Dick,
My comments were mostly about cylindrical stitching vs the Roundshot. It is a lot less clearcut when considering a high res single shot back.
...and the rest of us are talking about rectilinear.
I would be genuinely interested in seeing full res images produced by your system. I could be wrong, but my assumption is that:
1. Cristical focus will be very hard to achieve,
2. Perfect plan sliding of the back will be very challenging also.
Most of the equipment is not on the market yet (so I have not used it, and cannot send you any samples) - I hope to get the H4D-60 next month, and Kapture Group are working on the triple or hex stitch sliding back.
You need quality (yes heavy) e.g. Sinar kit to do this. Focus is always critical with very short lenses, but live view helps.
Per my experince cylindrical pano is actual somewhat faster.
Faster that what? ...are you saying that, in your experience, pano is faster that something of which you have no experience? ... like motorised automatic sliding backs?
I said that taking a picture of the house in your picture above with a single exposure with an H3D-60 might take one tenth of the time it took you to produce a 160Mpx picture, and still capture all the detail in that particular image.
There is no concern about accurate positioning, just a simple rotation that does not affect the relationship between lens and sensor.
... but, for best results, you have to accurately position the nodal point of the lens in the axis of rotation?
Well, you basically always use the center section of your lens when you cylindrical stitch. With correctly positioned lenses, the process of projection removes extremely little image quality, I would say less than 5%.
If you use a low-res DSLR with an anti-aliasing filter, you do not have much image quality to lose, and you do not miss what you never had, but any digital projection stretches pixels and wastes rez... I always try to print dot for dot, without scaling, which you can do with a 24 inch Epson and a 60 Mpx back.
I have never used a Digitar, but judging from the difference of sharpness between center and edge on my Schneider 110mm XL, one of the best wide 4x5 lenses produced, I would be very surprised if the corners of the digitars after sliding of your back were even close.
close to what? close to the quality of a film lens??? or close to their centre res? ...see the MTF charts. LF film lenses generally, I think, have about half the res of apo-digitars, so a 10 micron digiback would be ideal for stitching with LF film lenses. Most of the current wade-andgle apo-digitars (with the exception of the 47) do not have useful image circles, but there some in the pipeline.
Again I would love to see full size samples. I can provide you any time with a full size version of any of the stitched images I produce. Image uniformity is simply perfect all the time.
I would appreciate that - I have often thought that pictures that look good on a monitor would not necessarily look good printed five feet by four.
I believe that capture will be about the same,
I said that taking a picture of the house in your picture above with a single exposure with an H3D-60 might take one tenth of the time it took you to produce a 160Mpx picture, and still capture all the detail in that particular image.
Taking 3 * 60Mpx exposures less overlap ¿160MPx? would, I would have thought take, no longer than taking three pix with a 24 Mpx DSLR for cylinder stitching... but pano-cylinder stitching three 24 Mpx pix would ¿surely? give you less res than a single picture taken with an H4D-60
...post processing is indeed longer with cylindrical stitching, but it is mostly CPU time. nowadays I hardly ever spend time re-touching my panos, software does it perfect fully automatically 75% of the time, the rest of the cases typcally require only very limited manual work.
Anyway, again, I was mostly reacting to the Roundshot proposition.
Cheers,
Bernard
I appreciate that what I am saying is theoretical, and you can demonstrate that you technique works.
Dick
