticed, the time lag and system response timings could actually be quite significant as they are so obvious.

So, just after all these moaning and bashing (for what quite some particular Pentax fanboys which like to call me of being so from time to time), now think about the following two questions: 1. Was I able to take the photos? Yes. 2. But did I miss many photo opportunities? YES! So, I think this would answer many of the questions for those who like to attack me for either showing one or two of their "great" works or to ask me for my photos! (To those: there are nearly a hundred of photos shared and shown in my album above. Are these enough this time? :-))