How do you scan color?

[PatrickT]

Okay guys, seriously? You've pretty much ruined what could have been a very helpful thread for people.

Enough already. It's childish.

 

[edge100]

Okay guys, seriously? You've pretty much ruined what could have been a very helpful thread for people.

Enough already. It's childish.

There are lots of great tips on the first two pages; more than enough to make terrific colour scans from any source negative, and, should you wish to transgress against the laws of nature, to make them look however you want them to.

Everything beyond that is just sycophantry of the highest order.

 

[Photo_Smith]

But really, you're going to learn it some day. I imagine you'll forget all about this conversation at that point.

I imagine he'll never have the ability to grasp what has been stated, the position is one of intransigence.

In fact the 'PhD scientist' (lets hope for humanities sake he's not working on the cure for cancer) could try some colour experiments.

Take a red apple and a green one place them on a white tablecloth, take a picture on Portra film.

Develop the film normally.

Print the film yourself on RA4 paper.

What you'll find looking at the negative the tablecloth will be dark brown the red apple will be made up of cyan dye, it will look strange under the masking but cyan. The green one will have a magenta look that can't easily be judged; but nonetheless the colours in the negative will have the same relationship to each other as they do in the original scene, just inverse and with a mask.

When you print the negative on RA4 paper it will be possible to make the tablecloth white by adjusting the filter pack, the apple will be by a miracle red, and the green one will look green (within the limits of the dye set)

Of course as a user you might want to make the apple less red, so you put some Y+M into the filter pack. Your red apple now looks less red-but oh my the tablecloth is now cyan and the green apple is cyan green too.

Now try to make the green apple blue, or the red apple green...

What have we found?

The negative is part of a photographic system designed to be printed on RA4 paper; each colour in that negative is also designed so the user can take a scene and as long as we use the film under it's rated lighting type you should get a similar colour stimulus in the final output that you did in the original scene.

It isn't up to you what colour you make the apple, sure you can change the hue with filters and the density with light but they always will be referenced WRT the colour translation of the negative the rendering and the relationships between those colours will be the same.

PS whats 'sycophantry' is it a term for type of Pachyderm 😛

 

[Wburgess]

Thanks for the replies everyone. I downloaded a demo of ColorPerfect and Vuescan and was playing with it a bit...still not getting what I want. I was following this tutorial: https://www.youtube.com/watch?v=_pfQ61WTbug&noredirect=1

Previously, I was using EpsonScan with my V500 and it was really hit or miss. Sometimes it would give me a scan that was PERFECT. Color, exposure, everything. Bot more often than not, the result would be less than satisfactory.

One such "perfect" scan I got from Epsonscan I re-scanned using Vuescan + Colorperfect, and the Epsonscan version was much better than the V + CP version.

I guess I'll have to keep experimenting...

Just to help get this back on track... (Sorry Patrick)

I really struggle to get good colour from my v500. Even using cp and vuescan.

The problem is a linear scan isn't really truly linear. How the scanner sees colours and to what sensitivity really effects how close it gets to being linear.

I don't know if there is a way to calibrate the v500 to see colour accurately but for me, it certainly does not.

This is why I believe the epson software sometimes does better, because it knows the way it sees the negatives and adjusts them accordingly.

The biggest shock to me was when I got my fs4000 for 35mm, colour is just so bloody easy on it! If only I could afford a mf equivalent.

Also, with the v500, I don't think it's a problem that can be solved just by white balance, it's more deeply ingrained than that.

Hope that helps.

 

[Pete B]

I really struggle to get good colour from my v500. Even using cp and vuescan.


The biggest shock to me was when I got my fs4000 for 35mm, colour is just so bloody easy on it! If only I could afford a mf equivalent.

Also, with the v500, I don't think it's a problem that can be solved just by white balance, it's more deeply ingrained than that.

Hope that helps.

This is my experience with the V700. And again, getting the Plustek 8100 solved the problem.

I also got better results with EpsonScan using the V700.

Pete

 

[edge100]

Just to help get this back on track... (Sorry Patrick)

I really struggle to get good colour from my v500. Even using cp and vuescan.

The problem is a linear scan isn't really truly linear. How the scanner sees colours and to what sensitivity really effects how close it gets to being linear.

I don't know if there is a way to calibrate the v500 to see colour accurately but for me, it certainly does not.

This is why I believe the epson software sometimes does better, because it knows the way it sees the negatives and adjusts them accordingly.

The biggest shock to me was when I got my fs4000 for 35mm, colour is just so bloody easy on it! If only I could afford a mf equivalent.

Also, with the v500, I don't think it's a problem that can be solved just by white balance, it's more deeply ingrained than that.

Hope that helps.

Scene: a camera store

Wburgess: "Sir, I'd like to return this Jupiter 8 lens you sold me."

Salesperson: "Ok. Was there anything wrong with it?"

Wburgess: "Yes. It sucks. I went out shooting with my friend, and his lens was MUCH sharper than this."

Salesperson: "What lens was he shooting with?"

Wburgess: "50 Summiliux ASPH"

Salesperson: "......."

Wburgess: "So I want my $50 back, please. Photography is really hard."

Salesperson: "We don't usually say these things to customers, but this is so outrageously stupid that I have to say something. What on Earth made you think that a $50 lens - made over 50 years ago in a factory with Soviet quality control - could rival a $4000 lens made with computer-aided design and unrivaled quality control? Did it ever cross your mind that if you buy crappy equipment, you should expect crappy results? Also, did it ever cross your mind that if you spent just a little bit more on a decent lens - not $4000 mind you, but perhaps $500 - that you'd get 95% of the results your friend gets with his Summilux? You cheaped out and bought a piece of junk lens, and you expect it to perform miracles?"

Wburgess: "........"


Buy a decent scanner (a Plustek 8100 will work nicely), and follow my tutorial, which works.

 

[edge100]

This is my experience with the V700. And again, getting the Plustek 8100 solved the problem.

I also got better results with EpsonScan using the V700.

Pete

"Dedicated film scanner outperforms flatbed"

Film at 11.

 

[Photo_Smith]

I think that the default settings in Epson scan seem to clip highlight information and compress shadow detail somewhat.
I'm not sure the Epson scan software allows for a change in Gamma all I have read suggests that it is set at 1 normally which is either 1.8 or 2.2 depending on which OS you use.
In reality I use the scanner to digitise for web based images, I print most of my B&W and send out colour.
So for me just setting the black and white points to stop clipping gives an OK rendering which is flat looking I then adjust that in an image editor.

In the end the output from a cheap scanner for medium format 800 pixel wide web viewing with all the caveats of the viewers screen being accurate is going to be good enough.

If you want to make 20" wide prints on a regular basis investment in something better is a good idea.

 

[k__43]

Just to help get this back on track... (Sorry Patrick)

I really struggle to get good colour from my v500. Even using cp and vuescan.

The problem is a linear scan isn't really truly linear. How the scanner sees colours and to what sensitivity really effects how close it gets to being linear.

I don't know if there is a way to calibrate the v500 to see colour accurately but for me, it certainly does not.

This is why I believe the epson software sometimes does better, because it knows the way it sees the negatives and adjusts them accordingly.

The biggest shock to me was when I got my fs4000 for 35mm, colour is just so bloody easy on it! If only I could afford a mf equivalent.

Also, with the v500, I don't think it's a problem that can be solved just by white balance, it's more deeply ingrained than that.

Hope that helps.

I had that too with the V500 .. I later got the Plustek 8200 - not much better.
When I got my FS4000 my jaw dropped on the desk, but the LS-8000 was even a step up (also less shadow noise and more even scans with thin negs - my worst complaint about the canon). The Canon has still a slight edge in resolution. My theory goes like (important MY THEORY, not knowledge) .. the V500 and the plustek use white LED while the Canon has a cold cathode lamp and the Nikon mixes RGB leds, thus both get a purer white than the spectrum of a white LED gives.

 

[edge100]

I had that too with the V500 .. I later got the Plustek 8200 - not much better.
When I got my FS4000 my jaw dropped on the desk, but the LS-8000 was even a step up (also less shadow noise and more even scans with thin negs - my worst complaint about the canon). The Canon has still a slight edge in resolution. My theory goes like (important MY THEORY, not knowledge) .. the V500 and the plustek use white LED while the Canon has a cold cathode lamp and the Nikon mixes RGB leds, thus both get a purer white than the spectrum of a white LED gives.

Theory: "a coherent group of tested general propositions, commonly regarded as correct, that can be used as principles of explanation and prediction for a class of phenomena"

The Plustek works just fine. If you can't get good results with it, it's user error. Sorry, but I scanned perhaps 10000 frames on the Plustek, with excellent results on virtually every one.

Are there better scanners? Sure. But there's no excuse for anyone to be scanning 35mm film on a flatbed. If you can't afford a Plustek, you can't afford to shoot 35mm film.

 

[Pete B]

"Dedicated film scanner outperforms flatbed"

Film at 11.

The epson photo scanner v700 is marketed as a proper scanner and comes with holders for 35mm, 120, and at least one other, so, yes it was a bit of a shock that the linear scans from the V700 are not the same as those from a Plustek. Both are dedicated film scanners but the cheaper 8100 give better colour and, as might be expected, resolution.

I expect a lot of people are struggling to use the v700 to give representative colours, and think it must be something they are doing wrong. I certainly banged my head against a wall trying to get red to look red rather than magenta (as I was trying to avoid using silver fast or EpsonScan . For 35mm, there's no question that the 8100 is better, and I presume the Plustek 120 will give more accurate colour than the v700. If someone is wondering which to get for medium format, go for the Plustek if you can afford it.
Pete

 

[edge100]

The epson photo scanner v700 is marketed as a proper scanner and comes with holders for 35mm, 120, and at least one other, so, yes it was a bit of a shock that the linear scans from the V700 are not the same as those from a Plustek. Both are dedicated film scanners but the cheaper 8100 give better colour and, as might be expected, resolution.

I expect a lot of people are struggling to use the v700 to give representative colours, and think it must be something they are doing wrong. I certainly banged my head against a wall trying to get red to look red rather than magenta (as I was trying to avoid using silver fast or EpsonScan . For 35mm, there's no question that the 8100 is better, and I presume the Plustek 120 will give more accurate colour than the v700. If someone is wondering which to get for medium format, go for the Plustek if you can afford it.
Pete

The V700 is in no way a "dedicated" film scanner; no matter how Epson markets it. It's a remarkably decent film scanner, for a flatbed, but you still get what you pay for.

The only exception I've noted for this is the Polaroid SprintScan 120, which is dirt cheap but is the equal of a 9000ED.

 

[Wburgess]

The V700 is in no way a "dedicated" film scanner; no matter how Epson markets it. It's a remarkably decent film scanner, for a flatbed, but you still get what you pay for.

The only exception I've noted for this is the Polaroid SprintScan 120, which is dirt cheap but is the equal of a 9000ED.

Unfortunately, not everyone has this knowledge in advance. Especially when it comes to the ability of rendering colours, it's just not something everyone is going to think about. I bought a v500 as a first step into home scanning, it's fine for web scans but not great for much else (except mf, but the colours are still off).

My FS4000 cost me £95. SprintScan is also a great option, but they don't go for as cheap as they used to.

So, no, you don't always get what you pay for.

I found the noise on the FS4000 was mainly caused by the USB A/D, using it on SCSI is much quicker and also less noisy.

Can you try not to push the conversation into the realms for the petty again Edge?

 

[rover]

Can you try not to push the conversation into the realms for the petty again Edge?

Yes, to everyone, you again are showing a remarkable ability to be obnoxious with a complete lack of respect or curtesy towards each other. Please stay on the path unfortunately now seldom traveled.

Thank you

 

[TXForester]

Of course there is a standard: an ARBITRARY, manufacturer designed standard, as valid as any other because it has no basis in physical reality beyond what the manufacturer WANTS!

I agree with the standard being arbitrary. But for the purposes of achieving relatively uniform results for the consumer across a multitude of labs, that standard becomes the bull's eye in the target. When those labs consistently achieve those standards within a small tolerance (I assume the tolerance would mean differences are not readily detectable by the human eye) then those labs are both precise and accurate. So I disagree that there is no accuracy. The accuracy is based on the standard even if the standard is arbitrarily designated the manufacturer.

Outside of producing consistent results for the average consumer (say minilab or some other mass production user), then there is no right and wrong and the final print or scan is up to user.

 

[PatrickT]

Just to help get this back on track... (Sorry Patrick)

I really struggle to get good colour from my v500. Even using cp and vuescan.

The problem is a linear scan isn't really truly linear. How the scanner sees colours and to what sensitivity really effects how close it gets to being linear.

I don't know if there is a way to calibrate the v500 to see colour accurately but for me, it certainly does not.

This is why I believe the epson software sometimes does better, because it knows the way it sees the negatives and adjusts them accordingly.

The biggest shock to me was when I got my fs4000 for 35mm, colour is just so bloody easy on it! If only I could afford a mf equivalent.

Also, with the v500, I don't think it's a problem that can be solved just by white balance, it's more deeply ingrained than that.

Hope that helps.

I know what you're saying. I think for me, who is just really concerned with getting pleasing color, I can come close enough with the temperature slider in Lightroom. I realize that it's more complicated than that, but I think that's sufficient for me. Thanks!

 

[PatrickT]

I think that the default settings in Epson scan seem to clip highlight information and compress shadow detail somewhat.
I'm not sure the Epson scan software allows for a change in Gamma all I have read suggests that it is set at 1 normally which is either 1.8 or 2.2 depending on which OS you use.
In reality I use the scanner to digitise for web based images, I print most of my B&W and send out colour.
So for me just setting the black and white points to stop clipping gives an OK rendering which is flat looking I then adjust that in an image editor.

In the end the output from a cheap scanner for medium format 800 pixel wide web viewing with all the caveats of the viewers screen being accurate is going to be good enough.

If you want to make 20" wide prints on a regular basis investment in something better is a good idea.

Yeah, I think this is where I have ended up. Tweaking the histogram before scanning a tad to make the software not clip the highlights and shadows gives a more "flat" scan, allowing me to adjust slightly in lightroom.

Thanks!

 

[PatrickT]

I had that too with the V500 .. I later got the Plustek 8200 - not much better.
When I got my FS4000 my jaw dropped on the desk, but the LS-8000 was even a step up (also less shadow noise and more even scans with thin negs - my worst complaint about the canon). The Canon has still a slight edge in resolution. My theory goes like (important MY THEORY, not knowledge) .. the V500 and the plustek use white LED while the Canon has a cold cathode lamp and the Nikon mixes RGB leds, thus both get a purer white than the spectrum of a white LED gives.

I also own a Plustek 7600i and do notice that the scans ARE better than what I get from my V500 for sure. That said, I don't really shoot 35mm anymore, so it's a bit useless.

If I had the funds, I would for sure purchase a dedicated MF scanner...it'd be like heaven... 😛

 

[Wburgess]

...it'd be like heaven... 😛

I'm currently swearing at a drum scanner I can't get to work on an old G4, heaven is so close, but yet so far away!

 

[Wburgess]

I know what you're saying. I think for me, who is just really concerned with getting pleasing color, I can come close enough with the temperature slider in Lightroom. I realize that it's more complicated than that, but I think that's sufficient for me. Thanks!

Agreed, it is more than possible to get please colour out of the V500. It's just since the FS4000, I can see what is possible from a colour neg.

Ignorance is bliss, as they say.