Home Internegging, Part IV

( Jump to Part I | Part II | Part III )

Back To Great Slide vs. Print Debate

 Home-internegging in the previous rounds has shown to produce results so good that they can even fool a picky eye. But, how good is it, really, as compared to photos taken directly with print films?

To answer this question, I also duplicated the example photos shown in my Tale of Two Films, and made prints from these negatives. I only chose the first two photos since the last two examples in that comparison are used to show typical lab errors.

Again, the following are the side by side comparisons. The left is the original photos made from print films (Fuji New Reala, ISO 100), and the right are the photos made from the negatives duped from the slides (slide: Fuji Velvia, ISO 50; duping print film: Fuji Super G+ ISO 400).

My Observations:

  • First of all, the results are very similar, the overall difference is within the batch to batch variation of a lab.

  • Colors of the photos made from duped interneg are slightly less saturated and duller than that from original negatives. There could be several reasons: first is the film difference: the original photo are taken with Reala, which is quite saturated; the duping used the Fuji Super G+ 400, which is inherently less saturated (but the slides are made with Velvia, which is very saturated). I'll try to dupe into Reala later to see what the role the duping film plays. Anyway, color accuracy is still a big mystery in color photography. Second, flashing probably has washed out some highlight areas. Third, the flash light source has slightly higher color temperature (blue-tint) than daylight, thus makes the warm colors appear less saturated.

  • Close examination indicates that there is slight loss of sharpness of the photos made from the interneg. Again, this is not readily observable and is within the lab's tolerable focusing errors. Enlarged small portions of the above image scanned at 300dpi are shown in the following. Again, the left is the original photo, and the right is the photo made from interneg. Shown area is about 0.5 inch height in the photos. (NOTE: By saving the image to JPEG format, there are some lost of details. Unfortunately, due to the limit of my on-line disk space quota, I have to use this format.)
    For 4"x6" prints, such loss of sharpness is acceptable even for picky people. Feasibility for larger enlargements will be investigated later.

  • Overall, the aforementioned differences can only be observed when you are extremely picky, AND you are presented with two photos to compare. Otherwise, ....

Another Attempt

 In order to achieve a better color saturation in the interneg as compared to the original photos using print film, I attempted to duplicate the above photos, along with many others, into Fuji Reala, the same films used in those original photos (but not the "new" Reala, if there is any slight difference to note). The exposure is: +3EV compensation for normal exposure and -5EV for flashing. And, to ensure the best results, I used Kodak Premium Processing this time.

Although KPP does not necessarily guarantee the best possible results, one thing that is very comforting to know: on the back of the photos, the printer imprinted the exposure and color correction values used during the printing (PopPhoto did an explanation of these data early 96). And for all photos made from my duped negatives, no corrections on either color or exposure were made. This is an assurance that the exposure values I used are right on the mark!

The results of this batch, in terms of color saturation and sharpness, are quite the same as using Qualex, except a slightly different color cast. Color saturation is slightly improved. But the improvement could also be possibly due to slightly yellow green cast. Overall, the batch-to-batch color variation essentially overwhelmed the difference. The contrast was slightly lower, probably due to the characteristic of Reala. The following are the comparisons. Again, the left are the originals, and the right are made from interneg.

In the negatives, the -5 EV flashing leave a barely observable gray exposure. This indicates that in the previous batch (Round 3), the flashing is still overexposed slightly. And, it appears that, as long as the flashing leaves observable mark, the difference between -5EV and -2 EV in flashing is not significant, and this level of flashing, the contrast is the closest to the photos taken by the print film. If increasing the contrast is the goal, the flashing level need to be further reduced until no observable marks on the negative.

Conclusions

 Home-internegging is doable, and can be done cheaply. Results on the 4"x6" prints are almost as good as photos taken on negative film, and even a picky eye would have hard time to tell. Quality difference is within the range of variation of lab's processing.

For those who want to try this out at home, here is the summary of the procedures:
  • Hardware requirement:
    • Can achieve 1:1 magnification or slightly higher;
    • Optionally (but almost must) can perform multiple-exposure;
    • Optionally (but almost must) can perform TTL flash metering.
    • Optionally, can achieve variable magnification continuously from 1x to 1.5x if cropping is desired.
    I use Nikon N90s camera, which has TTL capability, fitted with MF-26 data back to provide multiple-exposure capability. Nikon Speedlight SB-26 as light source. Lens: Micro-Nikkor 55mm f/2.8 AIS fitted with PK-13 extension ring gives 1:1 reproduction. I further fitted with BR-2A and BR3 coupled together as a 20mm extension tube to achieve higher magnification. Use Nikon's slide copying attachment ES-1 to hold the slide in place. The set up gives continuous 1x to 1.2x magnification.
  • Film: any generic film; the finer grain the better; the higher color saturation the better; the lower contrast the better.
  • Exposure: use camera in manual exposure mode, but TTL flash metering. Shutter speed set at the highest sync speed. Aperture: use the best small aperture of the lens, such as f/11 or f/16 (and stay away from the smallest). Normal exposure: set tp +3 EV exposure compensation, one more stop if the slides is very thick. Flashing: set -5 EV exposure compensation.
    Note: these values are good for Nikon N90s + SB-26. For other cameras, use above values only as a reference, and repeat the experiment by yourself to obtained the best results. Bracketing by full stops. (If you tried out, I'd also like to hear your exposure data.) Still, the +3EV compensation remains unexplainable for me.
  • Things to watch out:
    • Best done at night when the ambient light is low. At the -5EV compensation, an ISO 100 film effectively becomes an ISO 3200. So, watch out the level of ambient light!
    • Focusing is critical, as well as using a small aperture!
    • N90s can warn a flash underexposure, but does not warn a flash overexposure. So, for flashing exposure, need to defuse the flashlight using one or two sheets of white paper to ensure the flash is weak enough for the camera's TTL system to respond.
    • If the camera cannot perform multiple-exposure, use an extremely low contrast film, or consider the interneg film;
    • If the camera cannot perform TTL flash metering, need to perform a similar experiment to determined the flash to camera distance, using the above exposure data as a reference.

Now I can go back shooting slides. My next project: slide dupin'.

Back to my Film Study Page
Last updated: July 19, 1997.