Aztek 8000 Premier, orange/yellow cast near image edge + scan lines

I feel like the red highlight around the image is give away to some kind of light leak or possibly a collimation issue in the optics.

If the colour bleed was blue/cyan, I would have thought it to be a processing issue.

Chester,

Please attempt to follow through with the suggestions being made both here and on the yahoo group. You need some kind of order and process of elimination to get down to resolving this issue, otherwise you will be overwhelmed to too many variables.

The 35mm film scan is the outlier here which has also not gone through an airport xray yet exhibits same symptoms. This can rule out the camera or processing (eg film holder or pressure plate issue) as it is extremely unlikely that these very specific issues would pertain across multiple formats and cameras.

Again, unless you inspect all this on very closely lightbox as originally suggested (the B&W neg should be easy to see this) or scanned same negs on another machine you cannot confirm anything yet. We are want to help you, but you are throwing us too many bones! Please do check the yahoo groups as a lot more specific instructions and information has since been posted there by people who deal with these machines on a daily basis unlike people on a forum like this.

Bests

Quote:

---

Originally Posted by Greg Davis

Have you considered the possibility that it is flare from the edge of the film holder while in the camera?

---

it could be...I can't recall if this was a holder that I carried through airport carry on x-ray while it was loaded, or if I loaded it after carrying the box through. My colleague who uses the drum scanner more than me, (although mainly only for black-and-white) also thinks that the issue is in the film, not the scanner.

OK, finally made it back out to the scanner. Spent an entire afternoon carefully re-reading this thread, and the one over on yahoo, and the email that Evan from Aztek sent me.

Like fiddling around with an enlarger – it's sort of a mixed bag. But I did improve my results somewhat.

Checking the alignment of the light source is confusing for me, I can see that the beam for the transparency is centered over the hole where it needs to be.

I thought that the bulb had been changed recently, but if it had, whoever did it had not bothered to reset the bulb hour counter. So I went ahead and popped in a fresh bulb. Again, opening up the bulb door on top of the scanner, both lights appear to be working and pointing where they should but it's still unclear to me exactly what I'm looking for to check if the bulbs are 'aligned'.

Opening up the scanner door, I followed Aztek's instructions for cleaning the FORI unit and lens. For anyone else attempting this and reading this thread, it took me 40 minutes to realize that the FORI unit can be pulled all the way off after loosing the allen screw on the left, at first it seemed like the fiber optic 'arm' was blocking it from coming off, I didn't realize that the arm simply bends back and is on a spring. The FORI unit was filthy from what looked like hundreds of hours of melted tape thwacking against it, sticking and adhering to it. I removed the worst of it, but because the unit is still attached to the wire and the gunk covering it was so extensive, I was unable to clean it totally, and not sure it would even be possible were one to remove it totally from the scanner interior.

I did carefully clean the lens behind the FORI (whatever that stands for) unit. It did have some dust and possibly a little haze from splattered fluid on it but it didn't seem that bad. I made sure to clean any big loose stuff from the front of the FORI unit that was anywhere near the opening where the light passes through into the lens. Poking around inside with a flashlight while the carriage was moved way over to the right revealed that two loose pieces of tape were hanging out in the rear top area, who knows how long they had been there. I removed them.

You can see that my results scanning the same 5x7 color negative are much improved although the edge flare is still there somewhat. I'm also getting way fewer scan lines (thin hard yellow lines in direction of drum spin) but am still getting some. (2 as opposed to 8 before) My colleague scanning 8x10 black and white at 2000 dpi aperture 13 is not getting scan lines. I made a test using aperture 16 and looking at it, confirmed what I already knew, that it is sharper, but that the noise and grain is less pleasing for enlarging then aperture 19. (for 400 large format color negative film anyway)... So the new one I'm posting below was made at Aperture 19 and 3200 dpi. Thanks to all of you who made suggestions.

1. old scan originally posted vs new scan after changes. 2. both new scans, one at aperture 16 and one at aperture 19 showing why I would intentionally oversample and make a slightly less sharp scan in exchange for less digital noise and more smoothness in the sky and highlights. 3. same thing but also showing the scan line that I'm still getting. I can see the subtleties are getting lost in these jpegs, but I know from experience that when enlarging up to 40 or 50 inches on the short side that less color noise and reduced appearance of grain is preferable to me. If you are making a 17x22 print or simply scanning to pixel peep on screen you might prefer 16 for color neg. For black and white or chrome, I would suggest aperture 10 or 8. Since we got into that.

Attachment 189724Attachment 189725Attachment 189726

Good to see making progress, appears mostly to be FORI lens (input to PMT)

BTW:

FORI - Fibre Optic Ring/Reflective Illumination
FOTI - Fibre Optic Transparency/Transmission Illumination

When I got my Howtek 8000 in 2001, it exhibited the flare from the rebate edge inward. I went around and around with Phil Lippincott who swore to me that it was the fault of Trident software because it only scanned in Log, in combination with an out of alignment optical system. I had Aztek align the optical system and replace the main bearing for the drum carriage and then bought and PC and a copy of DPL. After a month of side by side testing it was pretty clear that DPL was actually worse than Trident in regards to the flare, but according to Phil, it wasn't really flare but more of a smearing from the rotational speed of the drum combined with the latency of the PMT's that caused this. The optical alignment helped some but the only real cure was to tape around the edged of each scanned frame. At first I used 1/4" silver mylar tape and when that got harder to find I went to straight black electrical tape, taped over the mylar overlay sheet. I've only had to do that with color transparencies, not any neg film, color or black and white, but if the d-max of your rebate edge was high enough, I could see a need on color neg. It seems that this phenomenon only appears when a certain threshold between adjacent d-max and d-min is reached and blocking off all the light surrounding the frame effectively stops this. You almost never see any of this flare/smear within the image area as there is almost never enough differential d-min/d-max ajacently to cause it.

The lighting system in the 8000 in not as refined as in the 4500 (no extra lens to focus into the fori) and that 3 micron aperture doesn't help either but that's only used for true 8000 ppi scans. Interesting bug in Trident that I discovered is that when you set the resolution to 8000 Trident defaults to 6.35 microns. You have to use the manual override to get the smaller aperture and the highest res for those films that can use it. According to John Panazzo, that was on purpose to keep unknowing scanner operators back in the day from getting scans that were too grainy but it also explains why those Aztek resolution tests show the Howtek to be lower than the Aztek at 8000. Ha.

And for those manually setting their apertures, you can ONLY go larger than what the Auto setting gets to, never smaller, or you won't get the overlap that's built into the rotation of the drum and end up with what looks like rows that are slightly lighter and darker.

Huh, interesting Sasquatchian. I haven't scanned any 35mm film since I did the FORI cleaning to compare but I was getting real bad flare from the sprocket holes in both b+w and color neg a few months back when I made the initial post.. And yes when I used to run one of these scanners on a regular basis I would tape the sprockets holes on 35 mm slide film...never had to on 120 or sheet slide film though.

Appreciate you chiming in with your experience. I used to use black electrical tape too on 35mm black+white negs that had sky or continuous tone near the edges...

Chester. I first noticed this on 35mm sprockets holes "flaring" well into the frame. Retouchable, yes, but a pain in the ass. I went back at one point and looked at drum scan we had done in the mid 90's on Hell 3010's and Hell 3300's and they both exhibited a tiny bit of the flaring as well. Phil and I went round and round on this with hours of conversation on the phone but the best even he could offer was to tape off the film. He did acknowledge that the light system combined with the small aperture was the source, which makes sense. And remember that these Aztek Premiers and Howtek 8000's are about three times as fast for the same scan as the previous Howtek SM4500, which I had previously and never had to tape anything off. Or maybe it just never showed enough to matter. I have talked with Evan from time to time about adapting the lens from the 4500 that focuses the fiber optic light into the fori over to the 8000, which would help to address some of the issues. We also talked a long long time ago about switching out the 931B's with a higher sensitivity lower noise alternative from Hammamatsu. They were available at one time for about four times the cost of the 931B's but I never got around to actually buying and testing.

Ahhh....Yes it is actually a ton of work to get rid of the orange flaring cleanly without just cropping it. Wasn't aware of the speed difference of the earlier howtek model....good to know if I ever purchase one of these for myself. Taping seems to be the way to go for now, at least for 35mm...I was getting the flare even with taping on the initial post, but will try to test the same neg and see how much improved it is since I addressed some of the issues...appreciate the input.

Maybe this is part of the reason there were so few drum scanners that went down to a 3 micron aperture. I remember seeing ads for the Howtek 8000 over twenty years ago touting the speed and resolution. All the good stuff and none of the bad. Maybe that's why Hell limited the aperture stops as well, I don't know. At least we seem to have a workable solution for the time being. It's one that I'm pretty much going ride into the sunset. While scouring the internets a few years ago I came across a very interesting Howtek advertisement, talking about their new HR10000 scanner. They (maybe Phil) had apparently worked on developing at 10,000 dpi version of the 8000 but according to Evan, none were ever produced. That one would have a minimum aperture of 2.54 microns, getting in the range of some of the smallest usable pixels on digital sensors.