I grant that could very well be the case...
Tom C. ----Original Message Follows---- From: Paul Stenquist <[EMAIL PROTECTED]> Reply-To: Pentax-Discuss Mail List <[email protected]> To: Pentax-Discuss Mail List <[email protected]> Subject: Re: K10D aimed as D200 killer Date: Mon, 30 Oct 2006 18:41:13 -0500 You're reading a lot into brief and confusing remarks that have been translated from the Japanese. Paul On Oct 30, 2006, at 5:01 PM, Tom C wrote: > I agree with that in principle, but the issue here as I read it, is > one of > the sensor making this an issue because of high noise levels that > were not a > concern in the 6MP models. > > In other words, I expected that image qualiy would get better in > the new > body across the board, not that their would be tradeoffs. > > It seems Canon at least (don't know about Nikon) has been able to > continually increase sensor resolution while continuing to keep > noise to a > low level. > > > Tom C. > > > ----Original Message Follows---- > From: Adam Maas <[EMAIL PROTECTED]> > Reply-To: Pentax-Discuss Mail List <[email protected]> > To: Pentax-Discuss Mail List <[email protected]> > Subject: Re: K10D aimed as D200 killer > Date: Mon, 30 Oct 2006 16:31:02 -0500 > > The issue here is whether or not the camera should do any non- > reversable > processing such as sharpening or noise reduction. Nikon has decided to > make it configurable on the D80, and set the defaults to what's > appropriate for P&S use, while Pentax has chosen to make its defaults > more appropriate for post-processing. > > I prefer Pentax's approach. It gives me more control over > rendering, and > does not reduce the level of information in the image (which noise > reduction does). > > -Adam > > > Tom C wrote: >> In my mind it's difficult to understand the difference between >> sharpness/unsharpness/detail and noise. It seems to me that an image >> considered to be sharp, yet with a lot of noise, is in reality not >> sharp >> and/or contains less detail because the noise is itself replacing >> detail >> that would otherwise be there. >> >> Noisy picture = Yucky picture. >> >> >> Tom C. >> >> >> ----Original Message Follows---- >> From: Joseph Tainter <[EMAIL PROTECTED]> >> Reply-To: Pentax-Discuss Mail List <[email protected]> >> To: [email protected] >> Subject: K10D aimed as D200 killer >> Date: Mon, 30 Oct 2006 12:09:23 -0800 >> >> My comment at end. (Caution: Some of you wll hate it. You may not >> want >> to read it.) >> >> ----- >> >> Remember the reviews of the *istD? It got beaten up because Pentax >> decided to make soft pictures strait out of the box. I was not >> part of >> the list then, but I imagine many talking about this being better >> because it left the decision to the photographer. >> >> As I understand it, it's the same with noise vs. details. >> >> >> Tim >> Mostly harmless (just plain Norwegian) >> >> >> -----Original Message----- >> >> In addition, requiring the the user to do even more in post- >> processing >> to try to correct for what could be viewed as a camera short-coming, >> strikes me as a cop out. I already don't use the *ist D for anything >> serious over ISO 800. I don't want additional post-processing work, >> that may or may not correct the situation on an image-by-image basis. >> >> Tom C. >> >> ----- >> At dpreview I just found a translation of a interview with Hisashi >> Tatamiya, who has been leading the K10D project. >> >> http://forums.dpreview.com/forums/read.asp? >> forum=1036&message=20671456 >> >> ----- >> >> I believe Pentax made the right decision in regard to high ISO noise. >> Image sharpness is retained, if you can figure out a way to reduce >> noise >> in PP without softening the image. But Nikon may have been smarter >> marketing-wise. >> >> Pentax would have been criticized whichever way they went. Popular >> Photography's review of the D80 praised it for low noise at high ISO, >> completely forgetting to mention that the D80 achieves this at the >> cost >> of soft images. When Pop reviews the K10D, they will complain that it >> compares poorly to the D80 in high ISO noise. And readers who >> don't know >> any better will believe that that is the final word. >> >> Reading between the lines of the summary of the interview, Mr. >> Tatamiya >> is (it seems to me) saying two things: (1) there will be noise at >> high >> ISO and you may not like it, and (2) its your problem. None of >> this is a >> surprise. The sensor is known to be noisy at high ISO, and I >> suspected >> that Pentax would choose a middle course between Nikon and Sony. I >> just >> hope that images will be useable at ISO 800. If they are, I'll be >> satisfied. But I am not expecting this. >> >> Herb Chong contacted me off-list, and suggested something I had not >> heard before. According to Herb, the rule of thumb for good image >> quality is two steps above the base ISO. This matches my >> experience with >> the D, which is fine at 800, but (to my eye) not at 1600. If this >> rule >> of thumb holds for the 10 mp sensor, then ISO 400 will be the point >> above which we can expect image quality to decline noticeably due to > noise. >> >> (Actually, the paragraph above assumes that all else is equal--like >> pixel density. Since the K10D has a higher pixel density, one may >> expect >> the loss of an additional step due to inherently higher noise. >> Combining >> (1) lower base ISO, and (2) smaller pixel size, the K10D could >> conceiveably yield noticeable degradation in image quality above ISO >> 200. But Nikon seems to get good image quality without softening >> at ISO >> 400, so I believe we will too.) >> >> Joe >> >> -- >> PDML Pentax-Discuss Mail List >> [email protected] >> http://pdml.net/mailman/listinfo/pdml_pdml.net >> >> >> > > > > -- > PDML Pentax-Discuss Mail List > [email protected] > http://pdml.net/mailman/listinfo/pdml_pdml.net > > > > -- > PDML Pentax-Discuss Mail List > [email protected] > http://pdml.net/mailman/listinfo/pdml_pdml.net -- PDML Pentax-Discuss Mail List [email protected] http://pdml.net/mailman/listinfo/pdml_pdml.net -- PDML Pentax-Discuss Mail List [email protected] http://pdml.net/mailman/listinfo/pdml_pdml.net
