Ad
Ad
Ad
Pages: « 1 2 [3] 4 5 »   Bottom of Page
Print
Author Topic: Kodachrome Revisited  (Read 15516 times)
Tim Lookingbill
Sr. Member
****
Online Online

Posts: 1208



WWW
« Reply #40 on: January 25, 2011, 12:09:42 PM »
ReplyReply

Quote
...if not accurate color, then pleasing color.

And there is the rub in all of these discussions and what is causing the confusion. Accurate to how Kodachrome sees it or accurate to the original scene? Pleasing? What's that? Can you define it spectrally?

Who and how is the "Pleasing" color of Kodachrome defined and are their Lab/XYZ (pick a connection space) color coordinates associated with a color target IF there is no existing Kodachrome IT8 target?

And if there is, does anyone have the official reference file that describes the "Pleasing" color of Kodachrome.

Peter, that's a very good looking scan, but I don't see any uniquely "Pleasing" color characteristic that defines it as Kodachrome over any other film. If I were to see that out of context of this discussion I would've mistaken it for any type of film and even a DSLR capture.

It is clear from its reputation in American culture and from comments throughout the decades of its existence that Kodachrome has gained a reputation for a particular look and I haven't seen any image around that defines this look so it can be reproduced on a consistent basis.

Has anyone found an it8 target that faithfully and definitively shows the look of Kodachrome? We could at least build some type of spectral color definition maybe off a wide gamut display using Apple's Digital Colormeter to get Lab readouts.
« Last Edit: January 25, 2011, 12:13:58 PM by tlooknbill » Logged
Schewe
Sr. Member
****
Offline Offline

Posts: 5500


WWW
« Reply #41 on: January 25, 2011, 12:22:02 PM »
ReplyReply

Peter, that's a very good looking scan, but I don't see any uniquely "Pleasing" color characteristic that defines it as Kodachrome over any other film. If I were to see that out of context of this discussion I would've mistaken it for any type of film and even a DSLR capture.

Not me...that shot screams Kodachrome...the yellow intensity and that shade of blue in the sky is pure Kodachrome. The clean clouds with the warm ground in also a clue.

No doubt you could take a digital capture and make it look like this, but it would take effort. Good job Peter...
Logged
pfigen
Sr. Member
****
Offline Offline

Posts: 469


WWW
« Reply #42 on: January 25, 2011, 02:49:35 PM »
ReplyReply

Jeff is correct about the typical Kodachrome skies, which skew towards magenta, and the yellows as well. I did pull the skies back a bit as they were too magenta for my tastes. Tim's comments were also valid. This image takes advantage of the color contrast to make something visually appealing, but there is something different about Kodachrome vs. E-6 films. When you look at a large print of this image, and I have a 42 inch wide version in my girlfriend's music room, there is something more "meaty" about it that a similarly sized print from E-6 or digital that is hard to describe, other that to say it strikes at a guttural level. As with any visual medium, it always helps to choose subject matter that plays to the strengths of that medium.

My last rolls of Kodachrome came back from Dwayne's yesterday and I was disappointed to find that they had mounted ten roll of it instead of sleeving. That only makes my job more time consuming, but on a quick perusal, some the best images were from two rolls of K25 that expired in 1991 - twenty years ago. I'll try to get at least one on the scanner this afternoon.
Logged
crames
Full Member
***
Offline Offline

Posts: 210


WWW
« Reply #43 on: January 25, 2011, 04:38:01 PM »
ReplyReply

And there is the rub in all of these discussions and what is causing the confusion. Accurate to how Kodachrome sees it or accurate to the original scene? Pleasing? What's that? Can you define it spectrally?

I mean pleasing skin tones when projected. I'm not saying that it's accurate in reference to the original scene.

Hopefully this link to Google Books will open for a description of "preferred" reproduction starting at page 174 in Reproduction of Colour: http://books.google.com/books?id=nFtW4LG24fEC&pg=PA174&dq=%22preferred+colour+reproduction%22+OR+%22preferred+color+reproduction%22&hl=en&ei=XkM_TfveHYT68AbJxoT7Cg&sa=X&oi=book_result&ct=result&resnum=1&ved=0CCoQ6AEwAA#v=onepage&q=%22preferred%20colour%20reproduction%22%20OR%20%22preferred%20color%20reproduction%22&f=false

It's not specifically about Kodachrome, but covers projection viewing. Apparently a lot of the research was done at Kodak.

Quote
Who and how is the "Pleasing" color of Kodachrome defined and are their Lab/XYZ (pick a connection space) color coordinates associated with a color target IF there is no existing Kodachrome IT8 target?

And if there is, does anyone have the official reference file that describes the "Pleasing" color of Kodachrome.

A link to the Kodak Kodachrome target was given in one of the other threads. SilverFast has targets, too. At the Kodak link there is a QSP file that contains the spectral data if that helps you.

But what can you get out of a profiling target, anyway?

Quote
It is clear from its reputation in American culture and from comments throughout the decades of its existence that Kodachrome has gained a reputation for a particular look and I haven't seen any image around that defines this look so it can be reproduced on a consistent basis.

Without having a Kodachrome original to refer to, maybe reproducing the look on a random image won't be so simple. A detailed model would take into account the spectral sensitivity, the tone curves, and the colors of the dyes. Then you have to account for the perceptual effects of dark-surround viewing, incomplete adaptation, etc.

What has come out in these threads is that it's not that straightforward, even when you're starting from an actual Kodachrome scan! How do you want to reproduce the look of Kodachrome, with or without the blue cast?

Quote
Has anyone found an it8 target that faithfully and definitively shows the look of Kodachrome? We could at least build some type of spectral color definition maybe off a wide gamut display using Apple's Digital Colormeter to get Lab readouts.

If you want to definitively show the look of Kodachrome you will either have to 1. show an actual Kodachrome slide, projected or on light box, or 2. take a scan of a slide and modify it to match the appearance of the actual projected slide.

Cliff
Logged

Cliff
tokengirl
Sr. Member
****
Offline Offline

Posts: 360



« Reply #44 on: January 25, 2011, 04:41:54 PM »
ReplyReply


But then I notice how all four of those beach shots changed hue in sky blue and the color of the sand. They're the same beach and same time of day. Don't know if that's user error or the inconsistent nature of Kodachrome.

Those are my photos.  THey look that way because they were scanned on an Epson V750 with EpsonScan software in Full Idiot mode with the Color Correction feature turned on to get it done quickly.  In other words, User Laziness.  After fooling around with both Silverfast and VueScan, and getting horrific results, I just got plain ol' tired of dealing with it.

Honestly, I think the nicest Kodachrome scans I have seen to date are pfigen's.  Are they a "correct" representation of what Kodachrome is supposed to look like?  I don't know.  But they sure do look like the way I remember Kodachrome in the good old days.
Logged
davidh202
Sr. Member
****
Offline Offline

Posts: 551


« Reply #45 on: January 25, 2011, 06:24:06 PM »
ReplyReply

So, Should I save my $$
has anyone actually used Silverfasts Kodachrome "solution" effectively, and would you say it is worth the price  they are asking, or is there a better "solution"for scanning Kodachrome?

Can anyone give me a straight answer after 3 threads, as to a reasonably affordable scanning "solution" to achieve resonably good quality scans of my Kodachromes??
 
« Last Edit: January 25, 2011, 06:59:56 PM by davidh202 » Logged
pfigen
Sr. Member
****
Offline Offline

Posts: 469


WWW
« Reply #46 on: January 25, 2011, 07:02:44 PM »
ReplyReply

David,

Probably the best non drum scanned scan I've seen of Kodachrome were from a Nikon 8000 or 9000. That's the most affordable solution, but the CCD will still never record all the shadow or highlight detail that a drum scanner's pmt will, so it kind of depends on what your definition of good is. The Nikons can make some great scans particularly if you've got a slightly flat slide, but even the drums have a hard time digging all the way into Kodachrome's shadowy past. That's one of the hallmarks of Kodachrome - a very very dense black - higher than any normal pictorial film.
Logged
pfigen
Sr. Member
****
Offline Offline

Posts: 469


WWW
« Reply #47 on: January 25, 2011, 08:07:40 PM »
ReplyReply

I did get around to scanning a couple of pieces of film today. Here's a small scan from a piece of PKM Kodachrome 25 that expired in 1991, which means the film is at about three years older than that. Interesting enough, this batch came through being not nearly as blue in the shadows at other Kodachromes I've scanned, only underscoring the entire content of this thread. But even still it was about ten points blue in the shadows rather than twenty-five. Maybe the latest chemistry is better or maybe it's the old film. There was very little done to this in Photoshop aside from cleaning up some dust and a very minor overall curves adjustment and a very slight reddening of the lips. And it's very close to the film as well. The detail scan, for anyone interested, is from the original 8000 ppi scan. Yeah, it's grainier than digital and not quite as sharp, but it's got guts. This was shot right at dusk, so the sunsetty color is completely appropriate. Canon EOS1vn and a 35mm 1.4 at f/2. All in all, I spent about a minute tops in Trident and about the same for the Ps tweaks, aside from the minor dust. I almost forgot to give due to my model and girlfriend, Gee Rabe, LA's accordion diva, who was wearing no makeup aside from the lipstick.
« Last Edit: January 25, 2011, 08:20:10 PM by pfigen » Logged
pfigen
Sr. Member
****
Offline Offline

Posts: 469


WWW
« Reply #48 on: January 25, 2011, 08:21:00 PM »
ReplyReply

Not sure what happened to the full frame scan, so here it is...
Logged
crames
Full Member
***
Offline Offline

Posts: 210


WWW
« Reply #49 on: January 26, 2011, 12:33:08 AM »
ReplyReply

Beautiful shot, model, and scan. Great Kodachrome example, for sure.

It's amazing that the film held up so well, despite having expired so long ago.
Logged

Cliff
tgray
Newbie
*
Offline Offline

Posts: 39


« Reply #50 on: January 26, 2011, 09:24:54 AM »
ReplyReply

Well I just received my Lasersoft Kodachrome target slide in the mail.  It was a quick job to profile my Coolscan V with it and Vuescan.  The results might not match up with pfigens examples, but I'm satisfied.  The color straight looks pretty good right out of the scanner and with a couple curve tweaks in Photoshop, I'm getting the best scans I've ever gotten with Kodachrome.  Which is good, because I have about 10 new rolls to scan and thousands to cull and scan from my childhood.

I guess what I'm saying is this: I think it's worth the 50 bucks for the target if you have a prosumer scanner like a Coolscan.
Logged
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6976


WWW
« Reply #51 on: January 26, 2011, 12:10:41 PM »
ReplyReply

So, Should I save my $$
has anyone actually used Silverfasts Kodachrome "solution" effectively, and would you say it is worth the price  they are asking, or is there a better "solution"for scanning Kodachrome?

Can anyone give me a straight answer after 3 threads, as to a reasonably affordable scanning "solution" to achieve resonably good quality scans of my Kodachromes??
 

Have you read the two articles on this website where I deal with how SilverFast IT8 Auto profiling handles Kodachrome? See especially my Plustek Scanner Review. Is it worth the price? - up to you to decide that. I can only comment on my own technical findings. I only migrated to SilverFast after progressing from manufacturer's software, to Vuescan to SilverFast in order of increasing cost. All of it is usable in one way or another. Just depends on what you think gives you the best results, what you like best using, and how much you want to pay. All are available as free demo downloads to do your own testing.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6976


WWW
« Reply #52 on: January 26, 2011, 12:16:24 PM »
ReplyReply

David,

That's one of the hallmarks of Kodachrome - a very very dense black - higher than any normal pictorial film.

Yes, I agree with this observation, and I have found a really effective way of dealing with it is to scan in as much shadow detail as feasible without wrecking the rest of the image, then open the completed image in Lightroom and finish the job with Fill and Blacks. This is one example where deft control at the can stage and use of good complementary software at the post-scan stage can really help. I wrote this up in my review of SilverFast Ai6 on this website some time ago.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6976


WWW
« Reply #53 on: January 26, 2011, 12:21:26 PM »
ReplyReply

Those are my photos.  THey look that way because they were scanned on an Epson V750 with EpsonScan software in Full Idiot mode with the Color Correction feature turned on to get it done quickly.  In other words, User Laziness.  After fooling around with both Silverfast and VueScan, and getting horrific results, I just got plain ol' tired of dealing with it.


Software has "personality", and can behave differently, giving better or worse results depending on whether it is being "fooled around with" or used scientifically/seriously. :-) Joking aside, both programs can give good results - depends on the user.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
Tim Lookingbill
Sr. Member
****
Online Online

Posts: 1208



WWW
« Reply #54 on: January 26, 2011, 03:46:26 PM »
ReplyReply

Thank you! Thank you! Thank you! Mr. Figen. Excellent scan. Reminds me of the color palette of a lot of image's I remember seeing thumbing through Communication Arts magazine and other high quality magazines in the '70's and '80's.

I can definitely see the underpinnings/undercolor of the halogen lamp in that scan. I base this observation on my days as a painter and the first thing that was taught was to never start a painting on a white canvas. We had to apply a warm-ish undercolor/underpainting throughout the entire canvas to establish some kind of color reference other than white to prevent from going astray from what was considered "Pleasing" color.

Then I started studying the color effects of the Solux lamp and its halogen underpinnings and noticed the same kind of color palette characteristic though subtle in prints which happened to match to what I saw on my display. I also noticed this change of color examining the skin tones in my hand under halogen lamps at my local hardware store. The same subtle maroonish brown undercolor showed up in the shadows even though the overall skin color looked the same.

When I stair at Peter's Kodachrome scan for a length of time and compare it to the overall color palette of similar daylight scenes I've shot in Raw taken with my Pentax K100D DSLR and edited in ACR, all of a sudden they all look green-ish! I can tell you for certain they didn't look like that before I saw your scan, Peter.

I'm laughing at this because it proves something I've suspected all along and have demonstrated with the DSLR shot posted below which might explain why there are so many comments on why folks see a difference between "film and digital" but can't put their finger on it. It also explains why there's so many different looking versions of Kodachrome scans.

The sample shot below was taken early in the morning showing the golden color from the the sun shining through my window onto my wall. I use this shot for simplicity sake, but it proves an optical phenomenon explained above about undercolor painting. I've included color purity's and gradients I've sampled from Peter's scan in sRGB made on a blank layer set to "Multiply". The bottom dark squares of the purity's was set to "Overlay". The gray and black square is just 127RGB and 000Black as a reference.

Stair at the bottom image below and note the hues of each color. Then view the top version. Notice a difference to the overall look of all the colors especially the gray and yellow square? I see a subtle green cast.

The only thing that changed was I applied in ACR's Split Tone...360 Hue/50 Saturation-(Shadow) and +100-(Balance). Note the difference to the perception of the sun rendered golden color which if you sample doesn't change its RGB readouts but looks different. Also note on the bottom image's shadows RGB readout is R>G>B which should look warm but looks green-ish gray. Tell me if you see something different.

This same "undercolor" optical phenomenon is happening in TokenGirl's Beach scans of Kodachrome from an Epson scanner which uses a fluorescent light source. It's just not as amplified as other scans seen in the "Kodachrome Project" due to a combination of scanner light source and color temp difference and subject matter in each scene. This may explain why you can't rely on RGB readouts especially when establishing neutrality for Kodachrome.

It could be the scanner light source that's causing most of the problems.

« Last Edit: January 26, 2011, 03:56:16 PM by tlooknbill » Logged
pfigen
Sr. Member
****
Offline Offline

Posts: 469


WWW
« Reply #55 on: January 26, 2011, 09:06:55 PM »
ReplyReply

Tim,

I hear what you're saying. I should always wait a day before posting something. I think it's just a tad magenta right now or really just needs to be backed off in the skin saturation, which I did today and like better. It's all subjective in the end. I'm not sure I agree with your being able to tell that it was scanned with tungsten bulbs. I think it has more to do with manipulating the software. If it was the bulbs, then everything else would have a different look as well. The scanner profile takes into account the illumination anyway.

What I do notice is that this latest film has less blue in the shadows that the film that was processed by A&I back twenty years ago, but I also think that Dwayne's processing is somehow a bit less saturated overall than what I used to get from A&I. Whether this has anything to do with the new prepackaged chemistry and no onsite chemist, I have no idea, but there does seem to be a difference. And then there's the fact that this particular piece of film was over twenty years old.

The one negative Kodachrome attribute is that is really does not like to have shadows opened up very far. They immediately get extremely grainy, giving the look of almost having two pieces of film with drastically different ISO ratings merged into one. So it's either live with it, or merge exposure brackets if possible.

The drum scanner has been spinning all afternoon. Just for reference to those who complain about long scanning times, this scanner takes about 7 minutes to scan a 35mm at 4000 and 28 for an 8000 ppi scan. But you can gang as many as thirty on a drum, set up the prescans and walk away, go to lunch, or just take a nap.
Logged
Tim Lookingbill
Sr. Member
****
Online Online

Posts: 1208



WWW
« Reply #56 on: January 26, 2011, 11:47:04 PM »
ReplyReply

Peter, does the image you posted match to what you see on a light table under a loupe? Does the quality and color temp of light of your light table influence or change the perception of color in the slide compared to the Howtek's halogen light source?

Thought the whole point of your posting your scan sample was to show what Kodachrome is suppose to look like outside of the influences of other scanners that may have limits to their being properly profiled or other color idiosyncrasies.

I would think a Howtek could faithfully reproduce what you see on the light table over like say an Epson flatbed or Nikon scan. Not sure, I've never scanned slides, only negatives which of course there's nothing to match to. I can match to prints on an Epson flatbed but have never been able to get that kind of color I see in that Kodachrome you posted.

I still get a gut feeling the halogen is playing some part in delivering the depth of color and unique hues I see in your scan.

My little image demo was to show that light source plays a part in color perception in relation to what the RGB numbers should look like. Light source in a way influences the DNA of an image even profiling can't compensate for. IOW something that looks blue on the slide may not have the numbers that indicates it's really that blue in the shadows because of what the light source does to other colors influencing the perception of the appearance of blue. The bottom image I posted shows the RGB readouts in the shadows to be around 50,40,30 but it looks kind of greenish. How do you explain that?

Hope I'm making myself clear. 
« Last Edit: January 26, 2011, 11:49:20 PM by tlooknbill » Logged
pfigen
Sr. Member
****
Offline Offline

Posts: 469


WWW
« Reply #57 on: January 27, 2011, 12:38:05 AM »
ReplyReply

Tim,

The image I posted was VERY close to how the slide looked on a Just-Normlicht D-50 lightbox. I don't know if I buy into the idea that Kodachrome was supposed to be viewed under tungsten for proper viewing. Everyone I know has always, and this goes back to the late '70s, viewed on D-50 flourescent boxes, balanced to that with CC filters and shot away. Sure, we used to have slide shows as well, but no one ever really noticed a color shift between the box and the screen, but we weren't looking for one either.

I still have to think the real reason for the bluish cast to many Kodachrome scans is simply that most scanners see and interpret the K-14 dyes differently than they see E-6, and that most scanners have input profiles based on E-6 pigments only makes it all the more apparent. I'm only manually compensating for the deficiency in the profile I'm using, and as I've said previously, this latest batch of film is less blue in the shadows than the films I've scanned from decades past. It's really not that hard to do. It only takes a few seconds, but you have to know what to look for and maybe having done a lot of drum scans in the past thirteen or fourteen years, I've just learned what to look for. The really funny thing is, is that the monitor I'm using for scanning is really old and only sort of calibrated any more. It's a PressView 17 that was manufactured in '94, but still works. Not as good as the Artisan's but with the overall Hutchcolor profile and reading the numbers, the scans are really close when they get shoved over the network.

Here's one more from today. Again, shot late in the day in a not particularly great part of West Adams Street in Los Angeles, but in a neighborhood that has fantastic local color, and color that is very well suited to Kodachrome. This one is close to the film but I did pull some magenta out of the sky.
Logged
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6976


WWW
« Reply #58 on: January 27, 2011, 11:55:32 AM »
ReplyReply



The one negative Kodachrome attribute is that is really does not like to have shadows opened up very far. They immediately get extremely grainy, giving the look of almost having two pieces of film with drastically different ISO ratings merged into one. So it's either live with it, or merge exposure brackets if possible.



I think what you may be seeing here rather than film grain is scanner noise. Film grain is usually much more apparent in the mid-tones for obvious reasons, so it would be interesting if you could compare the pattern of noise/grain you see in the mid-tones with what you see in the opened-up shadows. If they look similar it may be different grain characteristics as a function of tonality, but if they look quite different I would expect the dark stuff to be mainly scanner noise.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
Tim Lookingbill
Sr. Member
****
Online Online

Posts: 1208



WWW
« Reply #59 on: January 27, 2011, 01:33:37 PM »
ReplyReply

Thanks for taking the time for the extra scan and confirming a match to your scanner and light box, Peter.

That kind of yellow lit by a low sun on that building seems like that familiar old school color from paint pigments I've worked with and what seems CMYK friendly for a commercial press to reproduce without gamut clipping and black and/or cyan entering into the mix. Photoshop CMYK readouts bares this out.

Just to confirm Kodachrome records yellow like this, was the yellow in the scene that golden (100%Y and 40% magenta) when you took the shot or more cadmium to lemon yellow which often DSLR's tend to render? My DSLR sees limestone lit by golden sunset in a clear blue sky such as yours with a weird slightly greenish but intense thalo yellow almost going orange, but actually that's exactly what it looks like.

I do get an overall feel of your scan of having a subtle patina of a pastel (CMYK) magenta permeating throughout the entire image compared to how my DSLR captures similarly lit scenes but I guess that's part of the look of Kodachrome since that's what it looks like on your light table. The sunset whites appear a pastel pinkish orange. It's far different than the other scans I've seen online especially on the "Kodachrome Project" site which have this overall dark blue to dark bluegreen patina but uncontaminated intense primaries.

Thanks for the feedback, Peter.
Logged
Pages: « 1 2 [3] 4 5 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad