Pakon Scanner

[jesus.stdm]

Hi guys,

Anybody who owns a Pakon and a Plustek 120?

I own a Pakon and i want to know if the Plustek worths the money and you recommend the upgrade.

At the moment i only shot 35mm and pano with Xpan but i want to be ready for bigger negatives just in case.

Thanks a lot!

 

[Pete B]

Yes, it's worth the upgrade for larger prints.
Pete

 

[pbo]

Just a heads up - someone on the FB group found a way to scan positives using PSI!

 

[jesus.stdm]

Yes, it's worth the upgrade for larger prints.
Pete

Thanks Pete.

At the moment i am able to print A4 with decent results (for my eye) using the Pakon scans.

My main problem is that Pakon gives me very confusing results with the same rolls. Some images very contrasty, others very flat... It's frustating and i don't know if it is due my development process or the scanner.

By the moment, i have ordered a Primefilm XA in Amazon (a.k.a Reflecta RPS 10m in my country) and i will be playing with it for comparing with the Pakon scans.

If i get the same, i will keep the Pakon and if i get consistent results, i will return the Reflecta and buy a Plustek 120.

 

[joeswe]

@Jesus

This might be a problem with your negatives --- even the best scanning software has its limits when it comes to counterbalancing the effects of gross errors in exposure and/or development.
Just buying new equipment is rarely the solution. Why not post some sample pictures here of the scans in question? How do the particular negatives you have problems with look like when you inspect them visually? Very thin or very dense? This should help to pin down the problem.

And one more thing: Can you please verify if by any chance you have activated the "Portrait" option check box in the "Setup"->"Scanner" dialogue in PSI?

 

[edge100]

Does anyone have this issue where they get a light band on the edge of the frame?
edgeband by WilliamJosephJefferson, on Flickr

I don't own a Pakon anymore, but yes, I did have this issue. Always on B&W and always on very dense negatives.

No clue what the cause is or what the solution is.

 

[alibosworth]

I put together a script that eases the pain of using 16-bit planar output from the Pakon 135, for people that want to capture as much data as possible using this hardware.

It does two things automatically to a directory full of .raw files:

1) Checks the file sizes and guesses which resolution you scanned and and whether you exported with or with headers. Based on this information it processes the raw files to 16-bit tiff files with ImageMagick. These will be inverted and dark and can processed with an external tool like VueScan or ColorPerfect if you want.

2) By default it runs these files through negfix8, which will give you a directory of color balanced images that look relatively normal, generally including a lot more shadow/highlight detail than the default 8-bit output from either PSI or TLXClientDemo, ready to import into lightroom, etc and batch-apply some contrast or curves to.

Here are some examples of the default output vs the scripts output: https://alibosworth.github.io/pakon-planar-raw-converter/comparison/

Here are some images showing the problematic quantization if you use PSI's 8-bit raw output: https://alibosworth.github.io/pakon-planar-raw-converter/8bit_raw_highlight_issue/

 

[brbo]

Thanks, alibosworth!

Do you happen to have PSI/TLX output files for the two shots where you demonstrate the limitation of the 8-bit PSI "raw" output. I downloaded one raw file and went through my regular ColorPerfect/Lightroom routine to arrive at this:

(full res tiff here)

I would love to compare it to the PSI interpretation.

I quite like the "minilab" style scans, but I'm afraid I'd quickly be looking for something more in my scans (well, that actually already happened years ago and now I have a number of scanners (among them (two) Minolta 5400 and a drum scanner)) at which point Pakon really stops being all that great. Just the time spent scanning the negative isn't that important any more when you do manual inversion and additional PP on the file. I see it's awesome for previews and to give you an idea or a push to be a bit more bold in PP. I mean, who need all the details in the highlights 😉

 

[alibosworth]

Hi brbo, that looks quite nice, and not too far from what I got via Vuescan batch processing of the 8-bit raw tiff file (which is what I did before I had made that script).

Here is PSI's default output, no adjustments before or after exporting:

(full download)

Here is my 8-bit raw tiff -> Vuescan processed file, no adjustments except right clicking in Vuescan to set the whitepoint:

(full download)

Before I got the Pakon in 2014, I had been scanning since 2006 with a Minolta 5400 also, using more or less the "advanced workflow suggestions" (locking down exposure and film base values after calibration on unexposed film) in Vuescan, with the addition of custom film profiles via Wolf Faust targets, which is why I found the default output from the Pakon software so limiting.

I agree that extra highlight/shadow detail isn't important if you have other tools at your disposal when you need them. As for the highlight issues because of quantization in 8-bit raw scans, 95% of the time they are not even noticeable, as this image shows (both your and my processing from the 8-bit version doesn't show them) but they led me down the path of wanting to optimize for the quickest workflow starting from the 16-bit file.

Do you have your regular CP/LR documented anywhere? If you do that on 5400 scans are you saving linear files out of the Minolta software, Vuescan, or something else?

 

[brbo]

Do you have your regular CP/LR documented anywhere? If you do that on 5400 scans are you saving linear files out of the Minolta software, Vuescan, or something else?

I make a linear scan with Vuescan. Then ColorPerfect and Lightroom. I used to lock the exposure on the unexposed part of the film but now I just lock exposure at 1.5 or less for negative film. I don't think you could actually clip the data of negative film on Minolta 5400. I usually set a low exp. time and disable IR channel, this way the scan times are quick - less issues with film curvature.

 

[alibosworth]

I usually set a low exp. time and disable IR channel, this way the scan times are quick - less issues with film curvature.

Tell me more! Film being not held flat is one of my big issues with my 5400!