Light Source and Smoothness

I am just wondering if anyone has made comparisons with their printer with several light sources to determine which one gives greater smoothness when printing digital negatives with alternative processes.

This question is of interest to me for two reasons. One, I just spoke to someone who suggested that one might get smoother tonal qualities by using a bank of fluorescent tubes rather than a collimated light source. OK, I have both and already tested that concept and both light sources gave similar results in terms of smoothness with the Epson 2200 and HP 9180.

The other question is, has anyone made comparisons where the useful exposing radiation was limited to certain bandwiths. I ask this question because in some recent tests I made with digital negatives made with both the 2200 and 9180, using a number of color possibilites, including composite black, it appears that my results with BLB tubes are smoother than with both BL tubes and with a collimated light source. The only logic I can see in this, if my preliminary tests are correct, is that the BLB tubes limit effective radiation to between 350 nm (below that glass cuts off UV radiation) and about 420 nm (above that the Wood's filter cuts radiation), which may reduce the contrast between dyes of low and high UV blocking.

Just curious to see if anyone else has results along this line of experimentation?

Sandy King

I had in mind to check out the diverence in tonality and smoothness with different light and I am in the process of setting up that experiment. My NuArc VFC unit has 2 filters that slide into place under the MH lamp. One filter looks black but when the lamp is on puts out fairly intense "black-light" and the filter turns deep purple in color. I've never used it nor the second filter that looks almost like a "skylight-polarizer". I am establishing the standard printing time for the 2 filters then will do print comparison of the 3 settings (2 filtered and 1 unfiltered). Should have that done this weekend.

Miles

Tuning the lights

Interesting. You're basically talking about diffuse vs. collimated though. Don't cold light heads and dichro heads give a softer look than a condenser head in B&W? I was always under the impression that this was a given. I think I even read in Post Factory.
I've been wondering about light frequency in the context of "tuning the light" to find a colour for the negative. I thought different tubes produced by different manufacturers probably would behave slightly differently. Question: Is it possible to change the actual frequency of UV bulbs?
~m

mkochsch said:

Interesting. You're basically talking about diffuse vs. collimated though. Don't cold light heads and dichro heads give a softer look than a condenser head in B&W? I was always under the impression that this was a given. I think I even read in Post Factory.
I've been wondering about light frequency in the context of "tuning the light" to find a colour for the negative. I thought different tubes produced by different manufacturers probably would behave slightly differently. Question: Is it possible to change the actual frequency of UV bulbs?
~m

Click to expand...

Diffuse versus collimated light is one issue. I already tested that and did not find any difference in image tonality and smoothness between the two. This was with light sources know to radiate in about the same nanometer range. And with vacuum frames on both units.

What I am specifically inquring about here is exposure with light sources that have a very narrow radiation band. For example, the BLB fluorescent tube radiates useful light only in the 350-420 nanometer range. Light below 350 is blocked by UV filtration in glass, above 420 it is blocked by Woods' filter. BL tubes, on the other hand, put out useful radiation from 350 nm up to 500 nm.

I am speculating, based on some comparison with the two UV tubes, that the narrow band BLB tubes may give smoother tones with digital negatives than the wider range BL tubes, when printing with the high UV blocking colors needed for alternative work.

Sandy

How smooth?

sanking said:

, that the narrow band BLB tubes may give smoother tones with digital negatives than the wider range BL tubes, when printing with the high UV blocking colors needed for alternative work.

Sandy

Click to expand...

Sandy,

You know Frank Sinatra's voice had smoother tones than Dean Martin's but I can't easily describe or define the phrase 'smoother tones' in that context.


The same is perhaps true when speaking of smoother print tones, that is, defining what is meant or implied by 'smoother tones'.

May we assume the differences you are comparing/observing are very large and glaring or are we spliting hairs here?

Don

Don Bryant said:

Sandy,

You know Frank Sinatra's voice had smoother tones than Dean Martin's but I can't easily describe or define the phrase 'smoother tones' in that context.


The same is perhaps true when speaking of smoother print tones, that is, defining what is meant or implied by 'smoother tones'.

May we assume the differences you are comparing/observing are very large and glaring or are we spliting hairs here?

Don

Click to expand...

Don,

By smooth I mean lack of grain. Ideally you would like a digital negative to print with the same smooth values you find in a LF negative. However, that is not always the case. Instead, you often see a kind of sand/grit. This is not splitting hairs at all as it is something that is very obvious. Just compare results from a green and red negative printed on the 3800 and you will see what I am talking about. Have Don Hutton show you the comparison he made.

Sandy

Variable Contrast Alternative

I think what I proved, to myself at least, when I shot silver paper with both visible (tungsten in the enlarger) and BL tubes (see website) was that frequency indeed was a major variable in which colour blocked best -- which makes sense. The test, so I thought, only worked with silver paper because as far as I knew "alt" processes only develop/print out with UV actinic light. I didn't realise some processes were sensitive up to 500 nm though. If that's the case that puts you in blue-green area of the visible light spectrum. Maybe it's possible to find a mylar or polyester sheet material for masking the fluorescent bulbs. But I don't know if you're going to smooth the grain out of the material right across the board though -- that would be nice if it happened. I'm thinking that you'll just end up using a different range of blocking colours, ones that happen to "print" smoother.
To that end, what I'm seeing in my own negs' these days is that "grain" is the typical flaw near the "black ink" areas. As I examine the RNP-HSB Array in an upward direction the new flaw becomes "banding" or Venetian blinding is some places. So it just becomes a game of finding a printer and ink set which isn't affected so much by either of these problems and ending up with a smooth path or, worst case, picking which flaw you want in your negative. I haven't as yet started playing with lowering the Saturation variable yet in my wedges, which would move the range of negatives colours towards the centre of the cone -- the CMY grey area. There's colours to be mined yet in them there ink carts'.
~m

I will say that it is curious to me that of the digital negatives I have made that looked just like pyro negatives (green-yellow)in colour printed by contact on silver and pd just as i would expect them to print as pyro negatives with the same "creamy-smoothness" that pyro negs are known for. I have not done other colors however. The black pigment negatives were ok and mostly would resolve the grain of the original negative if the scan resolution was turned up that high. The yellow-green pyro-look digital negatives however were always "smoother"... I might add that for me it was probably a serendipitous find rather derived find .


Miles

Sandy, what you are describing is the effective UV density differences between inks caused by the relative uv AND VISIBLE transmission characteristics of the various inks.

The more into the visible range the sensitivity of the process is, the more the visible filtering characteristics of the ink will be a factor.

I've done some testing on this WRT a nuarc unit and the separate inks, but I have not bothered to check other sources.

It does not surprise me that BL lamps will look a bit different than BLB, because with colorized negatives, the inks will respond differently to the visible light, and the portion of the process that is sensitive into the visible spectrum will be affected.

This is all simple color theory. There's nothing complicated about it. What is difficult, is that as far as I know there is no documentation of the real spectral sensitivity of FO (or of the transmission curves of the inks, for that matter) so that the overlap with the transmission of the various inks can be understood numerically.


---Michael

Michael,

I was interested in this primarily from the perspective of colloid spectral printing sensitivity, which is fairly well documented.

Regarding pt./pd. my recollection is that Mike Ware did some research on this and found that the maximum sensitivity of FO was around 370 nm. On the other hand, many people use the SA tubes to expose pt./pd. and these tubes peak at around 420 nm and put out virtually no radiation below 400 nm.

Sandy

sanking said:

What I am specifically inquring about here is exposure with light sources that have a very narrow radiation band. For example, the BLB fluorescent tube radiates useful light only in the 350-420 nanometer range. Light below 350 is blocked by UV filtration in glass, above 420 it is blocked by Woods' filter. BL tubes, on the other hand, put out useful radiation from 350 nm up to 500 nm.

I am speculating, based on some comparison with the two UV tubes, that the narrow band BLB tubes may give smoother tones with digital negatives than the wider range BL tubes, when printing with the high UV blocking colors needed for alternative work.

Click to expand...

This comes right back to what I've been talking about, as far as UV blocking ability. You've got different light sources with different spectra. One ink combination, by pure change, is smoother with one spectrum, another combination is smoother with a different spectrum.

You roll the dice, and sometimes you get lucky. An ink vendor may change the UV characteristics without warning, so another day, the "smooth" light source may suddenly be the "rough" one.

Even UV densitometers won't help all that much: they're useful when the negative's transition from visible light to UV is gradual and has some sort of correlation. This is the case when you're printing a negative only using carbon black, but not when you're mixing in yellow pigments.

sanking said:

This question is of interest to me for two reasons. One, I just spoke to someone who suggested that one might get smoother tonal qualities by using a bank of fluorescent tubes rather than a collimated light source. OK, I have both and already tested that concept and both light sources gave similar results in terms of smoothness with the Epson 2200 and HP 9180.

Click to expand...

If you've got good, high pressure contact between negative and print, then there's essentially no difference with an inkjet negative. Dye inks have colorants in solution, and pigment inks have such fine colorant particles that they can be treated as a uniform thin film of color. Both are bounded by the shape of the dots, which are considerably wider than they are thick. They're not like silver grain film where you a thick, tree dimensional distribution of large blocking particles in gelatin. That is what gives you the Callier effect, where condenser enlargers and point source contact printers give you more contrast than diffusion enlargers or light banks.

wiz said:

Even UV densitometers won't help all that much: they're useful when the negative's transition from visible light to UV is gradual and has some sort of correlation. This is the case when you're printing a negative only using carbon black, but not when you're mixing in yellow pigments.

Click to expand...

You are absolutely right about the limited use of UV densitometers. I have been working for the past couple of weeks trying to find the best RGB values for the HP 9180 and I eventually matched UV densities at about log 2.00 with three different combinations, one mostly green, another mostly red, and a composite black. When I printed them step wedges, however, the green record printed with much less contrast than the red and composite black.

Sandy King

This is what I was talking about regarding the spectral transmission of the inks and the spectral sensitivity of the process. If you know both, the effective density can be predicted.

Since the spectral sensitivity of the UV densitometers do not precisely correlate to that of the process, there will be effective reading errors when you use a densitometer with a negative that has uneven spectral filtration through the effective ranges of the UV densitometer and the process sensitivity.

As I said, it's simple color theory, and for someone with an appropriate lab, could probably be all figured out in fairly short order. But for most people, it's not worth the effort to go through the hassles to figure it out, even for a 'best guess' approach.

--Michael

michael mutmansky said:

This is what I was talking about regarding the spectral transmission of the inks and the spectral sensitivity of the process. If you know both, the effective density can be predicted.

Since the spectral sensitivity of the UV densitometers do not precisely correlate to that of the process, there will be effective reading errors when you use a densitometer with a negative that has uneven spectral filtration through the effective ranges of the UV densitometer and the process sensitivity.

--Michael

Click to expand...

I somehow missed this thread, hadn't seen it til this morning, but I find it most interesting. It touches on some things that I've struggled with in trying to understand the mechanism of the gum process.

I know it is generally believed that the printing spectrum for dichromated colloids peaks at 365 nm or so, and that it is the same for all dichromated colloid processes. But there is some research evidence that indicates otherwise; for example some work by Jorgenson in the mid-20th century that showed that the printing spectrum for dichromated gum arabic is shifted significantly "toward green or even longer" compared to another colloid, probably gelatin. I haven't been able to lay hands on the original research report to see if it is more specific than the summary citation as to the peak wavelength for gum, but it sounds from the citation as if it's not entirely clear that all dichromated colloids have the same spectral sensitivity. The citation added that "spectrophometric curves of the colloids showed that the difference is probably due to the difference in the spectral absorption of the colloids."

I haven't been able to find out the spectrum of the light I use, the EBV photoflood, but I've had the impression that it doesn't give out much radiation below 400, and yet it prints gum beautifully. That doesn't prove anything, since I don't know for sure what the spectral emission of the bulb is, but if true, it might provide some anecdotal support for the research report that gum prints more in the visible range than in the UV range.

What is the spectral range of a UV densitometer?
Katharine

Print Tone Smoothness Relative to Light Source Type

Sandy,

I know this is an old thread but I thought I'd throw in a few comments from experiments I did last winter. I do all of my own work with an Olite AL19 point light source which has a much smaller reflector than the Nuarc 26-1K making the light even more collimated. When I teach, we always use fluorescent light sources. I could not understand why I had problems with graininess in my prints but not when teaching at various locations. The only different variable I could isolate was the light source. So last winter I made a print with the Olite and then made a print from the same digital negative with a fluorescent light source. There was a noticeable reduction in the appearance of grain with the fluorescent light source. This phenomenon is worse with pigmented ink printers than with the 1280, a dye based printer. I can only guess that the dye inks tend to diffuse more in the coating on the Pictorico than the pigments. This especially noticeable with the Epson 2400, 4800, and 9800 printers all of which I have used and found the same problem. Negatives made on the 2200 do not seem to exhibit this problem even though it is also a pigment printer. So the newer technology has improved the sharpness of the prints, but this can be a bad thing when using those printers for making digital negatives. I have even gone so far as to use a small amount of Gaussian blur on the final image to reduce the appearance of grain while not losing image sharpness in the process.

Bob

sanking said:

Diffuse versus collimated light is one issue. I already tested that and did not find any difference in image tonality and smoothness between the two. This was with light sources know to radiate in about the same nanometer range. And with vacuum frames on both units.

What I am specifically inquring about here is exposure with light sources that have a very narrow radiation band. For example, the BLB fluorescent tube radiates useful light only in the 350-420 nanometer range. Light below 350 is blocked by UV filtration in glass, above 420 it is blocked by Woods' filter. BL tubes, on the other hand, put out useful radiation from 350 nm up to 500 nm.

I am speculating, based on some comparison with the two UV tubes, that the narrow band BLB tubes may give smoother tones with digital negatives than the wider range BL tubes, when printing with the high UV blocking colors needed for alternative work.

Sandy

Click to expand...

Katharine Thayer said:

What is the spectral range of a UV densitometer?
Katharine

Click to expand...

Missed this post for a long time.

Peak sensitivity is at about 370 nm, with bandwidth of about 30 nm.

Sandy

bobherbst said:

Sandy,

I know this is an old thread but I thought I'd throw in a few comments from experiments I did last winter. I do all of my own work with an Olite AL19 point light source which has a much smaller reflector than the Nuarc 26-1K making the light even more collimated. When I teach, we always use fluorescent light sources. I could not understand why I had problems with graininess in my prints but not when teaching at various locations. The only different variable I could isolate was the light source. So last winter I made a print with the Olite and then made a print from the same digital negative with a fluorescent light source. There was a noticeable reduction in the appearance of grain with the fluorescent light source. This phenomenon is worse with pigmented ink printers than with the 1280, a dye based printer. I can only guess that the dye inks tend to diffuse more in the coating on the Pictorico than the pigments. This especially noticeable with the Epson 2400, 4800, and 9800 printers all of which I have used and found the same problem. Negatives made on the 2200 do not seem to exhibit this problem even though it is also a pigment printer. So the newer technology has improved the sharpness of the prints, but this can be a bad thing when using those printers for making digital negatives. I have even gone so far as to use a small amount of Gaussian blur on the final image to reduce the appearance of grain while not losing image sharpness in the process.

Bob

Click to expand...

Interesting information, and thanks.

I work with a BLB bank of tubes and an Amergraph ULF-28 plateburner. The Amergraph is a collimated light source similar to NuARc. Definitely less of a point source than the Olex unit you use.

With these two light sources I have not observed any difference in grain appearance with my digital negatives in any of the alternative processes I use.

However, I would not be at all surprised to see more grain with a more pure point source light. After all, you definitely see this in silve printing when comparing a diffuse type light source to a point source light.

Sandy