[filmscanners] Re: Re:24bit vs more
Typically, when you get high bit data from the scanner, it's raw data. Raw data specifically means the setpoints have not been set, or the tonal curves applied. Having left the 8/16 hobbyhorse, but Im interested in the whole RAW data thing. My workflow for a neg roll is thus: (using Vuescan): scan neg profile (the vuescan film profile) loop scan frame save RAW data save scanner corrected TIF (16 bit) correct in Photoshop as required save PNg/JPG/small files for general viewing I archive the RAW files as Ive learnt (early in my scanning days) that Vuescans colour models improve over time, and I can get significant differences ... well, to me at least, with a review of particularly difficult frames. Can I assume that the RAW scan is just that, it will always be the same w.r.t a particular scanner and this negative, forgiving minor variations, but will vary on the scanning software. bert -- Linux - reaches the parts that other beers fail to reach. Unsubscribe by mail to [EMAIL PROTECTED], with 'unsubscribe filmscanners' or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or body
[filmscanners] Re: 8 bit versus 16
Austin Franklin wrote: I believe you're missing the point. It doesn't matter if you have a color file that has 100 bits/color, you simply aren't visually capable (because you are a human) of seeing a difference between that and an 8 bits/color file. It has nothing to do with the tools [of] tomorrow. Ahem, I'll clarify, as youve missed the point. The tools of tomorrow, be they better hardware, or better software, may allow me to manipulate the 16 bit data (14 bit in my case), better to produce a better looking image. If I have 256(8bit) greens in my file, and in the other I have 257(16bit), then I have more to work with to achieve an end. I may not be able to make any noticeable changes using the tools now, and produce noticeable results, (although I think they do, and I can, hell, I can, I can, Im still young, my eyes havent began to lose colour acuity so much yet) but Im optimistic that something better will come along to help get more from the image than I already have. I AM NOT looking at 16 bit files and saying, , thats lovely, far nicer than that 8 bit one - I am saying that if I scan it at 16 bit and store it thus, then I will be able to go back and get more out of it than if I scan it at 8 bit. You can now make a reasoned argument out of why 8 bit is enought to archive files, and why we pay for better hardware anyway. I need convincing completely in this case. But this is dull beyond compare. Good god its dull. You think you are right, I think Im right and were not talking of the same things. bert -- Linux - reaches the parts that other beers fail to reach. Unsubscribe by mail to [EMAIL PROTECTED], with 'unsubscribe filmscanners' or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or body
[filmscanners] Re: 8 bit versus 16
Arthur Entlich wrote: I don't think anyone is trying to talk you out of making and storing 16 bit scans. Good. Thats where I am. If you have the time to work with that large a file, and the disk space or other storage to do so, then go and do it. Thanks. wonder what you'll be doing when 32 bit ability becomes available Hopefully not rescanning all my negatives, and moping around listening to tired old arguments. As ever, I'll be hoping someone else tests it and finds it flaws. And someone else decrys 32 bit as too much, and 4 bit as just right. I think the problem is your theory doesn't actually hold any water, and since there are a lot of neophytes and newcomers to digital scanning on this list, who are impressionable, God yes, you are right. Lets make sure they get the facts, the facts and the facts. Not just your view which is pallid. You are telling me that there is no point in using 16 bit, yet working with grayscale there is! And the colour separations on RGB (the 256 colour 8 bit ones, are fine to work on) - yet not on BW. P.T. who? I see my job here is simply to warn them that the information you are suggesting is basically without merit and that they need not follow a path that just wastes their time and resources (unless of course, they want to). Or they might want to follow youre religion, and miss out on enlightenment. Sigh. Funny how technology and its use makes people descend into these petty mailing list melees. Far be it from me to tell someone so entrenched how to do their scanning. ;-) Entrenched .. no bloody way, Im just right ;) bert -- Linux - reaches the parts that other beers fail to reach. Unsubscribe by mail to [EMAIL PROTECTED], with 'unsubscribe filmscanners' or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or body
[filmscanners] Re: 8 bit versus 16
Money quote ... Yes, here we go again. You CAN bombard me with facts about 8 bit being fine. And people can 'talk up'/ 'talk down' their particular favourite, preferred or religious route. I will ALWAYS scan at 16 bit, and will always archive at 16 bit. Just because the tools today cant make my gold 100% pure, doesnt mean the tools tomorrow wont. Of course, I take everything I believe with a lump of reality, as, by the time I decide to review images that were scanned a long time ago, and realise that I couldve done better on the scan .. well: 1. The negs will have degraded ... colour lost. 2. The scanner I used will be a dusty relic with a wierd connector and wierder manual interface. 3. My new scanner will laugh at the low quality scanning I did ... Of course, The 8 Bitters are right, 8 bit is fine. But I dont think so. See above. bert Unsubscribe by mail to [EMAIL PROTECTED], with 'unsubscribe filmscanners' or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or body
[filmscanners] Re: 8 bit versus 16
Austin Franklin wrote: It really depends on if you are talking color or BW. For BW, there is no question, you need to use 16 bits for doing all but a minimum tonal curve adjustment, but for color, for most applications you won't see any difference using 8 bit data or 16 bit data. Have to agree on the BW front - 16 bit is essential - after scanning in a roll of old FP from some years ago and I forgot to set to 16 bit - I got a shock when doing curves - boom - highlights would just explode :) As for 16 bit, I cant agree. If you take a picture of a heavily red scene - autumnal sunsets and leaves etc are coming up for example, then your film is going to be using a much larger range of 'reds' than 8 bits can accomodate. Dithering with other colours will occur with the 8 bit scan to make up the difference in the digital scan vs the analogue film. Once you start messing with the curves on this, it will make matters worse. Having 16 bits of red to work with will leave much broader scope for manouvering in curves. Its analogous to the black and white issue above. 8 bits is only 256 possible reds/greens/blues. Theres no way I would rely on this for editing, although the final destination (print) might make no use of all that info. -- Linux - reaches the parts that other beers fail to reach. Unsubscribe by mail to [EMAIL PROTECTED], with 'unsubscribe filmscanners' or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or body
[filmscanners] Re: 8 bit versus 16
of each individual color, true, and it's also 16M colors. Also, you're not likely to get only one color out of three. Yes, but the 16M is just that, a mythical number that never appears in most images, the range of colours is typically more far restricted. For most images, there will be no visible degradation in the image using only 8 bits/color. If you haven't tried an experiment, and are only speaking of theory Ive noticed it in practice severally. Notably in shots with some very variable lighting across landscapes which contain numerous greens. Playing with the curves often results in problems when all 256 of the greens in 8 bit are used up. And as you point out - with grayscale - luminance can get blown out with too few bits, so why not chrominance. you really need to try an experiment for your self. Many people have done this experiment, and that's why they say that 8 bits/color works perfectly for most images. Yes - 8 bit does work fine for most images, but if you really want to throw an image into some editing, then relying on 8 bits is foolhardy if you can get more to work with. Remember - filmscanners work with an analogue medium that contains far more information than 16/8 bits can capture - now why not only use 4 bits? or 6/7? 8 Bits is no magic number - just as the 16 Million colours is a myth - in the sense that no digital image contains all 256*256*256 possibilities. Heyy - it might contain 3400*120*44 ... Its well documented in the 3D community that having 24 bit colour internally in 3D processing engines can result in banding in certain scenes, and thats why Nvidia and ATI have developed 32 bit engines, and more. I think that scanning to capture all the nuances and working from there is the sensible way. bert -- Linux - reaches the parts that other beers fail to reach. Unsubscribe by mail to [EMAIL PROTECTED], with 'unsubscribe filmscanners' or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or body
[filmscanners] Re: Digital Darkroom Computer Builders?
Anthony Atkielski wrote: Exactly the same thing was said of 32 bits, and 16 bits, and even 8 bits no doubt. Engineers _always_ get it wrong, and they _always_ refuse to believe that they should build in more capacity for the future. I doubt that, what was probably said was, ok, we can build an X bit design, but how can we stretch it for the current technology? Having worked in hardware engineering since the introduction of the 8088/86 architecture, I am much more amazed that we are still tied to its oddities. Its being stretched fairly thin. The problems of running out of physical memory are little to do with the limit of 2/4 Gig. Its far more to do with the MicroBloat OS and associated programs. Photoshop is a good example of something that comes from an atrocious memory model, and fails to ditch it for a better one - OSX should drag it into reality. We need 64 bit - not for better MM w.r.t programs, but addressability for massive file systems. Its arseholes like us who continue to accept bloat and poor coding - throwing more money and hardware at what we think is our problem. Its the Software Engineers fault, although I dread to think we can classify all programmers as SE's. You might note that Im a Unix/Linux user, who suffers far less than MSoft people. (As a non Mac user, whats the situation like regarding applications on OSX anyway?) AND! after watching the Greatest Briton program which looked at the work of Isambard Kingdom Brunel, I cannot believe you are referring to Engineers in such a way :) I regularly travel through his 1st tunnel (Wapping 1839), which has stood the test of being built for foot, and now carries tube trains. He was one hell of a man - and one staggering Engineer - gets my vote. bert -- Linux - reaches the parts that other beers fail to reach. Unsubscribe by mail to [EMAIL PROTECTED], with 'unsubscribe filmscanners' or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or body
[filmscanners] Latest developments in Scanners
Ok, something truly on topic. What are the latest developments in Film scanners that normal people might encounter in their filmscanning purchase options. Any real imporvements in dynamic range, bit depth, resolution (4000 seems to have been enough), low noise levels? Once Tony stopped reviewing scanners (did you?), there seems to be a hole in the 'review' market - and I would trust the list more anyway ... at least on this topic. bert -- Linux - reaches the parts that other beers fail to reach. Unsubscribe by mail to [EMAIL PROTECTED], with 'unsubscribe filmscanners' or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or body
Re: filmscanners: Go with Yahoo Groups
Hi Jack, Ive been offline for a while, and the Filmscanners archive is functioning even though I no longer work at the location of the server The archive exists at: http://phi.res.cse.dmu.ac.uk/Filmscan/ Its got 20,000 messages
Re: filmscanners: (anti)compression?
Jim Snyder wrote: [chop] you can stand a little bit of image quality loss, use ZIP [chop] H - this email list needs an FAQ - or some pointers to certain image FAQs on the web now and again. Image compression is a rather complex mathematical process that usually requires some 'dumping' of image data to gain good compression ratios - thus these compression schemes are 'lossy'. Non-lossy compression schemes use LZW type compressors which are good when there is a lot of replicated data in a file - but not so good for images that have a large variation of data components. The problem with most people is the mixup of file formats with compression schemes. For example, TIF can be compressed or uncompressed - it uses LZW to compress - but two TIF files are still called XXX.TIF and YYY.TIF even though one is raw data and one is compressed data. There is no such thing as an 'LZW' extension - only file formats that use it. Ive attached a small HTML doc with some specs. Not exact, but a guide - if anyone wants to add formats then do so. bert Filmscanners archive at: http://phi.res.cse.dmu.ac.uk/Filmscan/ Title: Compression File Extension Developed for? Compression Scheme Effect of compression % Saving for images Useable for? TIF image storage LZW or none lossless 15% archive copy ZIP general file store fancy LZW lossless 18% archive copy JPG (Joint Pictures expert Group) image storage JPEG lossy 80% non-archive (web!) GIF (good! interchange format) image storage LZW lossless 15% nasty 256 colour only PNG (portable network grahpics) image storage fancy LZW lossless 22% archive copy WIF (wavelet image format) image storage waveform mathematics lossy 95% proprietary FIF (fractal image format) image storage fractal mathematics lossy 90% proprietary PCD (Kodak PhotoCD) image storage fancy JPEG lossy 70% archive (but not perfect copy) FPX (Kodak Flashpix) image storage sortof JPEG / PCD mix lossy 80% non-archive web apps
Re: filmscanners: (anti)compression?
Lynn Allen penned: Although I haven't used it (some members have/do), PNG probably offers the best compression in a lossless format--according to the chart that Bert posted. Photoshop *does* offer that. Whether the format will be around in 20 years is another matter. :-) The classic question - will it be around. PNG is an open standard and offers a significant improvement for lossless compression over LZW with TIF files. Ive posted this example below, the reason is simple - the mathematics is more recent, so the compressor does better - every time. Will Photoshop be around? Or CD drives, or TIF? In my cupboard I have some 8 inch floppy disks ... but I moved stuff off of these when I saw that their end was nigh - TIF will go this way - as will PNG as will all :) As for support - it wont go away due to its growing use over GIF for non photographic images for web work (better features/ compression than GIF, and no software patent). Most web users dont even notice that an image is PNG - just a bit faster ... RAW TIF: 24532 Kb = 2500x3300@24bit LZW TIF: 20336 Kb = 17% smaller PNG: 16348 Kb = 33% smaller For a folder full of TIF scanned images (LZW) = 469 Megs. The same folder full compressed as PNG = 320 Megs. Two folders of images per CD ... Im no evangelist - just a conservationist :) bert
Re: filmscanners: ADMIN : PLEASE READ : MAIL BOUNCES
Tony, does your list account allow the use of 'procmail filtering? Or something of that ilk. I could knock up a script that filters the bounces and does something practical with them - (dump em, put them in a file, put them in another mailbox, parse them for a return path ... etc) bert
Re: filmscanners: Matrox G400 vs G450
NB: The G550 has just been released. One of the main considerations is the resolution you want to have in the secondary display - the G550 will allows a higher res on the second due to two RAMDACS (my G400 is limited to 1280x1024). The G450 is just a tidied up G400, the specs are a bit better but not great. bert
Re: filmscanners: Matrox G400 vs G450
= NB: The G550 has just been released. Whats the max resolution on main and second monitors? 2048x1536 and 1600x1200 for the G550 at v high refresh rates (100+) The G450 is the same, the G400 is lower. bert
filmscanners: FS2710 - Dying Scanner? Help!
Ok - my ever reliable Canon FS2710 seems to have developed a hardware problem and Im wondering if the cost of fixing etc is worth it. Ive only done about 1500 scans with it. Basically it powers up ok, and passes diagnostics in the CanoScan software, Vuescan recognises it, and does calibration ok, but it fails to do a scan in preview/ scan modes. Its out of warranty (Oct 99 cant remember if its one year, or two?) - so im wondering if anyone could hazarda guess as to how long/much a fix would take etc. Otherwise, what about cheap equivalents in the UK? Where to buy etc. Im thinking of 400-600 quid, and the Acer ScanWit 2740, the Dimage Dual II, refurb LS30 (where?). Im intersted in dynamic range/ depth rather than pure resolution - and occasional ICE (a wonderful drug!). bert Filmscanners archive: http://phi.res.cse.dmu.ac.uk/Filmscan/
filmscanners: List Archive (newbies monthly posting)
Just for any newbies etc. I keep a list archive (searchable) online for my own personal use - and its available to all. No adverts or profit - and nothing to do with Tony - except its his list. Its been archived since Jan 2000 - there are roughly 17700 messages in a nicely organised web setup - viewable by date/ thread/ author etc. http://phi.res.cse.dmu.ac.uk/Filmscan/ bert http://www.cse.dmu.ac.uk/~rl/
Re: filmscanners: FS2710
Did anyone sort out whay the FS2710 was producing such a washed out scan when using Vuescan? Well, it might have in the past, but using the latest version 7.0.24 on negs, I get a good full range of levels from say 5 - 250 on the scan straight out of VueScan, using default black and white limits. I have to agree - no problems with 7.0.24 here on both colour and black and white negatives. A small example here: http://phi.res.cse.dmu.ac.uk/~rl/Misc/estblackwhite.jpg bert
filmscanners: List Archive (newbies monthly posting)
Just for any newbies etc. I keep a list archive (searchable) online for my own personal use - and its available to all. No adverts or profit - and nothing to do with Tony - except its his list. Its been archived since Jan 2000 - there are roughly 16600 messages in a nicely organised web setup - viewable by date/ thread/ author etc. http://phi.res.cse.dmu.ac.uk/Filmscan/ bert http://www.cse.dmu.ac.uk/~rl/
RE: filmscanners: Canon FS2710 vs Minolta Dimage Scan Dual II
[Tim A wrote] I Have used the Canoscan with other scanners, but not the Minolta. I have found using Vuescan with the Canoscan does wonders. Even basic scans come out with less noise. Do a multi passes or the long exposure pass gives me great scans - adding shadow detail and getting rid of shadow noise. I have to agree - showed a mate the difference between Vuescan and CanoScan last night on one of his 'difficult' negs and he wept with joy (ummm). He spent about 2 hours with Photoshop trying to get a result to no effect. A bit more difficult to use, but a whole lot more effective. bert
filmscanners: List Archive (monthly posting)
Just for any newbies etc. I keep a list archive (searchable) online for my own personal use - and its available to all. No adverts or profit - and nothing to do with Tony - except its his list. Its been archived since Jan 2000 - there are roughly 14000 messages in a nicely organised web setup - viewable by date/ thread/ author etc. http://phi.res.cse.dmu.ac.uk/Filmscan/ bert http://www.cse.dmu.ac.uk/~rl/
Re: filmscanners: Canon FS2710 - any good?
Yes - its a nice scanner, and the Canon Software is ok for getting it to do simple scanning jobs. With Vuescan I can get a lot more out of it though. Its getting 'old' now and Ive been expecting a new model or two from Canon but Im not sure of its current price range w.r.t other better models. Easily set up too. This topic has been discussed a lot, and you can hunt for more detailed info on my archive of the list (no ads - just plain text, searchable etc). If you want some sample scans then I can stick them online for download - but at 20+ megs a pop they may be a bit too extreme for that. My favourite comparison is with PhotoCD, from where I got the desire to buy a scanner, and the Canon does better by a fair way. bert Archive: http://phi.res.cse.dmu.ac.uk/Filmscan/
filmscanners: Compression Formats (CD Storage)
Lossless: ie when you uncompress you get an exact of the original TIF LZW - compression 5-20% for photo images PNG - compression 10-40% for photo images Note that PNG will always be smaller due to a more efficient algorithm. Original: 24532 Kb = 2500x3300@24bit TIF LZW: 20336 Kb = 17% smaller PNG: 16348 Kb = 33% smaller Lossy: ie when u uncompress you get less than the original but much better compression JPG@15%: 1440 Kb JPG@25%: 990 Kb GIF: 6675 Kb PNG@8: 5172 Kb Note that JPG uses 24 bit colour and throws away data that the human eye/ brain ignores to some extent. GIF uses a 256 colour indexed system - for photo images it is terrible. PNG using 256 colour compresses smaller than GIF due to a more sophistcated algorithm. Conclusions (please contribute): TIF LZW is fine, but PNG is a lot better. Both are lossy but TIF has been around longer, and thus is better supported. PNG is realtively new, but it is open-source meaning that any software developer can use its code, which is freely available ... so if its not there then complain. Forget GIF. Its wa too lossy for photo images. For low colour images (256) it is fine, but PNG does a better job than GIF anyway for this type of image. Most browsers support JPG/GIF/PNG. PNG offers support for 8/24/48 bit colour, with better compression than TIF (significantly). Personally. I store in PNG format on CD, with a nice auto generated index.html file displaying thumbnail JPG images which lead to medium size 800xAAA JPG images and large 1400xAAA images. This gives me a full monitor view and a web capable view. bert unoffcial Filmscanners archive at: http://phi.res.cse.dmu.ac.uk/Filmscan/
Re: filmscanners: Re: List Archive (monthly posting)
Just for any newbies etc. I keep a list archive (searchable) online for my own personal use - and its available to all. No adverts or profit - and nothing to do with Tony - except its his list. Its been archived since Jan 2000 - there are roughly 11000 messages in a nicely organised web setup - viewable by date/ thread/ author etc. http://phi.res.cse.dmu.ac.uk/Filmscan/ bert http://www.cse.dmu.ac.uk/~rl/
filmscanners: Re: List Archive (monthly posting)
Just for any newbies etc. I keep a list archive (searchable) online for my own personal use - and its available to all. No adverts or profit - and nothing to do with Tony - except its his list. Its been archived since Jan 2000 - there are roughly 1 messages in a nicely organised web setup - viewable by date/ thread/ author etc. http://phi.res.cse.dmu.ac.uk/Filmscan/ bert http://www.cse.dmu.ac.uk/~rl/
Re: filmscanners: FS2710 questions
I noticed big decline is speed of scanning since I purchased Canon FS2710 some two weeks ago, especialy notable in high-res scans. When I got the scanner it was fairly fast, however, it now takes some time to scan full frame @ 2720 dpi. I know my system is rather slow and misses an big upgrade (K6/166MHz/64MB/2.0+3.2GB UDMA33 HDD) however what bothers me is that despite more or less same amount of free space on HDD for Win or PS swap scaning goes slower and slower. And I'm talking of just scanning, not time it takes to import into PS ... Yup - have suffered the same problem with my FS2710 which I realised was due to my hard disks being extremely badly fragmented with all the scan, edit, resize, reedit, copy, ... processes. This at the same time as my other applications were using the disk etc. On a slow machine, with limited memory, the disks also get used to create 'virtual memory' which means more fragmentation and slowdowns. Et voila, my disks go from one long spaghetti strand, to a chewed up mulch of chunks! (feeling hungry ..) These big scan files are notorious for fragmentation problems. I use a seperate disk (or partition) for scanning to alleviate this - then doing work which is saved onto other disks. Other things may be at issue - but once I started using a seperate disk my FS2710 was consistently fast (or slow - YMMV). bert