New venture - color film - Ektar & Portra

[Donald Qualls]

The car still has distinctly green shadows and very slightly magenta highlights, even to my rather accepting eye. The park has blue/cyan shadows, but that's to be expected in late day sunlight. I still see a hint of magenta in the hightlights on the tree trunks, so I suspect it's the same deal as with the car -- a magenta/green crossover shifting the blue shadows into cyan. Not so bad I'd be bothered by it if I saw it in a magazine or on a wall -- and it's just possible it's a difference in monitor calibration rather than a real crossover.

 

[pentaxuser]

Thanks Bormental Your exposures times does not suggest to me a gross over exposure time but I have no knowledge of the latitude of Ektar 100 to overexposure. 1/125 @ f8 suggests at a little less than 2 stops over and 1/500th no overexposure at all, if I have understood the "sunny 16" rule. They certainly look better to my eye but that is largely incidental as it is your taste that counts. What is important for me is whether you negatives are OK in the sense of being capable of being printed under an enlarger to look like your second set of scans or have you had to do something with the scanning tools that has enabled you to correct the crossover that has been suggested is there?

pentaxuser

 

[Bormental]

@pentaxuser I have no way of knowing, since I have zero experience printing in color! Just recently I've added color film to my bag and, to be honest, I am still having difficulties evaluating negatives on the light table!

@Donald Qualls Agreed. By the way, the car's paint had a faint green tint to it in real life. I wish I had taken a digital photo as well.

 

[pentaxuser]

One more question Bormental, if I may: Do you do your own processing or are these negs commercially processed? It looks like we may have already worked out that gross overexposure is not the problem in terms of crossover unless Ektar has almost no latitude at all in this respect but as processing is usually cited as the cause of colour crossover I feel it is worth finding out about your processing

Thanks

pentaxuser

 

[pentaxuser]

@pentaxuser I have no way of knowing, since I have zero experience printing in color! Just recently I've added color film to my bag and, to be honest, I am still having difficulties evaluating negatives on the light table!

Thanks I appreciate you have no way of knowing about printing under an enlarger and RA4 processing but if you have done something in the scan to correct crossover this might be important.It doesn't sound as if you have done any such thing but Donald sees green in shadows from the car as I do on close examination but it may be that the film picks up the green from the body of the car in the car's shadow which the human brain filters out.

So why is crossover important. Well it is the one thing that cannot be corrected in darkroom RA4 printing so it is important for those of us who still do darkroom colour printing. My whole enquiry was prompted by a comment by Lachlan that seemed to suggest that crossover had occurred so I am trying to get to the bottom of the possible causes.

pentaxuser

 

[Lachlan Young]

Lachlan what should I be looking for in the above examples that demonstrates colour crossover.

The yellow cast that results from attempts to correct the colour shifts produced by the crossover, but because it's crossover, not an overall cast, this attempted correction throws the colour balance of other areas off too.

If Ektar is correctly exposed, it's near dead neutral but saturated.

 

[Bormental]

One more question Bormental, if I may: Do you do your own processing or are these negs commercially processed?

These are my own scans using a digital camera. I usually order lab scans too, but I am rarely happy with their results. For some reason the contrast is always too high (and not recoverable) despite me asking not to do any editing to their 16bit TIFFs. Here're the lab versions. I think the color balance is truer, but too much detail gets killed in highlights and shadows:

 

[pentaxuser]

The yellow cast that results from attempts to correct the colour shifts produced by the crossover, but because it's crossover, not an overall cast, this attempted correction throws the colour balance of other areas off too.

If Ektar is correctly exposed, it's near dead neutral but saturated.

Lachlan, what do I need to look at to see the yellow cast and what has caused this, given it looks safe to assume that the Ektar negs were not grossly over-exposed and the development process was carried out as per C41 standard specs

Is a yellow cast the usual sign of crossover or just the sign of it in this scan? What has ensured that the middle picture of the car and house does not have a crossover or is this included in the yellow cast sign?

Thanks

pentaxuser

 

[DREW WILEY]

Lachlan has correctly expressed the dilemma. People try to correct for crossover by adjusting overall color balance afterwards, which just throws something else off. Crossover in Ektar generally goes cyan, so the steering wheel gets yanked the opposite direction (warm) afterwards, and potentially pulls the whole car off the road. Makes no difference whether we're talking about RA4 printing in a darkroom or digital reproduction. Either way, it's far easier to obtain "neutrality" using corrective filtration at the time of the shot if necessary, along with careful metering.

 

[pentaxuser]

Lachlan has correctly expressed the dilemma. People try to correct for crossover by adjusting overall color balance afterwards, which just throws something else off. Crossover in Ektar generally goes cyan, so the steering wheel gets yanked the opposite direction (warm) afterwards, and potentially pulls the whole car off the road. Makes no difference whether we're talking about RA4 printing in a darkroom or digital reproduction. Either way, it's far easier to obtain "neutrality" using corrective filtration at the time of the shot if necessary, along with careful metering.

The question is what caused the crossover, given the OP's camera exposure in terms of shutter speed, aperture and light conditions and whatever caused it in the top scene seems to be absent in the middle scene.

pentaxuser

 

[DREW WILEY]

I explained what causes crossover in the film itself, and how trying to post-correct CAN cause the kind of effect in image he's explaining about. But given all the intermediate variables of scanning and posting on a screen, each of these workflow steps would have to also be analyzed. The web is a helluva crude instrument for sorting such things out unless everyone's respective instrumentations and screens have been cross-calibrated. And I wasn't there with my own light meter and experience when he took those shots to begin with, so one more variable it's hard pin down. It is likely he hasn't learned proper color temp filtration yet, so that is the most probably thing to comment about, since it's the only way to really control crossover to begin with. ...Note that in the tree and lawn scene, at least on my monitor, the shaded part of the image in mostly in the trees above, yet also in the tree trunks below; and all these deep shadows have a characteristic blue-cyan bias, while the brightly sunlit lawn has gone quite warm, exactly as I would expect with Ektar. Shadows under a blue sky are indeed blue, just as the Impressionists learned. Ektar does not artificially warm these low values like Portra and other CN portrait films do, And it does risk a degree of cyan crossover into the blue at the extremes. It's inherent to this particular film. There are ways to tame it somewhat, which I have already suggested on numerous threads. But once the crossover has taken place, rather than being corrected at the time of the shot, it's going to be hard to untangle the overlapping dye curves. Argue all you want, or use that nonsensical "I can do anything in Photoshop" talk. Yeah, someone can probably clean up a train wreck that way too; but it would have been a lot easier just to stay on the rails to begin with.

 

[JWMster]

Just curious (since I've not used Ektar): Does correcting color temp in post do anything to correct the problem?

 

[Bormental]

Is there a problem? I am thrilled with the results. True-life colors are boring, and I'm staring at them all day anyway

 

[Lachlan Young]

@Bormental If you start from where the film's 'neutral' point (which may not be an actual neutral - see Portra 400) is supposed to be, it makes adjusting the colour to make the image say what you want very much easier and faster.

 

[MattKing]

Just curious (since I've not used Ektar): Does correcting color temp in post do anything to correct the problem?

Not if you have problems with crossover.
Crossover corrections would require different colour temperature corrections at different parts of the image.

 

[pentaxuser]

Not if you have problems with crossover.
Crossover corrections would require different colour temperature corrections at different parts of the image.

So if we go back to the scene in question, namely the green parked sportscar, where in this print do I need to look for the signs of colour crossover that are apparent there? Equally importantly what might be the cause and what might the OP have done in the middle shot( the car on a drive with a house) to have achieved a pic with no crossover unless there is also crossover here as well. In which case what do I need to look for here that says "crossover "

Clearly the OP has a different outlook on what he wants from a colour negative from what I'd want but this is largely irrelevant to the discussion on crossover but he appears to have answered honestly in terms of his exposure which doesn't seem to be much over-exposed and we know that his negs were commercially processed at what I assume to be standard C41 processing.

pentaxuser

 

[DREW WILEY]

It's tricky to balance Ektar when part of the scene is in deep shade, and another part in open sun, but involves truncating the cyan out of the blue at the time of the shot using a KR series warming filter. Which one is best all depends. I posted about how to do this numerous times. It's also possible to flash the film using a warming diffuser so that only the low values get differentially warmed, but that requires more experience. For Ektar, I always carry a mild pinkish 2B skylight filter, a KR1.5 for stronger correction, and often something even stronger still. Don't expect to post-correct it well. Matt just succinctly explained why.

 

[MattKing]

So if we go back to the scene in question, namely the green parked sportscar,

I think we may have sourced one of the problems.
I believe that car isn't green - it is grey.
If you are seeing a green car, it is unlikely that is either a result of a problem with your monitor calibration, or the effects of crossover.

Equally importantly what might be the cause and what might the OP have done in the middle shot( the car on a drive with a house) to have achieved a pic with no crossover

I don't see any clear indications of crossover in the uploaded version of this image - exposure seems to be good, and the scan appears good as well - at least as far as we can see here.
The most obvious signs of crossover are in the highlighted grass in the scene with the field and the arch of trees. The crossover makes that grass look like frost because of the cyan hue, rather than dried brown.

 

[JWMster]

"I think we may have sourced one of the problems. I believe that car isn't green - it is grey. "

Not sure how anyone saw the car as other than gray, but some monitors... ugh.

Okay, so I've got my B&W filters down. Now it looks like I have to collect color filters... $'s (ca-ching). More than that, it sounds like it would be more on target to focus on grad color filters. Did I say ca-ching? Yes, I think so. But that's from the outside looking in.
Surely you guys aren't carrying oodles of filters for color. So if you don't mind my asking, "What are you guys really carrying for daylight landscapes to deal with the possible crossover issue?"

 

[pentaxuser]

@pentaxuser I have no way of knowing, since I have zero experience printing in color! Just recently I've added color film to my bag and, to be honest, I am still having difficulties evaluating negatives on the light table!

@Donald Qualls Agreed. By the way, the car's paint had a faint green tint to it in real life. I wish I had taken a digital photo as well.

Donald Qualls mentioned a green look as well and unfortunately when I mention green I needed to add the word faint. I never saw this as other than just a hint of green. I agree that the grass under the trees in the middle shot looks strange but this is largely the result of what Bormental wants in his picture and this was corrected in the lab scans.

As I said none of this matters to Bormental and that's fine. Working out what has caused alleged colour crossover in the sportscar shot and trees in the park shot does concern me on several counts. One: I wanted to know what colour crossover and its causes are all about and equally importantly what they are NOT about

Two: Can overexposure of Ektar cause crossover as appeared to be alleged, and if so, how susceptible is Ektar to this? How many stops over does this start - half a stop, a whole stop etc

Three: In the parked sportscar shot on a very bright sunny day and should we ever get such bright sunshine in the U.K., do I need to apply a lens colour correction filter, as Drew suggests, to be sure of a negative that has no colour crossover in it? This has prompted JWMster to ask the same question

Bormental's valuable contribution in all of this was to give details of his exposure and confirm that the film was processed at standard C41 requirements. So in terms of processing deviations as a cause it would seem that we can eliminate that and unless Bormental was seriously mistaken about his exposure or is deliberately lying it would appear that anything like gross over-exposure can be eliminated

I think I have learned about as much as I am going to on colour crossover in this thread and it will be down to me to decide what the balance of probabilities are in terms of there being colour crossover.

I wish JWMster every success in getting good answers to his question in the last sentence of #69. I have already decided whether I need to carry oodles of colour correction filters to correct such a scene that Bormental shot in those sunny conditions

pentaxuser

 

[Lachlan Young]

"What are you guys really carrying for daylight landscapes to deal with the possible crossover issue?"

An incident meter (or a spotmeter that allows indexing to the IRE scale) - and bracket a roll or two until you arrive at an EI that works for you - you want to avoid overexposure, that's all. And remember that from the inherent properties of the film, you are going to get about an extra two stops of straight line down into the shadows compared to a transparency film. If you really need/ want to go beyond 7 stops of range (and remember that the film is supposed to give a look more like a transparency), you might want an ND grad or two. Optical wet darkroom printing adds a few more minor trip hazards, but nothing too major.

 

[GLS]

If you really need/ want to go beyond 7 stops of range (and remember that the film is supposed to give a look more like a transparency), you might want an ND grad or two

A polariser can also help, as the overexposure cast rears its head mostly with bright blue skies in my experience.

 

[MattKing]

So if you don't mind my asking, "What are you guys really carrying for daylight landscapes to deal with the possible crossover issue?"

A couple of different warming filters - the sort we used to use when we were using and projecting slide film.
The reason that Ektar shows crossover when over-exposed is that it has slightly more sensitivity in the cyan range than other films, so highlights plus over-exposure plus lots of blue-cyan light being available (like from a summer sky) ends up in a result where the cyan sensitive parts of the film have a different shape and slope than the other parts. When the slopes don't match, you get crossover. Put a warming filter in front and/or expose the film accurately and the slope of the cyan sensitive parts of the emulsion will match more closely the other parts of the emulsion - no crossover.
By the way, the lab scans that Bormental posted in this thread appear to have better colour and less crossover than his own scans. His scans have better contrast and detail. That tells me that he needs to adjust the exposure of his scans, plus how they are post-processed, in order to retain the good contrast and detail, but more accurately record the different colour components in the negative.
I expect his inversions are applying a different gamma to one or more of the colour components than to the others - thus the crossover.

 

[Bormental]

By the way, the lab scans that Bormental posted in this thread appear to have better colour and less crossover than his own scans. His scans have better contrast and detail. That tells me that he needs to adjust the exposure of his scans, plus how they are post-processed, in order to retain the good contrast and detail, but more accurately record the different colour components in the negative.

Agreed. I am alternating between lab scans, Epson V600 and DLSR "scanning". Sadly, each of the three methods beats the other two for some shots but loses with others. For example, the lab scans often introduce unpleasant green tint in the shadows (see the example from the same roll below). The world could benefit from better tooling for color inversion.

 

[Lachlan Young]

The world could benefit from better tooling for color inversion.

It can be done, but you have a choice of easy or good.