[HN Gopher] A Better Light Source for Scanning Color Negative Film
___________________________________________________________________
A Better Light Source for Scanning Color Negative Film
Author : eloisius
Score : 186 points
Date : 2024-08-06 04:58 UTC (2 days ago)
(HTM) web link (jackw01.github.io)
(TXT) w3m dump (jackw01.github.io)
| turnsout wrote:
| Productize this! Plenty of people would pay between $200--600 for
| this.
| findthewords wrote:
| This might explain why some new film scans on blu-ray look the
| way they do. Green-yellowish and strange.
| KaiserPro wrote:
| Blue-rays would have been either done by telecine
| https://www.ebay.co.uk/itm/283479247780 (From what I recall its
| nominally a real time machine). That used flourecent light
| sources (although I never worked on it, so that could be a
| lie.)
|
| or by this
| https://www.filmlight.ltd.uk/products/northlight/overview_nl...
| a non realtime scanner with "perfect" registration. Again I
| can't remember the light source, but I suspect its probably an
| arc gap like large projectors. I do know that it has a massive
| cooling chamber to make sure it doesn't heat the film though.
| That scanner is a non-realtime CCD slit scanner.
| jamesfmilne wrote:
| We offered an LED light source for both Northlight 1 and 2 as
| an upgrade over the previous metal halide bulb light source.
|
| https://www.filmlight.ltd.uk/pdf/datasheets/FL-NL-
| DS-0566-NL...
| KaiserPro wrote:
| sweet, I used to work at framestore(cfc) next to where the
| scanning lab was. I really liked watching the raw scans pop
| up
| Palomides wrote:
| doesn't this depend on matching the leds closely to the sensor?
| I'm not aware of camera manufacturers publishing details on the
| wavelengths their sensors respond best to
|
| maybe close enough is fine for this, though
| estebank wrote:
| > I'm not aware of camera manufacturers publishing details on
| the wavelengths their sensors respond best to
|
| It is relatively easy to experimentally find this out
| https://nikongear.net/revival/index.php?topic=10662.0
| flimsypremise wrote:
| Yes, you do actually need to do this. You basically need to
| calibrate every sensor to ensure that the correct wavelength of
| light ends up in the right channel.
| tecleandor wrote:
| I don't know if the creator is around here, but I guess if
| there's anything to consider on the proportion of green, blue and
| red power to adjust the curves.
|
| I think I still have an spectrophotometer around to check that...
| JKCalhoun wrote:
| I'm confused that in a diagram on the page, the BGR LED
| frequencies (particularly R) don't seem to all align with the
| peak sensitivity frequency of the film. It still seems like you
| would want a broader sampling of light. Post-processing,
| regardless of the math involved, is cheap.
| KaiserPro wrote:
| Personally I have found that using LED film softlights to be
| useful for scanning. I didn't have the time to do what this
| wonderful article does, which is research, design and build a
| decent softlight source.
|
| In the old days, you might have been able to use a florescent
| 5600k light sources, as rated ones have a known spectrum that can
| be counted on. Having those in a light table would get you 90% of
| the way to a decent scan.
|
| One thing I did note is that the second colour image appears to
| have nowhere near the aliasing or film noise of the first sample.
| Was its scanned at different settings?
| CrispyKerosene wrote:
| Amazing write up and research - We need more of this!
|
| My feeling is most people who are going to be interested in the
| slight increase in color accuracy are already drum scanning or
| using a virtual drum scanner like a Imacon flextight, and the
| team at Imacon has some crazy color scientists working on that as
| evidenced by the images it outputs.
|
| The quest for the most true colors from C-41 feels like a
| pointless exercise in ways. When i print RA-4 in the darkroom i
| am working with a set of color correction filters and spinning
| dials to mix color on my enlarger head. The resulting print is my
| interpretation of the negative.
|
| Back in the 1-Hour-Photo Minilab days, the tech was doing more or
| the less the same thing as well, or just hitting 'auto' and the
| Noritsu or Frontier was making adjustments to each frame before
| printing it.
|
| If i am scanning the negatives with a camera and light source and
| after inverting, a greenish mask is still present, as like in the
| first conversion example they give, a few tweaks of a few sliders
| in photo editing software is enough to correct it.
|
| The bigger factor at play here in my mind, is the availability of
| robust and consistent color developing services. Most indie labs
| these days are using C41 kits and at best a Jobo machine. There
| are very few labs even offering Dip and Dunk with a proper
| replenishment cycle with chemistry from the big players like
| Fujihunt or Kodak Flexicolor.
|
| A a half a degree off temp, or a developer that near its rated
| capacity is enough to megafuck the resulting negatives.
|
| There is an even worse trend of indie chemistry manufactures
| offering C41 kits with seemingly innocent replacements, that have
| huge consequences. For example one indie manufacturer in Canada
| is shipping there kits without a proper Color Developer (CD4) and
| instead using p-Phenylenediamine, which guarantees the incorrect
| formation of dyes
|
| Sorry if i sound negative and got on a rant, i really do love
| this sort of research.
| shiftpgdn wrote:
| Drum scanning is crazy time consuming and expensive. I shoot
| hundreds (sometimes thousands) of film photos per year and
| 99.999% of my scanning is done with a camera and a backlight.
| gorgoiler wrote:
| Cross polarised light (to eliminate specular reflection) and
| a home made vacuum bed is 99% of the way to a seriously pro
| scanning tool.
|
| A setup like that helped me get through 15k prints in no time
| with excellent results. The biggest barrier to success was
| after churning through the 7x5 and 6x4 shots, things got a
| lot harder with variable sizes of print. It really slowed the
| process down -- and conversely, uniform print sizes made the
| first 90% of the job almost enjoyable. I averaged one "scan"
| every 2s.
| eschneider wrote:
| This is interesting. I still shoot a fair bit of medium format
| film and I have to say that I'm not looking for _accurate_
| color so much as _attractive_ color.
| quercusa wrote:
| I worked in a minilab one summer. The Noritsu printer had,
| IIRC, a +/- 1,2,3 override for R,G, and B. So if you saw a head
| over a big blob of green (someone wearing a red shirt), you'd
| hit +2 Red to override the printer's attempt to "balance out"
| the colors.
|
| We never got any 'interesting' stuff. I suspect people would
| prefer a bit more anonymity than you would get from a 2-3
| person shop where the person who printed your stuff might also
| be the one ringing you up for it.
| aYsY4dDQ2NrcNzA wrote:
| Not only did we get interesting stuff, but we would routinely
| print a few extra prints for a photo album we kept in the
| back.
| js2 wrote:
| My father owned a photoshop that was a one-hour lab for the
| last decade of its life. I worked for him throughout my teens
| and have printed many thousands of photos. I've seen
| interesting stuff. Most of it is pretty boring.
| aYsY4dDQ2NrcNzA wrote:
| > Back in the 1-Hour-Photo Minilab days, the tech was doing
| more or the less the same thing as well, or just hitting 'auto'
| and the Noritsu or Frontier was making adjustments to each
| frame before printing it.
|
| This takes me back. I worked in a one-hour photo place way back
| in the day, operating a Noritsu. We had a film school in town
| and students would often come in with their C-41 or their Tri-X
| and complain about the colors or saturation of their prints.
| Which was totally fair, because tapping the right CMYK buttons
| on the machine was more art than science. Ah, memories.
| mauvehaus wrote:
| Tri-X is traditional gelatin silver black and white.
| anfractuosity wrote:
| Are new drum scanners still being made out of interest? It
| appears fairly hard to find used ones.
|
| They sound a bit awkward to use from what I've read, as I think
| you need to use liquid to adhere the film to the drum
| correctly?
| kkukshtel wrote:
| > If i am scanning the negatives with a camera and light source
| and after inverting, a greenish mask is still present, as like
| in the first conversion example they give, a few tweaks of a
| few sliders in photo editing software is enough to correct it.
|
| I think this is a major point. I applaud the effort of the post
| and would (as a Mamiya 7 shooter!) love a whole unit better
| than the Epson V600, but correcting a color cast in the film
| scan is trivially easy in an photo editing tools these days. I
| scan and get tifs and can tweak to whatever. More important are
| the iris/optics of the scanner itself and how flat the film is
| inside the bed.
| gorgoiler wrote:
| I'm not sure the indie, non megalab chemistry kits ought to be
| so easily dismissed. I have had fantastic results working with
| Tetenal Colortec in the past with really not that much more
| than a shift in the cyan direction. And this was using a
| kitchen sink for thermal stability.
|
| C41 is such a toilet process anyway -- everything is shades of
| brown?! -- that I can't imagine anyway would look for precise
| color work from it the same way I can't imagine anyone would
| look for resolution for 135 stock.
| whycome wrote:
| > Sorry if i sound negative
|
| Apt
| zokier wrote:
| If you want to get serious on this, get good quality color
| chart[1] and use that to compare different light sources etc.
| Just eyeballing resulting colors from random photos and
| guesstimating the various spectral curves gets you only so far.
|
| [1] e.g. https://www.silverfast.com/products-overview-products-
| compan...
| dvdkon wrote:
| Yes, but you'd want that colour chart on the type of film
| you're scanning, for reasons explained in the OP. Sadly all I
| found in a brief search were calibration targets on _slide_
| film, not negatives.
| musictubes wrote:
| Back in the early 90s I used a Noritsu printer. We had
| reference negatives on all the different film stocks, or at
| least all the ones we regularly saw there. We would group the
| jobs by film type. Set the calibration by the calibration
| negative and judge exposure and basic color correction by
| direct viewing of the negative. Another person would check
| the prints and flag any that needed to be redone for color,
| dust, exposure, etc. Then we'd change film types.
|
| The next time I touched a photofinishing machine in the early
| 2000s you looked at a screen to make adjustments and we
| offered digital services like scanning and printing from
| digital files. I still used my negative reading skills to
| talk to customers when we were troubleshooting results.
| Putting the negative on the light table to show them how thin
| they were or how wildly the color changed when you switched
| what kind of light the picture was shot in was the quickest
| way to resolve quality complaints.
| zokier wrote:
| I'd imagine that just grabbing the reflective target and
| shooting it yourself on film would get decent results?
| Assuming the target patches have good spectral coverage
| fallinditch wrote:
| I'm planning to do some negative scanning with a phone or iPad as
| a light source. I know I'll have to make some simple tweaks to
| the color balance of the scans. I believe it is totally normal to
| have to make some adjustments to scans, the side by side example
| in the article seems to show that a white light source is
| perfectly fine for this work. It's unlikely that an RGB light
| source would produce scans that don't require _any_ adjustments,
| so I 'm failing to see the benefit.
| cedricd wrote:
| I've scanned a few hundred images using an iPad as the light
| source. I've tried both a white screen and a bluish screen
| designed to basically invert the orange cast from the negative.
|
| Both seem to work well. The bluish thing works quite well, but
| it turns out that different rolls need slightly different light
| color to compensate, so it wasn't worth the trouble. In the end
| the best result was buying a license for Negative Lab Pro[0] to
| post process everything
|
| [0]: https://www.negativelabpro.com/
| fallinditch wrote:
| That's a good recommendation, thank you. It's amazing how
| complicated film photography has become in the digital era!
| My next task is to ditch Lightroom/Photoshop subscription.
| I'm going to give Darktable a go. Edit: aaah, I see
| NegativeLabPro needs Lightroom, hurrumph...
| kosma wrote:
| Seconded. NLP is well worth the money - not just in results,
| quality, and time saved, but also in finding joy in shooting
| and scanning color.
| zackmorris wrote:
| Even after working with colorspaces for decades in Photoshop and
| various game dev tools, I find color conversion mystifying. I've
| studied all of the equations and given it my best effort, but
| would not bet real money that the colors I'm displaying are close
| to real life. It's like the game of telephone, I just can't trust
| so many steps.
|
| So for this article, I don't see mathematical proof that the
| negatives have been inverted accurately, regardless of method,
| even though I'm sure the results are great. I suspect it comes
| down to subjective impression.
|
| Here's a video I found discussing monitor calibration:
|
| https://www.youtube.com/watch?v=Qxt2HUz3Sv4
|
| If I could fix everything, I'd make all image processing
| something like 64 bit linear RGB and keep the colorspace internal
| to the storage format and display, like a black box and not
| relevant to the user. So for example, no more HDR, and we'd
| always work with RGB in iOS instead of sRGB.
|
| Loosely that would look like: each step of image processing would
| know the colorspace, so it would alert you if you multiplied sRGB
| twice, taking the onus off of the user and making it impossible
| to mess up. This would be like including the character encoding
| with each string. This sanity check should be included in video
| card drivers and game dev libraries.
|
| If linear processing isn't accurate enough for this because our
| eyes are logarithmic, then something has gone terribly wrong.
| Perhaps 16 bit floating point 3 channel RGB should be standard. I
| suspect that objections to linearity get into audiophile
| territory so aren't objective.
|
| For scanning color negatives, the brand of film would be mapped
| to a colorspace, the light source would have its own colorspace,
| the two would get multiplied together somehow, and the result
| would be stored in linear RGB. Inversion would be linear. Then
| the output linear RGB would get mapped to the display's sRGB or
| whatever.
|
| My confusion is probably user error on my part, so if someone has
| a link for best practices around this stuff, I'd love to see it.
| lcrs wrote:
| Colour in the Photoshop/gamedev world is often handled pretty
| casually, but if you're interested the moving picture world
| gets a lot more rigorous about it and there's tons of
| documentation around the ACES system in particular:
| https://github.com/colour-science/colour-science-precis
| https://acescentral.com/knowledge-base-2/
|
| As you suggest storage in linear 16-bit float is standard, the
| procedure for calibrating cameras to produce the SMPTE-
| specified colourspace is standard, the output transforms for
| various display types are standards, files have metadata to
| avoid double-transforming etc etc. It is complex but gives you
| a lot more confidence than idly wondering how the RGB triplets
| in a given JPG relate to the light that actually entered the
| camera in the first place...
| qingcharles wrote:
| If anyone is doing this seriously, calibrate your monitor,
| calibrate your scanner:
|
| https://www.silverfast.com/products-overview-products-compan...
|
| BUT.. here's the rub: if your film is old, it has probably
| faded. So whatever you scan is going to be "wrong" compared to
| what it looked like the day it was taken. The only way to
| easily fix that is to try and find the white point and black
| point in the scan and recalibrate all your channels that way.
| Then you're really just down to eyeballing it, IMO.
| keepamovin wrote:
| This is cool. The original looks kinda green to me. Awesome.
| alnwlsn wrote:
| Anyone know if this is the right technique to use on 8/16 mm
| movie film (which is a positive instead of negative)? Modifying
| an old projector to go one frame at a time is the easy part, but
| you can't use the original halogen bulb since it will burn a hole
| right through the film at that speed.
| Finnucane wrote:
| Yeah, it would be interesting if this makes much of difference
| for slide film, either e-6 or Kodachrome (I am old and have
| both).
| tecleandor wrote:
| In the article they say that for positive film (slides, at
| least) white light is generally better than rgb, but it would
| be nice to test it.
| Finnucane wrote:
| Missed that. I guess that makes sense, since those films
| are designed for projecting more than printing.
| smogcutter wrote:
| There are also rotoscoping and analysis projectors designed to
| hold on a frame.
| qingcharles wrote:
| I scanned a lot of positives on an Epson V850 flatbed just
| fine. Except for the resolution and the setup being a bit
| finicky, there wasn't much between that and the $25K X5 scanner
| I had.
| felixhandte wrote:
| Awesome work!
|
| I get exactly that green cast and muted color range off of my
| flatbed scans (Epson v800). This is a really intriguing path to
| fixing them I hadn't considered.
|
| It seems like the writeup here doesn't specify what you're using
| for the actual imaging? A flatbed scanner? A camera?
| esafak wrote:
| After all that work I was expecting a chromaticity diagram to
| demonstrate the expanded gamut, but nice job regardless.
| flimsypremise wrote:
| RGB scanning doesn't actually expand the color gamut, but
| removes erroneous color information. If you use white light you
| end up recording color information from the dyes in wavelengths
| outside of those that RA-4 paper is sensitive to, and which the
| color engineers who designed the film never intended it to be
| used with.
| esafak wrote:
| Thank you for the correction. Then you can measure the color
| difference?
| assimpleaspossi wrote:
| Looking at the results, it looks to me that the print with the
| white light has far more detail while the RGB print has washed
| out ground under the tower.
| pimlottc wrote:
| Yeah, it's not clear to me either that the RGB image is
| obviously better, especially without knowing that actual
| conditions were like when the photo was taken.
|
| Perhaps the author could explain why they find one image
| superior instead of just putting two images side-by-side, with
| the implied message that "any idiot can see that <x> is
| better".
| throw0101b wrote:
| > _White light scan captured using 95+ CRI 5000K light source.
| RGB scan captured using custom 450nm+525nm+665nm light source._
|
| While high-CRI is better than low(er)-CRI, one criticism is that
| the 'score' is somewhat lacking in it measure an important
| component:
|
| > _Ra is the average value of R1-R8; other values from R9 to R15
| are not used in the calculation of Ra, including R9 "saturated
| red", R13 "skin color (light)", and R15 "skin color (medium)",
| which are all difficult colors to faithfully reproduce. R9 is a
| vital index in high-CRI lighting, as many applications require
| red lights, such as film and video lighting, medical lighting,
| art lighting, etc. However, in the general CRI (Ra) calculation
| R9 is not included._
|
| [...]
|
| > _R9 value, TCS 09, or in other words, the red color is the key
| color for many lighting applications, such as film and video
| lighting, textile printing, image printing, skin tone, medical
| lighting, and so on. Besides, many other objects which are not in
| red color, but actually consists of different colors including
| red color. For instance, the skin tone is impacted by the blood
| under the skin, which means that the skin tone also includes red
| color, although it looks much like close to white or light
| yellow. So, if the R9 value is not good enough, the skin tone
| under this light will be more paleness or even greenish in your
| eyes or cameras.[25]_
|
| * https://en.wikipedia.org/wiki/Color_rendering_index#Special_...
| twic wrote:
| The thing i found most interesting here is the brightness
| enhancing film:
|
| https://newhavendisplay.com/blog/brightness-enhancement-film...
|
| Basically, it's a collimator: it takes light going in all
| directions (eg from a lamp), and turns it into light all going in
| one direction.
|
| What does it look like to look through? Do objects appear
| brighter? I suppose they appear brighter but also smeared out?
| grvbck wrote:
| > What does it look like to look through? Do objects appear
| brighter? I suppose they appear brighter but also smeared out?
|
| Pretty much exactly so: https://youtu.be/ugkjNPH1J-4
| flimsypremise wrote:
| So I wrote an article about this a few years back and also
| developed a custom RGB light for my own scanning:
|
| https://medium.com/@alexi.maschas/color-negative-film-color-...
|
| There's also some proper academic research into this subject
| going on currently:
| https://www.researchgate.net/publication/352553983_A_multisp...
|
| One thing that's important to note about this process is that the
| idea is not to _image_ the film, but rather to measure the
| density of each film layer and reconstruct the color image from
| that information. This is a critical realization, because one of
| the most important things to know about color negative film is
| that the "color" information in the negative actually only exists
| relative to the RA-4 printing system. Negatives themselves don't
| have an inherent color space.
|
| Cool to see someone else working on this though. I actually
| considered those drivers for my build, but I ended up building a
| very high frequency, high resolution PWM (30khz/10bit) dimming
| solution with TI LM3409 drivers. It's very hard to get uniform
| light as well so I ended up getting some custom single chip RGB
| LEDs.
|
| https://i.imgur.com/BVM9p6Q.jpeg
|
| https://i.imgur.com/5oozHnN.jpeg
|
| I've been working on this for a few years, and what I will say is
| that there's actually another level of complexity beyond just
| implementing the light. There's a lot of testing to ensure that
| you're getting proper linearization of each channel, and there's
| still a color crosstalk problem arising from the misalignment
| between the color sensitivity of most modern digital cameras and
| the bands that are used to scan color negatives. It requires some
| additional tweaking to get all of the color information in the
| correct channel. You can also very easily end up saturating a
| channel without realizing it as well. Oversaturated reds are a
| common occurrence in RGB scanning.
|
| I'd also note that the wavelengths you should shoot for are more
| along the lines of 440nm 535nm 660nm, which correspond to the
| Status M densitometry standard. This standard was designed
| specifically for color negative film.
| m463 wrote:
| Is there a way to process dust/scratches? Like wavelengths
| outside the chosen r/g/b range?
| matthews2 wrote:
| Dust (and scratches?) can be detected with an infrared scan.
| The IR scan is only used to detect defects, and then
| something like a spot removal tool is automatically applied
| to the defect areas.
|
| https://www.hamrick.com/blog/digital-ice.html
| flimsypremise wrote:
| Technically, yes. I know a few people have done it. In a
| practical sense it is very difficult and you are unlikely to
| get it working without a lot of trial and error. The tricky
| part is that the IR image needs to be perfectly aligned with
| the rest of the image data, which introduces a number of
| difficulties.
|
| * You can modify a sensor for IR, though this is often a
| costly and difficult modification. But even if you do so, the
| IR focal distance is different from the visible light focal
| distance. So for every shot you need to refocus for IR, but
| also ensure that the refocussed IR image is exactly the same
| size as the visible image.
|
| * You can use another sensor that is sensitive to IR, but its
| probably not going to have the same resolution, you're going
| to struggle to somehow have both cameras see the target
| image, and then once you get both exposures, alignment
| becomes a problem.
|
| So yeah, doable but non-trivial.
| hoherd wrote:
| Maybe it's because I'm colorblind, but the top-right image looks
| much better than the bottom-right image to me. Can somebody
| explain why the bottom-right image is allegedly superior? I know
| there's a write up about what's going on and all the science
| behind it, but what I'm asking about is what you as a person with
| color receptive vision sees that is better.
| realreality wrote:
| The top photo has a blue-green cast, whereas the bottom photo
| has a magenta cast.
|
| Maybe the bottom one is a more realistic reproduction of the
| scene, but I also prefer the top one, which is more saturated
| and closer to what I associate as a film image.
|
| Each kind of film has its own character and color variations;
| it's silly to try to neutralize everything.
| asimpletune wrote:
| I looked at both before knowing which was which. Immediately I
| recognized the look of the top right photo, whereas the bottom
| right didn't quite seem to have "the look". So, I think it
| might be that it looks better to you because it looks more like
| how a photo looks. It's similar to how younger people may
| prefer 60fps or weird settings on TV shows that give it the
| "soap opera effect" vs how older people can't stand them.
|
| After switching back and fourth and really looking closely at
| each one I ended up deciding that I liked the bottom right
| photo, even though I could recognize the top right one had a
| more classic film look. For me it was just because there was
| more detail in the colors. The original scan was kind of washed
| out in the blues I guess, as well as being a little more red in
| the dirt area.
| mintycrisp wrote:
| To me, the bottom right image has a smoother more gradual range
| of colors while the top right seems like the saturation is
| turned up a bit too high so many of the same colors blend in
| loosing some of those color details. Like the typical blue sky
| present in the top right, in the bottom right version goes from
| a similar vibrant blue to light purple as the sky extends to
| the horizon. Similarly the bottom right's foreground
| trees/hills details of green/tan colors pop out more more as
| they sit together giving you a greater sense of detail to the
| dense foliage.
| anfractuosity wrote:
| Interestingly that doesn't appear to mention infrared from a
| quick scan, which is used to help remove dust as far as I
| understand.
|
| (I've got an old Canon FS4000, which uses IR)
| m463 wrote:
| I scanned negatives long ago without IR and it was horrible.
| Dust and fiber were a major headache, especially when scanning
| lots of film.
|
| But since we're living in the future, I suspect we could make
| AI models that would work practical magic.
| anfractuosity wrote:
| Heh, yeah good point, I imagine that would work well too,
| didn't think of that.
| zokier wrote:
| Btw regarding the camera sensitivity, if you shoot raw and just
| shoot the different colors separately, you can mostly ignore the
| spectral characteristics of the sensor. Debayering might end up
| being very different than standard though.
| mikewebkist wrote:
| It seems like an alternative would be a broad-spectrum white
| light source with narrow-band color filters that correspond to
| similar wavelengths to the LEDs mentioned. That would require
| simpler light source but more costly subtractive filtering.
|
| All those old-school minilabs pre-blue LEDs...they must have used
| white light sources and filters, right?
| strogonoff wrote:
| TL;DR Negative film is (obviously) intended not for viewing by
| humans, but for a specialized development process. Digital
| cameras are geared towards capturing images as humans would
| perceive them, and in regular photography using _full_ spectrum
| light supposedly makes metameric failure less likely. Thus, it
| may appear counter-intuitive to a seasoned photographer that
| using a specific narrow band RGB lighting can be preferable when
| digitizing typical negative film, working around the use case
| mismatch and improving colour reproduction.
___________________________________________________________________
(page generated 2024-08-08 23:00 UTC)