On the way.Originally Posted by nagairaj
On the way.
A quick note about the SMU in the HR8000 (same thing in the Premier). For wherever reason several drivers in that card are operating at a temperature point that indicates the need of heat sinks, IMHO. Some little clipons will help. A small fan on top of the power supply pointing towards the SMU even better. I hope no owner has had to replace theirs yet.
I noticed a few owners of 4000 and 4500 struggling with F702 errors. One of the sources for the error is related to the encoder in need of an index adjustment and many times the console shows a "check lumen" message right before showing the error. So in reality, the F702 is just the aftermath.
The procedure to perform the index adjustment is clearly explained in the Service Manual and the Field Service Tool helps with the task, showing an encoder reading counter and the option to move the drum to each end and center.
Better results (IMHO) can be achieve via the console using the manual focus option because the user is able to freely position the drum. This is important because in the real world the mounted drums are not exactly parallel to the carriage and the edge of the calibration strip will register slightly different values in the encoder in each end of the drum.
The location in the drum I prefer to use for the adjustment is in the middle between the beginning and the center (i.e., around 25% from the start). Most calibrations performed by the scanner at the start of the scans are done in the first half.
If any member need any assistance with the task please feel to get in touch.
Forgot to mention, the drum encoder used in the 4500 (also as replacement in some D4000) is not sealed. So it could be also a source for this type of error ( F702 ) as some dust gets in. Just remove the plastic cover and carefully clean it using canned air, keeping some distance, of course. Run the FST at the highest drum speed to verify proper operation.
Some electrical tape around it will help keeping the dust away.
After a few scans using the cree xml-color as the light source I feel quite confident it is a reliable option as a light source. Another couple of simple mods that improve the image are the use of sorbothane in the drum pads and in the step motor to leadscrew coupling. If you are only scanning film then a simple square mask with a small (3 to 4 mm) of black matte tape in the analyzer lens will work as an smaller aperture improving depth of field and cutting a lot of flare.
If interested in looking at some samples, I uploaded some images to howteks.com. Be aware, files sizes could be in the 100MB range.
Happy scanning!
Is anyone using DPL 8 for windows 64-bit? If so, what are your thoughts? What SCSI setup are you using?
-Adam
Glad you asked. I've been contemplating a message here to the group and your question now gives me a reason to post. Like you, I am very interested to hear from anyone else running DPL 8.0 on Windows 64-bit, especially if you are scanning at 4000dpi on a Howtek 4500.
Yes I am running DPL 8.0 on Windows 7 64-bit on a HP Z230 Workstation, 8GB RAM, using a 29160N SCSI card which I purchased from Aztek along with DPL 8. My scanner is a recently acquired refurbished Howtek 4500 (not from Aztek). I am new to high-end scanning but have been reading here a long time to research and learn from the collected wisdom as I develop my scanning and DPL skills. I will be scanning my 4x5 B&W negatives and perhaps later doing some scanning for other photographers.
Regarding the SCSI 29160N: This is a 32-bit PCI card, however there is a 64-bit driver which can be downloaded from Adaptec. So to use this card you will need an available PCI slot in your machine, or Aztek also sells a PCIe adapter/bridge card that would allow you to plug into a PCIe slot. In my case, in two different machines, I could not use the adapter card because the combined 29160N and adapter card were physically too tall to fit.
Regarding scanning @4000dpi: So far I cannot scan a 4x5 at 4000dpi because of "Buffer Transfer - Length Too Big" errors in DPL (saving as TIF 48-bit). Also I cannot do a scan at any resolution when saving to gray TIF format (different error: "Unable to create a buffer for one scan line"). I am in touch with Evan at Aztek who is working with me on trying to understand what is happening. Initially we thought there might be hardware problems with the IOC card in the scanner or somewhere in the SCSI communications, but after a lot of diligence in cleaning contacts and veritying the scanner HW, cables, etc. the failure pattern remains and is very consistent/repeatable. Our investigation is ongoing...
So my question back to the group: Is anyone using DPL 8.0 on Windows 64-bit with a Howtek 4500 and able to successfully scan a 4x5 negative/chrome @4000dpi?
Dave
Hi Dave,
Have you tried saving as a lower-quality TIFF, such as 24-bit? The 4500 can only scan at a max of 36-bit anyway, and any 48-bit scan is just interpolation.
Source for that claim is from the latest DPL User Guide that I have (v.7.86):
"The D4000 and SM4500 can produce variable optical scan resolutions without software interpolation up to 157 lines per millimeter 4000 dpi, 36-bit color images, and on an approximately 11" x 11" drum surface, with a .001 inch positioning accuracy."
For me, I asked because I am thinking about upgrading to DPL 8 for my Howtek 4500. I doubt it helps much, but I haven't had any problems scanning at 4000dpi, but that is DPL 7.83 and Windows 7 32-bit unfortunately.
I have been using the 4500 with an old SCSI card - Adaptec 2930CU - and Windows 7 32-bit, and DPL 7.83. I have to boot off of a separate hard drive for Windows 7 32-bit, and of course, since that operating system can only access 3gb of system RAM, it is pretty slow when editing large scans. So I have to change back to Windows 7 64-bit for editing, and can't really do much with the PC while scanning is happening. And maintaining two installations of Windows is pretty annoying. So... it may be worth spending the money for DPL 8.
-Adam
Hi Adam,
Yes, I'm aware that the 4500 is limited in hardware to 12 bits per pixel per color. However this does not (should not) prevent the software (DPL in this case) from saving a scan as a 48-bit (16 bits per color) TIF.
I have done a lot of testing using various resolutions and save file formats, and I cannot do a 4000dpi scan of a 4x5 negative even when saving as "ALL TIF" (which in DPL means a 24-bit [8 bits per color] TIF). Here's a quick summary of what I'm seeing (the errors appear in a DPL popup window):
2000dpi/13micron, ALL TIF: works
2000dpi/13micron, 48-bit TIF: works
4000dpi/13micron, ALL TIF: Buffer Transfer Length Too Big
4000dpi/13micron, 48-bit TIF: Buffer Transfer Length Too Big
1000dpi/13micron, 8-bit gray TIF: Unable to create a buffer for one scan line
2000dpi/13micron, 8-bit gray TIF: Unable to create a buffer for one scan line
2000dpi/13micron, 16-bit gray TIF: Unable to create a buffer for one scan line
4000dpi/13micron, 8-bit gray TIF: Unable to create a buffer for one scan line
In all cases, the error occurs immediately after scanning starts (after calibration has completed). When I get the "Buffer Transfer - Length Too Big" error, I can abort the scan in DPL and it recovers. However with the "Unable to create a buffer for one scan line" error, DPL hangs and the only way to recover is to kill the process in Task Manager. It seems strange to me that the 2000dpi ALL TIF scan works but the 2000dpi gray TIF (either 8-bit or 16-bit) do not.
In any case, as I mentioned above, this is still an open case with Evan at Aztek and we are still trying to pin down what is going on. But it would really help if someone running DPL 8 on Windows 7 64-bit could verify they are able to do a 4x5 scan at 4000dpi.
Dave
Allow me to make some comments.
...
"The 4500 can only scan at a max of 36-bit anyway, and any 48-bit scan is just interpolation."
...
Filling the lower 4 bits in each color is not related to interpolation, if I understand it correctly.
...
"The D4000 and SM4500 can produce variable optical scan resolutions without software interpolation up to 157 lines per millimeter 4000 dpi, 36-bit color images, and on an approximately 11" x 11" drum surface, with a .001 inch positioning accuracy."
...
I would check the exact dimensions in pixels after scanning a square image (an IT8 target can help). I think the pulley in the 4500 is smaller than the D4000 and it could be a little off from the expected 4000x4000dpi. Not sure to what the "0.001 inch position accuracy" is referring to.
...
"I have been using the 4500 with an old SCSI card - Adaptec 2930CU - and Windows 7 32-bit, and DPL 7.83. I have to boot off of a separate hard drive for Windows 7 32-bit, and of course, since that operating system can only access 3gb of system RAM, it is pretty slow when editing large scans. So I have to change back to Windows 7 64-bit for editing, and can't really do much with the PC while scanning is happening. And maintaining two installations of Windows is pretty annoying. So... it may be worth spending the money for DPL 8."
...
Why not using a dedicated old PC for scanning and another one with W7 for editing?
I use an old XP dedicated to scan and then upload the images to the server (an old W2K). In the W7, I download the images from the server and I can be editing while scanning.
Posting Permissions
Reply With Quote
Bookmarks