DPL Scanner software from Aztek for12 bit capture drum scanners

[onnect17]

Not sure if it's the lack of technical background or a skillful sales rep but I noticed a few Premier owners/fans repeatedly overstate the scanner and software capabilities and I want to make sure members are aware.

Let me start saying I enjoy working with the howteks. There great scanners.
All the models acquire data at 12 bit resolution. No 14, or 16, just 12. Nothing wrong with it. No matter what format the data is converted later still 12 bits are significant. This is clear in the specs in the back of the Premier User Guide (part number AZM092-04 Rev.B), section 7-3. No so clear in their web site. The only way I can imagine to expand the dynamic range is to get a second scan with the log mode enabled and merge/tune the files later.

Another claim is related to the access by DPL to the A/D converter (gain, voltages, etc.). All the parameters are controlled by the VDC module, not the app, and are set by the firmware during calibration. The only options to the app are to select the Log amp (instead of linear) or to populate the LUT. There also an overwrite to change the white point via the console in some models.

Regarding the number of apertures. The Premier user guide correctly lists the 17 apertures. Stills the specs in the web site shows 18. Same thing with the scanning area. The manual reads the correct scanning area (11.5"x12"). The site shows 12"x12".

I hope somebody in Aztek will take the time in some point to correct/clarify the info out there.

[Lenny Eiger]

The number of apertures is meaningless. There are only 4 or 54 of them that are useful.

Most film can not be scanned at 6, min is about 8. That leaves 8, 10, 13, 16 and 19.

Lenny

[onnect17]

I agree but that's not the point. I am referring to the inconsistency in the specs. Not reason for having different info between the Users Manual and the web site.

Taking about apertures, I had a similar experience with the first d4000 I purchased. 6 microns was by far too noisy so I used 13 microns and the effective scanning resolution is 2000 dpi instead of 4000. Nowadays is a different story. Power and alignment of the light path, updated model and condition of the PMTs got the Howteks where they should be. Even the 3 microns with the premier can be used.

Another part to check in the case of the HR8000 and the Premier is the aperture wheel. It's the perfect place to grow mold (dark and cold) and for some reason the rough surface in the wheel does not help. I remember the first time I pulled it out, looking under the microscope, I could see half of the 3 microns aperture blocked buy the mold. I am not saying that every scanner out there has some growing inside but it should be inspected.

[sanking]

Another claim is related to the access by DPL to the A/D converter (gain, voltages, etc.). All the parameters are controlled by the VDC module, not the app, and are set by the firmware during calibration. The only options to the app are to select the Log amp (instead of linear) or to populate the LUT. There also an overwrite to change the white point via the console in some models.

This claim definitely needs some clarification/proof from a person with the technical expertise to offer real insight.

Sandy

[onnect17]

This claim definitely needs some clarification/proof from a person with the technical expertise to offer real insight.

Sandy

Sandy,
I also would love to hear also the details from the guys that wrote the code, but more for the corporate drama associated with projects of this kind. That goes back to early 90's.

The reason I think the VRef in the ADC should not be touched after calibration is based on the working principle in which the design of the hardware is based. That includes a chain of at least 6 blocks, many not perfectly linear, including amplifiers and PMTs, which gain and offset is adjusted during calibration and recalculated/adjusted at the start of each raster line. Every step in the algorithm is detailed in patent 5'237'172.

When you run the FST app in the 4000 or 4500, the software needs to load a totally different firmware in the scanner to be able to handle and test each component directly.

If you are really, really interested in knowing, I can manage to wire the VAA and monitor the VRef.
Just for you, Sandy. :-)

[sanking]

Hi Armando,

Thank you so much for your tests of these Howtek systems. And I am interested in knowing, but unfortunately the electronic issues are well beyond my pay level!!

At this point I am still running the Howtek 7500 with Silverfast 6.6 on a Power PC MAC. I am happy with results but would consider DPL if I thought scan quality could be improved. So I pay close attention to your advice.

As for the claim in question, I have long suspected it could be similar to the beauty of the emperor's new clothes.

Best,

Sandy

[Lenny Eiger]

The only person who understood how everything was put together, Phil Lippincott, died a few years ago. His knowledge is gone.

I have done very well by DPL and people apparently like the scans, often saying things like they haven't seen scans from other places that were as good. Could it be the way I set up a scan, this piece of hardware, the software, are they just being nice (or hopeful)? Any of these things could point to this kind of comment.

I don't think you will find your answer. Aztek has done one thing very poorly, and that's to set their pricing on the high side. This pisses everyone off, and instead of looking for what the scanner will do (or the software) people have a tendency to look at what it won't do, or isn't doing. This
type of unhook this or unhook that to see what happens, to reverse engineer the thing, might be a fair process if it were done by people who were even-minded, vs looking to discredit.

You can not underestimate the contribution of a man like Phil Lippincott. He was a bit of a curmudgeon, to be sure, had strong and unpopular opinions, and was sometimes hard to understand, but his work goes all the way back to the MacBeth color chart, and his engineering can be seen, in some way, in almost every drum scanner. We would not have the kind of color control that we do now without him. He put his heart and soul into the building of the Premier, and the DPL software. No expense was spared. DPL was made to be amazing, as amazing as Phil knew how, and the contention that they just lied about the CMS feature, for marketing purposes, is not reasonable, at least not to me.

I am certain there are things in there that aren't being noticed, little choices that were made to adjust this or that part of the input, that don't makes sense, but just work, might be quite hidden, and do the job they said it would. Nothing is perfect, but every other software looks pretty simplistic by comparison to me. So, my answer is to give these guys their due. No, not everything, but at least most of it...

Lenny

[onnect17]

I never had the pleasure of meet him (Phil) but I am sure, based on his forum posts he liked the technical side and would enjoy to clarify many details in the scanner design without revealing too much. In that sense, we owe him to continue trying to improve the scanner in any way we can. I don't blame the guys in Aztek. They have a real business to manage and doing it good enough to survive when so many others are gone.

In the only part I disagree is "no expense was spare". Unless is Defense or Academic related there is always big pressure with costs. I can see it in the construction of all the scanners, specially when Howtek had the advantage in pricing ordering in quantities ten times what Aztek did. Just the aperture wheel would cost a grand nowadays. Aktek replaced the linear stage in the optical box with a trio of metal balls which is not the best solution over time. The list goes on and on.

Regarding DPL, I went back and read the product description again and it seems, if I understood correctly, they in fact improved the licensing. Now the product is divided in high resolution and legacy, ie hr8000+premier in one type and the 6500/7500/4500/4000 etc. in the other. That's a big step forward. Not a universal license but a lot closer.

The other "feature" (can't think of a better word) is the option to select the log mode. Trident behaves like is "always on" and (to my surprise) it seems Silverfast developers are not familiar with the basic Howtek scanner architecture and not aware there is a log amp in front of the ADC that can be used. Some users can really take advantage of the option and Aztec should list it in DPL specs.

In any case, Premier owners who paid big bucks to Aztek for a scanner deserve some kind of break with the new version, I think.

Just to be clear. I do not enjoy at all pointing out what it seems to me as a weakness or issue with this or that design or construction. I only see it as a possibility of improvement to obtain what we all want, a better image.

[onnect17]

Hi Armando,

Thank you so much for your tests of these Howtek systems. And I am interested in knowing, but unfortunately the electronic issues are well beyond my pay level!!

At this point I am still running the Howtek 7500 with Silverfast 6.6 on a Power PC MAC. I am happy with results but would consider DPL if I thought scan quality could be improved. So I pay close attention to your advice.

As for the claim in question, I have long suspected it could be similar to the beauty of the emperor's new clothes.

Best,

Sandy

Hey Sandy, I have a DPL 7.8 license for the 7500. I do not mind you trying it. Then you can decide if it's what you want.

Just borrow, OK? :-)

[Lenny Eiger]

In the only part I disagree is "no expense was spare". Unless is Defense or Academic related there is always big pressure with costs. I can see it in the construction of all the scanners, specially when Howtek had the advantage in pricing ordering in quantities ten times what Aztek did. Just the aperture wheel would cost a grand nowadays. Aktek replaced the linear stage in the optical box with a trio of metal balls which is not the best solution over time. The list goes on and on.

I spoke to Phil directly about this. "No expense spared" were his words. They went over every part, combined six boards into one main board (at no small expense, I imagine). Phil was very upset that Howtek did not follow his recommendations on the brass screw in the HR 8000 and made sure to update it to a steel one, then add whatever he calls ceramic bearings for the Premier. In describing his process, there weren't a few changes made here and there. They went after everything they could. He described it in detail to me and it took a very long time...

He had some number he had audited, for the creation of the DPL software, in the many millions.

It sounds like you have some great skills. It's too bad they didn't come to light back when Aztek was in the process of redesigning the machine that became the Premier. I am sure Phil would have gotten some enjoyment from speaking to you, and you would have been able to make a serious contribution.

However, you are starting with only half this story...

Further, you are casting aspersions at one of the best machines made for drum scanning. I am sure it has its faults. I am sure you could find design faults with each of the other scanners as well. I know someone who paid $80,000 for a Tango. Mine is quite a bit sharper than his, I have scans from both. (Who knows why, could be that Karl could fix it - or not.) Regardless, I am careful not to tell him that his expensive machine is a piece of junk. It's not, and even if it isn't quite as good, it can still make terrific scans. It has been said a million times that the best scans are made by the most talented operators, the machine is a very small factor.

Finally, I am trying to make a living from doing scans for others, or at least pay off a little of the scanner costs. Not so easy these days. Every time someone suggest that drum scans are not all they're made out to be, or the software is not quite what it should be, another person decides not to bother... Now, I don't want anyone to speak anything but the truth. I mean that. Yet I do think we need to be sensitive to others on this list.

Reading between the lines, it does sound to me that you have some issues with Aztek, and might enjoy seeing them taken down a notch. Unfortunately, we aren't going to get a better scanner from them, no matter what you come up with. They stopped making them last year. This is a world where no new drum scanners will be made. If you can make yours work better from your experiments, that's terrific. However, complaining about Aztek and their "unsupported claims", whatever they might be, or the veracity of either side, will get us nowhere.

Lenny