Printer experience survey

Hello, I would like to hear about your experience with making digital negatives. I have been using entry-level Canon and I feel that it's capabilities for digital negatives are limited. I do not print often and I am only interested in iron processes so not sure I really need any of the advanced models that are usually recomended. I would like get a general idea of current printer situation (including cheap models!) If you could post yor printer, process and general experience with it that would be appreciated. Maybe you have come across something specific, like, for example, not being able to reset waste ink counter.
I use Canon ix8650. It is dye based. It has usual max density problem of dye inks. I have found two inks that give enough density for cyanotypes and VDB (for VDB it is just enough, doubling exposure time starts producing highlight stains). Main issue with this printer is that all paper profiles when used on transparencies produce colored tints. That is different shades of gray have different color tint. And as different inks have different uv-opacity it leads to extra nonlinarity in correction curve. There are not that many contorls to counter that. I can't even print using black ink only (in photo channel). On the bright side the printer does not clog fast. Color prints with original paper and inks are decent.

vasya1945 said:

Main issue with this printer is that all paper profiles when used on transparencies produce colored tints. That is different shades of gray have different color tint. And as different inks have different uv-opacity it leads to extra nonlinarity in correction curve. There are not that many contorls to counter that. I can't even print using black ink only (in photo channel).

Click to expand...

What is the problem if there is extra non-linearity in the correction curve, I am not able to understand. As long as it is repeatable, correction curve can be non-linear so the print tonal curve doesn't have to. Can you post an example of your curve?

Also, in order to get the best out of the printer in terms of UV density, printing with different paper profiles may not be enough. I strongly recommend using Peer Mrhar's EDN ColorBlocker system. If it comes out to be all black, then fine but it might surprise you that some other color like orange or green is even much denser than just black. I had a HP B9180 for the longest time - it was a pigment printer, but the best UV blocking color was squarely in the green zone that was able to handle even salt prints.

:Niranjan.

Cheap HP printer user experience: dye inks work just fine. Yellow on its own gives a UV density of just under 2. Adding cyan (making green) brings it up to almost 2.4. CMY black barely manages 1.5.

Paper type selection makes a difference, as does colour profile. I find the highest colour density is with glossy brochure paper setting. ProPhoto is the best profile I've found so far.

As with low end Canon printers, the black channel is pigment and isn't compatible with most inkjet films, but it does work with non-waterproof film, if you can find it. The dot size is quite coarse and produces visible grain.

One other nice thing about HP is they let you refill their cartridges. You can therefore fill more than one channel with the strongest blocking colour, which gives you more options.

nmp said:

What is the problem if there is extra non-linearity in the correction curve, I am not able to understand

Click to expand...

Well results are not repeatable enought to finetune the curve to that level of precision.

nmp said:

Mrhar's EDN ColorBlocker system

Click to expand...

I does not remove the tints completely. Paper profiles seem to have more impact in this particular case. I can probably find a reasonably good combination of paper profile and negative color. But this is exhaustive, I basically have to do edn colorblocker runs for all the profiles, (maybe multiplied by other printing options). Maybe change the heuristic that chooses "optimal" color.

I made this thread not about this specific problem but just to get a general idea through what difficulties people have to go with digital negs. I am not ready to ask for help on this issue, need to do more expirements myself.
If you are curious I have attached examples. The curve has to not only be nonlinar, but also not monotone. This could be fixed easily if prints were perfectly reproducible, but they are not. I still haven't figured out the sources of variation.

revdoc said:

Cheap HP printer user experience: dye inks work just fine. Yellow on its own gives a UV density of just under 2. Adding cyan (making green) brings it up to almost 2.4. CMY black barely manages 1.5..

Click to expand...

What process are you talking about? I don't know what these density 2 and 1.5 mean. Like you can measure it with some instrument?
I also used cheap hp with printing head inside cartriges but only for cyanotypes. Density was enough with cmyk black, so I could choose a color that requires less corrections. But I could not see much difference in the prints.

revdoc said:

One other nice thing about HP is they let you refill their cartridges

Click to expand...

Canon too, I am refilling because original cartrides are 50eur and last for like 10 big negatives.

revdoc said:

You can therefore fill more than one channel with the strongest blocking colour,

Click to expand...

Oh, I have hard time holding myself from doing that. I am not sure that the printer can be restored to original state if the trick does not work.

vasya1945 said:

I does not remove the tints completely. Paper profiles seem to have more impact in this particular case. I can probably find a reasonably good combination of paper profile and negative color. But this is exhaustive, I basically have to do edn colorblocker runs for all the profiles, (maybe multiplied by other printing options). Maybe change the heuristic that chooses "optimal" color.

Click to expand...

Looks like the profiles may not be as they should be. Here is something you can try that might mitigate this situation. I would forgo using profiles all together using what is known as "null transform" which is what I do. If you are using Photoshop, first specify Printer Manages Color in the printer dialog. Then in the printer driver, pick Application Manages Color or No Color Management or whatever equivalent of it is in your printer. This way there is no profile used to transform the data before it goes to the printer. This might take out the kink in the tone that you are seeing using a paper profile. Of course, you would have to do your color blocking test under the same conditions as also the actual negative.

:Niranjan.

vasya1945 said:

What process are you talking about? I don't know what these density 2 and 1.5 mean. Like you can measure it with some instrument?

Click to expand...

Those numbers are log density. 0.3 on this scale blocks half the light passing through, 0.6 lets a quarter through, and so on. 2.1 only lets 1/128th of the light through, and it's what I would use for VDB. 1.8 is about right for classic cyanotypes.

vasya1945 said:

Oh, I have hard time holding myself from doing that. I am not sure that the printer can be restored to original state if the trick does not work.

Click to expand...

You can buy clear "ink" to flush the lines and print heads. It's used by people who make screen prints and use UV blocking inks.

revdoc said:

Those numbers are log density. 0.3

Click to expand...

Do you measure it with some instrument?

nmp said:

Looks like the profiles may not be as they should be.

Click to expand...

Some more tests are needed, but it is highly likely that it's the film and not the printer. I turned off the profiles but tinting was still there. In the meantime I came across a blog post with examples of grayscale tables that also showed tinting. And the author ised Epson P600 rather than a cheap entry level printer. He also did some interestings tests to rule out spikes.

vasya1945 said:

Do you measure it with some instrument?

Click to expand...

UV densitometers can measure this density, but it's simpler to just print a commercial step wedge and compare the result visually. For example: