Aztec or Tango scans

[coops]

My local lab has an Aztec photo lab 8000 drum scanner, and charges $40 for a 200 mb scan of 4x5. I have had one scan done so far, but not yet made a print. This is new to me and really have nothing to compare this too. West Coast Imaging claim the Tango is the best drum scanner (of course they would) and charge about the same. I would like to know if you guys have had experience with both scanners. Ideally I would have a scan made with each, but it gets kinda pricey.
My lab is slow, and takes about 7 days to have a scan made. I am not sure about the experience or dedication of the guys who work the Aztec. I live in a college town, say no more. That alone may be good reason to use someone else.
Any thoughts appreciated, thanks.

[Frank Petronio]

Well it is like Doug's post from a few days ago, the skill of the operator is probably the most important factor, even more than which scanner is ultimately "better". And you probably won't know what to compare unless you spring for a scan from both places.

[Ted Harris]

I scan for others and don't use either of those; I use two high-end flatbeds, a Screen Cezanne and a Kodak/Creo IQsmart 3. Frank is right. When you are dealing with high end scanners it is the person operating the machine and how much care and attention they give your scan that counts.

[jetcode]

I scan for others and don't use either of those; I use two high-end flatbeds, a Screen Cezanne and a Kodak/Creo IQsmart 3. Frank is right. When you are dealing with high end scanners it is the person operating the machine and how much care and attention they give your scan that counts.

and it helps to have really good equipment too - signed another flatbed user

[eric black]

That is the same scanner that I have and use. My feeling is the same as those from above- the skill of the operator is going to matter more than the particular scanner being used. I have georgeous results from my scanner that have been proven out in up to 40 x 50 inch prints but I have on occasion not executed a scan to the best of my ability and the results were less than spectacular and a repeat scan with more care was needed. Care and attention are the most important factors.

[Bruce Watson]

Much depends on the skill of the operator, true. Beyond that, much depends on the size of the final print, and on your film, exposure, processing, etc.

I own and drive an Optronics ColorGetter 3 Pro drum scanner with ColorRight Pro 2.0 software for the record. I've not driven any of the Azteks or the Heidleberg drum scanners, but their strengths and weaknesses aren't completely unknown to me as I did a lot of research prior to buying my scanner.

In drum scanning, aperture size is a determining factor in actual optical resolution. The Heidelberg tango scanners have a minimum aperture of 11 microns, which is large for drum scanners. My speculation based on my research is that the small Heidelberg was aimed at the advertising market. When your biggest enlargement is a two-page spread in a magazine, 11 microns can be sufficient.

The Aztek scanners were intended as more flexible scanners for a wider range of duties. As such even the lower end scanners have minimum apertures of around 6 microns, and the higher end scanners make it down to around 3 microns. That is, when push comes to shove, the Aztek scanner will easily out-resolve the Heidelberg. So, if you are making a large print, and your image is highly detailed, you'll get a sharper more highly detailed print from the Aztek scan. All other things being equal.

Then there's film. WCI doesn't like negative film. They used to say this right on their website. I think this is because the Heidelberg software for the tango scanner makes scanning negative film more difficult than scanning tranny film. Aztek's DPL software OTOH isn't biased for either tranny or negative film and can do either with about the same amount of effort. The same is true of the older Trident software for the Aztek scanners.

Bottom line? Like with most things it depends on what you want. Much of the quality of a scan is intangible. You have to make value judgments based on your image and your needs for your prints. And in the end, the only way to find out is to do it both ways, make prints (or small sections of what would be full sized prints), and compare the prints (comparing on screen isn't going to tell you what you need to know to make an informed decision).

[Lenny Eiger]

In drum scanning, aperture size is a determining factor in actual optical resolution. The Heidelberg tango scanners have a minimum aperture of 11 microns, which is large for drum scanners. My speculation based on my research is that the small Heidelberg was aimed at the advertising market. When your biggest enlargement is a two-page spread in a magazine, 11 microns can be sufficient.

The Aztek scanners were intended as more flexible scanners for a wider range of duties. As such even the lower end scanners have minimum apertures of around 6 microns, and the higher end scanners make it down to around 3 microns. That is, when push comes to shove, the Aztek scanner will easily out-resolve the Heidelberg. So, if you are making a large print, and your image is highly detailed, you'll get a sharper more highly detailed print from the Aztek scan. All other things being equal.

Then there's film. WCI doesn't like negative film. They used to say this right on their website. I think this is because the Heidelberg software for the tango scanner makes scanning negative film more difficult than scanning tranny film. Aztek's DPL software OTOH isn't biased for either tranny or negative film and can do either with about the same amount of effort. The same is true of the older Trident software for the Aztek scanners.
.

I actually have an Aztek Premier, 8000 dpi wonder. I have tested it numerous times against the Tango. The Aztek scanner is clearly superior.

There are three specific issues. As Bruce mentions above, the Aztek has a 3 micron capability. This means that it is going to be quite a bit sharper than the Tango. I have tested this with test targets, etc., against the one that Bill Atkinson uses. Suffice it to say, the operator was expert.

The Tango's 11 micron fixed aperture has another consequence. You can't modify it to scan larger or smaller. So if your grain is right around 11 microns you're in luck. The good news is that a lot of film falls into the 10-13 micron category. The bad news is that color neg film should be scanned at approx 19-22 microns. Scanning it at 11 microns will cause sever grain anti-aliasing. The word on the street - when everyone had Tango's - was that you couldn't scan negs. It's ludicrous, of course, you just can't scan them well on a Tango.

The Aztek has another advantage - its software. Standard scanning practice is to make an adjustment with a levels or curve type dialog so that the scan is a little closer to what you would want in the print - than a raw scan would be. The software then scans in raw and applies whatever correction is made after the scan is complete, just before showing it to you. This can comb the histogram, sometimes substantially. Aztek's Digital Photolab has the capacity to load the corrected values directly to the firmware inside the scanner and the scanner scans in the range that was selected, with all the corrections already applied, returning a healthy file every time.

Finally, if you are getting the type of scan that almost everyone gets, partially corrected, or something close to what you would want - this is going to depend on the operator. If its someone running the scanner that's new to photography - or just has a different idea what a good print looks like than you do - you might get a file that's corrected in a way that you hadn't intended.

This is why I have always advocated for working with a smaller shop, someone who will look at your work before they scan your film, and maybe ask you a few questions about what's important to you. And if they scan it in some strange way, they should have the integrity to scan it differently without another charge. This is the way I have set up my operation - these machines are quite sensitive and you have to know what direction you're going in or you will make the wrong scan.

BTW, 200mb's off a 4x5 is a tiny file, takes the scanner about 2 minutes to scan it... Price is reasonable, but I would get a little more, maybe a 4000 dpi scan... But that's just me. I wouldn't want to pay for another scan and do all the work all over again if I decided to make a larger print one day...

Lenny
eiger@eigerstudios.com

[Kirk Gittings]

Standard scanning practice is to make an adjustment with a levels or curve type dialog so that the scan is a little closer to what you would want in the print - than a raw scan would be. The software then scans in raw and applies whatever correction is made after the scan is complete, just before showing it to you. This can comb the histogram, sometimes substantially. Aztek's Digital Photolab has the capacity to load the corrected values directly to the firmware inside the scanner and the scanner scans in the range that was selected, with all the corrections already applied, returning a healthy file every time.

Interesting.....a question for everyone.

I would think that any software worth it salt would apply the adjustments inside the scan delivering a full histogram file, but I have been involved in many debates concerning this very point with some arguing that all scanning software works this way and some arguing that none do. I think, this is relatively easy to detect on 8 bit files where it is easy to break the histogram. For instance run an 8 bit file basically raw and apply significant clipping and a steep curve in PS to break the histogram vs. applying the same clipping and curve in the scanning software. On even an Epson 750 with Silverfast the first example, applying the adjustments in PS will significantly break the histogram where applying the same adjustments in SF will not, delivering a full histogram. Is that a valid test of whether a given scanning software is applying the adjustments in the scan? What else would explain it?

[Brian K]

When you're paying $40 for a drum, or other wet mount scan, think about the time involved and the amount of time it should take. I'm not referring to the file size and speed of actual scanning, but the amount of time needed to carefully clean the neg, wet mount the neg, scan, carefully remove the wet neg, and clean the wet neg. That's a lot of handling of the neg by someone who is only making $40 for the whole process and has to pay off expensive machinery. At this price the only way to make a living doing scans is by volume and the key to volume is speed and spending as little time as possible producing the product. Cutting corners and ganging up scans is the expedient way to save time here. This is not the type of business model that I would feel comfortable in sending my negs to.

This is why I believe that you should always consider the quality of the scan, ability and reputation of the scan operator and have cost as a secondary consideration.

As for Tango scanners, the only experience I have with them was with Nancy Scans ( a big mistake and one that was not consistent with the comments I made above). The scan was of a very subtle and highly saturated color image. They scanned it twice on a Tango and both scans were so unbelievably noisy and artifact filled as to be unusable. The same chrome scanned flawlessly on my Imacon 646. Was the failure of the Nancy Scans the operator or equipment failure? As they scanned it twice and therefore one would think that they paid more attention to it the second time that it was the equipment (Tango).

[Lenny Eiger]

When you're paying $40 for a drum, or other wet mount scan, think about the time involved and the amount of time it should take. I'm not referring to the file size and speed of actual scanning, but the amount of time needed to carefully clean the neg, wet mount the neg, scan, carefully remove the wet neg, and clean the wet neg. That's a lot of handling of the neg by someone who is only making $40 for the whole process and has to pay off expensive machinery. At this price the only way to make a living doing scans is by volume and the key to volume is speed and spending as little time as possible producing the product. Cutting corners and ganging up scans is the expedient way to save time here. This is not the type of business model that I would feel comfortable in sending my negs to.

I think that's what happened to you at Nancyscans. The Tango should beat the Imacon, every time. I hear tons of horror stories from my clients who finally make it here after dealing with a lab. When you consider processing a roll of film for a few bucks, it's all based on piecework, lots of volume. They hire people at the lowest rate to do the processing. Same with the scans, its how they think. Nancyscans is just like this, so is the factory in India.

I have a Masters degree in Photography, and the other person here who might touch the scanner also has a degree in Photography. It's very different when you have someone who understands what you are after with your image doing the work. You aren't going to find this in a factory setting.

Lenny
EigerStudios

[Kirk Gittings]

Lenny, I am interested in your thoughts on what I wrote above.


I have a Masters degree in Photography, and the other person here who might touch the scanner also has a degree in Photography. It's very different when you have someone who understands what you are after with your image doing the work. You aren't going to find this in a factory setting.


While I agree with this point in general, abstracted from an MFA (I have an MFA too, but more importantly I have taught graduate students at the two top graduate programs in the country), the important thing is looking at the personal work or client list of the person scanning, whether they understand the needs of a careful expressive artist. Unfortunately this kind of attention to detail and craft is not high on the priorities of most MFA programs these days.

[Lenny Eiger]

Interesting.....a question for everyone.
I would think that any software worth it salt would apply the adjustments inside the scan delivering a full histogram file, but I have been involved in many debates concerning this very point with some arguing that all scanning software works this way and some arguing that none do. I think, this is relatively easy to detect on 8 bit files where it is easy to break the histogram. For instance run an 8 bit file basically raw and apply significant clipping and a steep curve in PS to break the histogram vs. applying the same clipping and curve in the scanning software. On even an Epson 750 with Silverfast the first example, applying the adjustments in PS will significantly break the histogram where applying the same adjustments in SF will not, delivering a full histogram. Is that a valid test of whether a given scanning software is applying the adjustments in the scan? What else would explain it?

Kirk,
I didn't respond initially as I didn't have a definite answer...

I am very sure (I would have said absolutely positive if I meant so) that none of the other software systems other than Digital PhotoLab actually has that kind of control over the scanner, which only affects the Howteks and Premiers. One would have to know about the hardware registers and how to access them and I don't think that info was published. Phil Lippincott (owner of Aztek, died last year) did a lot of engineering for Howtek before he purchased the company and had a special relationship with them.

With regard to the testing of combing.... I think this is harder to test. This is why it comes down to operator so often. If anyone has a negative and a methodology to follow, I am happy to oblige. However, I have noticed that some of the moves I do in curves in Photoshop, for instance, will comb the histogram. Can I see it on the print? Not really. I don't utilize Levels at all in my workflow. The file is often close to what I want as I have made a lot of the adjustments in the scanner software. I am just moving a little here and there in the different zones.

I think one would have to comb a lot to see it. And I think one would have to have a special image, like one of mist rising off a lake, for instance, to affect the level of sensitivity. I have one client who had me scan things al sorts of different ways, raw included and decided he like the raw scan. I looked at the files and I think he's crazy - I don't come to the same conclusion. Including that not everyone is up to correcting a raw scan in Photoshop of a color neg... How would we judge who's right?

So, it sounds like you have an idea... and maybe it will work. I seem to be jsut rambling around the topic. If you want to send me something, I'd be happy to try something out...

Lenny
EigerStudios

[Ed Richards]

> charges $40 for a 200 mb scan of 4x5

This got lost in the discussion, but assuming this is a 16 bit RGB file, it about 1300 DPI. See what they want for a 600 mb scan, which would be about 2400 DPI, then decide if it is really cheaper than other labs.

[Lenny Eiger]

While I agree with this point in general, abstracted from an MFA (I have an MFA too, but more importantly I have taught graduate students at the two top graduate programs in the country), the important thing is looking at the personal work or client list of the person scanning, whether they understand the needs of a careful expressive artist. Unfortunately this kind of attention to detail and craft is not high on the priorities of most MFA programs these days.

I also had the opportunity to teach - Parsons, Cooper Union in NY and out here at SF Academy or Art. I loved it. Don't get me started on what gets taught today.... I am so tired of work that is purely cerebral, that is about the camera, or the process, and has no content in it at all. I guess I am a traditionalist. I love the work of the greats... I don't know why everyone is so gaga over the new stuff. Give me a Dorothea Lange anytime, a Weston, or a Julia Margaret Cameron, a Stieglitz Georgia O'Keefe portrait. Something with heart...

Printing for others is an entirely different skill. It doesn't matter how good you are at making a print in your own style, what matters is whether you can make a print in the other person's style. You make subtle changes and decisions. To use a California word, it's almost like you have to "channel" the other person while you are working. You certainly have to understand their aesthetic, regardless of whether you like it or not...

Lenny

[bob carnie]

Printing for others, you must be like a chamilion, you need to change your colour or shape with every client.


I also had the opportunity to teach - Parsons, Cooper Union in NY and out here at SF Academy or Art. I loved it. Don't get me started on what gets taught today.... I am so tired of work that is purely cerebral, that is about the camera, or the process, and has no content in it at all. I guess I am a traditionalist. I love the work of the greats... I don't know why everyone is so gaga over the new stuff. Give me a Dorothea Lange anytime, a Weston, or a Julia Margaret Cameron, a Stieglitz Georgia O'Keefe portrait. Something with heart...

Printing for others is an entirely different skill. It doesn't matter how good you are at making a print in your own style, what matters is whether you can make a print in the other person's style. You make subtle changes and decisions. To use a California word, it's almost like you have to "channel" the other person while you are working. You certainly have to understand their aesthetic, regardless of whether you like it or not...

Lenny

[rodinal1]

There have been comments about the key to a good drum scan being the operator. I totally agree with that. I recently found James Beck who has many years of experience with scanning (his company name is Independent Separation). His scans are the best I have seen. His grayscale scans are outstanding. What I really like is if you call James Beck you talk to him and can directly discuss your needs.
After going through a couple other companies that offer drum scan services I know it depends upon the operator.

[Nathan Potter]

Lenny you may already know this but Lippincott filed US Patent 6,459,825 B1 and was awarded it about 2002, I think. It details a strategy of software operation for the Aztek, (mostly the high level strategy) but is interesting. Go to the US Patent Office then to SEARCH and type in the patent number. For a printed copy hit images at the top of the screen. You might find it interesting at least at a high block diagram level.

Nate Potter

[Lenny Eiger]

Lenny you may already know this but Lippincott filed US Patent 6,459,825 B1 and was awarded it about 2002, I think. It details a strategy of software operation for the Aztek, (mostly the high level strategy) but is interesting. Go to the US Patent Office then to SEARCH and type in the patent number. For a printed copy hit images at the top of the screen. You might find it interesting at least at a high block diagram level.

Nate Potter

Nate,
Interesting reading....

Thanks,

Lenny

[Bruce Watson]

The Heidelberg tango scanners have a minimum aperture of 11 microns, which is large for drum scanners. My speculation based on my research is that the small Heidelberg was aimed at the advertising market. When your biggest enlargement is a two-page spread in a magazine, 11 microns can be sufficient.

Mea culpa. I was wrong. The correct answer is that the Tango has an aperture wheel with 25 apertures, the smallest of which is 10 microns. Doesn't change my conclusions much, but there's no point in leaving wrong information hanging out there in a thread like this.

My source for this information is Karl Hudson of Hudson Grafik Services, Inc. (http://www.hudsongrafik.com) The guy who works on and maintains these scanners should know.