Yes, ... film.

As you may remember, I recently had my Canon A1 35mm SLR serviced.

BTW, it's always a Good Idea  to run some kind of equipment test on new photographic equipment of just after a repair ... rather than to risk loosing an important shot.

I did shoot a test roll and had it processed at a nearby 1-hour lab -- this was both to test the camera and also to assess the capabilities of that lab.  In particular, I wanted to know what image resolution (pixel geometry) they could provide from 35mm film.

At that lab, I found some white balance problems in their print processing.  The scan resolution turned out to be a dismal 1545X1024 from a Noritsu Koki QSS-32_33 film processor.  That's only 1.5 megapixels!  Since 35mm film for still photography (135 film) has an image size of 24mm X 36mm, we know that the Noritsu Koki QSS-32_33 film processor at that location scans the film at 42.67 dots per mm or about 1000 dots per inch.

Contrast that with a Pacific Image Electronics PF1800 with 1800 dpi (about $80) or a Nikon Super Coolscan 5000-ED 35mm Film Scanner with 4000 dpi with 48bit color (about $1100).

On Monday, I took that same roll of film to a nearby camera store that also offers 1-hour film and print processing.  When I (innocently) asked scanning a previously developed roll of 35mm film and about the resolution of their scanner, I was offered some outlandish evasions instead of answers ... (a) "film doesn't have pixels -- only digital cameras ..."; (b) "Our film scanner has 'continuous tone processing'" ... whatever that is! ... the implication that the processing will be better than ... who knows what! Eventually, they said that the resolution would be 2000X3000 pixels which would be a 17MBy .TIFF file ... and then went on to say they couldn't provide that format as that software wasn't working.  They did know about JPEG file size -- even though I had said that I didn't care how big the file would be -- that my primary concern was to know the number of pixels in the scanned image.

I have discussed this question with other labs.  All of them complain about their equipment cost; technician cost and how digital is "killing film."

Based on my observations, I would say that it's the film processing labs who are killing film.

What do you think?

--Martin

[followup]

I picked up the CD with the scan done by the camera shop (the first one was done by a drug store with 1-hour processing). 

I need to study the 2 sets of images more carefully than I expected ...

OTOH, I noticed these differences right away ...

(1) the camera shop scan dropped 2 images;

(2) most of the images scanned by the camera shop have more green than the images scanned by the 1-hour drug store processing; Shadow detail is also a little diferent;

(3) one of the images scanned by the camera shop is sharper than the same image scanned by the 1-hour drug store ... must ask if they did any post-processing ... and then gently ask why.  After all, scanning at 2000X3000 pixels shouldn't give me an image that is sharper than one scanned at 1545X1024 ... any thoughts on that? 

I really will need to add my Pacific Image Electronics PF1800 film scanner to my current desktop configuration ... I'll rerun the scan on my own equipment ... and then I'll be at least 3 X as confused as now ... LOL 

I'll let you know what i see ... soon ...

--Martin

I've worked for a few labs that did digital scans (Noritsu and Fuji Frontier machines).

About the green cast - the machines that print digitally (onto the traditional photo paper that goes through chemistry - NONE of my descriptions will reference inkjet printing) color calibrate based on the roll of paper in the magazine and the mixture of chemistry right then and there. It's not a normal "greyscale" type of printing process - it's grey based on what colors will produce grey for THAT paper right then. It usually comes out a tad magenta, and so the machines scan typically a little green to offset it perfectly - to have rich accurate PRINTS, not SCANS. Get it?

Side note - I used to LOVE it when people would hand me a disc and say, "Don't make ANY corrections - I already did them in Photoshop and I don't want anyone screwing them up!" Came out horrible every single time...

About the sharpness - got no idea, really. I'd need to see them. Given that there is more room in the 2000x3000 image than the 1500x1000 (ish - I know there are some more pixels - I'm rounding down for simplicity), there is more "room" to define an edge. When reducing size, so that they are both 1500x1000, the "from the larger image" one just might look sharper and have more detail.

About the sizes - minilabs don't need super hi-res scans. Why is one of the images so small? Because all you need for a good photo-lab digitally printed image is 150ppi - a 600x900ppi image prints a beautiful 4x6 inch print. They can actually print very well based on 125ppi per output size. They tend to smooth things out a bit and fill in edges - so going in all gung-ho that you have 8x10 images at 600 dpi is pretty silly - not only is it overkill, the machines might not even be able to read it right. Really, don't worry about having to go above 200ppi - not kidding!

NOW - something to know about Frontiers! There are 2 flavors of scans, AND burning of CDs! Each labmight call them something different, but what you can do is have them scanned as 10x15, and burn the hi-res CD - if you do NOT denote you want the quality CD, the software automatically downconverts it and you will end up with a lower res, kinda chunky image on your computer. This is so it can lower its memory useage - it has nothing to do with quality, but based on the notion that as far as this image will go is over emaik to family and friends for your vacation photos.

The minilab scanners also aren't that good with shadow detail in general. They're made to go REALLY fast to handle a LOT of film in 1 hour - which is why your Pacific Image scanner will be able to do a much better job - especially with multiple sampling. They are meant to not have many controls - usually basic color and density corrections - which is why you can't do any fine tuning for highlight and shadow details. Make it brighter or darker, maybe a click more magenta. Looks good - next roll!

About the teenagers lipping you off with stuff like "film doesn't have pixels" - um... Yeah... DUH! Got no clue, but then again, they also probably can't spell doo 2 txting - U n0? Technology dumbs things down, and now you have dumbed down labs - sorry dude...

Hope this helps clear some things up-
-Lew ;-)

This gave me food for thought and I didn't even ask the questions.

:-)

About the sharp v. fuzzy scan, I have uploaded the scanned images to this folder DvP.  The file named 91300009.JPG was scanned at a 1-hour drug store; the file named 16A_0032.jpg was scanned at a camera store with 1-hour processing.  BTW, I had to leave the film overnight at the camera store's 1-hour processing ... LOL.

Looking at the images, I now see less of a difference.  I may have to scan the prints on a flatbed scanner to show the difference I saw. 

More later ...

--Martin

p.s. as to the "film doesn't have pixels" ... the guy at the camera store wasn't a kid. 

*"p.s. as to the "film doesn't have pixels" ... the guy at the camera store wasn't a kid."

...that's just sad...

I'll take a look at these after dinner - I'll be back in a little while.

I now have 4 images in the folder DvP. 

The images 91300009.JPG and 16A_0032.jpg are the 1-hour scanned images right off of the CDs. 

The images 91300009PS.JPG and 16A_0032PS.jpg were printed using the print services in Windows Explorere on winXP onto a Canon PIXMA IP5000 printer; those were scanned at 600dpi on my hp 3970 flatbed scanner into Photoshop and saved as medium quality JPEG images. 

--Martin

p.s. box.net offers free! network space for file sharing, collaboration, ... Of course that's for just 14 days ... Still, free! doesn't cost anything.  [grin]
 

@thundering1,

Would it help your analysis of the images if I rescanned the printed images at a higher resolution and saved them as TIFF files? 

--Martin

I haven't gotten back to you - I'm sorry!
Got lost over here - I can look at them when UI get home - when I get home I load email again and everything that came during the day also comes to my home computer at night. I downloaded them already - just haven't gone over them.
Sorry about that - more later this evening-

-Lew

Now it looks like the ones marked PS didn't handle grain as well, but other than that they seem to have about the same sharpness.

Also keep in mind that ALL scans and even digital camera images are a SAMPLING of color values that make up details. They need to be sharpened. Given the slightly chunkiness of the gtrain in the PS marked ones, the chunkiness might become exaggerated because of the sharpening.

Otherwise, the PS ones are simply a Levels adjustment away from being almost identical.

Either way would be good to make prints, especially from a lab printer like a Noritsu and Fuji Frontier. like I said, those types actually smooth things a bit - we're used to looking at them on computers where we can zoom in and see the actual pixels and get disgusted when we see noise and chunkiness. Not a problem for those printers, really.

You'd also be amazed at how much press-printing covers up a little shadow noise as well - camera manufacturers (and those reviewing cameras from said manufacturers) blow things up to see individual pixels to show how well their new cameras handle shadow noise. I've found that you're not going to see it in the final unless it's significant, and I'm pretty picky.

And the bigger you're going to make a print, the further away from it you're going to stand to look at it - not even within arm's length. Of course if you get up close you'll see it in big prints, but you don't really do that unless you're curious. You'll hang it on the mantle and be a good 4-5 feet away and it all blends together.

Hope this makes sense and helps-
-Lew ;-)

Something to note about home scanners - the reason the grain can get fairly chunky is because it's doing a continuous scan around the negative, highlighting and shadowing from one side to another each piece of grain.

This then gives each side a richness and tone - shadows. Now it's more pronounced - yay!

This is not a bad thing - like mentioned above, when using a lab it hides all kinds of stuff, and even inkjet printing blends a bit so it usually comes out fine in the wash, but looking at it onscreen when you can see it at 100% actual, we wonder why we paid so much for a scanner?

-Lew