Can anyone please help me identify the color problem in these these scanned TIFFs?

[Pushkal Arora]

Note:- Referring to attached G-drive Tiff link.

I recently sent some old developed negatives to a lab to be scanned and some of them turned out with green hue. I am able almost able to correct that hue even with my ameteur skills but in one roll there seems to be color glitches which seem digital. I believe those are digital because that color glitch/fade are covering around the edge of the block of walls in in the background. I have marked that image with Name "A".
In some other photos there is similar color glitch but I am not certain that's digital. In images D and E there is semicircle at the extreme left formed with group of small tiny circles. Both D and E are from totally different rolls and shot at time difference of almost an decade.

In image F, there are different colors over deep scratches which I believe is where scanner used Adobe ICE to correct them and I don't lnow if it is normal or not,
In image G, the faded colors inclding over the scratches are also following along the branch of tree on the right side which seems digital.
In most of the stripes/rolls that I gave for scanning there isn't problem like that except for a green tint on 1/3 of films which is correctable.

I have no good knowledge of photoshop and I want to make sure these negatives are scanned properly.
Can anyone help me verify the reasons of these errors? Is this just only because of incorrect color settings and is there some other problem?

These are scanned in 16 bit with Noritsu hs-1800 set to highest resolution.
G-drive Tiff link attached to the post.

 

[Les Sarile]

Only you will know what the original colors were at the time the picture was taken as color negatives are subject to anyone's interpretation. I work with a lab here and at the end of the day - depending on how much you are willing to pay, their goal is satisfy your expectations with their own interpretation of what you might think is pleasing.

I took a very cursory pass at your photo A. What do you think of my interpretation?

A by Les DMess, on Flickr

 

[Pushkal Arora]

Only you will know what the original colors were at the time the picture was taken as color negatives are subject to anyone's interpretation. I work with a lab here and at the end of the day - depending on how much you are willing to pay, their goal is satisfy your expectations with their own interpretation of what you might think is pleasing.

I took a very cursory pass at your photo A. What do you think of my interpretation?

A by Les DMess, on Flickr

Wow!! You almost nailed the true colors. I have attached a jpg of the print which was taken when the roll was developed. Sorry I should have share the prints too when uploading the tiffs.
The only error in your shared pictures is with the blues which is minor. Can you please share the settings which you used?

And do you think lab has done good work regarding extracting full information (Whatever that should be left now) from the negative?

I am asking this because in histogram of blue channel is quite narrow (lacking extreme bight and dark sides) compared to green and red.

 

[Les Sarile]

Wow!! You almost nailed the true colors. I have attached a jpg of the print which was taken when the roll was developed. Sorry I should have share the prints too when uploading the tiffs.
The only error in your shared pictures is with the blues which is minor. Can you please share the settings which you used?

And do you think lab has done good work regarding extracting full information (Whatever that should be left now) from the negative?

I am asking this because in histogram of blue channel is quite narrow (lacking extreme bight and dark sides) compared to green and red.

Unless I need layers, I find that ACDSee does all that I need.
I started out by adjusting the endpoints of each RGB color channel. Then I go into Advanced Color adjustment and turn down the yellow channel.
Like I said, very basic.

 

[Les Sarile]

I took a turn on F and all I can say is that is an extremely distressed film - scratches and color failure. I am sure there is no software based dust and scratch removal solution that would be satisfactory with it. I understand that wet mounting on a drum scanner is one possible way but I've never seen results from it with this type of sample. However, I've seen results from the Coolscan 9000 + Nikonscan ICE and as far as scratch removal is concerned, I believe this would greatly benefit from it.

F by Les DMess, on Flickr

BTW, since these are from a Noritsu, I think they did a good job restraining the use of excessive sharpening as well as over correction of auot levels and colors. Having a more "neutral" scan on such distressed film is a big advantage as it will offer the most adjustments in post. Dust and scratch removal however needs to use hardware based ICE or wet mounting. Coolscan 9000 + Nikonscan ICE being the best at this.

 

[Les Sarile]

For reference, this is how the different levels of Coolscan 9000 + Nikonscan ICE performs on dusty scratched negatives with no addional post work.

Kodak 160VC-06-36 K20D vs 9K ICE by Les DMess, on Flickr

 

[Pushkal Arora]

I took a turn on F and all I can say is that is an extremely distressed film - scratches and color failure. I am sure there is no software based dust and scratch removal solution that would be satisfactory with it. I understand that wet mounting on a drum scanner is one possible way but I've never seen results from it with this type of sample. However, I've seen results from the Coolscan 9000 + Nikonscan ICE and as far as scratch removal is concerned, I believe this would greatly benefit from it.

F by Les DMess, on Flickr

BTW, since these are from a Noritsu, I think they did a good job restraining the use of excessive sharpening as well as over correction of auot levels and colors. Having a more "neutral" scan on such distressed film is a big advantage as it will offer the most adjustments in post. Dust and scratch removal however needs to use hardware based ICE or wet mounting. Coolscan 9000 + Nikonscan ICE being the best at this.

I think Noritsu also have hardware based ICE. I didn't understood what is the conclusion.
Is Noritsu ICE inferior to Nikon coolscan ICE or the lab didn't used Noritsu ICE feature on this film?

 

[Pushkal Arora]

I took a turn on F and all I can say is that is an extremely distressed film - scratches and color failure. I am sure there is no software based dust and scratch removal solution that would be satisfactory with it. I understand that wet mounting on a drum scanner is one possible way but I've never seen results from it with this type of sample. However, I've seen results from the Coolscan 9000 + Nikonscan ICE and as far as scratch removal is concerned, I believe this would greatly benefit from it.

F by Les DMess, on Flickr

BTW, since these are from a Noritsu, I think they did a good job restraining the use of excessive sharpening as well as over correction of auot levels and colors. Having a more "neutral" scan on such distressed film is a big advantage as it will offer the most adjustments in post. Dust and scratch removal however needs to use hardware based ICE or wet mounting. Coolscan 9000 + Nikonscan ICE being the best at this.

I think they did a good job restraining the use of excessive sharpening as well as over correction of auot levels and colors. Having a more "neutral" scan on such distressed film is a big advantage as it will offer the most adjustments in post.

Thank you for the information. One thing that is bothering me about them is the histogram. On their green channel, the histogram is quite narrow.( Lacking extreme dark and extreme bright sides) compared to red and blue color channel. Is that normal for old negative films? I am little worried it didn't have to do something with not scanning properly.

 

[250swb]

All your colours are washed out that much is obvious. Using the scan of the print this is what simply using Auto Color and Auto Contrast in Photoshop gets you

 

[Les Sarile]

I think Noritsu also have hardware based ICE. I didn't understood what is the conclusion.
Is Noritsu ICE inferior to Nikon coolscan ICE or the lab didn't used Noritsu ICE feature on this film?

In your original post you said "Adobe ICE" which I assumed to be software based.
I believe I have a Noritsu scan of the same frame of film and it is not as good as the Coolscan 9000 + Nikonscan ICE. For that matter, I am not aware of any implementation that is as good.

 

[Les Sarile]

Thank you for the information. One thing that is bothering me about them is the histogram. On their green channel, the histogram is quite narrow.( Lacking extreme dark and extreme bright sides) compared to red and blue color channel. Is that normal for old negative films? I am little worried it didn't have to do something with not scanning properly.

Age old question when it comes to scanning color negatives - is it the film or the scanner. If you read enough posts here - and elsewhere, this is always the question. Myself, I have access to many scanners so I can always try a different one for comparison.

Here is an example of a perfectly well exposed frame of Kodak Gold 100 film scanned on the Coolscan compared to a Noritsu with all settings in auto and no post work applied except for combining the two and adding text.

Kodak Gold 100-7_30-36 Coolscan-Noritsu by Les DMess, on Flickr

I have scanned thousands of frames of various films with the Coolscans and occassionally compare results from many scanners. This doesn't happen to most scans I've had from the Noritsu but is one of the worst color failures I've seen over the years. If I didn't have the Coolscan, I would have been left with the same question you - and many others, have - is it the film or the scanner.

 

[Pushkal Arora]

Age old question when it comes to scanning color negatives - is it the film or the scanner. If you read enough posts here - and elsewhere, this is always the question. Myself, I have access to many scanners so I can always try a different one for comparison.

Here is an example of a perfectly well exposed frame of Kodak Gold 100 film scanned on the Coolscan compared to a Noritsu with all settings in auto and no post work applied except for combining the two and adding text.

Kodak Gold 100-7_30-36 Coolscan-Noritsu by Les DMess, on Flickr

I have scanned thousands of frames of various films with the Coolscans and occassionally compare results from many scanners. This doesn't happen to most scans I've had from the Noritsu but is one of the worst color failures I've seen over the years. If I didn't have the Coolscan, I would have been left with the same question you - and many others, have - is it the film or the scanner.

Okay, that's very very interesting to see difference like that on both of the scanners. By this we can conclude that there is a very high chance that it might be the scanner in the case of my films.

One question that is coming to my mind is about the "raw data" of both of these files.
Were you able to match the Noritsu image colors and quality like Nikon coolscan through post processing?

Because at the end of the day what I want is to archive my films or "film's raw data in digital form" for coming decades even if requires a little bit of post processing.

This doesn't happen to most scans I've had from the Noritsu but is one of the worst color failures I've seen over the years.

Out of all images that you scanned with Noritsu, what estimate percentage of them you will say the colors were accurate?
"Most scans" is a very wide term.

 

[Les Sarile]

Okay, that's very very interesting to see difference like that on both of the scanners. By this we can conclude that there is a very high chance that it might be the scanner in the case of my films.

One question that is coming to my mind is about the "raw data" of both of these files.
Were you able to match the Noritsu image colors and quality like Nikon coolscan through post processing?

Because at the end of the day what I want is to archive my films or "film's raw data in digital form" for coming decades even if requires a little bit of post processing.


Out of all images that you scanned with Noritsu, what estimate percentage of them you will say the colors were accurate?
"Most scans" is a very wide term.

I could not match the Coolscan results by post processing the Noritsu scan. Admittedly I only put minutes worth of work on it.
My experience here with cheap mass production Noritsu scans in minilabs are generally more overprocessed - too much sharpening and too heavy on auto levels, which doesn't look like your case. It's hard to say your results are due to bad film or poor scanning or a combination of both. If you're not satisfied with your results and have access to another scanner, it may be helpful to try a few to see.

 

[Pushkal Arora]

I could not match the Coolscan results by post processing the Noritsu scan. Admittedly I only put minutes worth of work on it.
My experience here with cheap mass production Noritsu scans in minilabs are generally more overprocessed - too much sharpening and too heavy on auto levels, which doesn't look like your case. It's hard to say your results are due to bad film or poor scanning or a combination of both. If you're not satisfied with your results and have access to another scanner, it may be helpful to try a few to see.

Here are a few examples with correct colors from same order but from different rolls.
Out of sexeen rolls, around four have slight green tint and one or two have worst color imbalance.

These rolls were developed from the span of a decade to 2.5 decade earlier. The color imbalance is totally random without anything to do when it was developed. If it has to do with the brand of roll, I can't tell now because the rolls are still with the lab and they have yet to ship them back.

If you're not satisfied with your results and have access to another scanner, it may be helpful to try a few to see.

It's getting rarer day by day to find a lab with coolscans. The labs that were offering in past are not offering that service anymore.
But the important thing is that in the case of negatives with slight green tint, I am able to correct them this way. I also have prints of them when they were developed and I am almost able to match the color. But I got worried if this have to do with bad scanning job, scanner or the roll also this method was not working in one or two of the pictures.

Do you think it will worth personally for me to get them scanned with coolscans?

 

[Les Sarile]

I did recall there was a scanning service in India that used Coolscans - ScanCafe.

Color is definitely a personal consideration. I know firsthand "accurate" may not be what a client may want. So if you can do the post work yourself then you're ahead of the game. Generally speaking, the more you ask of a lab, the more they will charge you.

 

[Pushkal Arora]

I did recall there was a scanning service in India that used Coolscans - ScanCafe.

Color is definitely a personal consideration. I know firsthand "accurate" may not be what a client may want. So if you can do the post work yourself then you're ahead of the game. Generally speaking, the more you ask of a lab, the more they will charge you.

They shifted to USA in 2019 and take orders from USA and Canada only. Anyways if I dig enough, I would definitely find a service with coolscans.
One last question that is coming to my mind is Fujifilm Frontier. Do you have experience how it compares to Coolscans and Noritsu?

 

[Les Sarile]

They shifted to USA in 2019 and take orders from USA and Canada only. Anyways if I dig enough, I would definitely find a service with coolscans.
One last question that is coming to my mind is Fujifilm Frontier. Do you have experience how it compares to Coolscans and Noritsu?

I did try a minilab running a Fuji Frontier for cheap scanning with film processing and found it to provide similar results to Noritsu as well as Agfa. Over processed - blownout highlights and over sharpening. Here's a comparison scan of Kodak 160VC from Fuji Frontier and Coolscan+Nikonscan.

Kodak 160VC2_01-32B by Les DMess, on Flickr

Of course their results are similar because their goal is to provide very fast results at an affordable price. I understand that the Noritsu can scan hundreds of frames per minute while the fastest Coolscan 5000 will take 30 seconds to scan one frame. It is possible that a lab can do custom work to provide better results but that will come at a higher price.

 

[grat]

I could not match the Coolscan results by post processing the Noritsu scan. Admittedly I only put minutes worth of work on it.

The highlights are blown out and it's under-saturated. Of course you can't match the results. I'd be curious what someone who knew what they were doing could do with the original negative on a Noritsu.

 

[MattKing]

I'd be curious what someone who knew what they were doing could do with the original negative on a Noritsu.

This is the crux of the matter.
While the quality of results is certainly affected by the nature of the scanner being employed, the results depend more on the combination of the scanner and software used and the skill and experience of the operator.

 

[Urs Gantner]

I don't have a Noritsu HS1800..... but i own a Frontier SP3000 since many, many years.

In my personal opinion for C41 the Fuji is colorwise unbeatable - for sure you have to do your color and contrast corrections live during scanning. This requires a lot of experience in color theories, every operator works and sees a little different. And a proper developed C41 negative within' specs is fundamental key for a good scan. If you want to see the difference between Fuji and Noritsu look at

https://carmencitafilmlab.com/blog/frontier-vs-noritsu-round-2/

https://www.richardphotolab.com/blog/post/its-back-noritsu-vs-frontier

Never had such shitty scans as Les Sarile shows since years here on this forum - sorry, but this is complete nonsense and will kill every serious discussion about the topic.

 

[Les Sarile]

The highlights are blown out and it's under-saturated. Of course you can't match the results. I'd be curious what someone who knew what they were doing could do with the original negative on a Noritsu.

This is the crux of the matter.
While the quality of results is certainly affected by the nature of the scanner being employed, the results depend more on the combination of the scanner and software used and the skill and experience of the operator.

I don't have a Noritsu HS1800..... but i own a Frontier SP3000 since many, many years.

In my personal opinion for C41 the Fuji is colorwise unbeatable - for sure you have to do your color and contrast corrections live during scanning. This requires a lot of experience in color theories, every operator works and sees a little different. And a proper developed C41 negative within' specs is fundamental key for a good scan. If you want to see the difference between Fuji and Noritsu look at

https://carmencitafilmlab.com/blog/frontier-vs-noritsu-round-2/

https://www.richardphotolab.com/blog/post/its-back-noritsu-vs-frontier

Never had such shitty scans as Les Sarile shows since years here on this forum - sorry, but this is complete nonsense and will kill every serious discussion about the topic.

As you can see from what I represented, these examples are all fully automatic and cheap scans conducted at the time of film processing. When I first started using these minilabs, these used to cost about $8 for a 36 frame roll. As I told the OP, the service he uses may well be able to deliver better results but it will come at a cost.

BTW, any of you ever scan such a distressed frame as the OP represented as F or my badly scratched Kodak 160VC? Got scans to show?

Urs, do you provide a scanning service with your Frontier?

 

[Pushkal Arora]

I don't have a Noritsu HS1800..... but i own a Frontier SP3000 since many, many years.

In my personal opinion for C41 the Fuji is colorwise unbeatable - for sure you have to do your color and contrast corrections live during scanning. This requires a lot of experience in color theories, every operator works and sees a little different. And a proper developed C41 negative within' specs is fundamental key for a good scan. If you want to see the difference between Fuji and Noritsu look at

https://carmencitafilmlab.com/blog/frontier-vs-noritsu-round-2/

https://www.richardphotolab.com/blog/post/its-back-noritsu-vs-frontier

Never had such shitty scans as Les Sarile shows since years here on this forum - sorry, but this is complete nonsense and will kill every serious discussion about the topic.

This is the crux of the matter.
While the quality of results is certainly affected by the nature of the scanner being employed, the results depend more on the combination of the scanner and software used and the skill and experience of the operator.

The highlights are blown out and it's under-saturated. Of course you can't match the results. I'd be curious what someone who knew what they were doing could do with the original negative on a Noritsu.

These are the samples of images with correct colors. I have picked out randomly from different rolls whose scans came out fine.
One question that is really coming to my mind is that, if we try to judge the quality of these scans by a professional eye (That's why I am asking from this community), can't we bring some conclusion about the photos with green tint and the
photos extremely damaged whether it is the scanning problem or the films or the combination of both?

If the scanner operater knows his/her job very well, why would s/he fuck up some of the rolls and some don't? Because in the scans the lab has given me, either the whole batch of roll has green tint problem or there isn't. (Out of 16 rolls, 5 has this type of problem and other 11 are fine.)

Please correct me if I am wrong may be missing something.

 

[MattKing]

can't we bring some conclusion about the photos with green tint and the
photos extremely damaged whether it is the scanning problem or the films or the combination of both?

Probably not - unless we have the original negatives in hand, plus the opportunity to re-scan them r print them optically.

 

[jtk]

ICE requires a specific scanner light source (and a specific copyright IMO)...it's not just dust/scratch and color guessing. There's an Epson flatbed version of ICE that will only scan prints, not film.

 

[Les Sarile]

ICE requires a specific scanner light source (and a specific copyright IMO)...it's not just dust/scratch and color guessing. There's an Epson flatbed version of ICE that will only scan prints, not film.

For the Coolscan (V, 5000 and 9000) Nikonscan ICE was licensed from ASF (Applied Science Fiction) which was acquired by Kodak shortly after. I recall that Ed Hamrick (Vuescan) didn't have the license and therefore couldn't use this in the software back then. I don't know if that has since changed.