Drum scanners - drum size / resolution

[AbsolutelyN]

Does anyone know if there a relationship between the true optical resolution of a drum scanner and the size of the drum being used? Do smaller drums allow higher resolution?

[Bruce Watson]

Does anyone know if there a relationship between the true optical resolution of a drum scanner and the size of the drum being used? Do smaller drums allow higher resolution?

What dictates the optical resolution of drum scanners is the size of the aperture and the size of the individual sample steps. Other things have small effects, including the optical efficiencies of the light path to the PMTs, electrical noise, vibration, etc. But drum size has no effect that I'm aware of.

[bglick]

I fully agree with Bruce.... resolution is a function of the design of the scanner... often you will see larger drums resolve less, but this is not always true, the two that come to mind are the ICG and the Tango IIRC...

[dwhistance]

Some of the Howtek models allow higher resolution on smaller drums, however this behaviour is specific to those models not a general rule for drum scanners.

David Whistance

[AbsolutelyN]

Thanks guys. It is a Howtek in question - the explanation from Aztek is: "5000 DPI will only be an optical resolution when the smaller 4” drum is used. This is due to the physics involved of the larger drum and rotation speeds ECT that make this limitation with the 8” drum". I was just curious if this was specific to Howtek only as I was very suprised this is the case.

[bglick]

> This is due to the physics involved of the larger drum and rotation speeds ECT that make this limitation with the 8” drum".


Argggg.... don't believe everything you hear :-) The Howtek 7500 or Grand had this limitation... IIRC, about 1/2 the optical ppi of the larger drum. In theory, a larger drum has potential to out resolve smaller diam drums, as the film plane has much less curvature on a large drum. The light source and PMT are dumb, they don't know what diam drum they are viewing...

[Bruce Watson]

Thanks guys. It is a Howtek in question - the explanation from Aztek is: "5000 DPI will only be an optical resolution when the smaller 4” drum is used. This is due to the physics involved of the larger drum and rotation speeds ECT that make this limitation with the 8” drum". I was just curious if this was specific to Howtek only as I was very suprised this is the case.

What physics would that be I wonder? I've only seen this limitation with Howteks myself. Still, it's conceivable that it's a drum curvature thing, although the curvature over a 6 micron pixel is miniscule and in all probability negligible. It's more likely a surface speed limitation -- that their processor isn't able to keep up with the higher surface speed and the smaller pixel size. IOW, it's a bit stream speed problem.

That said, the 4000/4500 won't give you a 5000 spi optical anyway. Won't resolve much more than 4000 spi optical because of the minimum aperture that's about 6 microns. This from Aztek's late great Phil Lippincott himself. He published some results under Scannerforum.com (http://www.scannerforum.com/), if you look at the 2002 DIMA Scanner Roundup link you'll find a presentation he gave at some point. IIRC around page 22 he gave results. My browser has problems reading the file and I no longer remember exactly what those results are.

[Jim Michael]

Wouldn't the speed be faster on the larger drum? It has a larger circumference which must pass through the sensor per revolution. Slower rotation speed? Also, interesting that there's no benefit from greater angular resolution from larger drum.

[bglick]

> Wouldn't the speed be faster on the larger drum? It has a larger circumference which must pass through the sensor per revolution. Slower rotation speed?


All correct.... the drums RPM is controlled, so anything is or "was" possible...


> Also, interesting that there's no benefit from greater angular resolution from larger drum.


In theory, there would be....except scanners had a development period of less than 7 years till the market started to cave in.... ending R&D and further developing products. I can't think of too many products that had such a short life span as drum scanners. In the early 90's, they were limited mostly to pre press houses, as the cost was too high. Graphics software was not at a very high level, till mid 90's, which then made high end drum scanners popular with photographers...... but then, a bit of market saturation, and the advent of digital capture, and the products life cycle was over before it hit puberty, a sad reality....

ICG in the UK still makes the best drum scanner in the world... and it has a large drum....very few people have them, so its hard to comment on comparisons, details, etc.

[IanMazursky]

bglick, Im sorry but drum scanners have been around even before the 70's. (http://en.wikipedia.org/wiki/Image_scanner#Drum)
It was direct to film, you scanned on one end and output on the other, no real retouching at all but they were amazing to me.
They were in continuos development until the end of the 90's but Aztek, ICG and one other that escapes me are still in the game.

I have seen those scanners and even worked on a few. They were a marvel then and they rival some of the best drum scanners today. Although they are the size of a large SUV and weigh as much. They also came with a $300-700,000 price tag.

The heyday for the scanner market was the 80's and the 90's. Many, many 1000's were sold. I have worked with some large prep houses that had 10 Howtek 7500's running 24/7 and others that had 10+ Hell 3000 series scanners running 24/7. They made the investment back in under a year in some cases. Imagine that! I even ended up with some of those scanners when they down sized.

On the large drum questions, most of the scanners i have worked on have to run them slower.
If you ran them at the same speed as the smaller drums, you could A. fracture the drum B. Wear out the bearings. C. Damage the end hubs to the point of failure. Thats my theory anyway but its seems to have proven itself out.
One of my 7500 large drums end hubs came off. Years and years of cleaners and overloading by the previous owner had caused it to fail.
Luckily this happened when i was inspecting the drum and not in the scanner. It would have destroyed the drum. I have taken precautions on all my drums to prevent this but it could happen to any drum on any scanner.
I had a colleague that had a Hell 3300 large drum fracture while spinning. He wasn't in the room but it could have killed him. He kept a shard for good luck.

Its the same problem when you overload a large drum. They become unbalanced and vibrate the whole scanner. This can knock them off the bearings especially when you have only 1 clamped area. You can hear the difference between a balanced and an unbalanced drum spin up.
Especially with the Hell 3000 series and to a lesser extent the Optronics, they only clamp on the drive end. There is so much rotational torque when they spin up that the scanner vibrates, too much and its a goner.

[Nathan Potter]

Just curious. What maximum RPM are we talking about with say a Hell, Aztek or other types? Can you ordinarily run them at less than maximum speed?

I'm thinking of speeds like a lab centrifuge say 10 to 30 thousand RPM. There must clearly be a speed at which the data capture rate can't keep up with the data from the film.

Nate Potter, Austin TX.

[IanMazursky]

I haven't used a Hell for many years so i dont have the exact numbers.
I found out that the Tango has a max rpm of 1800.
But the rest of them ie. 3010 to the 3900, I cant find the data.
They were fast enough to scare the crap out of me when i worked with them. Its one of those, make sure your hands and other appendages are inside the car at all times lest they be ripped off.

The Howtek 7500 is 600-1600 and the 4500 is 300-1200 rpm.

If the scanner went even close to the 10k rpm, the drum would shatter. After many years of use they have nicks and scratches that compromise the surface of the drums. The faster you go, the more likely they will fracture.
Its happened even at the slow speed of 1600 rpm. A small scratch can turn into a micro fracture then BOOM!!!