So, people are wondering what's up with the Fuji S5 Pro's image quality. The mere four official FujiFilm images are unremarkable.http://fujifilm.jp/personal/digitalcamera/finepixs5pro/portfolio.html
The PC Home web site's sample images made with the 24-120 VR Nikkor and a Tamron 28-75/2.8 and look very,very awful--just a drecky lens,their 24-120 VR,and the subject matter--Ack!. Will Crocket of Shoot Smarter University has recently put up a portrait sample taken with a Tamron 28-75 shot at f/5.6,and it looks very average in terms of resolution.Get the file here and examine it http://www.shootsmarter.com/downloads/S5_Wcrockett_01.zip
Crocket also has posted two ISO 1600 samples,one with the 28-75,the other with the 70-200 VR Nikkor, and while they do look nice, they do have pretty heavy,obvious Noise Reduction issues; there's a pretty high loss of detail,and quite pronounced, patterned noise reduction artifacts quite visible at 100 or 200 percent views, but when down-sized the images do look pretty good for 1600 ISO images in my opinion. According to Crockett, these two shots have fill-flash as 20% of the exposure and ambient light makes up 80%.
http://www.shootsmarter.com/downloads/S5_Wcrockett_02.zip
Crocket started a thread on the dPreview web site at http://forums.dpreview.com/forums/readflat.asp?forum=1020&thread=21958912&page=1
and his thread has quickly hit 70+ replies.If you stop by, you can see a couple of 100% crops showing the breakdown of hair detail in the S5 Pro photo Crocket offers as a sample; there's a lot of artifacting, where the individual hairs appear in the picture to just clump together,and are rendered as blotches,and not as discrete strands of hair. Image processing side-effect or noise reduction at work in the darker,hair tones, perhaps ruining the differentiation of the hair strands? Or lousy demosaicing of strong,closely-spaced lines, as found in things like hair, fabric,etc? I guess time will tell.
The DC Watch or Impress "Photogenic Weekend" samples from Japan are the BEST two sets I have seen so far from the S5 Pro, but there are some issues with the amount of detail resolved--the images are downsized to 2,300-pixels roughly, so we're not seeing original images and we're not being told what type of noise reduction,if any, was being used in-camera, nor are we privy to what types of post processing has been performed on the original captures. DC Watch has profiled the S5 Pro on the weekends of February 2 and February 9. Today's Photogenic Weekend is at http://dc.watch.impress.co.jp/cda/weekend/2007/02/09/5550.html
The four Photogenic Weekend shoots for January of 2007 were devoted to the Nikon D40,and I compared photos from those four shoots with all the two shoots' worth of Fuji S5 Pro images, all of which were made using the 35/2 and 50/1.4 AF-D Nikkor primes. The four January Nikon D40 shoots were made with a 28-70 AF-S Nikkor zoom, and the image quality of the files seemed to me to be fairly comparable between the D40 and the S5 Pro, but with the D40 prone to more color moire effects than the S5 Pro, but with the S5 Pro more prone to weird artifacts and jaggies,and a lower overall per-pixel sharpness level. There's no advanced EXIF information on these Photogenic Weekend shoots, but the S5 Pro images posted just today look okay to me. Not a lot of sharpness, but low noise,and pleasing colors,and generally agreeable image characteristics. Considering the type of lighting used on these shoots, which is typically window light or diffused daylight,augmented by not much more than a 4x6 foot reflector, the S5 Pro seems to me to be a decent camera for this type of photography. While I am concerned by the overall softness of the images, I realize that looking at d-slr images sized down into the 2,000-2,300-x pixel range is nothing at all like evaluating FULL-sized captures, which can take a lot more post than downsized files.
I'd love to see some well-lighted flash portraits done with a top-quality 85 or 105mm Nikkor lens using something in the f/4.8 to f/6.3 range with 800-1200 watt seconds pumped through a 3x4 foot softbox, just so we could see how a top-flight lens and electronic flash lighting will look when shot with the S5 Pro. So far, there are a number of unanswered questions, but today's February 9 Photogenic Weekend portfolio gives me a bit of hope for the S5 Pro. Overall, I have to say, the S5 Pro's sample photo "rollout" has been poor. So poor in fact that it can hardly be called a "rollout". The four sample photos FujiFilm posted have been very poorly-received samples among the installed Fuji d-slr user base. Let's keep in mind however that Canon's EOS 5D official samples were *EQUALLY AS LAME* as the Fuji S5 samples. Yes, Canon,mighty Canon had a paltry few sample photos available before the 5D's launch,and the pictures were LAME samples. As it turns out though, the sample photo fiasco Canon had with the 5D rollout was of no consequence, since the camera could actually deliver in spades. Let's hope the same holds true for the S5 Pro--that once the camera is in the hands of skilled shooters, that it too delivers.
No comments:
Post a Comment