Help Baldy shoot for a 120x180 peep-proof print

Baldy

I'm still waiting on my new 5D MKII to shoot the 72x720 SF pano. I think we figured out the parameters on that one: 600mm lens, the mother of all tripods, get the camera above the thermal effects on the surface of the water.

But what about this shot:



I have a printer that can print 600dpi Lightjet prints on one piece of paper, 72 wide by 120 high, so I guess the final print will have two seams (3 60x120 pieces).

I've watched every single person who has ever visited our office gasp when they see a big print, then step 6 inches away and start pixel peeping. We want to make this print pixel-peeper proof.

On this shot I'd guess we don't have to shoot with so long a lens. But to get to at least 300 dpi do we:

1. Shoot a bazillion 5d MKII shots with, say, a 200mm prime and the mother of all vibration-free mounts?

2. Shoot a bazillion/3 Hasselblad H3D50 frames, suffering the perils of ISO 50, and the mother of all vibration-free mounts?

3. Shoot one 8x10 sheet of film with the mother of all vibration-free mounts and do the best scan money can buy?

I just acquired a shot from Marc Muench to print something like 84x108 that was shot on 8x10 film, and it's so very lovely:



How well does 8x10 film hold up to pixel peepers at 180 inches?

Thanks!
Baldy

pathfinder

120 inches at 300 pixels per inch = 36,000 pixels

5DMkll is 5616 pixels in the long dimension of the frame, and 3744 pixels long across the short dimension.

36,000 / 3744 =~ 9.6 frames - I'll call it 10 - allow 50% overlap, so 15 frames is more than enough in portrait mode on the long dimension of the image


60 inches at 300 pixels per inch = 18,000 pixels but the 5DMkll frame is 5616 pixels long

18,000 / 5616 =~ 3.2 frames - allow 50 % overlap and call it 6 frames


Hence 6 x 15 = 90 total frames

I'll bet you can print your image at 160- 180 pixels per inch ( the number of ink dots per inch will be an order of magnitude higher of course ) and get by with "only " 50 frames from your 5D Mkll. Unless you want to bracket your exposures over and under by 2 stops......


I shot this pano with a 1DsMkll and a 90mm T&S and it is about 13000 pixels long - 7 frames in portrait mode - Not nearly as ambitious as yours, Baldy, but it is my largest to date.

DO we have any idea on the delivery date of the 5DMkll yet?

What is the resolution of the scan of 8x10 film? 5000 pixels per inch, at most? 5000 x10 = 50,000 for 180 inches, or 277.8 pixels per inch. In his print workshop, Marc said he frequently prints at 180 - 200 ppi for prints larger than 40 inches or so. Maybe he will give us an answer here also.


Consider using a wireless cable release such as Adorama's Wireless Radio Remote Release for Canon EOS 50D, 5D, 10D, 20D, 30D & 40D SLR Digital Cameras (3 Pin) - Mine seems to work fine. Why can't Canon build this into the body, like Nikon does anyway!

Andy

pathfinder wrote:

DO we have any idea on the delivery date of the 5DMkll yet?

next week, we think

Andy

Baldy wrote:

On this shot I'd guess we don't have to shoot with so long a lens.

Depends on the tides http://www.ot-montsaintmichel.com/horaires_gb.htm

Baldy

Thanks, Pathfinder!

pathfinder wrote:

What is the resolution of the scan of 8x10 film? 5000 pixels per inch, at most? 5000 x10 = 50,000 for 180 inches, or 277.8 pixels per inch. In his print workshop, Marc said he frequently prints at 180 - 200 ppi for prints larger than 40 inches or so. Maybe he will give us an answer here also.

He's gonna re-scan in the a.m. at 4,000 dpi. Beyond that he says you're just blowing up the grain. He didn't scan it that high the first time because he didn't think some freak like me would try to make so large a print.

It's very tempting to shoot Mont Saint Michel with an 8x10 camera because I fear the best light will be light that's changing rapidly. Getting 50 frames in a short time sounds hard.

Baldy

A clearer version, from here:

Marc Muench

Baldy wrote:

Thanks, Pathfinder!

He's gonna re-scan in the a.m. at 4,000 dpi. Beyond that he says you're just blowing up the grain. He didn't scan it that high the first time because he didn't think some freak like me would try to make so large a print.

It's very tempting to shoot Mont Saint Michel with an 8x10 camera because I fear the best light will be light that's changing rapidly. Getting 50 frames in a short time sounds hard.

Actually Adobe did not either:D as at that time Photoshop only opened files up to 2 gig. However, now it appears 6gig is no problem
What we must do for our drum scanner to complete this task is scan four vertical strips at 8x2 at 4000ppi, then stitch them together. It all has to do with the line length of the gathered information on the drum. At such high resolution the line length can not be greater than the scanners ability to absorb all the data. I will let you know how it goes today

As far as pixel peeping goes, I thought I was the freak as many have told me to just back off "Marc, don't get that close" Here are some interesting factors Tom and I have realized after 11 years of playing this game:
- I see dead pixels

- Best pixel peeping is done to printed files that have been properly sharpened.
- Only sharpen sharp images, remember we are discussing pixel peeping only
- Only enlarge optically SHARP images
- Do not enlarge beyond 100% of optical res
- Always let the best RIP money can buy do the enlargement for that specific printer, Yes that means Photoshop is not where you want to enlarge.
- No need to scan film greater than 4000ppi
- No need for input res greater than 300ppi x dimensions
- No need for output res on inkjet printers greater than 1440dpi unless printing on film or super high gloss.
- No need to print on lightjet/Lamda more than 300ppi x output size
- Lighjet/Lamda printers have the VERY best RIP, period, made by Symbolic Sciences!
*In fact we have made wonderfully peepable pixels on Lamda printers at 200ppi. 4x5 Velvia film, 12000 pixels in the long dim made a beautiful Lamda print at 50x60. The input res was only 200ppi The RIP on those printers interpolates to 300 no problemo. However, If you can shoot for 300ppi at size WHY NOT

For that matter then my 8x10 should make a great print at 24,000 pixels long or 120 inches or 10 feet. Final size for max pixel peeping would be 8x10 feet!

Marc Muench

Baldy wrote:

It's very tempting to shoot Mont Saint Michel with an 8x10 camera because I fear the best light will be light that's changing rapidly. Getting 50 frames in a short time sounds hard.

Regarding shooting this, Yes one sheet of film is quicker but!!!!! why not take 4 and stitch them together? Shooting 4x5 panos is something I have always thought of but never took the time.

I dont even know what 8x film can be bought

ian408

Hi Baldy! Have you considered Velvet Rope to keep the peepers at bay?


















:hide

pathfinder

Uh oh - shooting across water again also


The Hasselblad H3DII 50 has 50 million pixels 6132 x 8176 on a sensor 36.8 x 49.1 millimeters, with a typical RAW image size of 65Mbytes.


36,000 pixels / 6132 =~ 5.9 frames in portrait orientation 33% overlap for stitching --> 8 frames across.

18,000 pixels / 8176 =~2.2 frames - allow 33% for overlap or about 3 frames

Thus a total of maybe 24 frames = 65 Mpixels each will work, but still a lot more work than a single 8x10 film frame.

How about borrowing Stephen Johnson's scanning digital back - isn't it 11x14 or so, and just scan to memory, rather than going through film?

This sounds like a perfect place for a scanning digital back - trumps film grain, and is vastly faster to capture than multiple digital files to stitch, but a non moving subject that is perfect for the 60 sec exposure of a scanning back?

Here is a link to some of his projects http://www.sjphoto.com/project-tools.html and another http://homepage.mac.com/sjphotog/studio-photography-design.pdf

He lists a scanning back with almost 16,000 pixels in one dimension.