Bad processing or "film grain" as lab says

In my brief stint of getting paid for work back in the mid-2000s, I used a lab that printed optically. A lot of my more mundane stuff was handled by Wal-Mart, either in store or sent-off, which back then meant scanning on a Frontier and printing to RA-4.

Now, I don't shoot color negative film regularly, and when I want to print either it or slide film I scan it myself and have it printed whatever way is convenient(sometimes that means dye sub at the drug store, sometimes that means RA-4 at a photo lab, but always as a digital file). My B&W gets printed on an enlarger in the darkroom.

Regardless of how a negative is printed, though, a badly exposed negative will still print like crap. I would say that, in my experience, optical prints from an unexposed negative look different than ones that are scanned and brightened digitally, but the end result is that underexposed negatives tend to have accentuated grain and an overall lack of contrast.

As a side note, if I have a thin negative that MUST be printed, I can usually do better than a minilab at getting a scan that will print okay(note I said okay, not well).

I can use tricks like multi-pass scanning to my advantage to dig as much out of a negative as possible, but no scanner can find things that don't exist. Something like an Imacon or Hasselblad(or even better a PMT-based drum scanner) can probably do better than my Nikons, but then it costs a lot of money to pay someone who knows how to use one of those to spend the time getting a good scan from it.

The bottom line, though, is to start with a properly exposed negative and you'll have the easiest time getting good prints regardless of how the end result is actually printed.

It is pretty amazing sometimes what exists in the tail of the curve in C41 film.

I printed an underexposed negative once on Panalure, and there was so much more detail than you would see looking directly at the negative.

I suspect, though, that scanners don't do that as well, at least not with all the automatic modes on.

(Also, with overexposed negatives, it is hard to see visually what printing will bring out.)

I printed an underexposed negative once on Panalure, and there was so much more detail than you would see looking directly at the negative.

 

I suspect, though, that scanners don't do that as well, at least not with all the automatic modes on.

I won't dispute that, but it also confirms to me that dedicated film scanners still have their place.

I've never used a Konica, although I know they're highly regarded. Still, though, for an iffy negative I can definitely get better results on a Nikon than I can with an automated scan from a Frontier/Noritsu.

The Konica and Nikon scanners, even though they were(subjectively) expensive when new, are still firmly consumer scanners and they have nothing on what even an Imacon or the like can achieve, much less a "real" drum scanner that uses a PMT rather than a CCD. I deal with PMTs regularly(although not in imaging applications) and their close relative the electron multiplier on basically a daily basis. When you consider that even a single photon/ion can generate a measurable signal(although admittedly you are limited by the noise in the entire system) it becomes clear just how much better they are than a CCD/CMOS.

.....consumer scanners and they have nothing on what even an Imacon or the like can achieve, much less a "real" drum scanner that uses a PMT rather than a CCD. I deal with PMTs regularly(although not in imaging applications) and their close relative the electron multiplier on basically a daily basis. When you consider that even a single photon/ion can generate a measurable signal(although admittedly you are limited by the noise in the entire system) it becomes clear just how much better they are than a CCD/CMOS.

Hmmm. But a Photomultiplier tube has a photo-cathode target with about 1 million times bigger area than a CMOS or CCD photosite, and is monochrome sensitive (i.e. unfiltered). To make it work it needs a highly focussed spot of light that then defocusses into the PM tube, intercepted by a crude filter wheel whizzing around, or a prismatic beam-splitter. With no chance of using infra-red defect detection and correction.

That's like comparing Baird's 40 line Nipkow disk with a modern 4K video camera. Plus sensor sensitivity is a complete irrelevance if you have a bright enough light source to begin with, which LED lighting can easily provide.

Each system has its weaknesses, and in the case of the drum scanner, it's the 'flying' spot, which can't possibly be made small enough to warrant the ridiculously high PPI output figures claimed for drum scans.

Besides, there's no 'modern' 3 line sensor being produced using the latest CMOS technical advances. So we're stuck with nearly 100 year old opto-mechanical technology scanners, or those using at least 10 years outdated solid-state sensors.

I see no technical reason why a shoe-box sized dedicated film scanner using a 3-line or matrix CMOS sensor couldn't be produced, and with an output quality to rival or exceed that of the best drum scanners. I can also see every reason why such a venture would be doomed to abject economic failure.

As an analogy, it would be totally possible to utilise the same Laser technology used in CD players to read vinyl disks with improved fidelity and less noise. But the established way of doing things is to scrape a needle down the groove. You only have to read the number of unfounded objections to DSLR 'scanning' on this forum to understand how economically risky it would be mixing new and traditional (dare I mention the analogue word?) technology.

I nearly forgot the point of this thread!

Here's an example comparison I made years ago; between a 2700 ppi film scanner, and a 6000 ppi scan. The 200 ISO colour negative was wet-printed at 10x enlargement, and the print scanned at 600 ppi to give an effective 6000 samples per inch. This represented the 'grain' (dye clouds) as near visually perfectly to the print as makes no difference - unlike the obvious aliasing caused by sampling at a lower spatial frequency.

The interference between the digital sampling spacing and modal dye-cloud clumping is very obvious, and I think quite conclusive of the aliasing mechanism involved.

<p>AFAIK the entire Portra range has been designed as low grain and neutral colours; it is also very tolerant towards under- and over-exposure. For instance Fuji 400 Pro has more pronounced grain than Portra 400.</p>

<p>In my view the poor (medium resolution) scan is responsible for your dissatisfaction. High resolution TIFF in 16-bit mode run from around 75mb and over 200mb.</p>

<p>If you want to shot film perhaps you should buy a scanner and do the scanning yourself. No auto-settings are good for each negative, and I see no reason why such critical step as scanning should be left to chance.</p>

Portra is tolerant of over expose. But underexpose and you'll end up with a muddy mess.

 

(snip)

 

Besides, there's no 'modern' 3 line sensor being produced using the latest CMOS technical advances. So we're stuck with nearly 100 year old opto-mechanical technology scanners, or those using at least 10 years outdated solid-state sensors.

 

I see no technical reason why a shoe-box sized dedicated film scanner using a 3-line or matrix CMOS sensor couldn't be produced, and with an output quality to rival or exceed that of the best drum scanners. I can also see every reason why such a venture would be doomed to abject economic failure.

 

As an analogy, it would be totally possible to utilise the same Laser technology used in CD players to read vinyl disks with improved fidelity and less noise. But the established way of doing things is to scrape a needle down the groove. You only have to read the number of unfounded objections to DSLR 'scanning' on this forum to understand how economically risky it would be mixing new and traditional (dare I mention the analogue word?) technology.

Much of the advance in DSLR sensors is more sensitivity, which isn't needed for scanners.

But yes, I don't know about the supply or market for linear sensors for scanners.

Not so long after CDs closed the market for vinyl disk turntables, someone did start selling a laser based one.

The price tag was high, and I suspect no economy of scale. Maybe useful for archivists, but not for anyone else.

DSLR based scanning is reasonable if you already have a DSLR, and also have an appropriate lens and

film holder. (and light source.)

Dedicated 35mm scanners, with appropriate optics to image the frame width on a linear sensor,

do a nice job for a reasonable price. Flatbed scanners do a good job for larger formats.

These have some economy of scale from document (opaque) scanning users, which I suspect

is a much larger market than film scanners.

As an analogy, it would be totally possible to utilise the same Laser technology used in CD players to read vinyl disks with improved fidelity and less noise. But the established way of doing things is to scrape a needle down the groove. You only have to read the number of unfounded objections to DSLR 'scanning' on this forum to understand how economically risky it would be mixing new and traditional (dare I mention the analogue word?) technology.

There are laser "turntables" and they do have advantages when used with extremely clean vinyl.

There are laser "turntables" and they do have advantages when used with extremely clean vinyl.

Yes.

As I remember, for one, traditional stylii follow the sides of the groove, the laser follows the bottom, which

might be dirty, or otherwise not optimal. Also, the traditional stylus will push some kinds of dirt out of

the way, that the laser will read the dirt.

The laser video disk players that were around before DVDs, and maybe the early CD players, used

gas (HeNe) lasers that were big and expensive. The development, and economy of scale in production,

of CW semiconductor lasers made affordable CD and DVD players possible.

If not for CDs, it might have been that laser scanned vinyl disks would have become popular.

It is interesting to remember now, that computers with disk drives big enough to hold a whole CD

were not common when CDs first appeared. Along with CDs, a system for storing CD data as

a video signal on video tape was invented. With multi TB disks and multi GB RAM today, it is

easy to forget how things were not all that long ago.