RE: filmscanners: image samples of digital artifacts
Rafe wrote-- JPG doesn't produce topo maps Ah, but it does! I'd refer you to the Aniversary picture on Larry Berman's Compression page. I found (as Larry did) that getting the original image below 120mb without posterizing was impossible. :-) Topo maps are a result of extreme posterization (loss of intermediate tones.) Indexed color is, by definition, a severely posterized working space. Using that conventional wisdom, I was completely baffled when a picture I was working on in Photoshop suddenly posterized in a skin-tone area. I do not use a limited palette (except in Amiga graphics). The causes in that incident are still unknown--it was a program glitch of some sort that I corrected by using a different program to get the results I wanted. :-) [Indexed color is] *Entirely* unsuitable for any graphic arts work. That's also a bit too broad to be true. Indexed color *does* have its uses in output applications. I'd refer you to the book Real Life Photoshop. Limited color has limited applications, OTOH. The typical signature of JPG is little blocks (8x8 pixels) that are clearly discernable in the image. That's true enough. However, the little buggers are more recognizable by their shimmerey off-color than as patterns. The rule of thumb is to push the compression just that far, then back off a few clicks. You can only do this with a few programs, Picture Publisher 8 being one of them. Best regards--LRA From: rafeb [EMAIL PROTECTED] Reply-To: [EMAIL PROTECTED] To: [EMAIL PROTECTED] Subject: RE: filmscanners: image samples of digital artifacts Date: Thu, 19 Jul 2001 20:41:23 -0400 At 10:08 PM 7/19/01 +, Lynn Allen wrote: Hi, Dan-- That looks like Posterization to me (at least, tha's whut ah calls it! :-) --cf definitions (-:|:-) ). I'd say it's probably a result (in this case, anyway) of pushing the sizing and JPEG compression too far. A good reference is Larry Berman's Compression Comparisons (BermanGraphics--You can look it up--I can't access the URL without losing my link on this service). I'm willing to bet that Dan Honemann has his video set to 256 colors (indexed color.) JPG doesn't produce topo maps Topo maps are a result of extreme posterization (loss of intermediate tones.) Indexed color is, by definition, a severely posterized working space. *Entirely* unsuitable for any graphic arts work. To see posterization in Photoshop, go to Image-Adjust-Posterize, and select a small integer, say 10 or so. Some of the effects are quite nice, in fact, but hardly photographic. Amazingly, if the integer is over 50-100 on a well- adjusted image, you won't see the posterization at all. Which is one reason that I think all this talk about needing 48-bit color is... well, missing the point somehow. 16 million colors seems to do the trick for me. 256-color (indexed color) associates 256 triplets of RGB values, with the integers 0..255. Those 256 triplets are called a pallette. The video card can switch between pallettes quickly, and may be able to store several pallettes in its memory. But it can only *use* one pallette at a time. This is how color video was done, typically, about 10 years ago, before True Color became the norm. JPG doesn't cause topo map or posterization effects. The typical signature of JPG is little blocks (8x8 pixels) that are clearly discernable in the image. rafe b. _ Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp
Re: filmscanners: image samples of digital artifacts
Rafe wrote: The sky in the Prarie photo looks smooth as silk on my PC, with 24 bit video. With the screen set to 256 colors I get topo maps in the sky. and Bob wrote: Thanks Rafe. Mine looked smooth as silk too. I couldn't figure out what I was suppose to be seeing and wasn't. Now I get it. OK, I'm not exactly sure what's going on here, that one display set to factory specs (mine) shows posterization in an Internet JPEG, and two others (Rafe's and Bob's) do not. Should Internet picture postings come with the caveat, Warning, This Picture Must Be Viewed At 48-Bits!? That doesn't sound altogether realistic, to me. :-) Best regards--LRA From: Bob Kehl - Kvernstoen, Kehl Assoc. [EMAIL PROTECTED] Reply-To: [EMAIL PROTECTED] To: [EMAIL PROTECTED] Subject: Re: filmscanners: image samples of digital artifacts Date: Thu, 19 Jul 2001 20:09:24 -0500 - Original Message - From: rafeb [EMAIL PROTECTED] To: [EMAIL PROTECTED] Sent: Thursday, July 19, 2001 7:00 PM Subject: RE: filmscanners: image samples of digital artifacts Hold everything! Do you mean, Prairie, Northern Tibet? If you're seeing topo map effects in the sky, it's almost certainly because you have your video set to 256 colors. There's no way you want to attempt ANY image editing or capture with your screen set that way. The sky in the Prarie photo looks smooth as silk on my PC, with 24 bit video. With the screen set to 256 colors I get topo maps in the sky. Thanks Rafe. Mine looked smooth as silk too. I couldn't figure out what I was suppose to be seeing and wasn't. Now I get it. Actually, no-one COULD edit photos at 256 colors but they might try at 16 bit. At 16 bit the topo map effect is clearly visible too. I think you found the problem. BK _ Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp
Re: filmscanners: RE: filmscanners: image samples of digital artifacts
Well, this may be what Dan Honemann is up against on his notebook computer. I told him to ditch it. That's a little extreme, Rafe. :-) Granted that an LCD is not suited to *working* on graphics, it's viable for *viewing* them. Still, if Dan throws out his Dell Inspiron, I hope he throws it in my direction--I could use a portable backup, and could keep up with the List while I'm fishing or on vacation. g Best regards--LRA From: rafeb [EMAIL PROTECTED] Reply-To: [EMAIL PROTECTED] To: [EMAIL PROTECTED] Subject: Re: filmscanners: RE: filmscanners: image samples of digital artifacts Date: Thu, 19 Jul 2001 23:23:43 -0400 At 11:33 AM 7/20/01 +1000, Rob wrote: Rafe wrote: I'm willing to bet that Dan Honemann has his video set to 256 colors (indexed color.) Some video drivers in Windows (particularly the generic Windows ones as opposed to OEM) only display 256 colours despite being set to 16bit or 24bit. It was one reason I had to throw out a video card when I went from Win 3.11 to Win95. Well, this may be what Dan Honemann is up against on his notebook computer. I told him to ditch it. rafe b. _ Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp
Re: filmscanners: image samples of digital artifacts
On Fri, 20 Jul 2001 01:41:02 +1000 Rob Geraghty ([EMAIL PROTECTED]) wrote: I can't think of a meaningful picture of grain aliasing. It could be described with a drawing, not with an real life scan because by nature it is random. No, I have scans of the same neg showing the effect very strikingly. You'll have to wait a while longer though. Regards Tony Sleep http://www.halftone.co.uk - Online portfolio exhibit; + film scanner info comparisons
Re: filmscanners: image samples of digital artifacts
At 10:48 AM 20/07/01 +, Lynn wrote: OK, I'm not exactly sure what's going on here, that one display set to factory specs (mine) shows posterization in an Internet JPEG, and two others (Rafe's and Bob's) do not. Should Internet picture postings come with the caveat, Warning, This Picture Must Be Viewed At 48-Bits!? That doesn't sound altogether realistic, to me. :-) I also see a 'clean' sky in 32-bit but posterisation in 16-. Could it be that your browser is working in 16-bit mode..? I've seen some older browsers do this - check the options for any funny switches, or perhaps you should bravely upgrade from Netscape v1.0, Lynn? ;) I'm assuming you've checked all the obvious stuff, like your kids sneakily putting your video card into 16bit mode for a game. I've also noticed that flicking from one res/color depth to another can sometimes results in odd effects when you go back to the higher settings, so maybe it just needs a boot.. Failing all that, take your monitor/video card back to the supplier (at last, a non-Nikon fault!) MarkT.
Re: filmscanners: image samples of digital artifacts
At 10:48 AM 7/20/01 +, Lynn wrote: Rafe wrote: The sky in the Prarie photo looks smooth as silk on my PC, with 24 bit video. With the screen set to 256 colors I get topo maps in the sky. and Bob wrote: Thanks Rafe. Mine looked smooth as silk too. I couldn't figure out what I was suppose to be seeing and wasn't. Now I get it. OK, I'm not exactly sure what's going on here, that one display set to factory specs (mine) shows posterization in an Internet JPEG, and two others (Rafe's and Bob's) do not. Should Internet picture postings come with the caveat, Warning, This Picture Must Be Viewed At 48-Bits!? That doesn't sound altogether realistic, to me. :-) I'm viewing it at 24 bits and it's fine. The one video option that's not acceptable is 256 colors. This is also refered to as indexed color. Now it's also possible that the video driver or software in Dan's notebook is *using* an indexed-color mechansim while appearing to operate as 16-bit TrueColor. This is where you may need to dig deeper into the tech details of your video hardware. And of course the vendor/manufacturer of that notebook may not be altogether up-front with the necessary details. I think there's a good deal of evidence to indicate that notebook computers, with LCD screens, are poorly suited to viewing and/or editing color graphics. This is asking for trouble. In your case, Lynn, I'm more puzzled. Can you describe the hardware and software options on the system you're using that yields posterization on the Tibet jpg? rafe b.
Re: filmscanners: image samples of digital artifacts
Tony wrote (re grain aliasing): No, I have scans of the same neg showing the effect very strikingly. You'll have to wait a while longer though. I will wait, but since *you're* the one who sent us off in search of this Holy Grail, it's only appropriate that we see your examples, one day. :-) Best regards--LRA From: [EMAIL PROTECTED] (Tony Sleep) Reply-To: [EMAIL PROTECTED] To: [EMAIL PROTECTED] Subject: Re: filmscanners: image samples of digital artifacts Date: Fri, 20 Jul 2001 12:36 +0100 (BST) On Fri, 20 Jul 2001 01:41:02 +1000 Rob Geraghty ([EMAIL PROTECTED]) wrote: I can't think of a meaningful picture of grain aliasing. It could be described with a drawing, not with an real life scan because by nature it is random. No, I have scans of the same neg showing the effect very strikingly. You'll have to wait a while longer though. Regards Tony Sleep http://www.halftone.co.uk - Online portfolio exhibit; + film scanner info comparisons _ Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp
filmscanners: image samples of digital artifacts
Is there an online tutorial/FAQ/glossary somewhere that shows image samples of various digital artifacts (e.g., banding, grain-aliasing, jaggies, etc.)? I'm a newbie to all this, and Tony's glossary at halftone is a help but doesn't show pics. Here, I think, sample images would be worth a thousand words. Dan
Re: filmscanners: image samples of digital artifacts
Dan Honemann [EMAIL PROTECTED] wrote: Is there an online tutorial/FAQ/glossary somewhere that shows image samples of various digital artifacts (e.g., banding, grain-aliasing, jaggies, etc.)? I should put some examples of jaggies on my web site. Thankfully, Nikon finally seems to have fixed the problem with Nikonscan 3.1. Vuescan does better scans from my LS30 however. :) I'm a newbie to all this, and Tony's glossary at halftone is a help but doesn't show pics. Here, I think, sample images would be worth a thousand words. Wayne Fulton's scanning FAQ may have some of that sort of thing. I don't have the URL though. I can't think of a meaningful picture of grain aliasing. It could be described with a drawing, not with an real life scan because by nature it is random. The closest analogy is the moire patterns you get when scanning offset printed magazine pictures with a flatbed at certain ppi settings. Rob
Re: filmscanners: image samples of digital artifacts
On Thu, 19 Jul 2001, Dan Honemann wrote: Is there an online tutorial/FAQ/glossary somewhere that shows image samples of various digital artifacts (e.g., banding, grain-aliasing, jaggies, etc.)? You mean, like a Madame-Tussaud's wax museum of film scanner horrors? Sounds ghastly. Just buy one. Any brand. You'll learn soon enough. g rafe b.
Re: filmscanners: image samples of digital artifacts
Rob wrote (re grain-aliasing)-- The closest analogy is the moire patterns you get when scanning offset printed magazine pictures with a flatbed at certain ppi settings. This makes the exact point of my earlier post--that's not how I'd describe it, at all (and the Acer can grain-alias with the best of them)! :-) Best regards--LRA _ Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp
Re: filmscanners: image samples of digital artifacts
Dan wrote: Is there an online tutorial/FAQ/glossary somewhere that shows image samples of various digital artifacts (e.g., banding, grain-aliasing, jaggies, etc.)? I'm a newbie to all this, and Tony's glossary at halftone is a help but doesn't show pics. Here, I think, sample images would be worth a thousand words. Hoo, boy, that *would* be useful! Presently, every definition is about a half-click away from the next guy's definition. If I had a website, I'd give it a go (I've got *plenty* of examples!)--maybe some kind-sprited, web-savvy member will do it? Best regards--LRA _ Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp
RE: filmscanners: image samples of digital artifacts
Lynn, I would be glad to contribute the web space and storage for this - I would love to see examples of the terms used by everyone! /fn (email to: [EMAIL PROTECTED]) -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On Behalf Of Lynn Allen Sent: Thursday, July 19, 2001 10:11 AM To: [EMAIL PROTECTED] Subject: Re: filmscanners: image samples of digital artifacts Dan wrote: Is there an online tutorial/FAQ/glossary somewhere that shows image samples of various digital artifacts (e.g., banding, grain-aliasing, jaggies, etc.)? I'm a newbie to all this, and Tony's glossary at halftone is a help but doesn't show pics. Here, I think, sample images would be worth a thousand words. Hoo, boy, that *would* be useful! Presently, every definition is about a half-click away from the next guy's definition. If I had a website, I'd give it a go (I've got *plenty* of examples!)--maybe some kind-sprited, web-savvy member will do it? Best regards--LRA _ Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp
RE: filmscanners: image samples of digital artifacts
Lynn, Rafe, Rob and others: One thing I've always been curious about is what causes the topographical map type of lines you see in the blue sky portion of this image: http://www.chebucto.ns.ca/~taiji/gallery/t21.htm ??? I see this sort of artifact a lot in jpegs on the web. Is this what is called jaggies? Do they show up in prints? Thanks, Dan
RE: filmscanners: image samples of digital artifacts
Frank wrote: I would be glad to contribute the web space and storage for this - I would love to see examples of the terms used by everyone! Count me in for samples! (even though I'll have to go back through and retrieve the originals--stuff I've fixed doesn't count). :-) I haven't had time to learn much about web presentation--set your parameters (file sizes, etc) and I'll try to comply. Best regards and luck--LRA -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On Behalf Of Lynn Allen Sent: Thursday, July 19, 2001 10:11 AM To: [EMAIL PROTECTED] Subject: Re: filmscanners: image samples of digital artifacts Dan wrote: Is there an online tutorial/FAQ/glossary somewhere that shows image samples of various digital artifacts (e.g., banding, grain-aliasing, jaggies, etc.)? I'm a newbie to all this, and Tony's glossary at halftone is a help but doesn't show pics. Here, I think, sample images would be worth a thousand words. Hoo, boy, that *would* be useful! Presently, every definition is about a half-click away from the next guy's definition. If I had a website, I'd give it a go (I've got *plenty* of examples!)--maybe some kind-sprited, web-savvy member will do it? Best regards--LRA _ Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp _ Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp
RE: filmscanners: image samples of digital artifacts
Hi, Dan-- That looks like Posterization to me (at least, tha's whut ah calls it! :-) --cf definitions (-:|:-) ). I'd say it's probably a result (in this case, anyway) of pushing the sizing and JPEG compression too far. A good reference is Larry Berman's Compression Comparisons (BermanGraphics--You can look it up--I can't access the URL without losing my link on this service). No, it's not jaggies. Jaggies are usually those obvious stair-steps you sometimes see on contrasty diagonals in the picture, a result of not enough anti-aliasing or too few colors (posterization is also a result of too few colors). Rob G, OTOH has all sorts of dagger-shaped jaggies produced by his LS30 stepper and/or software. Here again, same term, different visual appearance. Best reagards--LRA From: Dan Honemann [EMAIL PROTECTED] Reply-To: [EMAIL PROTECTED] To: [EMAIL PROTECTED] Subject: RE: filmscanners: image samples of digital artifacts Date: Thu, 19 Jul 2001 17:44:41 -0400 Lynn, Rafe, Rob and others: One thing I've always been curious about is what causes the topographical map type of lines you see in the blue sky portion of this image: http://www.chebucto.ns.ca/~taiji/gallery/t21.htm ??? I see this sort of artifact a lot in jpegs on the web. Is this what is called jaggies? Do they show up in prints? Thanks, Dan _ Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp
RE: filmscanners: image samples of digital artifacts
--- Dan Honemann [EMAIL PROTECTED] wrote: One thing I've always been curious about is what causes the topographical map type of lines you see in the blue sky portion of this image: The old JPEG (not JPEG 2000) does code three channels Y, Cr, Cb. The channels Cr and Cb are downsampled. Then each channel is divided in blocks of 8x8. For each such block you do a Discret Cosinus Transform (DCT), devide each of the 64 resulting values by one of 64 numbers defined by the quantization table (higher frequency values are divided by higher numbers then low frequency values), and then Huffman (arithmetic coding is also possible but is less common) entropy encoded. This is true for lossy compression. Now if you do a high compression you divide the values after the DCT by higher factors so you get more 0s. Because of that the transition of one 8x8 block becomes less smooth and you see 8x8 block in the final image. Robert __ Do You Yahoo!? Get personalized email addresses from Yahoo! Mail http://personal.mail.yahoo.com/
filmscanners: Re: filmscanners: image samples of digital artifacts
Lynn wrote: Hoo, boy, that *would* be useful! Presently, every definition is about a half-click away from the next guy's definition. If I had a website, I'd give it a go (I've got *plenty* of examples!)-- maybe some kind-sprited, web-savvy member will do it? I'd be happy to put things online provided the examples are appropriately sized. I already have a page about scanning to explain the work which was being done on looking at film types. It would be good to have examples of things which are problematic about scanning. Rob Rob Geraghty [EMAIL PROTECTED] http://wordweb.com
filmscanners: RE: filmscanners: image samples of digital artifacts
Dan wrote: One thing I've always been curious about is what causes the topographical map type of lines you see in the blue sky portion of this image: http://www.chebucto.ns.ca/~taiji/gallery/t21.htm I see this sort of artifact a lot in jpegs on the web. Is this what is called jaggies? Do they show up in prints? You need more colours. This looks fine in 24bit on my work computer. You may be running less than 24bit colour. Depending on the OS some video drivers don't display a full palette of 24bit colour even though the driver claims to be set to it. So no, it's not jaggies exactly - it's your video card dithering the colours down to what fits in your palette. Rob Rob Geraghty [EMAIL PROTECTED] http://wordweb.com
RE: filmscanners: image samples of digital artifacts
I hate to admit this and invite pressure :), but I have been collecting some bits and pieces for exactly this purpose.. My initial plan was to use microphotographs as well as scan samples to show how the grain-aliasing on my Acer is indeed 'set off' by real grain, and also to show how grain patterns vary in different colours/densities on negs. I once found a web-site with some of this, but do you reckon I can find it now..? I still plan to do that, but may as well toss in other defects as well.. (Although I don't have a Nikon ;), so I'll need to get permission from some kind soul to add some banding/jaggie samples.) If anyone else wants to contribute or make suggestions on other defects I should try to document, feel free. In the meantime, Pete's site at http://www.photoscientia.co.uk/Grain.htm has some good g-a samples and explanations. Mark T. At 02:29 PM 19/07/01 -0600, you wrote: Lynn, I would be glad to contribute the web space and storage for this - I would love to see examples of the terms used by everyone! /fn (email to: [EMAIL PROTECTED]) -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On Behalf Of Lynn Allen Sent: Thursday, July 19, 2001 10:11 AM To: [EMAIL PROTECTED] Subject: Re: filmscanners: image samples of digital artifacts Dan wrote: Is there an online tutorial/FAQ/glossary somewhere that shows image samples of various digital artifacts (e.g., banding, grain-aliasing, jaggies, etc.)? I'm a newbie to all this, and Tony's glossary at halftone is a help but doesn't show pics. Here, I think, sample images would be worth a thousand words. Hoo, boy, that *would* be useful! Presently, every definition is about a half-click away from the next guy's definition. If I had a website, I'd give it a go (I've got *plenty* of examples!)--maybe some kind-sprited, web-savvy member will do it? Best regards--LRA _ Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp
RE: filmscanners: image samples of digital artifacts
At 05:44 PM 7/19/01 -0400, you wrote: Lynn, Rafe, Rob and others: One thing I've always been curious about is what causes the topographical map type of lines you see in the blue sky portion of this image: http://www.chebucto.ns.ca/~taiji/gallery/t21.htm ??? I see this sort of artifact a lot in jpegs on the web. Is this what is called jaggies? Do they show up in prints? Hold everything! Do you mean, Prairie, Northern Tibet? If you're seeing topo map effects in the sky, it's almost certainly because you have your video set to 256 colors. There's no way you want to attempt ANY image editing or capture with your screen set that way. This is something you'd change (on a PC) using Control Panel-Display-Settings. What you want is True Color, most likely 24 bits. Using 24 bits with a high resolution requires a video card with a decent amount of video RAM. So you may find that some of the higher resolution settings are grayed out when you select 24 bit color. The sky in the Prarie photo looks smooth as silk on my PC, with 24 bit video. With the screen set to 256 colors I get topo maps in the sky. Get yourself an up to date video card, with at least 8 or 16 Mbytes of video RAM. Matrox is a decent pick for graphic arts and 2-D images. Jaggies are an altogether different matter; they're a consequence of scanning and/or printing at too low a resolution. For example, if you were to try to grab this little image off the web, and print it as 8x10 on your Epson, you'd get jaggies. There are ways to smooth out jaggies, but they invariably involve softening the image. rafe b.
RE: filmscanners: image samples of digital artifacts
At 10:08 PM 7/19/01 +, Lynn Allen wrote: Hi, Dan-- That looks like Posterization to me (at least, tha's whut ah calls it! :-) --cf definitions (-:|:-) ). I'd say it's probably a result (in this case, anyway) of pushing the sizing and JPEG compression too far. A good reference is Larry Berman's Compression Comparisons (BermanGraphics--You can look it up--I can't access the URL without losing my link on this service). I'm willing to bet that Dan Honemann has his video set to 256 colors (indexed color.) JPG doesn't produce topo maps Topo maps are a result of extreme posterization (loss of intermediate tones.) Indexed color is, by definition, a severely posterized working space. *Entirely* unsuitable for any graphic arts work. To see posterization in Photoshop, go to Image-Adjust-Posterize, and select a small integer, say 10 or so. Some of the effects are quite nice, in fact, but hardly photographic. Amazingly, if the integer is over 50-100 on a well- adjusted image, you won't see the posterization at all. Which is one reason that I think all this talk about needing 48-bit color is... well, missing the point somehow. 16 million colors seems to do the trick for me. 256-color (indexed color) associates 256 triplets of RGB values, with the integers 0..255. Those 256 triplets are called a pallette. The video card can switch between pallettes quickly, and may be able to store several pallettes in its memory. But it can only *use* one pallette at a time. This is how color video was done, typically, about 10 years ago, before True Color became the norm. JPG doesn't cause topo map or posterization effects. The typical signature of JPG is little blocks (8x8 pixels) that are clearly discernable in the image. rafe b.
Re: filmscanners: image samples of digital artifacts
- Original Message - From: rafeb [EMAIL PROTECTED] To: [EMAIL PROTECTED] Sent: Thursday, July 19, 2001 7:00 PM Subject: RE: filmscanners: image samples of digital artifacts Hold everything! Do you mean, Prairie, Northern Tibet? If you're seeing topo map effects in the sky, it's almost certainly because you have your video set to 256 colors. There's no way you want to attempt ANY image editing or capture with your screen set that way. The sky in the Prarie photo looks smooth as silk on my PC, with 24 bit video. With the screen set to 256 colors I get topo maps in the sky. Thanks Rafe. Mine looked smooth as silk too. I couldn't figure out what I was suppose to be seeing and wasn't. Now I get it. Actually, no-one COULD edit photos at 256 colors but they might try at 16 bit. At 16 bit the topo map effect is clearly visible too. I think you found the problem. BK
RE: filmscanners: image samples of digital artifacts
Robert, The old JPEG (not JPEG 2000) does code three channels Y, Cr, Cb. The channels Cr and Cb are downsampled. Then each channel is divided in blocks of 8x8. For each such block you do a Discret Cosinus Transform (DCT), devide each of the 64 resulting values by one of 64 numbers defined by the quantization table (higher frequency values are divided by higher numbers then low frequency values), and then Huffman (arithmetic coding is also possible but is less common) entropy encoded. This is true for lossy compression. Now if you do a high compression you divide the values after the DCT by higher factors so you get more 0s. Because of that the transition of one 8x8 block becomes less smooth and you see 8x8 block in the final image. I guessed as much, but I had thought it was Guffman, not Huffman, and I think I forgot to carry a 1. ;) Dan
RE: filmscanners: image samples of digital artifacts
Thanks, Lynn! I look forward to whatever artifact samples you care to share. :) Dan
filmscanners: RE: filmscanners: image samples of digital artifacts
Rafe wrote: I'm willing to bet that Dan Honemann has his video set to 256 colors (indexed color.) Some video drivers in Windows (particularly the generic Windows ones as opposed to OEM) only display 256 colours despite being set to 16bit or 24bit. It was one reason I had to throw out a video card when I went from Win 3.11 to Win95. Rob Rob Geraghty [EMAIL PROTECTED] http://wordweb.com
RE: filmscanners: image samples of digital artifacts
Rafe, I'm willing to bet that Dan Honemann has his video set to 256 colors (indexed color.) It was set to 16-bit (True Color), so I changed it to 24-bit (High Color) and rebooted. Still see the lines in the sky, but this is only a Dell Inspiron 3500 notebook PC with a NeoMagic MagicMedia 256AV card and a 14 LCD screen. No doubt something in that mix isn't up to snuff. Dan
Re: filmscanners: RE: filmscanners: image samples of digital artifacts
At 11:33 AM 7/20/01 +1000, Rob wrote: Rafe wrote: I'm willing to bet that Dan Honemann has his video set to 256 colors (indexed color.) Some video drivers in Windows (particularly the generic Windows ones as opposed to OEM) only display 256 colours despite being set to 16bit or 24bit. It was one reason I had to throw out a video card when I went from Win 3.11 to Win95. Well, this may be what Dan Honemann is up against on his notebook computer. I told him to ditch it. rafe b.
Re: filmscanners: image samples of digital artifacts
I had this problem for a while and nothing seemed to fix it until it went away by itself. Doesn't seem to have to do with color bit-depth. Wish I could be of more help. John M. It was set to 16-bit (True Color), so I changed it to 24-bit (High Color) and rebooted. Still see the lines in the sky, but this is only a Dell Inspiron 3500 notebook PC with a NeoMagic MagicMedia 256AV card and a 14 LCD screen. No doubt something in that mix isn't up to snuff. Dan
filmscanners: RE: filmscanners: image samples of digital artifacts
Dan wrote: It was set to 16-bit (True Color), so I changed it to 24-bit (High Color) and rebooted. Still see the lines in the sky, but this is only a Dell Inspiron 3500 notebook PC with a NeoMagic MagicMedia 256AV card and a 14 LCD screen. No doubt something in that mix isn't up to snuff. Try hooking up a proper monitor. The LCD is probably very limited in the number of colours it can display. Rob Rob Geraghty [EMAIL PROTECTED] http://wordweb.com
