Hi Ernst, I don't remember being vague about QTR. QTR resamples all input files to 720x720 pixels. Then dithering is done to whatever dpi you've selected -- mostly 2880x1440 or 1440x720. A couple printers also have 1440x1440. The resampling to 720x720 is done by the OS on the Mac so I'm not sure of the algorithm. On the PC it's a bi-linear algorithm -- basically two linear interpolations -- horizontal then vertical. As far as I know all the Epson drivers do either 360x360 or 720x720 resampling. In general the smaller printers tend to do 720x720 and the larger ones do 360x360. If the driver has a Finest Detail switch that gives the user a choice between these resamplings. ---------- While I think this "small text experiments" are interesting and definitely show some differences, you can read too much into the results. If you look at the actual input files that Paul, Jon and others are using you can see that one-point text can look very different even before you print it. Anti-aliasing for text is complicated balance between sharpness and smoothness. Readability is also a tricky evaluation. One eyes fill in details much of the time. Choice of font makes a huge difference -- some lend themselves to small type much better than others. The angle has an effect too. BTW, I agree with Paul that for some reason the size of his 1pt text is different than Jon's 1pt text. Jon's font appears to be a simpler font, too. BTW again, take Paul's file and just insert a 1pt Verdana, no aliasing and the difference is very dramatic. (not sure why Paul did 7200dpi and reduced it). Roy On Tue, Sep 2, 2008 at 8:10 AM, Ernst Dinkla <edinkla@...> wrote: > Paul Roark wrote: >> Jon Cone wrote, in part: >> >>> ... I posted a 1/2 inch square file ... >>> if you would like to test it ... >> >>> Download at http://www.inkjetmall.com/piezoTypetest_tif.zip >> >> I just tried printing this on an Epson 1400 with QTR and the Epson driver. >> It's very sharp with either. The text actually looks a little larger than >> the "1 point" text I had printed earlier. >> >> I'm not sure what QTR does with a 1080 dpi file. I think the Epson driver >> downsizes it to 720. That might actually be a source of difference that >> could affect the K3 example. I've experimented with very high resolution >> files in the past and noticed that the information thrown out by the Epson >> driver (I forget what printer model) made the very high resolution file >> print less well than one that was "only" 720. That is, there may be a >> difference in how the approaches treat very high res files. >> >> At any rate, while high resolution scans are wonderful, I can't imagine >> anyone wanting to deal with or needing a file in excess of 720 ppi at the >> print size. The 20,000 pixel wide panorama file I just printed was such a >> nuisance to deal with and print at its full resolution that I downsized to >> 240 ppi for the 51 inch print. >> >> Paul >> www.PaulRoark.com > > > > Drivers have improved on upsampling but few on downsampling. > Good anti-aliasing routines could improve the output of > downsampled images like they do in Qimage. Not only is the > detail compromised but contrast can be exaggerated without > the right anti-aliasing, depending on the detail/texture of > the image content. If you like to scan once at the highest > resolution you can afford (equipment or third party scan) > you will be faced with the fact that downsampling has to be > done from time to time for smaller prints. If the driver has > to do it or just an image editor you may encounter problems, > you better check what the quality of anti-aliasing is in the > program you use and whether it does the up- and dowsampling > to the native resolution before the driver does it. > > For text it has been my experience that even with a rough > driver like the Epson 9000 one the quality of the characters > can be very high if you let Photoshop rasterise fonts (for > example a PDF with vector fonts dropped in Photoshop) to 720 > PPI and print that raster image. In the 9000 driver it will > be downsampled to 360 PPI native resolution. most likely > with a nearest neighbour routine. That's very nice for that > content but not for photographic images. On the other hand > Qimage doesn't do it as nicely on fonts or you switch the > extrapolation routine to the worst one or simply off. > > Roy has always been vague about the native resolution of > QTR. More or less writing that it doesn't exist. > > -- > Met vriendelijke groeten, Ernst > > > | Dinkla Grafische Techniek | > | www.pigment-print.com | > | ( unvollendet ) | > > ------------------------------------
Message
Re: [QuadtoneRIP] Piezography Type Test File
2008-09-02 by Roy Harrington
Attachments
- No local attachments were found for this message.