snipped-for-privacy@isbd.co.uk wrote
Whoops - I know :) It was very good. I used kodak, agfa slide film. I think Provia 100 is as good as Kodachrome 25, without the need for a tripod...
Peter.
-- Return address is invalid to help stop junk mail. E-mail replies to snipped-for-privacy@peter2000XY.co.uk but remove the X and the Y. Please do NOT copy usenet posts to email - it is NOT necessary.
Pete C wrote
Colour cast.
Peter.
-- Return address is invalid to help stop junk mail. E-mail replies to snipped-for-privacy@peter2000XY.co.uk but remove the X and the Y. Please do NOT copy usenet posts to email - it is NOT necessary.
news:comp.periphs.scanners would be the most appropriate NG.
Is there some reason you do not want to store them, or hide them away in a box somewhere? It is not like slides are that big, even several thousand (I should know, I have more than that in my office).
Almost all of the toaster shaped scanners are closer to 12 to 14 bits, and not much better with dynamic range. This is true, despite wild claims in manufacturer brochures for CCD film scanners. This is also an area that is somewhat hindered by software, and like I stated previously, the Canon software lets down that scanner.
You would need to spend a great deal more to get any noticeable difference in hardware. Spending a little more on some software, would be a more economical big step to improve image quality.
Could be the settings, or again the software.
There update was not a great improvement, and the preview image often does not match the final scan. This is unfortunately true of the latest Canon Scanners, and those from several other companies.
Okay, that is what I was confusing about your previous explanation.
Zooming in is still judging by the limitations of the monitor. It is just not a real representation of reality until you get that image printed. Never judge image quality on a monitor, unless you never intend to print that image. Of course, with many years in graphics, I am surprised people did not tell you that, though the long time of doing web based work could have got you into that habit (sorry for the criticism).
All monitors are slightly different, and then each operating system is set for a given range of parameters on how to display colours. Recall that you can go into a store with many television sets, and the image can be slightly different on all of them. Without going into science, or software and hardware interaction, the reality is that Windows based monitors are often closer to television, while MacOS (and SGI IRIX) are closer to paper. You might be familiar with display gamut, likely a value of 2.0, 2.2, or 2.4 that you might have noticed when you calibrated your monitor; and that gives an indication of what you see. Hopefully, if you are more interested in that, someone else will chime in and add some more details.
Not even all printers are even close, even the presses that do various offset printing. While there are ISO standards, the reality is that to get the absolute best, the pre-press specialist needs to set each image file to get the best results on each device. I have well over 400 different device profiles information ready to use, just for that reason of achieving the optimum quality. It is possible to use a set of parameters that will get you close on many devices, though the results will still be better outputs on some, and worse outputs (prints) on others.
Not at all . . . I have heard of many individuals frustrated by the lack of some stricter standardization. However, what prints well off a press, might look truly awful on newsprint, so there is need for variation.
The autofocus does not work well on some films. It seems to land either at the very front, or very back of some emulsions, which can make apparent grain worse. It is better to manually focus, and sometimes a step or two in front of, or behind, the spot the autofocus wants.
Manual exposure control is another area. The autoexposure is okay, but to really pull out the details, it is often better to overexpose. Then what you do is adjust the gamma setting lower, near 1.4 if possible, and rarely above 1.8. The gamma setting in the scanner software is different than your monitor settings.
You can also adjust the exposure curves, helping to retain some highlight detail, without losing shadows. One problem here is that it is tough to tell subtle changes with the Canon software.
SilverFast AI is the absolute best for this. I suggest you download the trial version, and try doing some trial runs on some difficult slides. I have many night images that are tough to scan with CanoScan, though very easy with SilverFast. This software is so good, that you rarely need to do any further adjustments in PhotoShop. It is also possible to print directly from SilverFast, which means easier proofing on your PostScript proofer.
There are several batch type scanners on the market, Braun, Nikon, and even Kodak. The reality is that no matter how much you spend on these, batch scanning will often not give the best results. It is possible to get better scanners to do many scans at a time, and at high quality, though these are very expensive.
Check out Creo iQSmart and EverSmart scanners to see the high end. These go for around $US 18000 to over $US 45000 (yes, three zeros). These even have an oil mount station to improve colour capture, though they are usually used dry.
Stepper motor, lamp, and something with the converter. Not a huge change, but it made it faster, and easier to use. I don't know the details, but I trusted him when he said he could improve on it.
I guess that cost per slide is what you should compare for your time, and the cost of any scanning gear. Unless you want to provide scanning services, don't get into a Creo system .
SilverFast is not too bad on cost, though there is still the time issue of just loading and scanning that many chromes. You could also look into ViewScan, and see if that lower cost software give you enough boost in quality. Best of luck.
Ciao!
Gordon Moat A G Studio
Yes. You can also use LZW compression with TIFF files, and that is lossless too.
Not in some green, anything near pure Cyan, and anything near pure Yellows. There are also some Reds that will not display properly on any monitor.
CMYK to HexaChrome, or also to HiFi colour, is often a choice based upon cost. While CMYK is lower cost, some images work better with Hexachrome, which does an even better job with greens, and near reds (HiFi colour is similar). However, a newer method is replacing one of the CMYK inks with a Pantone, often at the same cost from some printers. There is also an in between cost advantage to going with a five colour print, and adding one Pantone to a CMYK run. I think cost is the greatest issue of why most things are still done CMYK, though I have noticed more eight and ten colour presses lately.
Ciao!
Gordon Moat A G Studio
John Rumm wrote
I will check out Silverfast - thank you.
In the meantime, I posted a fresh scan of a pic here
whose 35mm original I know will be as sharp as I can do with Provia
100 in the OM4 camera, and with no camera shake.This was scanned on the FS2710, with and without auto colour correction (can't see any difference), direct into Photoshop in 36-bit mode, had the 2710 colour profile applied to it (not sure why this absolutely necessary step isn't automatic...) and saved as a high quality (#9 quality) 24-bit jpeg.
Do you think this is of reasonable quality?
I realise that I have no independent way of checking the colours because I cannot get away from the colour profiles of my screen (19" CTX LCD) or my printer (Canon i850).
For some reason, perhaps from getting a load of slides scanned a few years ago onto a photo-cd (back in the days when Kodak thought the photo-cd was going to change the world) I think the resolution alone ought to be lot better than this.
Even looking at resolution alone, my scanned slides seem to be no sharper than 4MP digital camera images. Scanning a 35mm slide at
2700dpi should yield a lot more pixels...Peter.
-- Return address is invalid to help stop junk mail. E-mail replies to snipped-for-privacy@peter2000XY.co.uk but remove the X and the Y. Please do NOT copy usenet posts to email - it is NOT necessary.
BTW, do you know of any scanner software that can scan at 2 exposure levels and combine the two images for best results? Do VueScan and Silverfast do this?
cheers, Pete.
Looks fine, though the white roof of the building looks a little bleached out, does the projected slide look any better?
I wouldn't worry too much about colour, as long as the all the information in the image is scanned the colour can be altered later.
All the dust and hairs look well focused ;) Seriously though, if there is something in the image that is 1 pixel wide, it will only be scanned perfectly if it lines up exactly with the pixels in the scanner.
So ideally you want the scanning resolution to be somewhat higher than the maximum resolution of the slide. A good way to check the focussing of the scanner would be to scan a slide of a line pair chart, look at where the lines match the pixels of the scanner and see how tight the focussing is in these places compared to a projected image.
cheers, Pete.
Not really a function implemented in scanning software. The most important capability would be to scan in 16 bits per color per pixel (*sometimes* called HDR scanning).
Method One: Try making two scans, and combining them in PhotoShop, Picture Window Pro, Corel PhotoPaint ... and possibly several other image editing programs. There are several tutorials 'on the Web' that you can easily find with your friendly local search engine.
Method Two: Do an Internet search on 'Contrast Masking' ...
The scan does not look too bad to me, having said that I have not seen the original and hence have no frame of reference.
Reasonable certainly, some noise on the darker areas, and the contrast may be a tad over done (could just be lack of dynamic range) but no glaring problems obvious. With some of the other scans you posted (like the one of the field with side by side scan of a print), the slide scan seemed quite low on red and green, but about the same on blue in comparison to the print.
Aha! Note that the LCD monitor may be accounting for a portion of your problems. They typically have colour gamut that is narrower than a CRT and can produce images that look a little posterised as they often loose the ability to resolve the difference between subtle shades of the same colour. Your scan above showed a nice smooth graduation of blues over the sky on my 22" CRT aperture grill monitor. How does it look on yours?
I would expect you would start to see the film grain in the scan at that resolution, from the file its not that obvious but then again the jpg artefacts make it harder to see anyway.
If you wanted to post me the slide, I could scan it for you on the Nikon and also on an Epson 1680 pro flatbed to give you something to compare against. Drop me a private email if you are interested.
Pete C wrote
On the slide, there is no texture in the roof either.
Where can I get a line pair chart from?
Thank you very much for your comments,
Peter.
-- Return address is invalid to help stop junk mail. E-mail replies to snipped-for-privacy@peter2000XY.co.uk but remove the X and the Y. Please do NOT copy usenet posts to email - it is NOT necessary.
Normally, that is something you would do with a few difficult images, and then use your editing software to combine the results. I still use LivePicture for this, since it allows this manipulation in 16 bit mode.
Vuescan and SilverFast both have multi pass scanning, though the passes are not varied in exposure. Basically, the multi pass feature allows for the scanner converter to average out possible errors and could result in a smoother scan with less noise. In practice, I rarely use that feature, due to the extra scan time.
I have found that SilverFast is so good, that I rarely would need to double scan at different exposures. I could not accomplish that same level with CanoScan, or Vuescan software.
Ciao!
Gordon Moat A G Studio
John Rumm wrote
The sky does not show any sign of posterisation on the LCD monitor. I am going to look at it on a 19" Iyama later today also.
Do you think 2700dpi would reveal 35mm film grain?? Also can you see jpg compression artefacts in that image? I could not see any, even around the text of the roadsign (which is an area where jpg falls over first, I think).
Thank you very much for the offer. I will do this exercise with a local camera show that sells slide scanners :)
Peter.
-- Return address is invalid to help stop junk mail. E-mail replies to snipped-for-privacy@peter2000XY.co.uk but remove the X and the Y. Please do NOT copy usenet posts to email - it is NOT necessary.
I just did some experiments. The answer (with my kit) seems to be "sometimes". On a very fine film like Velvia, no chance, with Provia 100 I can see very slight grain, and on some of the "consumer" grade tranny films it is quite obvious. It is also more obvious on negatives than on colour reversal film.
Only if I zoom right in with photoshop - you can start to see the cell boundaries. As you say, jpg works best on contiuous tone images rather than things with slabs of solid colour and high contrast.
Sounds like a good plan. You should be able to try out some more recent kit that wasy as well.
How best can they be combined, are there any tutorials on the web at all? I've tried blending them 50/50 or 30/70, but wonder if there's a better way.
Noise isn't such a problem as I'm scanning negatives.
SilverFast doesn't support my scanner :(
cheers, Pete.
"Pete C" posted: "... How best can they be combined, are there any tutorials on the web at all? I've tried blending them 50/50 or 30/70, but wonder if there's a better way. ..."
One such tutorial is "Extending the Dynamic Range of Film" Written by Jonathan Sachs Copyright © 1999-2003 Digital Light & Color
Another would be Making fine prints in your digital darkroom Tonal quality and dynamic range in digital cameras by Norman Koren
Check the next message for some links to tutorials. The reality is that you should rarely ever need to do this. Also, there is no such thing as a method that uses the same approach each and every time, so it would not be a good idea to automate, and better to treat these few special images individually.
It tends to vary by type and brand of film, with some films being worse than others. Like I stated previously, I rarely have need to worry about noise in scans.
Canon CanoScan FS2710? SCSI connection?
This is the page about SilverFast AI support for the Canon FS2710. It lists Windows XP, 2000, and
98SE/ME, as well as several versions of MacOS X, and Mac OS 9.Are you running Linux, or the older Windows 98, or Windows NT? If that is the situation, then look into Vuescan. You can download a demo at:
it is also lower cost than SilverFast.
Sorry I forgot the Vuescan software link in my earlier message. Hopefully that will help.
Ciao!
Gordon Moat A G Studio
Thanks for those, will have a trawl on Google too.
cheers, Pete.
True, the scanner has most trouble scanning flash photos, if I could improve the dynamic range and get rid of the blue colour cast the scans would be much better. Photos with shadow detail don't come out too well too.
No it's a 5000F on XP, Silverfast support the 9000F though.
Thanks I'll check it out. Maybe someone will develop some good software for combining scans in the future, in the meantime I'll scan at different exposure levels where required and combine as necessary.
cheers, Pete.
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.