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

[Chester McCheeserton]

Hi all,

I was once a moderately competent operator of these scanners. Recently I've gained access to one again and am trying to work out some rusty kinks. I keep seeing orange/yellow density changes at the edge of the image (scanning C-41 color neg, developed by LTI lab in nyc) and simply darker/more density at the edge of B&W film (mostly 35mm, processed by me).

I know my negs are probably generally slightly overexposed, but they're certainly not bulletproof - I don't remember having to perform this level of surgery at the edges all the time. Any tips? Besides putting down ruby lith tape around all 4 edges? that's the only thing that seemed to help, at least with the 35mm, I've never had to do it before with sheet film, but now I'm thinking might have too. Cropping is not an option. ))

I'm also getting scan lines, like you can see in image #4. Pretty sure they line up with the 1 inch pieces of stretchy tessa tape I use on the bottom of the mylar to pull it tight. But again, I never recall having this many issues with scan lines, it seems like I'm getting them on every 5x7 I do, this last one I scanned tonight at 4000 dpi and there are at least 8 scan lines that will need to be tediously retouched out if I'm going to use this scan.

Thanks
188737188738188740188742

[Chester McCheeserton]

and how it looks on 35mm B&W:
188744

[Mark Sampson]

I've no experience with that machine, but all that comes to mind is what the optical engineers call "stray light". Which can manifest itself in many ways... study your whole setup carefully. could be from some kind of reflection that's not obvious.

[Steven Ruttenberg]

How about a black mask and wet mount.

[Bruce Watson]

How about calling Aztek and asking them? If anyone knows for sure, it would be Aztek.

[Chester McCheeserton]

How about a black mask and wet mount.
these are wet mount.

[onnect17]

...

[Chester McCheeserton]

How about calling Aztek and asking them? If anyone knows for sure, it would be Aztek.

Evan emailed me some tips, he's always very helpful. But I thought I'd try the hive mind here.

[onnect17]

....

[Chester McCheeserton]

My guess also is you are not scanning using SilverFast.


Sent from my iPhone using Tapatalk

That's correct - I was trained to use and was using in this case the DPL software on the Aztek machines. Which is not perfect and has it's problems, but I don't think anyone who has learned it would say that silverfast is better.

I actually prefer good ole epson scan when I'm using a flatbed, silver fast I find incredibly difficult to control...although I'm sure many of you have figured it out.

[Chester McCheeserton]

Outside the box I would check the illuminator and the lamp alignments.

Thanks – that's what Evan from Azrek also suggested. I will try this next time I drive out to where the scanner is.

[onnect17]

....

[calebarchie]

Is this THE aztek Premier aka top of the line 8000 DPI?! I've read that this machine suffers from underpowered light source, people have gone as far as to modify the source with higher powered LEDs for scanning 8000 and smallest aperture.

This is like opposite, too much light or too much in uncontrolled way (re possible fibre alignment issue or other). I think obvious question here is what aperture is OP using?

Perhaps the engineers had this in mind when designing machine, trying to optimise it for use with 8000 DPI and smaller apertures yet still had problems and compromises? Design is about compromise after all.. :)

[Chester McCheeserton]

I think obvious question here is what aperture is OP using?


19

[calebarchie]

19

19, as in smallest or largest? Sorry not that familiar with DPL or howtek/aztek in general. Unless you are talking about 19microns...

[onnect17]

....

[calebarchie]

Aperture isn’t the same than sampling. Scanning with a 19 microns aperture limits the optical resolution to around 1333 dpi. You can save a file at 4000 dpi but still the resolving power will be limited to 1333 dpi.



Sent from my iPhone using Tapatalk

Hi mate, sorry not sure where you draw that from my post. I know aperture more or less directly correlates to resolving power among other things. I said 8000DPI @ smallest aperture aka 3microns which in theory should be maximum resolving/sampling rate of machine which I respect is way above any normal film.

My comments re aperture purely relate to light throughput, sorry for the confusion. If the scans were made at the largest aperture, I suspect issues would be much worse then if done at smallest aperture. I do hope OP gets to bottom of issue (I was looking at an aztek) just curious if it is in line my current conspiracy theory :)

[onnect17]

....

[calebarchie]

I was referring to the OP mentioning the scan been done 4000 dpi.


Sent from my iPhone using Tapatalk

Ahh, apologies for the confusion all good. Just re-read the OPs post for clarities sake too...
Well, do you know what the OP means by "19" since you have some experience with the machine and DPL itself? Would really like to know

[onnect17]

...

[Chester McCheeserton]

Scanning with a 19 microns aperture limits the optical resolution to around 1333 dpi. You can save a file at 4000 dpi but still the resolving power will be limited to 1333 dpi.

I don't think that's accurate. I understand what the aperture does. (caleb it goes from 3 being sharpest to I think 27 being the softest)

I've made thousands of scans and prints with other Aztek premier machines, working closely in collaboration with other artists and technicians testing different apertures on different film stocks for different printing processes...

The fact that you are suggesting using silverfast on this machine causes me to infer that you do not have much direct experience with this machine.

And don't get me wrong, I have absolutely zero interest in debating with you the 19/1333 number claim...I'm sure you can find people eager to debate things like that in another thread.

Even the people at Aztek do not make prints and consider them on the wall...There's a whole thing built into the software that automatically ties the aperture to the dpi setting, anyone who knows what they're doing will manually override that all the time.

I was hoping Lenny Eiger or someone who's actually used this machine would chime in here, but he's probably making scans, not reading stuff on internet forums.

[onnect17]

Trust me, I do not need to post in your thread. Actually, I don’t think you would find another idiot like me, who had such a curiosity in those scanners to purchase each model and I’m kind of familiar with all of them.

And no.
27 microns isn’t the softest. I don’t think you even opened the cover ever.

And no.
I’m not suggesting the use of Silverfast over DPL.

If you’re incapable of understand the difference between sampling and optical resolution then you’re another perfect Aztek customer. I’m sure Lenny (or wherever his name is) will share the alignment tool with you and get you back in business.

Forgive me for my input. I’ll make sure to delete my writings.




Sent from my iPhone using Tapatalk

[calebarchie]

Hi guys, I think it is unfortunate what has happened here, probably got too personal. I don't think deleting stuff is the best way to go, all information and discussion is useful in future especially in such a niche market like this.

Chester, I am reading on the aztek spec page that there is only 18 aperture steps so I am not sure where the 27 is coming from. In that case you would be using it fully open @ 19 (aka softest) along with potential misalignment/lightbox problems I think that may be the root of your problems. Perhaps try scanning at 3microns and inspect difference?

Unless they took out the wheel of your machine and added more holes (with some kind of laser precision) and modified your individual software to suit but that is more unlikely conspiracy theory content ;)

[Chester McCheeserton]

Chester, I am reading on the aztek spec page that there is only 18 aperture steps so I am not sure where the 27 is coming from. In that case you would be using it fully open @ 19 (aka softest) along with potential misalignment/lightbox problems I think that may be the root of your problems.

I don't think so. onnect17 is correct 27 microns isn't the softest, it's just the softest I ever used, and that was only for something like a pure blue sky with no detail. guy I worked with used to scan his 8x10 landscape color neg at two apertures, 16 or 19 for the ground and 22 or 25 for the sky (provided there was no detail like clouds, etc), and put the two scans together.

3 is not usable in my opinion, unless it's techpan developed in technidol, and even for that I found 6 was better if I recall correctly.

[Chester McCheeserton]

Trust me, I do not need to post in your thread. Actually, I don’t think you would find another idiot like me, who had such a curiosity in those scanners to purchase each model and I’m kind of familiar with all of them.

And no.
27 microns isn’t the softest. I don’t think you even opened the cover ever.

And no.
I’m not suggesting the use of Silverfast over DPL.

If you’re incapable of understand the difference between sampling and optical resolution then you’re another perfect Aztek customer. I’m sure Lenny (or wherever his name is) will share the alignment tool with you and get you back in business.

Forgive me for my input. I’ll make sure to delete my writings.


I wasn't digging the tone in your last few posts, it didn't sound to me like you were trying to be helpful. It's fine with me if you want to delete your previous posts.

I have no doubt that you know way more than me about the hardware of these machines. You're right I'm not sure I have opened the cover except to change a bulb or put the drum in and out.

I never really did understand sampling and optical resolution at least in terms of what those exact terms meant in practical plain language. But I do understand the difference and relationship between dpi and aperture on these machines at least.

Not an Aztek customer, never bought a single thing from them. Maybe you make prints to evaluate your scans or maybe you just buy used scanners and tinker with them and start forums over on Rangefinder.

But I did not mean for this thread to go south like this. Many people have been generous and tried to be helpful to me on this forum. I would like to keep it positive and try to maintain the spirit of goodwill that someone like Q.T. Luong had back in the day.

[Chester McCheeserton]

I certainly don't know everything about the scanner. And was merely trying to improve my results, which obviously need improvement. Maybe onnect17 is correct, and I've been doing it all wrong for years. I'd welcome any thoughts others might have. Here's a section of the same neg, scanned at 888 dpi, also aperture 19 on the same day. (not sure why the color shifted, another mystery)

Also these 5x7 negs went through the airport carry on scanner at least 3 maybe 4 or 5 times. Could that be the cause of my edge issue?

188945

[calebarchie]

That is so strange, why are there so many numbers if there is only 18 steps? If anything Im going to download and study DPL user manual. Coming from scanmates, the numbers do not correlate to the micron size and there is only 4!

Scanning at 3 microns is not practical in 90% cases but I just wanted to see if it will make a difference with the flaring results, that could give us more of an idea I think.

So all these negatives you have scanned with flaring were from same batch, which got developed together and/or gone through airport scan? I think it is unlikely, at first I thought developing or camera issue, the issue being on the film itself. Much less likely now, you have shown different formats that have the same issue - I would stick with looking towards scanner.

PS I think attachment is not working

[Chester McCheeserton]

I don't know man. I think the scanmates are pretty different animals, I've never used one.

In my experience, studying the manual will only get you so far.

Similar to reading the manual for an enlarger or a camera vs making prints or pictures.

All my jpegs I posted are the same color negative. with the exception of that one 35mm black and white - that one did not go thru airport scanner.

[Chester McCheeserton]

For instance, when onnect17 mentioned sampling, I think what I'm doing on a regular basis is called oversampling, using a softer aperture then the true optical resolution, because in the trail and error tests that I did, that made the grain structure in the print more appealing, for me at least. I don't think the manual advises you to do this. I'm sure there are other threads somewhere that discuss this, and explain it better then I can.

[Chester McCheeserton]

You have not seen this with any other negative scanned on the same scanner at the same settings?

I have - here's one from a year ago:188957 this also has the lines, and the problem edge. And also went thru the airport scanners. Altho on a different trip. and this was scanned at 3200 ap 19.

Here's one made on a different scanner:188958 still an aztek premier, this was at 2666 ap 19. this one did not go thru the airport x-ray. ( it doesn't seem to show the same issue)

I don't think the dpi or the aperture is causing the edge thing...but I realize there's lots of variables I'm throwing out there.

[mactastic]

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.

[calebarchie]

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

[Chester McCheeserton]

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.

[Chester McCheeserton]

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.

189724189725189726

[calebarchie]

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

[Sasquatchian]

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.

[Chester McCheeserton]

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...

[Sasquatchian]

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.

[Chester McCheeserton]

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.

[Sasquatchian]

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.