The Price of 8x10 Color Film Out of Control

[138S]

in short, if it were really that straightforward, then why does it seem to be so difficult to get consistent/good color negative scans?

Epson V850 makes fantastic color conversions, it is a cheap scanner but it is backed by a company sporting top notch color science, not by chance they have that strong position in Pro inkjet printers.

The Fuji Frontier and the Noritsu also had a lot of color science backing the conversions. Have you ever seen a Frontier print looking bad even with auto settings? Throw to the frontier the film you want, consumer or Pro... no problem.

...but if having problems: 3D LUT Creator of equivalent. With that you do what you want with color.

Please see this (it's only the basic/cheap version):



Me, I've no problem with color, beyond my own artistic shortcomings.

 

[Adrian Bacon]

Epson V850 makes fantastic color conversions, it is a cheap scanner but it is backed by a company sporting top notch color science, not by chance they have that strong position in Pro inkjet printers.

The Fuji Frontier and the Noritsu also had a lot of color science backing the conversions. Have you ever seen a Frontier print looking bad even with auto settings? Throw to the frontier the film you want, consumer or Pro... no problem.

...but if having problems: 3D LUT Creator of equivalent. With that you do what you want with color.

Please see this (it's only the basic/cheap version):



Me, I've no problem with color, beyond my own artistic shortcomings.

I have a V850 Pro. It’s not the hardware that does the color conversions, it’s the software. Yes, I’ve seen plenty of frontier scans that are not up to snuff. In fact most of them are not. The fact that you don’t seem to have much problems with the color output tells me everything I need to know. Your definition of acceptable color quality is quite a bit more lax than a fair number of other members here on Photrio. There is nothing wrong with that, however, you would do well to recognize that if somebody else says you’re simplifying things, there’s a very high probability that you are, so continuing to try justify/prove your point and/or repeat yourself instead of figuring out where you can extend your knowledge is counter productive.

 

[Sal Santamaura]

...continuing to try justify/prove your point and/or repeat yourself...is counter productive.

Particularly when feeding a troll.

 

[Adrian Bacon]

Particularly when feeding a troll.

I just couldn’t resist... I did good and refrained for a while though.

 

[Alan Edward Klein]

The first time I accidentally walked into a Kinkade gallery I burst out laughing (not deliberately - couldn't help it). And when Iater I accidentally walked in to Lik gallery I literally felt sick to my stomach, and walked straight back out. Seems like he dosed kindergartners with LSD and then handed them cans of fluorescent spray paint. Kitchy, kitchy, kitchy.

But all Lik's female salespeople are hot. When they're selling to you, your brain melts. It was a good thing my wife was with me to pull me out of the gallery.

 

[jtk]

What is a Kinkade and what's this got to do with the price of 8X10 film?

My first serious camera was an Exacta 66 (6X6). I needed a view camera so an Italian Buddhist sold me an Agfa Ansco 8X10... I dumped it that for a 4X5 Linhof.

I can't imagine a woman who wants to "pull me out" of anywhere.

 

[DREW WILEY]

You mean you have an art background and never heard of the greatest painter of all time? He took paint by numbers into the big leagues. People would pick the colors they wanted to match their sofa and whatever, then he had assembly lines of assistants fill in the numbered areas with those colors; and finally he'd put a tiny spot or two of paint on with his own hand to dodge art fraud laws. But he eventually got indicted for business fraud. In this topic, Peter Lik followed both his snake oil business model as well as abominable coloration, but in his case using 8x10 film (either color or B&W), then horrendously colorizing it in PS. I really don't know why he uses a camera at all, but he is a consumer of 8x10 color film who has made a lot of money with it. In both Kinkade and Lik galleries the gals walk up in their high heels trying to convince you they're selling something that can turn lead into gold. I once told one of them they'd have to pay me $50,000 just to put one of those abominations on my wall, and then I'd reserve the right to place another layer of drywall and plaster over it.

 

[Alan Edward Klein]

I have a V850 Pro. It’s not the hardware that does the color conversions, it’s the software. Yes, I’ve seen plenty of frontier scans that are not up to snuff. In fact most of them are not. The fact that you don’t seem to have much problems with the color output tells me everything I need to know. Your definition of acceptable color quality is quite a bit more lax than a fair number of other members here on Photrio. There is nothing wrong with that, however, you would do well to recognize that if somebody else says you’re simplifying things, there’s a very high probability that you are, so continuing to try justify/prove your point and/or repeat yourself instead of figuring out where you can extend your knowledge is counter productive.

My tastes are simple. So much of this thread is beyond me. I shoot Velvia 50 because I'm a sucker for punch. All processing is done in a pro lab. I use to make color prints in a lab. I think they used 4x5 internegatives for the 16x20 prints. Printed on R paper, I think. However, I'm not doing that right now, I'm scanning with a V600 using Epson scan software. I adjust in Lightroom or Elements. I also find Velvia easier to scan than negative color film. I had trouble with Ektar 100 getting the colors right. Plus chromes let you know immediately which of the bracketed shots is correct.

When I adjust the colors after the scan, I don;t compare to the original Velvia film. I adjust until it looks right to me on my calibrated monitor. I figure, if it looks right to me, it will look right to others. Plus no one, including me, sits there comparing the monitor screen with the original colors on the film. In any case, the Fuji film designer's taste may be different than mine. Who says his are better? Here's two Velvia shots taken minutes apart on the same roll with two different lenses on an RB67. I think they were 5t0mm and 90mm. Which represents the original film better? Does it matter? Some might like the more contrasty version and others might like the less contrasty wide angle version. My point is aesthetics are internal and everyone got an opinion.

As an aside Adrian, I've just bought a 4x5 camera, something I never shot before and some Velvia 50 direct from Japan. (They don't sell it in America in that format). So I need a new scanner, the V600 I have only scan medium format max. Do you use the V850 Pro with Epsonscan or Silverfast? And would you recommend the V850 or another for home use?
https://www.flickr.com/photos/alanklein2000/5270429762/in/album-72157625476289859/
https://www.flickr.com/photos/alanklein2000/5270637805/in/album-72157625476289859/

 

[Lachlan Young]

in short, if it were really that straightforward, then why does it seem to be so difficult to get consistent/good color negative scans? Have you actually seen color negative scans from most labs lately?

...still, you should be able to get a fairly consistent and colormetrically accurate scan from one lab to the next, and you can’t, despite the fact that most their equipment is the same. That says a lot.

What's even more fascinating to me is if you can get the un-inverted data from some of those scans & invert them yourself you quickly discover that the problems are often a mix of software written by people who don't understand how colour neg is supposed to work & operators who assume that the software was written by someone who does know how colour neg works... All too often I've found that making a new scan which when inverted will actually place the values in correct relationships to each other is easier and faster than trying to de-f*** the strange colours etc in files supplied for colour correction.

Or you get certain minilab systems designed to present semi-skilled operators with a range of 'look' options that are difficult to completely screw up & allow the file to be comfortably placed inside the gamut of whatever paper is in the exposure unit - at the cost of an actual representation in latitude and gamut of what the film really recorded.

 

[Adrian Bacon]

My tastes are simple. So much of this thread is beyond me. I shoot Velvia 50 because I'm a sucker for punch. All processing is done in a pro lab. I use to make color prints in a lab. I think they used 4x5 internegatives for the 16x20 prints. Printed on R paper, I think. However, I'm not doing that right now, I'm scanning with a V600 using Epson scan software. I adjust in Lightroom or Elements. I also find Velvia easier to scan than negative color film. I had trouble with Ektar 100 getting the colors right. Plus chromes let you know immediately which of the bracketed shots is correct.

When I adjust the colors after the scan, I don;t compare to the original Velvia film. I adjust until it looks right to me on my calibrated monitor. I figure, if it looks right to me, it will look right to others. Plus no one, including me, sits there comparing the monitor screen with the original colors on the film. In any case, the Fuji film designer's taste may be different than mine. Who says his are better? Here's two Velvia shots taken minutes apart on the same roll with two different lenses on an RB67. I think they were 5t0mm and 90mm. Which represents the original film better? Does it matter? Some might like the more contrasty version and others might like the less contrasty wide angle version. My point is aesthetics are internal and everyone got an opinion.

As an aside Adrian, I've just bought a 4x5 camera, something I never shot before and some Velvia 50 direct from Japan. (They don't sell it in America in that format). So I need a new scanner, the V600 I have only scan medium format max. Do you use the V850 Pro with Epsonscan or Silverfast? And would you recommend the V850 or another for home use?
https://www.flickr.com/photos/alanklein2000/5270429762/in/album-72157625476289859/
https://www.flickr.com/photos/alanklein2000/5270637805/in/album-72157625476289859/

I use the 850 with vuescan, and only use it to scan 4x5 and up. If you need to scan large format, your options are pretty limited without doing any stitching.

 

[Adrian Bacon]

What's even more fascinating to me is if you can get the un-inverted data from some of those scans & invert them yourself you quickly discover that the problems are often a mix of software written by people who don't understand how colour neg is supposed to work & operators who assume that the software was written by someone who does know how colour neg works... All too often I've found that making a new scan which when inverted will actually place the values in correct relationships to each other is easier and faster than trying to de-f*** the strange colours etc in files supplied for colour correction.

Or you get certain minilab systems designed to present semi-skilled operators with a range of 'look' options that are difficult to completely screw up & allow the file to be comfortably placed inside the gamut of whatever paper is in the exposure unit - at the cost of an actual representation in latitude and gamut of what the film really recorded.

Yep. I'm the first to admit that I don't consider myself to be the foremost expert when it comes to color neg film. My scans from a couple of years ago don't look nearly as good as scans that I do today, and I'm sure another couple of years of the same vigorous study I've been doing will result in scans that look that much better than my scans today, though, today they're at a good enough level in terms of color that most everybody who uses my services won't willingly go back to scans from other labs. The combination of providing floating point DNG files and as accurate as possible tone scale and color has proven to be a combination that many are looking for.

 

[138S]

Yep.

What's even more fascinating to me

For an advanced edition, today we have very powerful tools to edit color in the way we want without limitations.

In particular Photoshop layers/masks combined with a plugin like 3D LUT Creator offers total freedom to get what we want, single limit is our aesthetic criterion and our creativity .

Going back to theory, the real color capture a film makes is the silver/halide content in the 3 layers after 1st developer. Those silver/halide contents are abstract color. Dyes resulting after color Developer are only an interpretation of the real color capture, but with a tool like 3D LUT Creator we deformate the color space like we want. Additionally with Ps layers/masks we can apply a different color space deformation (or interpretation) to each separate subject in the image, this allows some limited edition of the film spectral signature.



This is better understood when one has practiced Tri-Color photography, we depart from 3 BW images, color filters for each image are critical, later we assign colors and channel cross-talk and color space deformations to the abstract 3 BW color capture.

Every photographer wanting to master color technique should practice some tri-color to start learning deep concepts. It can be done with a SLR/DSLR with static scenes, to not waste film.

At all, those deep concepts are not necessary to edit color, of course, but if wanting to master color technique...

___________

Guys, try 3D LUT Creator demo version... and you won't complain anymore about color.

 

[Alan Edward Klein]

I use the 850 with vuescan, and only use it to scan 4x5 and up. If you need to scan large format, your options are pretty limited without doing any stitching.

The V850 Pro is advertised to come with Silverfast, not Vuescan. Did you have to buy Vuescan separately? Do you prefer Vuescan over Silverfast?

 

[Lachlan Young]

For an advanced edition, today we have very powerful tools to edit color in the way we want without limitations.

In particular Photoshop layers/masks combined with a plugin like 3D LUT Creator offers total freedom to get what we want, single limit is our aesthetic criterion and our creativity .

Going back to theory, the real color capture a film makes is the silver/halide content in the 3 layers after 1st developer. Those silver/halide contents are abstract color. Dyes resulting after color Developer are only an interpretation of the real color capture, but with a tool like 3D LUT Creator we deformate the color space like we want. Additionally with Ps layers/masks we can apply a different color space deformation (or interpretation) to each separate subject in the image, this allows some limited edition of the film spectral signature.

View attachment 242233

This is better understood when one has practiced Tri-Color photography, we depart from 3 BW images, color filters for each image are critical, later we assign colors and channel cross-talk and color space deformations to the abstract 3 BW color capture.

Every photographer wanting to master color technique should practice some tri-color to start learning deep concepts. It can be done with a SLR/DSLR with static scenes, to not waste film.

At all, those deep concepts are not necessary to edit color, of course, but if wanting to master color technique...

___________

Guys, try 3D LUT Creator demo version... and you won't complain anymore about color.

I don't think you realise how much you're revealing about how little you actually know or have useful experience of. Most of the popularity of these 'creative' LUTs you are so keen on exist for two reasons: to try and season bland digital imagery; and to cover up faults, errors and shortcomings in a vaguely 'creative' way with a semi-detached resemblance to foggy memories of assumed film behaviour in a past that never existed. They are every bit as much a bad cliche as the Velvia-ised landscapery you seem so desperate to defend. Very often the most striking film grading in cinema and photography today is the result of actually letting the materials be themselves either via actual optical printing or inversion approaches that aim towards the same end - and even then there is plenty of freedom to alter colour, but in a way that (for whatever reason) seems more convincing to the viewer's psychological perception. 3D LUT creator is essentially a simplified GUI for things that can already be done by other means.

 

[138S]

3D LUT creator is essentially a simplified GUI for things that can already be done by other means.

I guess that you are not aware about what 3D LUT edition can do. Photoshop can use 3D LUTs but it cannot edit 3D LUTs.

A 3D LUT maps any source color to any destination arbitrary color with no restriction. It's the tool doing that. Do you know other means doing that ? This is Maping each color from the source space to totally arbitrary colors in the destination space !!

Anyway you need an edition tool generating that map, if not wanting to rely in canned 3D LUTs, and you need a really advanced GUI to make the 3D LUT that suits your desire, this is 3D LUT Creator. Other tools may edit 3D LUTs, but other means won't be able to make what all you may make with 3D LUT, by a far extent.

 

[Adrian Bacon]

The V850 Pro is advertised to come with Silverfast, not Vuescan. Did you have to buy Vuescan separately? Do you prefer Vuescan over Silverfast?

I bought Vuescan separately. The V850 isn’t my only scanner and Vuescan basically works with almost everything and has a lifetime license. My workflow is not typical. I use Vuescan to get a raw scan of the negative and then feed it to my software to create a color positive image in Adobe’s DNG format. I’ve never used the bundled software that comes with any of my scanners, so I can’t speak to how well they do or don’t work, though if you’re scanning chromes, they should work just fine. It’s scanning color negatives is where Things tend to get bumpy for most people.

 

[Adrian Bacon]

I guess that you are not aware about what 3D LUT edition can do. Photoshop can use 3D LUTs but it cannot edit 3D LUTs.

A 3D LUT maps any source color to any destination arbitrary color with no restriction. It's the tool doing that. Do you know other means doing that ? This is Maping each color from the source space to totally arbitrary colors in the destination space !!

Anyway you need an edition tool generating that map, if not wanting to rely in canned 3D LUTs, and you need a really advanced GUI to make the 3D LUT that suits your desire, this is 3D LUT Creator. Other tools may edit 3D LUTs, but other means won't be able to make what all you may make with 3D LUT, by a far extent.

I’ve written software that generates 3D LUTs. No need to buy/use some lut creator that doesn’t do what I want. 3D LUTs are actually pretty limiting as they can only contain 256 points per dimension. The rest of the points are interpolated. They also only work in RGB space, which is fine if you use them for how they’re intended (mapping RGB source to RGB destination), but if you want to do something in another color model, like XYZ or a cylindrical color model, good luck.

 

[138S]

3D LUTs are actually pretty limiting as they can only contain 256 points per dimension. The rest of the points are interpolated ... They also only work in RGB space

> A 3D LUT, in general, can map any color model, and of course it can map form a color model to another one, for example it can map from HSL to RGB for technical applications, or to speed up conversions.

> See in the posted video min 2:50, 3D LUT Creator edits in any color space, not only RGB: https://www.photrio.com/forum/threa...lm-out-of-control.173528/page-13#post-2261842

> Even in the case the LUT has only 32 points per dimension color grading is totally accurate even if applied to a 16 bit/channel image, because the LUT can be used to determine the color space deformation and spline interpolation works perfectly.

As an example when you multiply all values in a 16bits/channel image by 0.8f to decrease global brightness you loss nothing in the quality, and you used a single value. Or when you adjust the tonal curve in a BW image... you may use only two points in the curve but tonality adjustment is perfectly smooth from the spline interpolation, having total control with not many points. Editing more than 32 points per dimension is an overkill that complicates edition, usually you do it perfectly with way less points, the 3D LUT Creator Pro version allows to edit 96 points per dimension which is a total overkill. It is a surprise you missed that...

 

[Adrian Bacon]

> A 3D LUT, in general, can map any color model, and of course it can map form a color model to another one, for example it can map from HSL to RGB for technical applications, or to speed up conversions.

> See in the posted video min 2:50, 3D LUT Creator edits in any color space, not only RGB: https://www.photrio.com/forum/threa...lm-out-of-control.173528/page-13#post-2261842

> Even in the case the LUT has only 32 points per dimension color grading is totally accurate even if applied to a 16 bit/channel image, because the LUT can be used to determine the color space deformation and spline interpolation works perfectly.

As an example when you multiply all values in a 16bits/channel image by 0.8f to decrease global brightness you loss nothing in the quality, and you used a single value. Or when you adjust the tonal curve in a BW image... you may use only two points in the curve but tonality adjustment is perfectly smooth from the spline interpolation, having total control with not many points. Editing more than 32 points per dimension is an overkill that complicates edition, usually you do it perfectly with way less points, the 3D LUT Creator Pro version allows to edit 96 points per dimension which is a total overkill. It is a surprise you missed that...

*sigh*

yes, you can make a “lut” for pretty much anything, assuming the software that uses it knows what it’s for. I was referring to Adobe’s lut cube format.

mapping HSL to RGB... seriously? You truly don’t know what you’re talking about. I’d never use a lut for that, especially given the fact that RGB to HSL/HSV and back is already a well documented algorithm. You claim it’s possible. It’d be interesting to see how you’re going to map 65536 values for each RGB channel to 360 hue degrees, saturation values, and luminance values in a lut that Photoshop knows what to do with without losing any color fidelity. Go start a thread over in the hybrid forum demonstrating that and we can talk.

 

[Alan Edward Klein]

I bought Vuescan separately. The V850 isn’t my only scanner and Vuescan basically works with almost everything and has a lifetime license. My workflow is not typical. I use Vuescan to get a raw scan of the negative and then feed it to my software to create a color positive image in Adobe’s DNG format. I’ve never used the bundled software that comes with any of my scanners, so I can’t speak to how well they do or don’t work, though if you’re scanning chromes, they should work just fine. It’s scanning color negatives is where Things tend to get bumpy for most people.

I found it hard to scan negative color with Epson scan too. One of the reasons I stick with chromes.

 

[miha]

Guys,

I'm impressed by your scanning / mapping / etc show off, but please respect the fact that this is an analogue photography forum only. Thank you.

 

[138S]

mapping HSL to RGB... seriously? You truly don’t know what you’re talking about.

Of course !!!! I'll show you a technical application for it.

You have realtime 200FPS video and Machine Vision software has to analyze each image, your blob detection (segmentation) needs to find what connected pixels have certain hue, saturation or lightness so you need to convert RGB to HSL realtime while consuming the least computing possible. Then instead calculating HSL values from RGB levels for each pixel you make 3D LUT. From each RGB entry you get a pre-calculated HSL value,

It can be just a 3D table of precalculated values, but that RGB to HSL conversion 3D LUT may contain image adjustments, like gamma, brightness, saturation modifications.... so in a single operation you get image conditioning and conversion to HSL to feed the segmentation engine.

This is one of the real cases I've personally implemented...

LUTs in general have been always used for realtime color space conversions, like ancient NTSC-PAL-SECAM conversions.

Guys,
I'm impressed by your scanning / mapping / etc show off, but please respect the fact that this is an analogue photography forum only. Thank you.

OK, you are right, sorry for the off-topic, I end it here

Let's return to sheet film price...

 

[Sal Santamaura]

...Let's return to sheet film price...

Let's not.

 

[Sirius Glass]

For an advanced edition, today we have very powerful tools to edit color in the way we want without limitations.

In particular Photoshop layers/masks combined with a plugin like 3D LUT Creator offers total freedom to get what we want, single limit is our aesthetic criterion and our creativity .

Going back to theory, the real color capture a film makes is the silver/halide content in the 3 layers after 1st developer. Those silver/halide contents are abstract color. Dyes resulting after color Developer are only an interpretation of the real color capture, but with a tool like 3D LUT Creator we deformate the color space like we want. Additionally with Ps layers/masks we can apply a different color space deformation (or interpretation) to each separate subject in the image, this allows some limited edition of the film spectral signature.

View attachment 242233

This is better understood when one has practiced Tri-Color photography, we depart from 3 BW images, color filters for each image are critical, later we assign colors and channel cross-talk and color space deformations to the abstract 3 BW color capture.

Every photographer wanting to master color technique should practice some tri-color to start learning deep concepts. It can be done with a SLR/DSLR with static scenes, to not waste film.

At all, those deep concepts are not necessary to edit color, of course, but if wanting to master color technique...

___________

Guys, try 3D LUT Creator demo version... and you won't complain anymore about color.

Among other things you have no comprehensive how nonlinear color film is and how interlinked the color crossover is. A wise person would have backed out of a discussion which is over their head; you are not and will not. Sadly for the Photrio.

 

[DREW WILEY]

Color mapping was invented long before computers, clear back in the 1920's. It was just very slow using graph paper and analytic geometry. But there were real advantages to setting a continuous spectrum plot per unbroken axis rather than just interpolated points, provided you weren't in a hurry. I worked for awhile with a huge continuous spectrum IBM spectrophotometer. It was engineered very differently than those which break up the spectrum into discrete points using xenon flashtubes etc; but that kind of improvement is necessary for modern simplified programming. Mapping opaque colorants like inkjet inks is based upon a 4-axis model, with tone (whiteness) and tint (blackness) reading opposite directions from the axis centerpoint. The industrial readers for these kinds of applications involve about 13,000 points! Yet they still don't eliminate problems of metamerism, among others issues. That's because no colorants or dyes are perfect. At the end of the day, it's the human eye that adjudicates the result.