I don't think it is as rare as NK believes, it depends more on the scanner imho. I used to scan with a Nikon L5000 at 4000 DPI and the grain was much more obvious that with comparable pro scans done outside. Then I realized that the grain got more intense as I scanned at a lower resolution (to save time as I didn't always need 22mp files), which didn't make sense to me at all. I have optically printed color negative film at 70x100 cm a few times in the past and have never seen grain similar to what the scans show on screen even when I put my nose in the print (is that sufficient pixel peeping?)

Lotusm50 wrote:
edwardkaraa wrote:
Many don't realize that the noise they see in scans is not film grain but actually grain aliasing.


You probably should define "grain aliasing" for the uninitiated. If you've got samples it would be helpful to show them to illustrate.

Further what you see isn't necessarily grain aliasing, as your comment suggests, it merely can be. It is far from a certainty and it really is not that common.

See the discussion here: http://www.normankoren.com/Tutorials/MTF8.html
Norm Korens summary comment on this is as follows: "Film images are composed of grain, and film (especially negative film) can have more grain than most people realize, especially if they haven't seen sharp prints. Film grain is often mistaken for scanner noise or grain aliasing (a real effect, but somewhat rare)."

I see actual grain (not merely aliasing) in my scans -- certainly with the 5400 dpi scans, but also, depending on the film with the 4000 dpi scans. Scan high enough to resolve the grain and you limit the potential for grain aliasing.

chez wrote:


Only reason I moved to digital was convenience...not image quality.


Amen to that.

Scanned film is a poor representation of what film is capable of when printed directly to paper.

edwardkaraa wrote:
I don't think it is as rare as NK believes, it depends more on the scanner imho. I used to scan with a Nikon L5000 at 4000 DPI and the grain was much more obvious that with comparable pro scans done outside. Then I realized that the grain got more intense as I scanned at a lower resolution (to save time as I didn't always need 22mp files), which didn't make sense to me at all.


I've noticed the same thing. Grain looks to be about the same size at a 100% view whether I scan at 1000 dpi or 4000 dpi (on traditional B&W negs). I guess it makes sense to me - the silver grain is going to take up a whole pixel, so when those pixels represent a larger area (as in the 1000 dpi scan), the grain appears larger. There is a noticeable drop in grain if I scan at 4000 dpi and then downsize.

shirozina wrote:
Scanned film is a poor representation of what film is capable of when printed directly to paper.

This is true, but in color printing digital has significantly increased the flexibility that photographers have to execute an image as they intend. In the past, this has only been possible through the use of very complex masks. In that regard, I think hybrid digital workflow has greatly helped photographers.

mrladewig wrote:
shirozina wrote:
Scanned film is a poor representation of what film is capable of when printed directly to paper.

This is true, but in color printing digital has significantly increased the flexibility that photographers have to execute an image as they intend. In the past, this has only been possible through the use of very complex masks. In that regard, I think hybrid digital workflow has greatly helped photographers. Certainly and I was using such a workflow for a good while before a viable totaly digital system came along. However when comparing the ultimate qualiteis of each system I think it's still a compromise - as grain is exagerated and subtle tonal quality is lost.

Lotusm50 wrote:
Actually, if you really want good B&W, get a a CCD sensor without a RGB filter (a Foveon style filter is irrelevant for B&W, just like the Bayer filter, you don't need 3 different color sensitives at each photosite or alternating photosite to capture B&W). Monochrome versions of most bayer filters are made by sensor manufacturers.

Kodak produced a monochrome version of one of it's early DSLR's and it is, from what I've seen, spectacular and very sharp. Noticeably more resolution that the equivalent bayer-filtered version. Of course, you would have to go back to using color filters to control the tonalities of the image (because you are not capturing colors that you can select later). Of course, the best thing about digital, color digital that is, is that for B&W you don't need to bring a half dozen red, orange yellow, green and blue filters with you, as well as assess filtration and constantly change them out to optimize a particular image before you capture it.


Interesting discussion, and my only comment is the related in color, astrophotos made from B&W CCD chips. They use RGB fiilter wheels to capture the three colors and then use stacking software to get the color file. Granted these are long exposures, but it is well known the B&W chips used are typically more sensitive and higher resolution than the color chips. The process can get more complicated with luminance channels and Halpha channels, but point is 100% specificity of the selected wavelengths through filters.

chez wrote:

In my experience, though, the colors I get with my calibrated cam are *far* more accurate than any print/slide I've seen.

....

If you were getting better images from your D30 than from you 6x7 Mamiya slides...then you must have been murdering your slides.


Yes, no question, the resolution is there in 6x7. But most prints that I see from film - including the "wonderful" Ansel Adam prints that I saw last year in Chicago - look like crap. We are just so used to overlooking the smudging, dirty, random noise that comes from grain.

I never uprez. The maximum I would print with my 1DsII was 18x27 at 180 dpi - mostly for portraits, etc. I consider that the largest native size for that camera, maybe push it to 20x30 for portraits. The D30 falls apart above 8.5x11

I still shoot 4x5 film when I need larger prints (I have 2 Epson 7600's and a 9600 - I like to print large.) No question that digital is limited in resolution, and limited in size if you can't afford a P45+ or 65+ (I can't right now, not working due to neck/back issues.)

But even my best 4x5 scans and prints have that damn grain/dye crud, especially in smooth areas liek the sky. And taht serves primarily to "dilute" the color clarity, because the gain artifacts are multi-hued little color blossums, or dull grey "additives", like mixing some grey paint into your saturated hues.

Now, going to something like an Astia slide, you are talking about much less noise than traditional Kodak films, or especially B&W. And a digital process flow can give much better prints than I ever got in the darkroom, no question.

I shot film for 20+ years. I went to all digital output in 1998, think I bought a Minolta Dimage Scan Multi then for my 6x7 and 645. I had long ago given up 35mm, just too noisy and grainy and crappy. Same for any medium format iamges above 100 iso.

The Dimage was not quite adequate for my 645. Went through a few more scanners. Started with digital capture in 2002. Never intended to stop shooting 6x7 film, it just got replaced in practical terms. As I said, I am stull shooting 4x5. That is the only place where the resoultion outweighs the noise for me today.

I just think we are so biased to tolerating film artifacts that we don't even see them, as mentioned above.

Back to the original question, which raw converter produces in your opinion the best and purest reds that are similar to what one can get with color reversal film such as Fujichrome Velvia. As I mentioned earlier, portrait style in DPP seems to give good results in some cases. Also someone mentioned C1 and I do remember from memory having tried it long time ago that it did produce some nice pure reds. It seems to me that nice pure reds usually show high R values, and very low G and B values, while DPP neutral and standard seem to have all RGB values quite high which results in pinkish or washed out reds. Any recommendations?

Same question for blue.

mmurph wrote:
Film is "dirty" - the grain and/or dyes are really noise. The substrate is really poor. We are fooled into thinking there is more detail beacuse of the random artefacts that hide real low-level detail. But blow it up and it looks like what it is - garbage.


Your experience with film must be very out of date; the latest emulsions capture remarkable amounts of detail, and I'm only shooting on 4x5, not even 8x10... to match the detail in the first 40x50 print I got made, you'd need a 150 megapixel full-color sensor.

You can look at it from "normal" viewing distance as well as up close, and it holds up to both.

JimBuchanan wrote:
Interesting discussion, and my only comment is the related in color, astrophotos made from B&W CCD chips. They use RGB fiilter wheels to capture the three colors and then use stacking software to get the color file. Granted these are long exposures, but it is well known the B&W chips used are typically more sensitive and higher resolution than the color chips. The process can get more complicated with luminance channels and Halpha channels, but point is 100% specificity of the selected wavelengths through filters.

All CCD sensors are monochromatic. They only measure luminosity, never color. The ones with filters just record the amount of light that the filters pass, as far as the photosites are concerned, the light could still be white for all they care.

That alone is a large part of how the CCDs made for astrophotography and other non-portable imaging applications can outperform dSLRs; the engineers designing them don't need to worry about usability concerns that would cripple dSLRs... can you imagine a sport photographer trying to capture color images by taking three pictures, one in each color, and combining them after the fact?

edwardkaraa wrote:
Back to the original question, which raw converter produces in your opinion the best and purest reds that are similar to what one can get with color reversal film such as Fujichrome Velvia.

I have only used Lightroom or ACR recently. But, you need to calibrate for your camera!

There are online tools available, or use an i1. You will need a ColorChecker card. The high-end i1 comes with the ColorChecker DC I think (?) - extended version for digital, a $260 card.

The paradigm has recently shifted with Lightroom 2.0 and the DNG ICC Profile editor though.

Reds are pulled in best via calibration, especially for Canon. Really nails tough color matcxhes for product.

Tamerlin wrote:

Your experience with film must be very out of date; the latest emulsions capture remarkable amounts of detail, and I'm only shooting on 4x5,


Latest revision Fuji Astia in 4x5, bought fresh and frozen. Plus Kodak Portra 160 NC. Been using those 2 films only for 10+ years (except for some Agfa.)

But we were mostly talking about smaller film, 6x7, 645 and even 35mm. There is a point where a quantitative difference becomes a qualitative difference. That point is certainly reached with 4x5.

I am shooting a 1DsII, 40D, 5DII, and 5 different 4x5's. Plus a Mmaiya 645 that is sorely neglected. Bought a new zoom for it this summer and never even tested it. It was waiting for a digital back, but ....

FWIW, I have a 6 Sigma Black Belt certification in quality. Designed propriatary digital asset management systems since 1992 (large scale systems, starting at $2 million plus back in 1992.)

mmurph wrote:

Yes, no question, the resolution is there in 6x7. But most prints that I see from film - including the "wonderful" Ansel Adam prints that I saw last year in Chicago - look like crap. We are just so used to overlooking the smudging, dirty, random noise that comes from grain.



I guess that's where we differ. I just printed some TMZ 3200 negs in the darkroom at 8x10, and I think they look fabulous. I know thats a small size for the digital crowd, but I'm amazed at how NON-grainy the prints came out. I kind of wish it was a bit grainier...

mmurph wrote:
Yes, no question, the resolution is there in 6x7. But most prints that I see from film - including the "wonderful" Ansel Adam prints that I saw last year in Chicago - look like crap. We are just so used to overlooking the smudging, dirty, random noise that comes from grain. {snip}

I just think we are so biased to tolerating film artifacts that we don't even see them, as mentioned above.


I don't know what you're talking about. I've seen "wonderful" Ansel Adams prints, as well as other large format prints from that era, and don't see the "smudging, dirty, random noise" that you suggest you are seeing. They just look great. It will be great when digital can approach the overall quality to these images (it is getting closer, however)

This is not becuase I am accustomed to "overlooking" it or "biased to tolerating" it. Most people will disagree with you on the quality of these prints and you can't try to effectively dismiss their views with such preemptive throw-away comments.

mmurph wrote:
Latest revision Fuji Astia in 4x5, bought fresh and frozen. Plus Kodak Portra 160 NC. Been using those 2 films only for 10+ years (except for some Agfa.)

But we were mostly talking about smaller film, 6x7, 645 and even 35mm. There is a point where a quantitative difference becomes a qualitative difference. That point is certainly reached with 4x5.

I am shooting a 1DsII, 40D, 5DII, and 5 different 4x5's. Plus a Mmaiya 645 that is sorely neglected. Bought a new zoom for it this summer and never even tested it. It was waiting for a digital back, but ....

FWIW, I have a 6 Sigma Black Belt certification in quality. Designed propriatary digital asset management systems since 1992 (large scale systems, starting at $2 million plus back in 1992.)

You must be doing something pretty horrible to your film then. I've seen a lot of scans from film, and although their quality depends a lot on the skill of the scanner as well as on the skill of the photographer, they've been exquisite... and nothing shot in 35mm format digital is even in the ballpark with medium format film. I've seen lots of people claim otherwise, but no one's ever been able to back it up with anything other than claims of authority.

The question in this thread is about color rendition and, by extension color purity and noise. It has nothing to do with resolution, which you all continue to return to again and again.

We have two ways of rendering an image of the world, both using sensors. One renders directly from the world. The other renders from an intermediate physical substrate.

There is no analog/physical process that does not involve the introduction of noise. Noise decreases the "purity" of the color signal.

mmurph wrote:
We have two ways of rendering an image of the world, both using sensors. One renders directly from the world. The other renders from an intermediate physical substrate.


Which is more "real"? The physical emulsion on a substrate or the 1010110101001000...

Photography has not always relied on digital sensors and processing to produce a print.

Just because color rendition is accurate and precise does not mean that the result is pleasing. This is one component of the issue at hand. It seems that Adobe have finally realized that there is more to color than accuracy and now they are back tracking a bit to create profiles more like the ones the camera manufacturers have provided. Anyone who has used a Canon camera with ACR and DPP for some time will know that the resulting converted images will look significantly different depending on the raw conversion software choice or further in DPP it will change by the "style" applied. Often for reds, oranges and yellows, Canon's software produces a more pleasing rendition than ACR. There is a visible difference in the files.

There is an saying in photography that a man with 1 light meter knows the correct exposure, but a man with 2 light meters cannot be certain. Which is the "right" conversion?

On the other hand of the argument is film. Film manufacturers have spent time creating films that provide pleasing color in various palettes and contrast ranges. I wouldn't normally choose to use Portra 160NC for a landscape shot, but it produces gorgeous portraiture. The contrast, response curve and color balance built into the film is specifically tailored to this purpose. Fuji Provia and Kodak E100G are two films with a similar purpose (general purpose ISO 100 slide films). They are fine grain, provide good neutral colors and have good saturation for stronger colors. Yet the two films look different. The choice of subtle differences in palette and color rendition and contrast and saturation are the things that film photographers like. I can choose a film with a built in response and I know that I'll get a certain result. I can do the same thing in digital, true enough, but this does not take away from the qualities that photographing with film provides. Film itself is not trapped within the Prophoto RGB colorspace. Its isn't trapped by 256 or 4096 shades of red, blue and green.

When you capture digitally, you are then left to cook the shot yourself. You've got to come up with an appropriate development to match the purpose. This is stuff that the film companies did for you. Granted when you scan film yourself, you've got to do very similar work to get the scanning software to produce a good result, but the film still provides a baseline. You can choose to use the film scanner simply to replicate the film or you can choose to use it as an artists tool.

There is more to photography than clinical recording.

This discussion started off with the question:

Why can't I get color response from digital capture like the color I got from film?

There have been some constructive comments and suggestions that the OP might consider using camera calibration. But the underlying question is still there.