CLPEX Chat Board

Any takers??

All help would be appreciated, I always used the seat-of-my-pants method but would like to know if anyone could help me with a precise way of doing this.

Much appreciated

My first question is why do you want to print 1:1?

If it's for comparison purposes, then you may want to consider printing enlargements of at least 3:1 to preserve resolution.

Are you asking how to calibrate an image to 1:1?

If so, there are several methods including using the crop tool - cropping to a known length (ruler in image) across the height or width of the image.
There's using the measuring tool while having the rulers of your image window scaled by pixel and then stretching the measuring tool across a known distance (may require some math and converstion from cm to inch if you use ppi but metric rulers).
Someone showed another method at the IAI conference that they said was so easy they don't use the others. Yet it required as many steps as they do, so I didn't pay a lot of attention.
There are also software packages that incorporate these tools into an easier to use window interface.
You could probably have someone create a script for you that would do the same thing too.

However you do it, I recommend using as much of the scale as you have available to define your distance. With lens distortions and such, you will be more accurate using 3cm as opposed to 1cm.

There are two aspects to this. Making scans and digital photos 1:1, then printing them once they are 1:1. I'll write my approach, with a bit of opinion thrown in while doing so.

Regarding scans. If you scan at 100 percent, then those images are 1:1 already. My suggestion (which is different from SWGFAST) is to scan at a native scanning resolution that provides the level of detail required. For a scanner that has a maximum native resolution of 1600 PPI, the scan will be best at 1600 PPI, 800 PPP, 400 PPI, etc. Scanning at 1000 PPI or 1200 PPI will require image interpolation and some loss of sharpness - thus it is potential that your 1000 PPI scan gives a larger file but less true resolution than the 800 PPI scan. Other factors play a role, but they get too far outside of your question.

Regarding digital photos. Be sure to include a ruler on the same plane as the print, and that they are parallel with the camera back. I suggest filling the frame with the print and a portion of the ruler. Open that image in Photoshop and choose the Ruler/Measure Tool, be sure Rulers are showing (View>Show Rulers), and be sure the rulers are displayed in Pixels (this can be changed by right clicking in Photoshop's Ruler area and choosing Pixels from the drop-down menu). If using CS3 or newer, uncheck the "Use Measurement Scale" box in the Option Bar. Okay, that was the set-up, and if you don't change preferences, or change these settings, Photoshop will always be ready to go with future prints. Now to make the image 1:1. With the Ruler/Measure Tool, measure either 1 inch or 1 cm on the ruler in the photo. The "L1" or "D1" value shows the number of pixels measured. That is how many pixels there are in 1 inch or 1 cm. So, choose Image>Image Size and be sure that the Resample box is NOT checked (this will make sure that your image retains it's full quality and that it gets calibrated to 1:1). Now, set the resolution to Pixels/Inch or Pixels/CM based on the ruler units you measured. Then, enter the length you measured into the Resolution text box. For example, if you measured a 1 inch length and it was 1,203 pixels, the resolution setting should now read 1203 Pixels/Inch. Click OK, and your image is now calibrated.

Printing. I don't see any reason to ever print a fingerprint 1:1, and when you do so, your result will not be as good as the image on your monitor. The same was true with film - printing papers did not have the same latitude as film, they were not a sharp, etc. For comparisons, doing them on-screen provides much better detail, etc. When comparing to ten print cards, just put the latent on screen and hold the ten prints in your hand with your loupe. For AFIS entry, do direct electronic entry. Okay, but you insist on printing, so to get the best print, I suggest using an ink-jet printer that uses quad tone inks (four shades of black). If you can't do that, then a high quality ink-jet printer will be fine (Epson, Canon or HP). To prepare the print, choose Image>Image Size, and check the Resample box. In the Resolution setting, set the PPI to the optimal PPI setting for your specific printer (probably in the 300 to 360 PPI range, depending on the printer you are using). Click OK, and you're ready to get your tiny 1:1 print.

I hope this helps.

George

Steve

Thanks for the response.

I need to print latents at 1:1 at first for comparison and AFIS purposes but the 3:1 is also very handy as we use this level of enlargement for the bulk of our courtroom evidence and presentations.

Will try with the methods you describe and see how I do.

George

Reading your response now, Thanx!!

George

Thanks very much, the comments in combination helped a great deal.

How do I find out what the native resolution of a scanner is? Is it the maximum resolution a scanner will do and then halves and quarters of that?

Yes, the native resolution of your scanner should be the maximum and then halves, quarters, etc. However, you should check the manual, or on-line documentation to verify that you are seeing the maximum optical/native resolution, not the maximum interpolated resolution.

Also, Steve's point about using the longest measurement possible for best precision is a good point that I failed to mention in my post.

George

George yes I agree that was a very valuable point Steve made.

I would be very happy if I could just snap my fingers and learn all the Photoshop tricks in an instant. I see and understand the power of Photoshop but cannot figure out how to do those things. My work schedule also doesn't really allow me enough time to fiddle and learn enough

I have been doing film photography and darkroom work since the early 90's though, so setting up the camera square to the print, getting in as close as possible and all that is practically second nature to me. Therefore calibrating a distance over the greatest known distance made all the sense to me.

As I'm sure I have mentioned to you in the past is that the 5% or so of Photoshop that I do understand I learned editing wedding pics.

The whole world of Photoshop as a forensic tool is only opening up to me now. I think I can now believe that I understand 7% of Photoshop...

antonroland wrote:Steve

Thanks for the response.

I need to print latents at 1:1 at first for comparison and AFIS purposes but the 3:1 is also very handy as we use this level of enlargement for the bulk of our courtroom evidence and presentations.

Will try with the methods you describe and see how I do.

Anton,

Understand that when you print out 1:1 for comparison, regardless what the resolution you send to the printer, you're limited by the printer's ability to render those pixels using a limited number of different inks. As George mentioned, you're really looking at 300-360ppi being represented in a 1:1 latent print. That's why I recommend at least 3:1 if it's going to be your source for comparison. That'll put you in the 900 - 1080 range. Also, different printers like different resolutions. I posted something on this earlier. Using our Epson 2200, a 360ppi resolution image comes out truer than 400ppi in a straight and diagonal line test.

So regardless of what resolution you send to the printer, over a certain point (which is rather low) you're wasting it anyhow.

Steve

Thanks for that. I do my own printing on a Canon Pixma IP 5000 and I have always been happy with the quality but admittedly, as you also point out, no printer can output that resolution in 1:1.

I am also frustrated by the magenta cast I often get in B&W prints and this has made me wanting to go to Epson as their printers' software have dedicated B&W fine tuning pug-ins...if you could call it that.

With all the years I have worked in a proper B&W darkroom I want those exact results in digital and that is not always easy or possible to attain...your thoughts on this?

Anton

While I was away in Lousiville, we had our new computers installed. We have CS3 and I spent much of the last day or so getting things set up. Luckily I have George's book to guide me in the right direction for my General, Bridge, and printing preferences. I've been fighting a bit of a magenta cast to my grayscale images with our Epson 2200. This is a bit older printer now, but it still does a decent job. I've found that I can use the advanced print properties box to pull out some magenta to even things out a bit. I tried taking it all the way out, but it went to green. I found using a -6 on the magenta looked much better. I think there's still some experimenting to be done. Once I get it figured out, I'll be saving the specifications so that I don't have to rely on the memory.

My first recommendation is to pick up George's book and use it as a starting point for your preferences.

One thought - do you have the ability in your printer software to allow Photoshop to manage colors? That's one problem that I was having - it defaulted to letting the printer manage colors. I still get the light magenta cast, but I had a severe green cast before making that change.

Steve Everist wrote:
My first recommendation is to pick up George's book and use it as a starting point for your preferences.

One thought - do you have the ability in your printer software to allow Photoshop to manage colors? That's one problem that I was having - it defaulted to letting the printer manage colors. I still get the light magenta cast, but I had a severe green cast before making that change.

Working on that and

will look into the printer to Photoshop conversation - I have seen something resembling some set-up / preferences dialogue SOMEWHERE in the mix but ignored it at the time because I was thinking of going Epson.

Based on what I have now learned from you it looks like this printer / Photoshop setting might be more important - off to have a looky...

Anton

antonroland wrote:With all the years I have worked in a proper B&W darkroom I want those exact results in digital and that is not always easy or possible to attain...your thoughts on this?

I feel your pain. After a lifetime of learning to do exacting work with film, including formal professional training, working in a pro color lab, and meticulously slogging through the Zone System with 4x5 and 8x10 views, I now tackle digital. Now ICC profiles are in the mix and an (so far) inability to tune a printer the way I used to tune the filtration on a color enlarger. It's pretty clear I need to replace the cheap LCD monitor I bought as a hurried replacement for one that died so that I have an actual manufacturer with tech information and support. For now, I'm at the stage of keeping notes on the color balance adjustments I need to produce correct balance from my printer. All those years of judging the needed filtration for color correction is at least helping me judge when my usual corrections need tweaking before printing.

I found that AdoramaPix makes their production profiles for the four papers they use available for download, and I want to see if I can really get the monitor and software balanced to match the display to their output. I've also been playing with grayscale convert alternatives for black and white by breaking out the color layers and adjusting them. But I still find myself wanting three color controls and a contrast knob on the printer.

On the other hand, the good dynamic range of the better sensors and things like the ability to automatically capture a sequence of three bracketed frames that can be manipulated and combined in the software offer a lot of possibilities.

I have a handy dandy document put together that has the step by step process to calibrate images 1:1 in Photoshop. Its basically the same method George Reis outlined earlier, but in a step by step document with screen shots of Photoshop to guide you through the process.

Its probably the same one that Steve didn't listen to at the IAI conference....tsk..tsk..tsk.. But once you get the process down, it takes about 15 seconds to calibrate a digital image 1:1.

Anyone interested in the document can email me at schriever.andrew@gmail.com and I will reply with the document.

Andrew

or, if Steve wants to post the whole document on clpex.com that is fine also. As long as its OK with George also, since I learned the technique from him and just put it into my own words for a class I taught.

No, the measuring tool version I mentioned in my reply above is the same as what George outlined. The one from the IAI was something that seemed like more effort than I'm currently using, yet was still "quicker/easier than all the rest."