Is it possible for a particular color to fall outside of the Portra 400 gamut?

[faberryman]

Here's another anecdote: on another forum one member stated that he had set up a prism to spread sunlight into a spectrum. Then he photographed it with a handful of his digital cameras. He says that none of them could record the spectral "yellow" which he himself could clearly see.

Of course I was skeptical and sorta tried it myself, but using a CD vs a lamp filament. This is pretty easy to do, and I think some of the people in this thread will be a little surprised if they tried it.

And yet you see yellow in digital images. It must be a miracle.

 

[koraks]

And yet you see yellow in digital images. It must be a miracle.

Yes, yellow apparently is a really funny color indeed. I guess it has to do with the 'red' cones really being yellow-orange cones, but that we apparently use the trailing end of their sensitivity to see red. I figure this also has something to do with the fact that there are three ways we can see yellow: either as yellow light (e.g. low pressure sodium), as a combination of red + green, and as white light from which violet is removed. What I take from this is that innately we're probably not so good at seeing yellow, so we've evolved a couple of cheats to fake the sensation.

There's like much more to it, but yes - yellow is a miracle in a way.

 

[DREW WILEY]

As a slice of my own specific experience, I can recall how a specific species of deep blue-violet wildflower in similar lighting would turn out wonderfully and accurately (to my eye) color-saturated on chrome Provia film and past Portra 160VC color neg films, but appear disappointing using current Porta films or even Ektar. The wonderful violet reproduction I mention earlier in reference to Ektar was a fabric color within the overall scene. And even things like deep yellow and golden flowers, as we perceive them, might come out quite differently on different color films.

It's quite a complex topic. Among the most difficult hues to reproduce are those of fluorescent algae and lichen. I once shot an old grainy pre-E6 Agfachrome 50 which rendered that wonderfully, which no other color film hence could do.

Yellows? Since most color neg films have skintones in mind, they have a tendency to dump any warm color into the same category, and do a poor job resolving yellows from yellow-oranges or warm tans and earthtones. Everything deliberately ends up a muddy fleshtone. Most of the entire genre of 70's color photographers like Stephen Shore and Misrach relied on that very color error, especially in its exaggerated form in old Vericolor L. Chrome films do not do that, nor Ektar quite so much.

As far as a print medium, Cibachrome could render clean yellows and oranges wonderfully. It has nothing to do with an inability to see yellow precisely. Where does that notion come from? Apparently neither Van Gogh nor your house painter ever heard of it.

 

[pentaxuser]

Sure:



Any further summary would involve a (re)treatment of the subject matter @Mr Bill and I have discussed in some depth.

On topics where nuance is the essence, summarizing doesn't always help much.

If you feel a better summary is necessary, I'd suggest writing one.

So in summary: 1. He is stuck with what he has got using Portra and with what appears on my VDU at least, to be an appalling rendition of the hanging flowers compared to his digital camera? I assume in that of course that the digital rendition is a correct rendition of the hanging flowers

2. There is nothing he can do in scanning terms to improve on this

Laser said: "The result is that current Porta Films work well but it is not possible to get the absolute best results when using dissimilar components. That said Portra Films are likely to yield the best color reproduction of difficult colors. By design, Portra is much better than films that preceded it. "

What's a better colour film to use for Steven's difficult colour, Laser and is Steven's colour the only one that Portra struggles with to the extent that it does?

Thanks

If I have read Steven's post correctly it is solely about what his scans look like. If so, then just out of interest would an RA4 print be no better than his scan? It doesn't sound as if it would be any better but I may as well ask

One of our members called Fred used to frequently show us his Portra prints and I can't remember seeing any colours so badly represented as this colour that is as Steven said somewhere between pink and purple. Maybe Fred's shots never contained this colour?

It would certainly seem one to be avoided. So bad in fact that if I were a wedding or portrait photographer using Portra film I'd want to inquire before the shoot if any of the subjects were wearing this colour rather than risk my shots being compared to the many digital iPhones that get used at such gatherings


pentaxuser

 

[DREW WILEY]

When timed commercial shoots were involved, which can't be redone, like weddings, the only smart thing was to survey the lighting circumstances and colors involved, so that you'd be properly equipped in advance, including the right film. Portrait studios tended to work production-line style, and concentrate on certain films and controlled lighting scenarios compatible with that. But if clients were willing to pay a premium, certain studios also offered the option of shooting the portrait in chrome film and having it dye transfer printed.

Today, both color neg films and chromegenic printing papers have come a long ways. Nothing is perfect or ever will be.
But significant progress is also a fact.

I've commented on the shortcomings of Portra. But my gosh, there are a lot of other suspects milling about which you might investigate before lynching the film itself in this case - lighting idiosyncrasies, scanning errors, post-processing nonsense, whatever. ...

 

[4season]

The film photos look about right for photos taken in open shade: Bluer, with more muted colors. It's probably pretty close to the truth.

The better phone cameras (actually a whole lot of software wizardry involved) do an amazing job at capturing the world as we perceive it, but our brains are powerful signal-processing engines, and what we "see" is not always the literal truth! Your phone is doing a pretty good job of replicating what you think you are seeing, but the film camera can only reveal what's actually there.

 

[faberryman]

Your phone is doing a pretty good job of replicating what you think you are seeing, but the film camera can only reveal what's actually there.

Films do not reveal what is actually there. Every film sees the world differently.

 

[BMbikerider]

But is/was the problem in the printing or the recording stage? If the latter, how was this established?
See my rather lengthy post on the issue. It's for instance very likely that the saturation cost of representing an intense violet using a cyan and magenta dye (e.g. in RA4 paper) will result in a disappointingly unsaturated violet color.

I first came across it in the early 1990's when I was temporarily attached attached to a police scenes of crime unit in UK where an offence had taken place in an area of woodland in the South of England and photographs had been taken using a Pentax 67 using Kodak film and printed on Kodak paper. I don't know what the film was, but the paper was the same Kodak that I used in my darkroom at the time, although that came off a roll and I used cut sheets..

We saw when the prints were produced that the blues of the Bluebell flowers were wrong, they had reproduced as almost a shade of purple varying in density depending if they were in sunlight or shadows. The printing (machine) was recalibrated and all this did was correct the blues (almost), but threw the other colours out of what they should have been and looked quite false.

In the event when the case went to court using the original prints, the defence put a case that the film images were wrong and had been tampered with. The case was adjourned until one of the senior technicians from the Kodak UK HQ in Hemel Hempstead, England was asked to look at the pictures and have the same negatives printed under controlled conditions by Kodak and give his assessment which was along the lines of what I have been saying in my previous posts and I am only repeating what was said there under oath in court.

The defence lost their argument and the offender found guilty. In the words of a certain court prosecutor "I rest my case"

 

[koraks]

What's a better colour film to use for Steven's difficult colour, Laser and is Steven's colour the only one that Portra struggles with to the extent that it does?

I don't know; as soon as our wisteria decides to man up and actually start flowering I might give it a go. With a little bit of luck that would be late spring 2024. If the darn vine knows what's good for it, that is.

 

[DREW WILEY]

Depends what you meter for. Portra 400 is more contrasty than Portra 160, but is still fairly low contrast. Let's take korak's example of wisteria. My own wisteria blossoms are fairly transparent, and a relatively light shade of purple. But if I spot metered those, and placed the exposure about half an EV, or perhaps no more than one EV above midtone, I'm sure I'd get a decent rendition of that color using most pro color films, either CN or chrome.

If on the other hand, if metering something else in the scene caused the wisteria blossoms to be overexposed, then of course there's more risk of not only the hue saturation being washed out somewhat, but of crossover with an unwanted hue tinge. But I'm speaking as a darkroom color printmaker having every step of the process under my own control. Gosh knows how an amateur level scanner might interpolate things.

 

[albireo]

This can also be the limitation of my scanning. At the moment I sold all of my scanners and using the Sony A7R. I have access to 3 automatic color inversion tools, but I prefer to invert manually. No matter how I go about the inversion, I can't get that color to be even close (without destroying every other color in a photo). In case you're wondering, I have zero complaints about the colors I normally get, as long as it's not this one

A few questions

• could you provide more details on your manual inversion method?
• related to the above - as far as you're aware, do you suffer from any flavour of colour blindness and if so which?
• could you show a side by side comparison of the colour of interest including if possible a control colour in the same frame (one you're happy about), as rendered by all of the 3 automatic inversions methods you have access to?

 

[Xylo]

A very long time ago I read about a strange effect that makes some plants impossible to photograph properly. If I could just remember the name...
But in a nutshell, colors on film or a sensor doesn't work like it does for the eye and these colors fall to that in between spot. I know that it almost always happens with purples. There's a type of clematis and also the British bluebells that produce this effect.
The human eye is mostly sensitive to green tones (the bane of graphic designers), much less to the blues; so that says a lot about color rendition. 'Cause lets face it, when we take a picture we're always looking at the 3rd generation of the original.

 

[pentaxuser]

I don't know; as soon as our wisteria decides to man up and actually start flowering I might give it a go. With a little bit of luck that would be late spring 2024. If the darn vine knows what's good for it, that is.

Yes me neither which was why I had framed this question to Laser who had said as follows: "The result is that current Porta Films work well but it is not possible to get the absolute best results when using dissimilar components. That said Portra Films are likely to yield the best color reproduction of difficult colors. By design, Portra is much better than films that preceded it. "

I am simply interested to ask him what other colour films, if any, are better than Portra in respect of the problem colour that Steven identifies and what other colours may give those films problems I must admit that from what Laser has said the answer in terms of what better film there is overall would seem to be none

However Steven's colour seems to be particularly badly reproduced hence my comment on portrait and weddings v those doing it digitally

pentaxuser

 

[MattKing]

laser's reference to "dissimilar components" is a reference to the fact that the optically printed, matched papers of the time are no longer employed or even available.
With the current disparate approaches to end product, it probably isn't possible to identify one best film for all purposes.

 

[DREW WILEY]

We're really talking about color neg films which cannot be readily examined right atop a light box after processing like chrome film can be, but which must jump through intervening hoops to analyze - whether a proof print, or a scanned viewing option.

And the notion that Portra will yield the "best reproduction of difficult colors" is vacuous - which Portra, which difficult colors, what do you mean by "best", and what on earth do you possess as real evidence behind that statement? Yes, the Portra lineup has solved some of the color repro problems characteristic of earlier color neg films, but has inherited others, deliberately. It's just another stage in the evolution of the same things. And only Ektar realistically fills in the gap between that and what chrome films do way better. There simply is no silver bullet.

If you need to reproduce very specific hues, test, test, test film, and then after going insane trying, and sitting on the porch of the insane asylum in your rocking chair, with a set of watercolor pigments for therapy, you'll finally realize you should have been doing it that way all along. In my case, I go insane either way, so it's easier to blame the camera and film than my own direct mess. But I'm obsessed with capturing and rendering certain complex hues in nature.

 

[Steven Lee]

Thank you everyone! There are so many thoughtful responses here, apologies for not mentioning all of you by name. Some of the raised questions will steer the conversation directly into the digital territory and the analog/digital separation rules here are fairly strict. Let's just say I will re-scan these negatives on the Flextight X1 next week. Happy to share my findings in a separate thread in the scanning forum (CC @albireo).

I will say though that @BMbikerider comments blow my mind. Can't wait for some free time to do more experimentation.

 

[reddesert]

I would also disagree with their assertion "it was just not possible..." but... perhaps the digital cameras they have tried were not capable in specific situations.

Here's another anecdote: on another forum one member stated that he had set up a prism to spread sunlight into a spectrum. Then he photographed it with a handful of his digital cameras. He says that none of them could record the spectral "yellow" which he himself could clearly see.

Of course I was skeptical and sorta tried it myself, but using a CD vs a lamp filament. This is pretty easy to do, and I think some of the people in this thread will be a little surprised if they tried it.

And yet you see yellow in digital images. It must be a miracle.

I don't want to get too deep into the digital-images part since this is an analog area, but in fact this miracle is the crux of the problem. Both analog color film and digital sensors are designed with (usually) three color channels/layers that have fairly broad and usually slightly overlapping wavelength sensitivity. The sensitivity and the output colors of the channels are designed so that the output combination gives natural color renderings of most natural subjects. And natural subjects typically themselves have fairly broad color reflectance (or emission, for light emitting sources).

This means that rendering could get weird for light sources that have a quite restricted or nearly monochromatic spectrum, because the way that the film or digital color behaves isn't literally replicating what our eyes and brain do. In the case of the purple dresses or blue flowers, they may be out on the edge of both a color channel and the eye's wavelength range of sensitivity.

For the yellow light, dispersing the spectrum with a prism or diffraction grating (like a CD) creates a spread of fairly monochromatic light, where only a small range between red and green actually looks pure yellow to our eyes. For some cameras, that might do weird things with the RGB of a Bayer sensor. (For the record, I tried photographing the spectrum of a continuum light source reflecting off a CD with a digital camera and the yellow looked normal, but with a different camera, sensor, algorithms one might get different results.)

 

[BMbikerider]

All this chat about yellow in digital images is nothing new. Most people looking at a scene in a digital image the spring with new grass growing and the usual comment would be along the lines of 'look at that green field' Now if you were using a programme like photoshop or similar to change the green if it appears to be to intense in the picture you do not reduce the colour green but adjust the but the yellow to make it more mellow. Even a slight tweak of the yellow when photographing a forest of mixed conifers then it can change the whole outlook.

What is acceptable to one person may be totally wrong with someone else. I have seen many prints from a digital image of sunrises taken over the North Sea on the coast of Northumberland in England where the reds are completely and utterly way over the top for what we get in that area. Especially over the sea where the air in that region is particularly clear. You can tell what has gone wrong by examining the blacks and they will on close examination have a detectable hint of red. Other colours will not be affected because there is no red in the mix of dyes to make up the tone, but the black will in almost every case show a hint of red (sometimes more than a hint!)

 

[faberryman]

Mr. Bill stated that a variety of digital cameras were unable to to capture yellow from a spectrum of light. I remarked that that was curious because you see yellow in digital images and prints. It is just the usual disparagement of digital we see here every day. Meanwhile, the OP can't capture purple on his Portra 400 film.

 

[Mr Bill]

Mr. Bill stated that a variety of digital cameras were unable to to capture yellow from a spectrum of light.

No, that is not what I said. Perhaps you just didn't read carefully enough.

In the meantime I realized that this thread is in an analog only forum so I don't plan to explain further in this thread. But I would be happy to do so privately if you want to open a "conversation" or in a new thread.

Meanwhile, the OP can't capture purple on his Portra 400 film.

Fwiw I feel confident that the OP's Portra film actually DID "capture purple," and that the issue is actually in properly recovering that data. Digital cameras typically have much wider spectral sensitivity bands for each color than a scanner typically does, and my initial suspicion is that this is related to the color problem. At any rate the OP is apparently gonna rescan with a high-quality machine and this will hopefully answer the question.

 

[faberryman]

Mr. Bill stated that a variety of digital cameras were unable to to capture yellow from a spectrum of light. I remarked that that was curious because you see yellow in digital images and prints. It is just the usual disparagement of digital we see here every day. Meanwhile, the OP can't capture purple on his Portra 400 film.

No, that is not what I said. Perhaps you just didn't read carefully enough.

Here's another anecdote: on another forum one member stated that he had set up a prism to spread sunlight into a spectrum. Then he photographed it with a handful of his digital cameras. He says that none of them could record the spectral "yellow" which he himself could clearly see.

 

[pentaxuser]

Fwiw I feel confident that the OP's Portra film actually DID "capture purple," and that the issue is actually in properly recovering that data. Digital cameras typically have much wider spectral sensitivity bands for each color than a scanner typically does, and my initial suspicion is that this is related to the color problem. At any rate the OP is apparently gonna rescan with a high-quality machine and this will hopefully answer the question.

That's at least hopeful in that the extent of the poor purple rendition in Steven's scans and the one of the flowers in particular, may be (a) correctable to at least an extent. As we have only seen scans we have no way of knowing how well or badly the colour will be reproduced when exposed optically and printed on RA4

Given how difficult it seems to be judge colours from digital pictures of colour negatives there is probably no point in asking Steven to give us these

I still find it difficult to believe that the Wisteria as identified by koraks cannot be either scanned or "RA4ed" to render a better reproduction of it. It just looks so desaturated. Could it in fact have been this desaturated in reality and its our eyes/brains and his digital camera that have got it wrong to the extent that it appears to be wrong, surely not?

Maybe my memory is fooling me as well as my brain's translation of the colours I have seen on my RA4 prints but I just cannot recall any prints that had anything as apparently so badly reproduced as the Wisteria

Pity in some ways that we haven't seen any Fred's prints recently. He might have been able to shed some light on this pinkish-purple issue

pentaxuser

 

[jmrochester]

The issue that the OP raised is a bit simpler, I believe, than we are making it out to be. The principal reason for the inaccurate reproduction of blues and violets in particular is due to the spectral sensitivity of the film. Portra at initial introduction utilized a technology that was designed to improve color reproduction for difficult colors, e.g., fabrics and flowers, that traditionally suffered from metameric failure in order to benefit commercial fashion and product photographers. This technology was incorporated in the red-sensitive layers of Portra to more closely mimic human vision sensitivity than did most other films. Original Portra probably would have done a much better job of reproducing Steven Lee’s dress correctly. Apparently at some point during Portra product lifetime, this technology was abandoned in favor of more common red sensitivity whereby the newer version is more sensitive to longer wavelengths. This can be seen from published spectral sensitivity curves from different time periods.

As koraks pointed out, some bluish colors begin to reflect at longer wavelengths that our eyes can’t see, but the film does. Consequently, those colors will be captured by the film incorrectly and won’t be fixable.

 

[pentaxuser]

As koraks pointed out, some bluish colors begin to reflect at longer wavelengths that our eyes can’t see, but the film does. Consequently, those colors will be captured by the film incorrectly and won’t be fixable.

At the risk of appearing dim can I ask that having looked at both scans, the girl's dress and wisteria plant you see no hope of fixing either to any extent at all. The wisteria one looked so bad that I harboured a thought that there must be something that can be done

pentaxuser

 

[DREW WILEY]

Part of the problem with the discussion here is that, yes, you have representative spectrograms of the dye sensitivity of the film in question, but you DON'T have any representative spectrograms of the light reflecting off this or that flower or sample of fabric, since there can be a multitude of potential variations of those, which the human eye is capable of differentiating in a different manner than the given film itself, or even differently from pollinators like insects and their own kind of vision, which is really the priority with blossoms, and not our pictures.