Photo Forum / Digital Photography / Digital Photo / November 2005
TIFF OR JPEG
|
|
Thread rating:  |
stevenjon1 - 23 Nov 2005 07:57 GMT I have to scan several thousand old 35mm negs. 90% of them are in good shape, the rest are a bit faded or otherwise damaged. All of them would need 10mins to 1 hour work before printing to a standard that I would be happy with. Theoretcially they should be saved in a lossless format like TIFF but that needs a lot more storage space than I have or can afford. Saving them on cd's at high resolution means having an awful lot of cd's. Can someone give me an idea of just how much the image of a jpeg suffers if its opened, cropped, smudges removed, levels altered then sharpened before closing? Is it something that would be noticed on a monitor, a 6"*8" print, a photo publised in a guide book? None of my photos are ever going to be enlarged much more tha 6*8" and exhibited as 'art'. Thanks for any comments.
Pete D - 23 Nov 2005 09:01 GMT Don't use CD's use DVD's, cheap and hold lots more data.
By the sound of it as long as you save in the highest quality jpg it will work okay, of course you should do a few tests to see if the result is acceptable. Remember a 400Gb drive is only $150.
>I have to scan several thousand old 35mm negs. 90% of them are in good >shape, the rest are a bit faded or otherwise damaged. All of them would [quoted text clipped - 8 lines] > going to be enlarged much more tha 6*8" and exhibited as 'art'. > Thanks for any comments. Captain Blammo - 23 Nov 2005 13:07 GMT I am still shocked by the quality losses with even the gentlest JPEG compression. You can save a lot of space by applying lossless compression to your TIFF file (either through photoshop or just using a file compression utility). I just popped an image out of my camera and tried photoshop's TIFF compression when saving. It pared a 20.2MB TIFF file down to 9MB, which all things said and done isn't that much bigger than the 3.8MB JPEG option on a quality settings of 12.
If you had 4000 9MB compressed TIFF files, that's only 36GB; which is eight 4.7Gb DVD-Rs. That would have been 80GB and 18 DVD-Rs with no compression.
Making 3.8MB JPEG files of them would only get you down to 15.2GB, which is still 4 DVD-Rs, and you take a huge hit in quality that you can never get back. I'd stick with TIFF. Even if you can't notice the difference in quality, it's entirely possible that you will later and end up kicking yourself. The additional 4 DVD-Rs will cost you about $4. That seems like a pretty reasonable investment!
CB
> Don't use CD's use DVD's, cheap and hold lots more data. > [quoted text clipped - 14 lines] > > going to be enlarged much more tha 6*8" and exhibited as 'art'. > > Thanks for any comments. Dirty Harry - 23 Nov 2005 21:42 GMT > I am still shocked by the quality losses with even the gentlest JPEG > compression. You can save a lot of space by applying lossless compression to [quoted text clipped - 35 lines] > > > going to be enlarged much more tha 6*8" and exhibited as 'art'. > > > Thanks for any comments. Does Photoshop compress tiffs by default?? I only get compression options for the layers.
Captain Blammo - 23 Nov 2005 23:53 GMT > Does Photoshop compress tiffs by default?? I only get compression options > for the layers. I only saved a single layer, and when I chose "TIFF" as the file type and pressed save, the resulting dialog had an "Image compression" section that let you chose none. LZW, ZIP or JPEG. It was set to "none" by default.
I'm using PS 7.0
CB
Dirty Harry - 24 Nov 2005 04:54 GMT > > Does Photoshop compress tiffs by default?? I only get compression options > > for the layers. [quoted text clipped - 6 lines] > > CB Thanks I guess i just needed to open my eyes. I'm thinking of batch processing all of my tiffs to compressed tiffs, i imagine I'd have 10 gigs of free space extra.
Martin Brown - 24 Nov 2005 14:35 GMT > I am still shocked by the quality losses with even the gentlest JPEG > compression. Bullshit. The difference between a first generation JPEG compressed photographic image at its highest quality and the original cannot normally be seen by the human eye. You need to demonstrate one of these "shocking" quality losses if you wish to retain any credibilty.
Multiple load and save of JPEGs has an additional price but the initial losses are miniscule for all but the most pathological test cases.
The only situations where it is really possible to see JPEG artefacts at highest quality are when line artwork is compressed. And even then with careful choice of the quantisation table it is possible to do a lot better than most people would imagine.
Regards, Martin Brown
netkiller - 25 Nov 2005 01:50 GMT don't used TIFF.pls used Raw.
Captain Blammo - 25 Nov 2005 08:06 GMT > You need to demonstrate one of these > "shocking" quality losses if you wish to retain any credibilty. To be fair, I never had any credibility to retain. I'm just some random bloke off the internet :-)
> The only situations where it is really possible to see JPEG artefacts at > highest quality are when line artwork is compressed. And even then with > careful choice of the quantisation table it is possible to do a lot > better than most people would imagine. That would actually explain a lot, the nasty effects I saw were on graphics tablet drawings. The lines became dull and unsharp. That said, I also noticed that some subtle tonal transitions get quite butchered (made using large brushes set to a low flow rate). It's not that it makes them look *bad* as such, but it *does* make them look rather different.
I tried a few tests on some photos, and I indeed couldn't see the difference at all, not a pixel's worth of it. Of course, I assume that there are certain types of image that will exhibit the nasty changes I saw in the drawings, if the content is right? That possibility alone is enough to keep me off the JPEGs now that I've seen how bad they can get.
Well, thanks for setting me straight on the conditions for a nasty compression! Er, how does one choose a quantisation table?
CB
Martin Brown - 28 Nov 2005 09:32 GMT >>You need to demonstrate one of these >>"shocking" quality losses if you wish to retain any credibilty. > > To be fair, I never had any credibility to retain. I'm just some random > bloke off the internet :-) Fair enough. Too many people blame JPEG unfairly for lots of things...
>>The only situations where it is really possible to see JPEG artefacts at >>highest quality are when line artwork is compressed. And even then with [quoted text clipped - 6 lines] > using large brushes set to a low flow rate). It's not that it makes them > look *bad* as such, but it *does* make them look rather different. I am a bit surprised that PhotoShop level 12 has failed on them. I would genuinely like to see a small example of these data. An small example of line art and the damage JPEG with/without chroma subsampling is at:
http://www.nezumi.demon.co.uk/photo/jpeg/2/jpeg2.htm
> I tried a few tests on some photos, and I indeed couldn't see the difference > at all, not a pixel's worth of it. Of course, I assume that there are > certain types of image that will exhibit the nasty changes I saw in the > drawings, if the content is right? That possibility alone is enough to keep > me off the JPEGs now that I've seen how bad they can get. The only types of image where serious damage will result belong to the large hyper space of pseudo random pathological noise like images. I have a few examples designed to break JPEG, but you will never ever photograph a target that looks even remotely like one.
> Well, thanks for setting me straight on the conditions for a nasty > compression! Er, how does one choose a quantisation table? Wizard mode on the IJG codec. A few of the most recent cameras are actually using all 1's in their JPEG quantisation table (eg Pentax istDS). I think this is overkill.
Regards, Martin Brown
Captain Blammo - 30 Nov 2005 14:42 GMT > I would genuinely like to see a small example of these data. > An small example of line art and the damage JPEG with/without > chroma subsampling is at: No probs, I'll post it in a couple of days when I can dig out the files and find 5 minutes free. Stay tuned :-)
CB
Ron Hunter - 23 Nov 2005 09:44 GMT > I have to scan several thousand old 35mm negs. 90% of them are in good > shape, the rest are a bit faded or otherwise damaged. All of them would need [quoted text clipped - 8 lines] > enlarged much more tha 6*8" and exhibited as 'art'. > Thanks for any comments. DON'T! Scan to .TIFF, then edit, THEN save to .jpg. Do it in batches, then, if you must, delete the .TIFF files. It is probably cheaper to get a really large external disk drive than to mess with all those CD/DVD backups.
 Signature Ron Hunter rphunter@charter.net
Ian Tindale - 23 Nov 2005 10:41 GMT Back them up onto cassette.
traction - 23 Nov 2005 10:54 GMT >I have to scan several thousand old 35mm negs. 90% of them are in good >shape, the rest are a bit faded or otherwise damaged. All of them would [quoted text clipped - 8 lines] > going to be enlarged much more tha 6*8" and exhibited as 'art'. > Thanks for any comments. How many do you expect to use in this book? Why scan them all at this standard, when perhaps 1% will ever be used. Scan them all at a nominal 300DPI at 6x8, view them , decide which you want , and spend your time scanning the quality ones. Been there, seen it, done it, and now I have shoeboxes full of CD's that never even get opened.!!
Martin Brown - 23 Nov 2005 12:33 GMT > I have to scan several thousand old 35mm negs. 90% of them are in good > shape, the rest are a bit faded or otherwise damaged. All of them would need > 10mins to 1 hour work before printing to a standard that I would be happy > with. Theoretcially they should be saved in a lossless format like TIFF but > that needs a lot more storage space than I have or can afford. Saving them > on cd's at high resolution means having an awful lot of cd's. Keep work in progress as TIFF files and save the final result as high quality JPEGs. Any that are really good keep the TIFF as well.
> Can someone give me an idea of just how much the image of a jpeg suffers if > its opened, cropped, smudges removed, levels altered then sharpened before > closing? Is it something that would be noticed on a monitor, a 6"*8" print, > a photo publised in a guide book? None of my photos are ever going to be > enlarged much more tha 6*8" and exhibited as 'art'. > Thanks for any comments. Unless you know exactly what to look for and there is suitable detail in the target photographic image the difference between a first generation JPEG and the original is beyond the ability of the human eye to distinguish. Line artwork and pathological patterns can be constructed that will give JPEG a headache but even then the losses are small.
Regards, Martin Brown
Steven Feinstein - 23 Nov 2005 14:50 GMT > I have to scan several thousand old 35mm negs. 90% of them are in good > shape, the rest are a bit faded or otherwise damaged. All of them would [quoted text clipped - 9 lines] > going to be enlarged much more tha 6*8" and exhibited as 'art'. > Thanks for any comments. You will see a lot of advice on the "best" way to do this. But, there is always a cost/benefit to doing anything and you need to decide if the extra cost of the file size is worth the difference in image quality (as an example, people will look for lower cost gas to save money, but will they drive 55 instead of 65 to get better mileage? Cost vs benefit).
I've done some tests on my own images. There is obviously a difference between JPG and TIFF (otherwise why would everyone recommend TIFF). The question is, is there enough of a difference. For my own images, I decided no. I'd rather save the space. In my opinion, you will have a difficult time noticing the differences between a TIFF and a 2nd generation JPG printed at 6x8. Is there a difference, sure. Side by side you will say the TIFF looks better. But as you said, these are for a guide book not for hanging on a wall. Will you be happy with the JPG version? I'm guessing yes, but only you can tell.
So, my recommendation is, scan a negative twice, saving one as TIFF the other as JPG. Make the same change to both. Save the TIFF as TIFF and the JPG as JPG again. Then print both final images 6x8. See if you care about the difference.
When you decide what you want to do, I recommend while working on the image to save in the format of your image software until you are done making changes. The one hour changes may take a couple of sessions.
BTW, another option would be to scan at TIFF and convert to PNG (or scan to PNG if your scanning software allows it). PNGs are about 1/3rd the size of TIFFs. Downside is less software supports PNG and they are much slower to open and close.
Steve
David Dyer-Bennet - 23 Nov 2005 16:30 GMT > BTW, another option would be to scan at TIFF and convert to PNG (or scan to > PNG if your scanning software allows it). PNGs are about 1/3rd the size of > TIFFs. Downside is less software supports PNG and they are much slower to > open and close. Um, with what options? Yeah, if you reduce to an 8-bit palletted PNG, then it'll be about 1/3 the size of the TIFF. But a 24-bit PNG isn't notably smaller than a 24-bit TIFF that I've noticed (using the same compression options in each).
 Signature David Dyer-Bennet, <mailto:dd-b@dd-b.net>, <http://www.dd-b.net/dd-b/> RKBA: <http://noguns-nomoney.com/> <http://www.dd-b.net/carry/> Pics: <http://dd-b.lighthunters.net/> <http://www.dd-b.net/dd-b/SnapshotAlbum/> Dragaera/Steven Brust: <http://dragaera.info/>
Steven Feinstein - 23 Nov 2005 18:00 GMT >> BTW, another option would be to scan at TIFF and convert to PNG (or scan >> to [quoted text clipped - 7 lines] > notably smaller than a 24-bit TIFF that I've noticed (using the same > compression options in each). I was talking about 8-bit. I don't think 24-bit is necessary for photos destined to be printed 6x8.
Steve
David Dyer-Bennet - 23 Nov 2005 18:12 GMT > >> BTW, another option would be to scan at TIFF and convert to PNG (or scan > >> to [quoted text clipped - 10 lines] > I was talking about 8-bit. I don't think 24-bit is necessary for photos > destined to be printed 6x8. Well, 8-bit is thoroughly inadequate for screen display; so I haven't tried printing it.
 Signature David Dyer-Bennet, <mailto:dd-b@dd-b.net>, <http://www.dd-b.net/dd-b/> RKBA: <http://noguns-nomoney.com/> <http://www.dd-b.net/carry/> Pics: <http://dd-b.lighthunters.net/> <http://www.dd-b.net/dd-b/SnapshotAlbum/> Dragaera/Steven Brust: <http://dragaera.info/>
Steven Feinstein - 23 Nov 2005 18:31 GMT >> >> BTW, another option would be to scan at TIFF and convert to PNG (or >> >> scan to [quoted text clipped - 13 lines] > Well, 8-bit is thoroughly inadequate for screen display; so I haven't > tried printing it. Sorry, I totally mis-read your post (that's what I get for being in a hurry). I was thinking 8-bit vs 16-bit scanning not pallette. I agree 8-bit pallette is not suitable. I got smaller PNGs when I compressed mine, but then I don't use compressed TIFFs that often since I have not found support for compressed TIFFs to be consistent across applications. Some do and some don't.
Steve
David Dyer-Bennet - 23 Nov 2005 20:46 GMT > >> >> BTW, another option would be to scan at TIFF and convert to PNG (or > >> >> scan to [quoted text clipped - 20 lines] > support for compressed TIFFs to be consistent across applications. Some do > and some don't. I've had the same experience with compressed TIFFs. But my current set of primary tools all handle them, so I'm using them more at the moment. I'm guessing support will only increase over time.
Anyway, the confusion and mystery appears to be cleared up, thanks!
 Signature David Dyer-Bennet, <mailto:dd-b@dd-b.net>, <http://www.dd-b.net/dd-b/> RKBA: <http://noguns-nomoney.com/> <http://www.dd-b.net/carry/> Pics: <http://dd-b.lighthunters.net/> <http://www.dd-b.net/dd-b/SnapshotAlbum/> Dragaera/Steven Brust: <http://dragaera.info/>
Scott W - 23 Nov 2005 18:37 GMT > Well, 8-bit is thoroughly inadequate for screen display; so I haven't
> tried printing it. > -- That is a bit odd, that 8-bit is "thoroughly inadequate for screen display"
Can you tell buy just looking at this photo which side is 8 bits and which side is 16? http://www.sewcon.com/temp/half_and_half.psd
I know it is easy to tell by looking at the histograms, but do you claim to be able to tell by just looking which side is which?
Scott
David Dyer-Bennet - 23 Nov 2005 20:51 GMT > > Well, 8-bit is thoroughly inadequate for screen display; so I > haven't [quoted text clipped - 3 lines] > That is a bit odd, that 8-bit is "thoroughly inadequate for screen > display" Not 8-bits-per-channel; 8-bit palletted, i.e. 256 colors total.
> Can you tell buy just looking at this photo which side is 8 bits and > which side is 16? > http://www.sewcon.com/temp/half_and_half.psd Probably not, especially if it wasn't an extreme adjustment from the original...downlowding...starting photoshop...........nope, I can't.
 Signature David Dyer-Bennet, <mailto:dd-b@dd-b.net>, <http://www.dd-b.net/dd-b/> RKBA: <http://noguns-nomoney.com/> <http://www.dd-b.net/carry/> Pics: <http://dd-b.lighthunters.net/> <http://www.dd-b.net/dd-b/SnapshotAlbum/> Dragaera/Steven Brust: <http://dragaera.info/>
Scott W - 23 Nov 2005 20:58 GMT > > > Well, 8-bit is thoroughly inadequate for screen display; so I > > haven't [quoted text clipped - 5 lines] > > Not 8-bits-per-channel; 8-bit palletted, i.e. 256 colors total. Or sorry, your right 8 bit palletted is not normally enough.
Scott
David Dyer-Bennet - 23 Nov 2005 16:25 GMT > I have to scan several thousand old 35mm negs. 90% of them are in good > shape, the rest are a bit faded or otherwise damaged. All of them would need > 10mins to 1 hour work before printing to a standard that I would be happy > with. Theoretcially they should be saved in a lossless format like TIFF but > that needs a lot more storage space than I have or can afford. Saving them > on cd's at high resolution means having an awful lot of cd's. This is why they invented DVDs :-). The latest 16x drives (or have they gone up another level in the week since I last looked?) can be had *amazingly* cheaply -- like under $50 for a quality internal drive.
I do recommend having multiple optical media copies, ideally on different brands of media, before you even *think* of relying on them as your only copy of anything important.
Your time estimates look spot on to me. If you're seriously thinking of spending that much *time*, I'd suggest you work *very* hard to find the money to avoid making other compromises. The time is generally the most expensive part of this sort of project.
> Can someone give me an idea of just how much the image of a jpeg > suffers if its opened, cropped, smudges removed, levels altered then > sharpened before closing? Is it something that would be noticed on a > monitor, a 6"*8" print, a photo publised in a guide book? None of my > photos are ever going to be enlarged much more tha 6*8" and > exhibited as 'art'. Thanks for any comments. It varies tremendously based on the compression level chosen. At Photoshop jpeg levels of 4 and below, you can often see very obvious posterization in areas of even tone, especially background areas. At levels up to 7 you can often see differences when you click preview on and off, but they're quite minor, around the edges of things.
At Photoshop level 10, you're not likely to notice anything, and that still compresses fairly effectively.
(The levels I cite are rough impressions; I'm not claiming there are "magical changes" as you move the jpeg level through those values, just trying to give a rough impression of the kinds of things you'll see in various ranges.)
It's easy, of course, to grab a relatively small number of pictures and play around with the jpeg levels, then load them up and look at them and really see for yourself. I do find that the Photoshop preview often looks much better than that same photo does when I view it on the web later. I can't tell if this is a failure in preview, or if photoshop simply does a better job of displaying pictures on the screen. But it's worth using other viewers to check your results, in my experience.
Given the range of uses you talk about, sharpening before saving isn't a good idea, by the way; because those uses will have rather different optimum sharpening parameters most likely.
 Signature David Dyer-Bennet, <mailto:dd-b@dd-b.net>, <http://www.dd-b.net/dd-b/> RKBA: <http://noguns-nomoney.com/> <http://www.dd-b.net/carry/> Pics: <http://dd-b.lighthunters.net/> <http://www.dd-b.net/dd-b/SnapshotAlbum/> Dragaera/Steven Brust: <http://dragaera.info/>
Scott W - 23 Nov 2005 18:03 GMT > I have to scan several thousand old 35mm negs. 90% of them are in good > shape, the rest are a bit faded or otherwise damaged. All of them would need [quoted text clipped - 8 lines] > enlarged much more tha 6*8" and exhibited as 'art'. > Thanks for any comments. First off let me say that there is no big problem with saving the scans as tiffs. Having said that in reading some of the post here one would think that jpg is no capable of storing anything close to a excellent photo, this is simply not true. I this photo below I started with a raw file from a 20D and converted it to a 16 bit / color psd format using photoshop elements. I then took this file and converted it to 8 bits and saved as a jpg at quality setting 12 (again in PSE). I then copied half the jpg image on top of the 16 bit psd photo.
http://www.sewcon.com/temp/half_and_half.psd
Now anybody who is any good at photoshop at all can tell in a minute which side is which by checking the color depth. But can you do any kind of edit to the photo that will made the difference show up, at least any kind of an edit that you would ever really do to a photo?
For anyone who cares here are the jpg version and psd versions of the photo http://www.sewcon.com/temp/jpg.jpg http://www.sewcon.com/temp/clean.psd
I should note that no all software will allow saving to a very low level of compression and I should also point out that scans of film don't compress well as jpgs because of the noise in them from grain (or dye clouds if you prefer). But this means that you will end up with a pretty large jpg, not that you will end up with a poor jpg, that is if the software that is saving as a jpg is doing its job right.
Captain Blammo - 25 Nov 2005 08:06 GMT > http://www.sewcon.com/temp/half_and_half.psd Ok, I can hardly see any difference at all, but I'm going to guess the left is the JPEG, because the tonal transition from the centre of the clouds to the edge looks *ever* so slightly harsher than on the right. Did I get it right?
By the way, how do you check colour depth in PS7? I used the dropper and the histograms, but they both give the same ranges of values. Is that because it can't handle 16-bit files? I assume it just converted it all to 8-bit for viewing.
CB
Scott W - 25 Nov 2005 11:30 GMT > > http://www.sewcon.com/temp/half_and_half.psd > [quoted text clipped - 9 lines] > > CB Checking color depth is easy, select an area and using level set the high end of the input level to something like 64. Now look at the histogram, if it was 16 bits deep the histogram will still be smooth, if it as 8 bit deep you will see a bunch of lines in the histogram.
It is the left side that is 8 bits deep, but when I compare the 8 bit version to the 16, flashing between them, I can't see any differance, here is the full frame at 16 bits, see if the image looks any different to you when you can look at the same area in both 16 and 8. http://www.sewcon.com/temp/clean.psd
Here is the whole image as a jpg http://www.sewcon.com/temp/jpg.jpg
Scott
Arnold - 23 Nov 2005 18:56 GMT If you are going to spend all that time, I would save in the best possible format possible. Hard drive space is relatively cheap, the worst part is the time involved in scanning.
> I have to scan several thousand old 35mm negs. 90% of them are in good > shape, the rest are a bit faded or otherwise damaged. All of them would need [quoted text clipped - 8 lines] > enlarged much more tha 6*8" and exhibited as 'art'. > Thanks for any comments. tomm101 - 23 Nov 2005 21:26 GMT > I have to scan several thousand old 35mm negs. 90% of them are in good > shape, the rest are a bit faded or otherwise damaged. All of them would need [quoted text clipped - 8 lines] > enlarged much more tha 6*8" and exhibited as 'art'. > Thanks for any comments. If you are printing a book, the publisher is going to want final size x 300dpi (ppi in the photoworld) in uncompressed tif format. Scanning is a big time grabber, you would do best by editting the images first, triage the edit, book 1st (especially if you are still selling your idea to publishers), stuff you still like 2nd, the rest third. Scan in packs if "the rest" don't get scanned fine. You can keep the slides anyway. Scan the ones for the book twice to size in the book layout and at 8x10 @ 300dpi. Why 8x10, it is what publishers want to see. It is a big file, about 30mb, but worth doing if you want to go in the publishing direction. Also save everything in uncompressed tifs, printers have an unrational fear of anything else, even compressed tifs. The last book I worked on, about two years ago, I was told by the printer that lzw compressed tifs screwed up the image. Printers are a tough bunch, but when your work depends on them, you do what they say. They can make you look very bad if they don't like you. To catalogue your files set up a simple database for each of your CDs. If you want something more visual use Extensis Portfolio.
Tom
Charlie Self - 24 Nov 2005 00:29 GMT > > I have to scan several thousand old 35mm negs. 90% of them are in good > > shape, the rest are a bit faded or otherwise damaged. All of them would need > > 10mins to 1 hour work before printing to a standard that I would be happy > > with. Theoretcially they should be saved in a lossless format like TIFF but > > that needs a lot more storage space than I have or can afford. Saving them > > on cd's at high resolution means having an awful lot of cd's. Also save everything in uncompressed tifs, printers have an
> unrational fear of anything else, even compressed tifs. The last book I > worked on, about two years ago, I was told by the printer that lzw > compressed tifs screwed up the image. Printers are a tough bunch, but > when your work depends on them, you do what they say. They can make you > look very bad if they don't like you. Not arguing here, as I don't know, but did any printer ever tell you HOW a compressed--LZW--TIFF screwed up a file?
I hadn't bothered, but just tried today to save a TIFF into LZW compressed TIFF: 17.3 MBs down to 7.2 MBs. If that is effective all the way, when coming back to uncompressed, it is less than twice the size of the JPEGs put out by my Pentax *istD, and only 40% or so the size of the original TIFF. I've got a spare 160 gig hard drive for photos, so I can figure on getting an awful lot of LZE TIFFs on it at those sizes, probably more than I'll ever need, though you never know. I can recall when I thought a 32 MB hard drive was the ultimate, and now I've got a few dozen or hundred files that large.
Whoops. My wife's new computer came. Time to set that up.
tomm101 - 24 Nov 2005 13:35 GMT > Not arguing here, as I don't know, but did any printer ever tell you > HOW a compressed--LZW--TIFF screwed up a file? [quoted text clipped - 10 lines] > > Whoops. My wife's new computer came. Time to set that up. Well the printers I have dealt with have always been crusty guys who did everything the same for their first 20 years in the business, then they were dragged kicking and screaming into the digital age. Now they have a way of doing things that works for them and again they don't want to change, if all the only thing they impose on you is uncompressed tifs, not a big deal. One printer said lzw tif damage the image, I say that is not what I here, then they ask to show them the evidence right there, and if you want them to do your job you get the feeling, just shut up and agree. The reason maybe that 10 years ago RIPs and typesetters didn't accept lzw tif, some still don't. The first version of a RIP I had for a wide printer I bought 5 years ago didn't accept lzw compression. There first upgrade changed that. I think 10-25% reduction with lzw compression, sometimes you get the results you did but not very often.
Tom
Charles Self - 25 Nov 2005 00:50 GMT >> Not arguing here, as I don't know, but did any printer ever tell you >> HOW a compressed--LZW--TIFF screwed up a file? [quoted text clipped - 26 lines] > I think 10-25% reduction with lzw compression, sometimes you get the > results you did but not very often. OK. But I just tried another file, same size to start 17.3 MBs, and the resulting LZW TIFF was 7.69 MBs.
Years ago, I had to work with printers, not too long after the change to offset from hot type. It was barely long enough for a few newspapers to go broke because they set in cold type for offset, but had to pay linotypers to set the same material in hot type, which was then discarded. They were, I guess, a lot like the railroad firemen who kept their jobs even though diesel was the new fuel and there was no need to shovel coal and tend the fires.
|
|
|