regarding the D1 and the D1H...

[StevenMCurrie]

Is anyone shooting with the 2.7Mpixel D1 or D1H? How are your prints?! I heard someone claim they get great prints at 11x14 via inkjet with images taken with a D1. Fact of fiction?

[Roger_Martin]

http://www.assuredphoto.com/Events/M...2k/img0495.htm This photo just went nearly photographic 13X19 with an Epson 1280 printer and Pro Gloss paper. It was shot with a D1H using a 80-200 AF-S lens and fine jpeg compression.

Fact!
Gav

[JeffKardas]

You have to be careful how you print, and how you manipulate the image (i.e. using Genuine Fractals or the Stair Interpolator to blow it up), but I've gotten killer 13x19 prints from my D1 images at 240 dpi on a 1280.

[StevenMCurrie]

I personally haven't been impressed with GF, and all the other up-sampling techniques I've seen. They don't seem to do much better than vanilla bicubic upsampling in most cases. Lack of detail in the original (somewhat obviously) can't be reconstructed with snazzy upsampling methods. I could believe it if someone claimed a particular method handled edges (high-contrast areas) better, but the "little things" will still not be present.

Really, there are two variables in the relationship I'm pondering: the size of the sensor, and the number of pixels in that sensor. Each can be independent of the other, but I'm interested in the how they interact. E.g., if you increase the sensor size, but keep the number of pixels constant, how does that compare to the case with the original sensor size but more pixels. It's not a clearcut "it's always better to have X" answer because of things like distortion/aberations present in lenses, etc.

Getting to the example above, assuming 13x19 print size at 240 dpi, you're basically upsampling about 500% (~14M div ~2.7M), right? Does it really improve the print to upsample ahead of time? Fine details (hair, eyes of a small subject, texture in surroundings) are still not tack sharp, so how do they look in a non-upsampled print compared to an upsampled print?

I personally haven't used a D1* yet, hence the line of questioning, but I have a hard time believing some of the claims I was hearing. Hearing more people discuss it (here) has been helpful in the sense that I've increased the number of opinion data-points I have. Thanks for you contribution.

[Andys]

I believe the reason people upsize the images before printing is that basically, if you are making a large print, you don't want the result to be pixelated.

Standard resizing will not add any detail to the picture, it will simply interpolate the pixels to create smooth graduations. Up close, it will look blurry instead of being high definition - this is probably why you haven't been impressed with it.

Most people create these enlargements for hanging on a wall or a mantlepiece, and when they have been resized on the computer, it makes them look smooth and hiresolution when viewing at an appropriate distance, instead of being blocky or pixelated.

What programs such as Genuine Fractals try to do above and beyond standard interpolation, is help create extra detail on areas of repeating patterns or repeating geometry.

If there is a program that could help create extra detail on lines and contrasting edges, in addition to standard or fractal interpolation - I would love to hear about it!


- Andrew