ScanMate drum scanner DIY maintenance, troubleshooting, mods

brbo said:
I hope it's OK to also ask/post some Howtek related stuff in this thread...
...
Is there any benefit in canceling C-41 mask at calibration vs. later in software? How do you mount your film on the drum so that your calibration strip is covered with blank film?
Click to expand...

Yes. During auto calibration the Howteks will search for the density corresponding to the white point, before the A/D conversion. You should be able to see a couple of areas/strips 3/4" wide marked in the drum. The first one is typically covered by a white tape, used to calibrate the reflective work. I do not scan any reflective media so I removed it. The second one determine the white point for transparency mode. This area should be covered with the blank film. This helps a little if DPL (and log mode) is not available.

brbo said:
Maybe I've asked this before, have you done (or know of) any comparison of negative scans between DPL (in log mode) and other software (I have Silverfast)? I'd like to get an idea of what to expect from DPL.
Click to expand...
Yes. The easiest way is to ask Aztek for a demo, so you can compare by yourself. After all, you are the one paying $1300.

brbo said:
Also, how much work does it take to check the optical path alignment in 4500? Any special tools required?
Click to expand...
It's quite simple but you need the tool. I posted a drawing of it somewhere... can't remember now. I can post it again if you are interested.
 
I started a similar thread titled "Howtek drum scanner DIY thread: maintenance, troubleshooting, mods", so we don't hijack this one.
 
onnect17 said:
And the A/D converters are pretty fast too. You will be able to see the jitter in you images. Any ripple in the output bigger that one bit will impact the quality from the get-go.

The halogen bulbs are pretty slow, which is helping in this case. It's not the case with leds. Some can switch pretty fast. Besides, I am sure the lamps in the scanmates are getting a nice and clean DC feed.
Click to expand...

No way it can catch a led blinking at 1Mhz. That would be just insane. The drum would have to be rotating at insane speeds for it to be able to sample at that high rates.

The jitter you see in the images is just the pixels not aligning perfectly. It's normal, as you can not expect it to hit every pixel at 100% accuracy as the drum can not rotate at 100% constant speed. Things just does not keep 100% perfect when moving fast & trying to make microscopic enlargements at the same time.

This is why there is a softening applied to the TIFF image that colorperfect outputs. To smooth things out.
 
monkeyfist said:
No way it can catch a led blinking at 1Mhz. That would be just insane. The drum would have to be rotating at insane speeds for it to be able to sample at that high rates.
Click to expand...

It does not need to. For example, rotating at 300 rpms, the Howtek 4000/4500 needs to complete the conversion in less than 5 microsecs. So a signal of 1 microsec (1 MHz) should not be an issue. The problem is we do not know when and for how long the sample and hold is capturing the data.

monkeyfist said:
The jitter you see in the images is just the pixels not aligning perfectly. It's normal, as you can not expect it to hit every pixel at 100% accuracy as the drum can not rotate at 100% constant speed. Things just does not keep 100% perfect when moving fast & trying to make microscopic enlargements at the same time.
Click to expand...

The jitter I am referring to is not the pixel-shift due to inaccuracy of the encoder. I am referring to a jitter in the luminance component.
 
I am getting better with wetmounting my negatives and now almost always wetmount the film. Everything seems to be OK but I wanted to ask if you "wash" the film after wetmounting it? I use the Kami fluid (tried Scanscience but prefer Kami) and thought it dries clean, it does leave small streaks that I don't like especially on the E6 slides.

Can you please share your tips and tricks on wetmounting and how you keep your negatives clean after scanning?
 
Mounting Fluid

Mounting Fluid

I use the Sunnyside Odorless paint thinner for mounting/cleaning the film. It's similar to the Lumina, which I prefer (vs Kami, too aggressive IMHO). Be careful with the type of plastic you are using to store/apply the fluid. Some are not resistant enough to the solvent and can produce a residue that will end up in the film.
 
Thank you both.

I tried to switch to Gamsol myself but I find that Kami is easier to mount with because it seems less prone to bubbles.

Do you recommend wiping with scanner wipe and fluid for cleaning the negative?

Pali
 
Netsoft2k said:
Thank you both.

I tried to switch to Gamsol myself but I find that Kami is easier to mount with because it seems less prone to bubbles.

Do you recommend wiping with scanner wipe and fluid for cleaning the negative?

Pali
Click to expand...

I would avoid wiping the emulsion. You can spray it or dip it in a small tray using mounting fluid. Hang it to dry if needed.
 
I've installed suggested UV-IR cut filter and used 8 bit TIFF RAW - improvement is visible.
 

Attachments

  • IMG_5515.jpg
    IMG_5515.jpg
    19.1 KB · Views: 1
Anyone with non working Scanmate in Australia? I'm going to send one to engineer for making new electronics (AD, laser light, raspberry pi).

Update: one person wrote to me he can send Scanmate 5000 for making new electronics!
 
Kamph said:
I can't speak to the Howtek, but I tried to calibrate on the negative mask on my SM5000 and it resulted in an error.
Click to expand...

It almost always says that the white balance calibration was NOT successful when you try to calibrate it on the C41 mask, but still if you scan it.. it has calibrated it. I don't know why it does that, but it works.
 
Yes, that is most likely the case. Voltages not only have to be proper but they have to be in balance across three channels and within the 10% tolerance from each other. Also keep in mind that white calibration is performed for each individual aperture and if one fails, you get a full failure message.

Pali
 
Interesting. Did any of you try it yet? Was the result noticeably better? I imagine that it could add some levage when scanning 8bit.
 
Kamph said:
Interesting. Did any of you try it yet? Was the result noticeably better? I imagine that it could add some levage when scanning 8bit.
Click to expand...

I use it all the time. It does give me a scan that just needs to be inverted. And sometimes it even says SUCCESSFUL on the white calibration. So it's probably just on edge.

But i do scan at 16bit always, mostly to have room in the edit.. i'm not sure why to scan at 8bit? 11000 is 14bit and 5000 is 12bit in the hardware side.

Also, i think my printer can output more than 8bit colors even if my display could not. 8bits is really few color, simply not enough to produce quality prints. In b&w it means literally 256 shades!

8bit = 265
12bit = 4096 (1/16th of full 16bit file)
14bit = 16384 (1/4th of the full 16bit file)
16bit = 65536

it's a quite substantial leap, even from 8 to 12bits.

People who compare 8bit files to 16bit raw files do not understand, that as the scanner does no output 16bit information, the file is not "full". It's just fills small part of the 16bit file and will look crappy directly from the scanner compared to 8bit file that is full. You can see that a histogram of 16bit shows that it has only small part of the histogram that actually has info. This is because 12 or 14bit is so much less than the full 16bits.

You need to use levels to move the information correctly on the histogram in 16bit file.

I would not waste time on a scanner that would be hardware limited to 8bit color space. It's just too few colors for modern workflow or for modern inkjets. 8bit scanning is a bitch, you need to do all the curves etc before scanning. And pretty much leave the file untouched after that. In b&w it's unable to produce smooth gradients.
 
You must log in or register to reply here.
Back
Top Bottom