> > At first look, the edge still has more jaggies than Photoshop, but > definitely better than the 'black halo' of PhotoLab. The softness is > the initial impression, but this can soon be sorted, and colour brought > into line. There seems to be a little less detail in your image than > either PhotoLab of PhotoShop at the final stage though - is this just > the sharpening method they employ creating the impression of detail? > Maybe if I were better at this sharpening lark I could sort that out.
The jaggies are an artifact of the very simple box filter I am using in the interpolation step (which is a fancy way of describing just averaging the neighbouring pixels). I'll investigate better filters later, once I'm happy with the overall colour balance. The softness and/or lack of detail are, indeed, because I'm not doing any sharpening at present. Again, that's a topic for later investigation. I'd be very interested to know what corrections you felt were necessary to "bring the colour into line". I think I'm missing a step somewhere in calculating the weights, and any additional data would be valuable. > Thanks for the comparison, makes one wonder how Pentax could have gotten > it soo wrong... > > I am gonna try Genzo later to have a look at that too, but the product > is a GUI nightmare! > > What plans do you have for your conversion? Seems like a lot of people > will eventually looking for 3rd party options before long... Oh, I'll be making my stuff freely available to anyone on the Pentax list. But at present all I have is a DOS command-line utility that loses all the EXIF data, and only writes PNGs or JPEGs. Not all that much use. I'd like to produce: o A command-line converter that can write TIFF or JPEG o An interactive program to replace PhotoLaboratory > > > > > -----Original Message----- > > From: John Francis [mailto:[EMAIL PROTECTED] > > Sent: 12 March 2004 18:34 > > To: [EMAIL PROTECTED] > > Subject: Re: RAW Conversion comparison > > > > > > > > > > To do the test quickly I did not try to colour correct any of the > > > images, merely clicked on a grey point and left it at that. > > Because > > > the Pentax software doesn't allow for accuracy of cliking > > in the tiny > > > image, it is quite likely that the colour differences have > > something > > > to do with my not clicking on exactly the same spot in each image. > > > > > > The main thing I was trying to show was the bad bayer > > interpolation of > > > the Pentax Software, as many were not believing there could be that > > > much difference. I think it is definitely worth > > investigating other > > > products whether that be CS, DCRAW or custom made > > software... Maybe > > > if you do, you will be enabled after all! > > > > OK - here's as close to the RAW sensor data as you are likely to get. > > > <http://panix.com/~johnf/temp/pef1127.jpg> > > The only processing done here to go frow raw sensor data to RGB is: > > o Fill in missing sample values (average of closest neighbours) > > o Scale the R/G/B values based on what I believe to be the white > balance scaling stored in the raw file by the camera. > > o Convert from linear to sRGB colour space. > > >
