On 2005-09-16 10:43, Godfrey DiGiorgi wrote: > The report that caused a lot of hubbub about JPEG was Phil Askey's > review of the DS on DPReview.com.
Yes - and it's one of the best documented and usable setups I know. > His test procedure set the camera > to its defaults in Auto Picture or Program mode, I feel, this is very justified, too. Systems should be at a reasonable default setup. Only few people have the knowledge to alter this default in order to get the optimum out of it. Most people would not mind (e.g. combining the top-of-the-line SLR with a super zoom lense 24-300). But although people could obtain better results by taking a better setup, for comparions it's very reasonable to use the 'default' setup. There's a German testing organisation ("Stiftung Warentest") which is blamed many times for 'crazy' test conditions. Maybe people know better. There are many things that I could critize, where I might no better. But there are some general testing rules for them: - they have no advertisment in their journal 'test' - they buy the products from the shelf (various real market source instead of triple-tested models from the manufacturer himself) (... and that's one of the explanations why their results always are very late, while the model may have disappeared from the market - they test the user interface as a typical user with little experience would expect - they test typical configurations So a superb camera may obtain poor results for image quality, since the camera comes with a rather poor set lense. You may find this by reading the rather good explanation which caused the fault - but the dealer may use the total test certificate only (such as 'sehr gut' = 'very good'). > which means > "Vivid" (highly saturated) color rendering, a high degree of > sharpening and contrast, sRGB colorspace. This is documented within the manuals, as well as the effect of other settings? > This degrades image quality > when you're measurbating and comparing image quality at the > pixel:pixel level, other DSLR manufacturer defaults are often a bit > softer, but Pentax spiked up the defaults for the intended consumer/ > snapshooter audience who mostly output to 4x6 prints. All other > reports I've seen about "JPEG problems" were directly traceable to > this DPReview report. > > Like parrots in a bordello... ;-) Agreed - but did Pentax offer any update or info which setup would be suited best for which operation? I'll have a look at http://www.pentax.co.jp/english/support/man-pdf/istds.pdf ... no occurrence of 'vivid' within this manual. Function reference 107ff tells me little about different default setups and recommendations. >From page 176 I read: default saturation: 0 (of 5), sharpness 0 (of 5), contrast 0 (of 5), quality level *** (best, of 3: best/better/good), Color Space sRGB (or AdobeRGB instead) You mentionned elsewhere 'natural' image tone instead of 'bright' (default and no override for Picture modes) > I normally save exposures in RAW format. dpreview stated here, that Pentax' RAW is huge, since it's uncompressed. This is true by now? > The key is Natural color mode and Adobe RGB color space. Natural > color mode *ONLY* works in P, Tv, Av, and M exposure modes; Auto > Picture and all the program presets override the setting and go to > Vivid. So you don't have any other choice there? I feel this qualifies for taking 'vivid' for comparisons, as well as for critisizing the poorer default quality. Do you know any other comparison documented online, comparing the jpeg quality of different cameras, which is of better quality than dpreview looks like? - Martin