Megapixel to film equation - DPChallenge Forums

01/18/2006 04:38:15 PM · #26

Originally posted by Brent_Ward:

Originally posted by nsbca7:

Originally posted by petrakka:
anyways, for most occasions I think it's kind of a moot point, because it's not too often that you print 40x50. It's my opinion that at sizesl ike 11x14 or 16x20 sometimes a digital print can look better than scanned film.

Perhaps it is not too often that you print to that size. I love to print big and have had no problems printing sharp images to 4 feet wide from the 1Ds and 1Ds Mk II.

Do you use a pro rip?

No. I just up-size in several steps in PH CS and print directly to the 7600.

01/18/2006 04:42:09 PM · #27

What do you print your 40x50's on?

01/18/2006 04:45:32 PM · #28

Originally posted by Brent_Ward:

What do you print your 40x50's on?

I don't print 40x50s. I print cropped images 48 inches long or wide depending on the orientation. I use an Epson 7600 with Ultrachrome.

01/18/2006 04:51:49 PM · #29

Above is an image taken with an 8Mp 1D Mk II

This is a detail from that image. Clean. Compare this to the pug scan crop and imagine the possibilities.

01/18/2006 05:15:08 PM · #30

Originally posted by nsbca7:

Above is an image taken with an 8Mp 1D Mk II

This is a detail from that image. Clean. Compare this to the pug scan crop and imagine the possibilities.

What do you think that pug scan would like if I scanned it in @ 24mb? That pug file is around 250mb.

I own the 1D MArk II, I'm well aware of the pros & cons of it.

01/18/2006 05:28:46 PM · #31

Originally posted by Brent_Ward:

What do you think that pug scan would like if I scanned it in @ 24mb? That pug file is around 250mb.

What is the difference? At 30x45 this image is 695Mb.

01/18/2006 05:49:51 PM · #32

Originally posted by nsbca7:

Originally posted by Brent_Ward:

What do you think that pug scan would like if I scanned it in @ 24mb? That pug file is around 250mb.

What is the difference? At 30x45 this image is 695Mb.

do a 100% crop of that file and I bet it looks like shit. All noisey and jagged.

01/18/2006 06:01:10 PM · #33

Originally posted by Brent_Ward:

Originally posted by nsbca7:

Originally posted by Brent_Ward:

What do you think that pug scan would like if I scanned it in @ 24mb? That pug file is around 250mb.

What is the difference? At 30x45 this image is 695Mb.

do a 100% crop of that file and I bet it looks like shit. All noisey and jagged.

That is a crop from that file. I blew it up to make sure it would stand up to the enlargement. The cropped image is 6.44 Mb by itself before I downsized it in Save to Web so that is could be uploaded. No noise or jagged edges either way. If you are having that problem perhaps something in your workflow is off.

01/18/2006 06:06:41 PM · #34

Originally posted by nsbca7:

Originally posted by Brent_Ward:

Originally posted by nsbca7:

Originally posted by Brent_Ward:

What do you think that pug scan would like if I scanned it in @ 24mb? That pug file is around 250mb.

What is the difference? At 30x45 this image is 695Mb.

do a 100% crop of that file and I bet it looks like shit. All noisey and jagged.

That is a crop from that file. I blew it up to make sure it would stand up to the enlargement. The cropped image is 6.44 Mb by itself before I downsized it in Save to Web so that is could be uploaded. No noise or jagged edges either way. If you are having that problem perhaps something in your workflow is off.

Don't downsize it. Post an actual 640x640 crop of the image at 100%. Thats' what I did with the pug.

01/18/2006 06:13:15 PM · #35

Originally posted by Brent_Ward:

Don't downsize it. Post an actual 640x640 crop of the image at 100%. Thats' what I did with the pug.

Apples to apples? What size was your pug image?

01/18/2006 06:13:38 PM · #36

Originally posted by nsbca7:

Originally posted by Brent_Ward:

Don't downsize it. Post an actual 640x640 crop of the image at 100%. Thats' what I did with the pug.

Apples to apples? What size was your pug image?

250mb.

01/18/2006 06:32:43 PM · #37

Originally posted by Brent_Ward:

Originally posted by nsbca7:

Originally posted by Brent_Ward:

Don't downsize it. Post an actual 640x640 crop of the image at 100%. Thats' what I did with the pug.

Apples to apples? What size was your pug image?

250mb.

01/18/2006 06:38:30 PM · #38

You have to understand too that at 38x25 (the size of a 251Mb image at 300 PPI) this crop would in actuality be 2.25x2.25 inches wide, so you can imagine the close up detail in an image printed that size. And to me it looks a whole lot less noisy then the pug crop.

01/18/2006 06:43:24 PM · #39

Originally posted by nsbca7:

What camera is that from? Remember the pug crop was from 160 ISO neg film.

Canon EF-S 18-55mm f/3.5-5.6 IS

01/18/2006 06:50:34 PM · #40

Originally posted by Brent_Ward:

Originally posted by nsbca7:
You have to understand too that at 38x25 (the size of a 251Mb image at 300 PPI) this crop would in actuality be 2.25x2.25 inches wide, so you can imagine the close up detail in an image printed that size. And to me it looks a whole lot less noisy then the pug crop.

What camera is that from? Remember the pug crop was from 160 ISO neg film.

That was just taken about an hour ago.

Canon EOS-1Ds Mark II

2006-01-18T16:06:10-06:00

1/1000 sec

f/1.8

ISO 160

85.00 mm

justin_hewlett

Canon EOS-1000D Rebel XS

01/18/2006 07:36:53 PM · #41

Originally posted by Brent_Ward:

Remember the pug crop was from 160 ISO neg film.

On your 6x7?

01/18/2006 07:38:03 PM · #42

Originally posted by justin_hewlett:

Originally posted by Brent_Ward:
Remember the pug crop was from 160 ISO neg film.

On your 6x7?

yes.

01/18/2006 07:38:54 PM · #43

Originally posted by nsbca7:

Originally posted by Brent_Ward:

Originally posted by nsbca7:
You have to understand too that at 38x25 (the size of a 251Mb image at 300 PPI) this crop would in actuality be 2.25x2.25 inches wide, so you can imagine the close up detail in an image printed that size. And to me it looks a whole lot less noisy then the pug crop.

What camera is that from? Remember the pug crop was from 160 ISO neg film.

That was just taken about an hour ago.

Canon EOS-1Ds Mark II

2006-01-18T16:06:10-06:00

1/1000 sec

f/1.8

ISO 160

85.00 mm

I'll work up one from a velvia scan & post it later. I gotta run, Going to the nuggets game tonight.

01/18/2006 10:31:13 PM · #44

Don't bother. I'll take your word. My original point has been more then proved: That is that the info from Clark Vision is flawed to say the least. His graphs and comparisons show 35 mm film faring well in comparison to top end digital when you are having trouble showing me medium format that can keep up. I realize that our little comparisons here have not in the least bit been scientific, but I contend that neither have his.

I have proved this for myself almost two years ago through practical application and am more then satisfied with the current levels of tools I now employ and the work output they produce, where I had never been satisfied with the quality output of my former tools. That in itself is proof enough for me.

01/19/2006 01:51:40 AM · #45

35mm film gives frames 24 x 36mm, or .9449 x 1.4173in. Scanning that at 4000dpi gives 3779.6 x 5669.2px. Multiply that out, you get 21.4mp. So 35 millimeter film scanned at 4000 dots per inch technically equals 21.4 mega pixels.

However that does not tell the whole story. Mrladewig was first to point out the Clumps and Chumps article. If a film grain is 2 micrometers, and you need 30 of them to represent the full tonal range (or the abilities of a single pixel), then the film equivalent of a pixel is 30 square micrometers . 36000x24000 / 30x30 = 1200x800px. That would mean that 35mm film is physically equivalent to .96 megapixels. That seems wrong.

Needing 30 of them to represent the full tonal range, I think is wrong. That 2 micron grain is binary. So is a digital image. 8 bit depth means 8 binary units are organized to represent every shade from black to white. If we use the same idea for film. We need 8 grains to represent a pixel. The math is the same and yields 13.5 megapixels. That seems better.

However that is still not the final answer. I'm assuming that the Grains are perfectly uniform with no space between them. They are not. They come in a verity of shapes, and have random spaces seperating them. So the actual space they take up is maybe 10 or 12 micrometers instead. That makes 8.64 or 6 megapixels respectively. That is more in line with what people are seeing. So the best 35mm film in the world might be as good a 5D. But probably not.

Also, for comparison that same impossibly perfect film in medium format, would match 65.6 megapixels. More realistic would be 42 to 29.1 megapixels.

Sorry if this is hard to follow. I think I will write it up better and make a real post out of it some time.

Posted on another thread.

Nikon MF 180mm f/2.8 ED AIS

01/19/2006 02:19:31 AM · #46

Originally posted by Brent_Ward:

If it was all that easy to calculate that would be great. The fact is it is not. You can scan a piece of cardboard in at 21.4Mb that does not make the image quality better then a file from an 8Mp camera. There are too many variables to be able to figure this out the way you are attempting to.

The only accurate way to calculate all of this is to evaluate the final product: The print. Because in the final analysis that is really all that matters (to me).

mcmurma

Canon EOS-300D Rebel

01/19/2006 02:31:44 AM · #47

Fellows, this has been an interesting thread to try and follow. I have enjoyed all the posturing and point>>counterpoint discussion.

Personally though, as a non pro who had to send 35mm slides to the lab to get prints made, I was more than sold on digital the day my first 3.2mp image (from the canon s30) came out of my 5 year old HP printer.

If I recall right, my initial response was something like, "holy crap! F*%k film!"

So for me a lowly 3.2mp image was the breakpoint. One look at the print and I knew there was no going back.

I still waited a year before I sold off my 35mm gear... wished I had waited a year longer, as I sorely regret letting go of some of the glass I sold off to fund my first DSLR purchase.

Nuno

Nikon D700

Nikon AF-S NIKKOR 24-70mm f/2.8G ED

01/19/2006 04:28:56 AM · #48

Originally posted by Brent_Ward:

Thanks for the simple yet effective and realistic comparison and explanation.

I don't know what get's in to people's head to get so impressed with film scans just because this way they get huge files. What does this have to do with qualaty? Sertainly everyone con remember the first threads in th enet between film and digital an how the film lovers claim that the digital images had so much grain (back in that first stage). Know they are complaining that the images don't have grain, it's unnatural. Well some people will always find excuses to justify their lobbys. I was reading one of these days in the portuguese photojournalists page some people complaining of editors of newspapers/magazines because some refuse to conttract photographers who doesen't work digital. I've also seen some wedding photographer in some exibicion swear to customers that they would be doing a crucial mistake in having a photog that works digitally, because the supposed qualaty of their film cameras aer much more high.

Each to his own. I think we will be in to a big surprise when respected sites/persons will post some serious tests in digital medium format versus medium format film. And then some will say that big format is what is best! I've seen some people used to medium format drool over a digital 16mp back from kodak over scanned medium format film. And know with 32mp back? Probably they will kill to get one!

Let's wait and see.