The 1000 ppi "standard"??

[David Fairhurst]

Who wins?

Who gets more detections?
Who's detection rate will benefit most from the new equipment?

[antonroland]

Tazman, valid point but I reckon you could have twice your wish list for the same value as Sgt Miller's squad car...

[Tazman]

I don't know how it is in South Africa, but in the US many jurisdictions have no large crime lab. Crime scene and fingerprint services, if they exist at all, are performed by small police departments or county sheriffs offices. In many of those small agencies, the crime scene and fingerprint people are low paid civilian employees who are second class citizens within the department. Their function is seen as a necessary evil by police administrators who view the agency mission as "real police work," not civilian support functions. Because the Crime Scene and ID Unit in such a department is a low priority, it fights an uphill battle each year at budget time. When money is short, new cameras and superglue tanks compete with new guns and squad cars in the budget debate. I have fought those battles in two small town police departments. You want a superglue cabinet? Here's $200, go buy some glass and some plywood and build one.

Not all jurisdictions are like that. Most of the big state and federal labs have good equipment. Many of the big city labs do, too. The fingerprint sections in those labs, for the most part, enjoy new cameras and scanners with great resolution. They have computers with sufficient memory to save all the images they want.

I wonder what percentage of the US population is served by a modern, full service lab and what percentage is served by a small town ID unit without up-to-date resources. Anybody know?

[George Reis]

There are a couple of issues that are in this thread that I thought I'd chime in on.

One potential problem with the 1000 PPI standard is that most scanners do not have a native resolution of 1000 PPI, so when one sets their scanner for 1000 PPI, they are getting an interpolated resolution. This breaks the standard (which should be a guideline, BTW, IMHO), and should be based on native resolutions of available equipment.

It also seems that the guideline should tell us what minimum sampling rate is needed for level 1, level 2 and level 3 detail (easy to determine if we know the size of these features and just apply Nyquist). The guideline could then explain native resolution, and the other factors that affect resolution (many of which have been mentioned in this thread, from optics, mechanics, CCD or CMOS quality, noise, bit-depth, etc.). Sounds like a lot, but it could be done in a couple of short paragraphs. From this information, they could then recommend a minimum, non-interpolated, resolution as a guideline, and anyone reading the document can understand the basis for that recommendation, and how to apply it to their scanner. I think the guideline should also emphasize that it is a guideline, not a standard, and that if a print has been captured at a lower resolution, it only means that there will be less resolution, not that the image is unusable.

George

[ER]

The standard is clear that 1000 ppi is a minimum standard. If 1000 doesn't fit with your printing standards, then use 1200. I typically scan at 1200 because it is evenly divisible by the native resolution of the scanner. I don't worry about the printer, cause I never print the latents. I typically photograph so that the latent fills the shot (while still maintaining at least 1000 ppi). The standard as written also allows for lower resolution if a reason is given for not meeting or exceeding 1000 ppi. If you have a palm and your camera can't capture the whole thing, take one shot of the whole palm at less than 1000, then capture each section of the palm at 1000+.

Even if this remains a 'standard' and not a 'guideline', just use 1200 when you scan (it's virtually impossible to hit an exact number resolution when you photograph), and remember this is a minimum standard.

[Neville]

Hi Tazman
Why not go for a second hand Nikon with good quality lens and less than 10mp, you will be better off than having a point and shot and pay the same.

I still think some are missing the point, it is not about a figure pulled out of thin air by a group of people who maybe misguided. Once you start the interpolation software, information is being created in the image that is not actually there, you then start fiddling with your wonderful Photoshop and this information starts playing a part in the image and is false. Far better by far to use the best quality scanner you can afford and only use the settings at the top scan rate without introducing interpolation. Scanners are not that expensive. Murphys car may just not be able to have leather seats this year.
Neville