pass around scanner test : 1 neg, many scanners

Great Job! I normally scan using Vuescan, how may passes do you want me to make of each negative? When I convert the TIFF's to JPEGS for posting on the web, what size should they be?
 
OK, here are the instructions for scanning and posting :


1. scan one frame from each media/size at FULL RESOLUTION on your scanner

You may do anything to fine tune the quality of your scans in terms of tone and color, but NO SHARPENING or any other similar effect.

Use glass holders if you have them. If you don't, that's just part of the test - proof that glass holders are very effective. Or not.

2. crop out a 100% zoom section that shows the detail of the scan - and downsize to below 300x300 pixels.

2.5. crop out a section that shows the shadow performance of the scanner. Same size.

3. downsize the full-res scans to 1000 pixels on the longest side and 300dpi - this shoud be possible. I got a test down to 250K at reasonable JPG quality. The point of this image is to show tone and color quality and all that.

whatever modifications you make in tems of resolution, curving, contrast, color correction, please note that in your post.

Post all the images to this thread.
 
Interesting test and looking forward to seeing the results - but are you sure you want to 'allow' for curves, tone and contrast adjustments?

If you are also looking for how well the scanner handles shadow areas, etc such adjustments have a direct impact on how well they are presented. Any such adjustments would also add an element you don't control which is how good the person is at doing these sort of adjustments and the comparison would be more of who is best in PS :)

Are you 'allowing' for single-pass scanning only?

If scanning through a software like vuescan should the testers allow for colour balance through it (whitebalance, neutral, etc) which could also effect how the scanned file looks like.
 
I've noticed ssome differences in color balance between Vuescan and Minolta software. I may do some of each.

I have a 4490 as well-might put that to the test!
 
If we all use Vuescan we could agree on a common set of adjustments, that would make the scanner the only variable.
 
You could even create ONE ini file for vuescan and save it as an attachment here and everyone scans using those defaults :)
 
Rich Silfver said:
Interesting test and looking forward to seeing the results - but are you sure you want to 'allow' for curves, tone and contrast adjustments?

If you are also looking for how well the scanner handles shadow areas, etc such adjustments have a direct impact on how well they are presented. Any such adjustments would also add an element you don't control which is how good the person is at doing these sort of adjustments and the comparison would be more of who is best in PS :)

Are you 'allowing' for single-pass scanning only?

If scanning through a software like vuescan should the testers allow for colour balance through it (whitebalance, neutral, etc) which could also effect how the scanned file looks like.


good point.

Let's have one totally flat, virgin scan of each frame as well - smallish - like 500 pixels on the longest side.
 
fgianni said:
If we all use Vuescan we could agree on a common set of adjustments, that would make the scanner the only variable.


Is Vuesan free? I don't have it and can't afford to buy anything. But working with the same settings is a good idea.

I figure just having one virgin scan and the others all together will show the performance accurately.

What people really want to know is "How well can this scanner do?". That means, what is the best image I'm going to get from this scanner? So, I figure curving and color adjustment is allowable on those terms.

Sharpening is something people generally do, but that would skew the results too much to be allowable.
 
Rich Silfver said:
Interesting test and looking forward to seeing the results - but are you sure you want to 'allow' for curves, tone and contrast adjustments?

If you are also looking for how well the scanner handles shadow areas, etc such adjustments have a direct impact on how well they are presented. Any such adjustments would also add an element you don't control which is how good the person is at doing these sort of adjustments and the comparison would be more of who is best in PS :)

Are you 'allowing' for single-pass scanning only?

If scanning through a software like vuescan should the testers allow for colour balance through it (whitebalance, neutral, etc) which could also effect how the scanned file looks like.

no. I didn't say anyything about how many passes you take or anything. BUT, and this is VITAL : no layering, no multi pass/multi exposure scans designed to give impossible DMAX values. You can multi-pass enough to kill noise in shadows, but that's it.
 
Rich Silfver said:
You could even create ONE ini file for vuescan and save it as an attachment here and everyone scans using those defaults :)
That sounds the right thing to do!
 
do whatever works


for the non-virgin scans :
-no sharpening or other similar filters
-curving and color correction is allowed

basic idea : people want to know how well the scanner CAN perform. That means produce the best scan you can without doing technical acrobatics like mutli exposure/multi layer images with insanely high dynamic range.
 
Last edited:
Bump so everyone gets a look at the guidelines, can give input, and can read some input.

Please read a few posts back to see the list of who I mailed the kits to and what the guidelines are . . .so far.
 
Konica Minolta Dimage Scan Multi Pro FULL RES

Konica Minolta Dimage Scan Multi Pro FULL RES

Here is 120 6x4.5cm frames of Neopan 100 Acros, Fuji Velvia 100, and Fuji 160s (B&W, E6, and C41)

scanned and posted without any adjustments at all. Scanned in glass holder

these are the full frame images resized, of course.
 

Attachments

  • 120B&Wfull.jpg
    120B&Wfull.jpg
    338.1 KB · Views: 0
  • 120E-6full.jpg
    120E-6full.jpg
    329.1 KB · Views: 0
  • 120C41full.jpg
    120C41full.jpg
    339.8 KB · Views: 0
Last edited:
Konica Minolta Dimage Scan Multi Pro FULL RES

Konica Minolta Dimage Scan Multi Pro FULL RES

Here are the crops of the 120

same settings - same images, of course
 

Attachments

  • 120B&W.jpg
    120B&W.jpg
    184.7 KB · Views: 0
  • 120E-6crop.jpg
    120E-6crop.jpg
    231.1 KB · Views: 0
  • 120C41crop.jpg
    120C41crop.jpg
    359.7 KB · Views: 0
Last edited:
Konica Minolta Dimage Scan Multi Pro FULL RES

Konica Minolta Dimage Scan Multi Pro FULL RES

Here are the 35mm ones

B&W, E-6, and C41 in that order, with crops to follow

35mm B&W to demonstrate the benefits of scanning in POSITIVE B&W instead of the default NEGATIVE. The first is negative, the second positive. Huge grain reduction and a tonal benefit as well, I'd say.

If scanning in positive, you need to recurve the image to get proper contrast.
 

Attachments

  • 35grain.jpg
    35grain.jpg
    189.5 KB · Views: 0
  • 35grainPOS.jpg
    35grainPOS.jpg
    162.8 KB · Views: 0
Last edited:
Konica Minolta Dimage Scan Multi Pro FULL RES

Konica Minolta Dimage Scan Multi Pro FULL RES

Here are 35mm B&W, E6, and C41 full frame

Neopan 100, Fuji Velvia 100, and Fuji 160s

no adjustments

glassless holder
 

Attachments

  • 35mmB&Wfull.jpg
    35mmB&Wfull.jpg
    313 KB · Views: 0
  • 35mmE-6full.jpg
    35mmE-6full.jpg
    330.4 KB · Views: 0
  • 35mmC41full.jpg
    35mmC41full.jpg
    296.6 KB · Views: 0
and the 35mm crops
 

Attachments

  • 35mmB&Wcrop.jpg
    35mmB&Wcrop.jpg
    121.4 KB · Views: 0
  • 35mmE-6crop.jpg
    35mmE-6crop.jpg
    202.7 KB · Views: 0
  • 35mmC41crop.jpg
    35mmC41crop.jpg
    342.8 KB · Views: 0
Back
Top Bottom