Good news for photographers who yearn for film like images

[student]

"Silver halide" prints are free of halides after development and fixing. What you see is metallic silver, not halides (which are mostly colourless in the first place, apart from a slight yellowish tinge).

Thank you for the correction. It is the reflection of the silver that I was talking about.

Silver halide prints predate platinum prints. The standard print medium since the days of Fox Talbot was "salt paper", which is essentially a silver chloride medium.

I will check this out. In any case, platinum prints almost came to a standstill after WW1. But revived later.

 

[LittleWolf]

In any case, platinum prints almost came to a standstill after WW1. But revived later.

I understand that, apart from being expensive, platinum was restricted during WW1 (at least in the US), presumably because it was needed for military purposes.

 

[student]

I understand that, apart from being expensive, platinum was restricted during WW1 (at least in the US), presumably because it was needed for military purposes.

If I remember it correctly, the main reason was that Russia restricted the sales of platinum because it was needed for military purpose. I do not think US mine platinum.

 

[lsisaxon]

You're not totally wrong, it is just a misperception.
That article from Kodak was talking about graininess, it does not mention about the varying thickness of the emulsion after development. But it did indirectly tell that the grains are not all located at the same plane, but randomly within the thickness of emulsion.
This is the small words quoted from that article:
"Note that surface grains are in focus while grains deeper in the emulsion are out of focus. The apparent "clumping" of silver grains is actually caused by overlap of grains at different depths when viewed in two-dimensional projection"

I drew some diagram below (this is simplified, B&W film, just to illustrate the logic):

The unexposed film, will have all silver-halide intact in its emulsion layer.

If the film was exposed to a light with varying brightness, some of its silver halide will be activated.

After development, the non-activated silver halide will be dissolved, and gone. The rest of silver particles will create a "clump" of particles proportional to the light brightness.

Those clumps of particles create different thickness of the remaining emulsion.
If you see the film from top (two dimensionally), then you'll see the grains form the image, this is what you said as "dithering". And this is what that Kodak's article describe.
Film manufacturer usually use the word "density" instead of "dithering".

For those wondering whether this kind of knowledge is necessary for a photographer, I can say: no. You don't need this knowledge to create a good photograph.

Yeah.. but the idea given by your first article was that the optical density on the film was more continuous than discrete. The fact that the film is made of silver-halide crystals suggests that it is discrete and the tonal gradations are formed when the silver is washed away after the reaction with the dye also has to be discrete.

You are definitely not wrong in that there will be modulation of the thickness of the emulsion after development, I do not dispute that, but I dispute that the intensity recorded is continuous rather than discrete. Which is why I used the term dithering. Even in modern inkjet technology, the droplet size is modulated but that does not mean that the image has become continuous, it is still dithered. Dithering is employed when the number of gradations available is not sufficient to represent a certain tone, therefore, it has to be represented by the average optical density over an area. Film, certainly behaves this way. Otherwise, it would simply become lithographic film, which at this point in time still has the highest available resolution but it is just truly black and white, nothing in between.

Even for digital, the intensity is recorded in steps depending on how many bit is used to represent a pixel. If you used less bits to represent, then dithering will still be required. One difficulty to directly compare, however, is that there is no direct relation to film grains because grain sizes are different for a sensitivity and there is a whole random mixture of different size crystals for the film to be sensitive to different levels of intensity, while in a digital sensor, it is a uniform array of photosites, so in space, it is also different.

 

[student]

But my point is, for people who do colour photography and send to lab to print, why would there be a need to go through a film medium, then send to the lab for processing and at the end still scan the negative at 6mp before sending it to a laser(?) writer to write the image at 300dpi to the colour paper? And at the end, claim that it is still superior to a DSLR image at 10mp? How many people making that remark is doing just that? Oh and one more thing is that the laser writer writes inverted RGB, so it doesn't go into the CMY phase like transparencies and negatives will go through. So that would be about as accurate as what the sensor receives. Here is some information. http://www.photosig.com/articles/342/article

Your point is very valid.

If the end point is an inkjet print, then I do not see much value in using film and then convert it to digital file. And that is why for me, when I do color, I will just stick to my DSLR.

However for my B&W, I do have an alternative. An alternative which to me, gives a result that pleases me more. And an alternative which is obsolescence-proof. Which will still go on and on after today's iMac and Epson 4990 had been buried and forgotton.

Let's see... the last time I did B&W was.. I think at least 15 years ago. Silver halide but I have never seen a platinum print yet.

If you have time, and wishes to see some really good silver prints and platinum prints, I will be glad to show some pieces by masters. MIne are very amateurish.

 

[lsisaxon]

Your point is very valid.

If the end point is an inkjet print, then I do not see much value in using film and then convert it to digital file. And that is why for me, when I do color, I will just stick to my DSLR.

However for my B&W, I do have an alternative. An alternative which to me, gives a result that pleases me more. And an alternative which is obsolescence-proof. Which will still go on and on after today's iMac and Epson 4990 had been buried and forgotton.



If you have time, and wishes to see some really good silver prints and platinum prints, I will be glad to show some pieces by masters. MIne are very amateurish.

Sure it would be nice to appreciate some real B&W prints again. I didn't really got serious into B&W because I find the film developing a hassle because of the temperature. Colour film processing temperature is 40degC and it is easy to heat the chemicals up but B&W needs cooling down to 20degC.. Something which is more troublesome. Anyway, the standard lab C-41 process is good enough for me, so I mostly let the lab do the film and I just print.

For colour prints, I did drum processing because of the lack of space in the tiny HDB toilet and although I could use B&W chemicals in the drums, I chose not to because if they weren't washed properly it could affect the colour printing which I do more often. Drum processing is very prone to streaking if the drums were not washed properly between prints. I stopped when the toilet started leaking on the enlarger.. Plus since I shoot a lot of digital now, it means that short of getting laser writers, I have no way to do prints from digital myself. Like yourself, I find inkjet prints somehow lacking.. I still prefer the photochemical prints even though I can't do it myself anymore.

 

[tsdh]

Yeah.. but the idea given by your first article was that the optical density on the film was more continuous than discrete. The fact that the film is made of silver-halide crystals suggests that it is discrete and the tonal gradations are formed when the silver is washed away after the reaction with the dye also has to be discrete.

You are definitely not wrong in that there will be modulation of the thickness of the emulsion after development, I do not dispute that, but I dispute that the intensity recorded is continuous rather than discrete. Which is why I used the term dithering. Even in modern inkjet technology, the droplet size is modulated but that does not mean that the image has become continuous, it is still dithered. Dithering is employed when the number of gradations available is not sufficient to represent a certain tone, therefore, it has to be represented by the average optical density over an area. Film, certainly behaves this way. Otherwise, it would simply become lithographic film, which at this point in time still has the highest available resolution but it is just truly black and white, nothing in between.

Even for digital, the intensity is recorded in steps depending on how many bit is used to represent a pixel. If you used less bits to represent, then dithering will still be required. One difficulty to directly compare, however, is that there is no direct relation to film grains because grain sizes are different for a sensitivity and there is a whole random mixture of different size crystals for the film to be sensitive to different levels of intensity, while in a digital sensor, it is a uniform array of photosites, so in space, it is also different.

ok, I know what you mean.
If you want to compare the continuity of gradation levels between film and digital, it is hardly possible.
The closest possibility is to sample the same area size of both mediums.
For example:
Canon 1Ds' sensor size which is the same as 35mm film, has 4992 x 3328 pixels (monochromatic). Meaning each pixel is 0.0072mm rectangular (7.2 micron).
Then we cut a B&W film at 7.2 x 7.2 micron. It has a volume 7.2 x 7.2 x emulsion thickness.
The size of each silver particle as accroding to Kodak, is between 2 micron to 0.2 micron. But I don't have the information on how many particles are there inside that volume.
The digital using 12-bit ADC will have maximum of 4096 levels, with minimum step of 1 bit. While the film will have minimum step of 1 particle, and the maximum levels as many as the number of particles at that volume (assuming that we ignore the overlapping of particles).

The differences between those two mediums are actually more pronounced on the continuity of image shape. Let see the picture below.
If we cut 4 x 3 pixels, and strike it with light shaped as shown on the first picture.
The film, second picture, will render image shape more faithfully than digital. While the digital sensor, third picture, will result with some incorrrect halftone value due to the averaging of light intensity. This issue will be amplified while using color due to the Bayer masking.
Which one is better? well, up to the photographer, each has its own merit.

 

[lsisaxon]

ok, I know what you mean.
If you want to compare the continuity of gradation levels between film and digital, it is hardly possible.
The closest possibility is to sample the same area size of both mediums.
For example:
Canon 1Ds' sensor size which is the same as 35mm film, has 4992 x 3328 pixels (monochromatic). Meaning each pixel is 0.0072mm rectangular (7.2 micron).
Then we cut a B&W film at 7.2 x 7.2 micron. It has a volume 7.2 x 7.2 x emulsion thickness.
The size of each silver particle as accroding to Kodak, is between 2 micron to 0.2 micron. But I don't have the information on how many particles are there inside that volume.
The digital using 12-bit ADC will have maximum of 4096 levels, with minimum step of 1 bit. While the film will have minimum step of 1 particle, and the maximum levels as many as the number of particles at that volume (assuming that we ignore the overlapping of particles).

The differences between those two mediums are actually more pronounced on the continuity of image shape. Let see the picture below.
If we cut 4 x 3 pixels, and strike it with light shaped as shown on the first picture.
The film, second picture, will render image shape more faithfully than digital. While the digital sensor, third picture, will result with some incorrrect halftone value due to the averaging of light intensity. This issue will be amplified while using color due to the Bayer masking.
Which one is better? well, up to the photographer, each has its own merit.

Oh yes. Definitely, maybe previously I have not made myself clear on what I meant by dithering. At least we see each other's point now and this clearly shows the differences between film and digital and how they actually compare. And I think your illustrations are great.

 

[Witness]

maybe this is the prototype for the s4 5 6 pro hahaha

 

[lsisaxon]

maybe this is the prototype for the s4 5 6 pro hahaha

Must see how long the organic sensor can last first.. got lifespan one...

 

[tsdh]

If the end point is an inkjet print, then I do not see much value in using film and then convert it to digital file. And that is why for me, when I do color, I will just stick to my DSLR.

Some of the film properties does not change just because of scanning.
Example:
There are two pictures below, the first one was taken using Nikon D70s. Output is RAW, converted using Nikon Capture, saved as 16-bit TIFF.
The second one was taken using Nikon F90 with color film, scanned using Minolta ScanDual III, saved as 16-bit TIFF.
Both pictures were shot toward the setting sun, using the same lens. The scanned image resized to match the 6.1 Mpixel output of D70s. Then cropped at the portion with sun in it.
Both pictures still show the properties of each medium; the abrupt clipping of the sun's over-exposed image on the D70s', and the smooth transition to white on the film's image. A linear vs non-linear characteristic of the mediums.
Therefore, scanning the film will still give benefits by retaining some of its properties.

 

[lsisaxon]

Some of the film properties does not change just because of scanning.
Example:
There are two pictures below, the first one was taken using Nikon D70s. Output is RAW, converted using Nikon Capture, saved as 16-bit TIFF.
The second one was taken using Nikon F90 with color film, scanned using Minolta ScanDual III, saved as 16-bit TIFF.
Both pictures were shot toward the setting sun, using the same lens. The scanned image resized to match the 6.1 Mpixel output of D70s. Then cropped at the portion with sun in it.
Both pictures still show the properties of each medium; the abrupt clipping of the sun's over-exposed image on the D70s', and the smooth transition to white on the film's image. A linear vs non-linear characteristic of the mediums.
Therefore, scanning the film will still give benefits by retaining some of its properties.
http://gallery.clubsnap.com/data/500/over1.jpg

Sigh... Compressed NEF.. Maybe I must introduce some highlight compression into the tone curve. Very much the same as a CD recording of a tape recording with the inherent compression.

 

[tsdh]

Sigh... Compressed NEF.. Maybe I must introduce some highlight compression into the tone curve. Very much the same as a CD recording of a tape recording with the inherent compression.

I thought Nikon said that their compressed-NEF is using lossless compression, right?
I agree that highlight compression curve will help to reduce the harsness of clipping. My previous comparison is just to underline that scanning a film will retain most of its properties, even if the end result is digital file.

 

[lsisaxon]

I thought Nikon said that their compressed-NEF is using lossless compression, right?
I agree that highlight compression curve will help to reduce the harsness of clipping. My previous comparison is just to underline that scanning a film will retain most of its properties, even if the end result is digital file.

From what I read, they have some form of dynamic compression, so for highlights, a portion of the LSBs are discarded.

I think at the end of the day, it's because of the dynamic curve characteristic of the film vs the characteristics of the digital sensor and the strengths of film are very evident at both ends of the curve, but not so much evident in the linear region. This is very hard to be replicated digitally unless a lot of headroom is put in. Very much the same for audio recording.

 

[vince123123]

I was wondering, do you have like samples of the same image, taken with a DSLR and then with BW Film, same conditions. Then that image is printed out the DSLR on inkjet, and the BW film print on BW paper.

I would love to see the differences in inkjet vs the silver prints - and better yet, you could help point out to me how the silver print is nicer than the inkjet print. I've never really had a chance to do a comparison or fully appreciate the quality of film, especially under the guidance of a BW Film enthusiast as yourself and would love the eye opener

If you have time, and wishes to see some really good silver prints and platinum prints, I will be glad to show some pieces by masters. MIne are very amateurish.

 

[yanyewkay]

i think it's 2 different things here.. the lossless NEF compression is something like winZip where you compress the file and yet you can recover the exact same file later after decompression.

They don't really do hilight compression in the NEF to reduce file sizes.

 

[vince123123]

There was a thread talking about the lossiness or losslessness of compressed NEFs some time ago:

http://clubsnap.org/forums/showthread.php?t=178784

 

[student]

I was wondering, do you have like samples of the same image, taken with a DSLR and then with BW Film, same conditions. Then that image is printed out the DSLR on inkjet, and the BW film print on BW paper.

I would love to see the differences in inkjet vs the silver prints - and better yet, you could help point out to me how the silver print is nicer than the inkjet print. I've never really had a chance to do a comparison or fully appreciate the quality of film, especially under the guidance of a BW Film enthusiast as yourself and would love the eye opener

1 This is one "study" I will have to do for myself, when I have the time.

2 I NEVER said that silver prints are better than inkjet. I said that they are different. I said that I PREFER the look of silver prints over inkjet prints, having seen what I consider to be very good silver prints and inkjet prints. I said that esthetics are different for different people, and some will prefer the look of inkjet over silver.

 

[lsisaxon]

i think it's 2 different things here.. the lossless NEF compression is something like winZip where you compress the file and yet you can recover the exact same file later after decompression.

They don't really do hilight compression in the NEF to reduce file sizes.

I think the dynamic compression is inherent in the NEF format in the D70. Then the NEF is compressed somemore using lossless compression. I can't remember where I read that from..

Edit: Here's one of the article. http://www.majid.info/mylos/weblog/2004/05/02-1.html
It says that the 4096 levels after the 12-bit ADC is reduced to 683 levels which is only about 9.4 bits.

 

[lsisaxon]

I was wondering, do you have like samples of the same image, taken with a DSLR and then with BW Film, same conditions. Then that image is printed out the DSLR on inkjet, and the BW film print on BW paper.

I would love to see the differences in inkjet vs the silver prints - and better yet, you could help point out to me how the silver print is nicer than the inkjet print. I've never really had a chance to do a comparison or fully appreciate the quality of film, especially under the guidance of a BW Film enthusiast as yourself and would love the eye opener

The quality of the 'black' looks radically different. Say, when you print B&W on colour paper, it also looks different.