| Author | Message | ||
     Peter Bowyer New member Username: Reywob Post Number: 13 Registered: 09-2002 |
I've noticed that TIFF files I export from Photoshop CS and then resave from PWP (PWP's crop tool is so much easier to use, it justifies the whole program's cost) are often 2-3 times bigger than the Photoshop-produced files. I'm guessing this is because you don't do LZW or Zip compression on the file - is this correct? If so, is there any way to enable it in PWP? It's nice to halve the disk space I need. | ||
| Kiril Sinkel New member Username: Ksinkel Post Number: 111 Registered: 08-2002 |
Peter, Currently PWP supports reading of LZW-compressed tiff files but not writing them. LZW compression of tiff files has been on our wish list and may be added in a future upgrade. Kiril _______ Kiril Sinkel Digital Light & Color | ||
| James Winthorpe New member Username: Cobaltgreen Post Number: 3 Registered: 03-2005 |
LZW compression would singlularly be the most desirable addition to Pic Win pro. at this juncture. Realistically speaking, LZW comp. should be gainsaid for any Pro. version of a serious image editing application, whatever the otherwise excellence or merit. I realise there are cost implications here over patent rights usage and deeply respect the cost conscious, value for money philosophy of Picture Window and its authors commendable commitment to excellence .... but would it really be so much of a problem though, or add so much to the cost to implent? I Loathe JPEGS for good reason...I want to store the highest quality files possible....after working hard to get good images no less will do, but no matter how much storage space I procure it is soon eaten up by uncompressed Tiffs from modern high resolution cameras . I don't see this problem easing ...Lzw only saves a little, but every little sure helps, especially if it can avoid the dreadful JPEG. | ||
| Jonathan Sachs New member Username: Jsachs Post Number: 409 Registered: 08-2002 |
FYI, the LZW patent has expired. Jonathan Sachs Digital Light & Color | ||
| Tarlika Elisabeth Schmitz New member Username: Tarlika Post Number: 109 Registered: 05-2003 |
How much of a saving could one expect of the lossless compression of a 48bit TIFF? In other words, would it be worth it at all considering the cheap storage prices these days? A DVD coss £0.15 GBP ($0.25). Lossless JPEG2000 compression seems to REALLY save a lot of space. If I am not mistaken, a TIFF file might be compressed to as little as 10% [non-lossy]. There are some tools out there, none of them are free for commercial use. It might be worth looking into buying a dedicated piece of software if it is not overly expensive. And I have a feeling lossless JPEG2000 compression would save way more space. Some links can be found here: http://datacompression.info/JPEG2000.shtml Regards, Tarlika Elisabeth Schmitz, Scotland | ||
| Tarlika Elisabeth Schmitz New member Username: Tarlika Post Number: 110 Registered: 05-2003 |
In any case I often find it the better philosophy* to have several small dedicated programs which do their job but do it well rather than one "egg laying woolly milk sow", i.e. one big, sluggish program that attempts to do everything. *as you find it on Unix/Linux Regards, Tarlika Elisabeth Schmitz, Scotland | ||
| afx Unregistered guest |
Cheap storage might be available, but when an uncompressed 16bit TIFF is 130MB and a compressed one goes down to 70, it is obvious, that there is still a huge benefit. cheers afx | ||
| Tarlika Elisabeth Schmitz New member Username: Tarlika Post Number: 111 Registered: 05-2003 |
The other thing you need to take into consideration is the cost at which the compression comes, i.e. the time it takes to compress/uncompress. Depends on the software, of course. I doubt that you can compress a TIFF losslessly with LZW to 50% of its size. More knowledgeable people might be able to tell you what compression rates can be achieved with LZW. Regards, Tarlika Elisabeth Schmitz, Scotland | ||
| James Winthorpe New member Username: Cobaltgreen Post Number: 6 Registered: 03-2005 |
TIFF IS,IS, IS...lossless compression....I have never heard otherwisw, that's the whole point of TIFF...but the price is the limited amount of compression and I guess that depends on the tonal or whatever qualities. Maybe we need an expert to illuminate this in detail. But in use terms, anyone serious about storing quality image files tha I know of uses it ubiquitously. | ||
| John Lodge Unregistered guest |
Here is an example at 48 bit: CR2 =10,748,547 (ISO 1600 it would be less at lower settings.) Unique colors =3,865,130 Uncompressed Tiff =47,817,600 LZW Compressed TIFF =57,249,792 ZIP compressed TIFF =39,473,152 JP2 lossless = 35,418,112 and no I did not get the tiff numbers reversed; LZW can make it worse if there is a lot of detail. Note the ZIP compressed TIFF does not mean a Zipped tiff file, it means ZIP used internally as LZW would be used. There is not much software that can read it!! But the result seems to be on par with lossless JP2. I would say that LZW is not worth the effort, since more and more Camera's deliver 48bit (or 36bit really) color, and in practice many color scenes have two to four million colors, the choices here become lossless JP2 or ZIP TIFF. But since the improvement is only 25% by either of those choices, whereas keeping the RAW is a 75% improvement, and both these choices (JP2 and ZIP TIFF) seem to be technology byways, neither seem reasonable. Given that all this came from a 10Meg RAW file, there is presumably some transform that can correctly leverage the redundancy in the three color channels, but it does not seem to be in general use. | ||
| Peter Bowyer New member Username: Reywob Post Number: 14 Registered: 09-2002 |
Funnily enough I also carried out some tests yesterday. The results are at http://peter.mapledesign.co.uk/writings/TIFFsize.html Quick conclusion: if like me you shoot in JPEG and save the modifed files as TIFF (to prevent any further loss of quality) then compression is worth it. Otherwise LZW generally increases the size, while ZIP reduces it slightly. | ||
| James Winthorpe New member Username: Cobaltgreen Post Number: 7 Registered: 03-2005 |
Peter, thankyou for the diligence of your test in respect of file size results with LZW and ZIP TIFF...most interesting food for thought, to see that a "compression" methodology can actually increase a file size.I haven't come across this in any references to LZW TIFF previously. SO a rather strange eye opener. Obviously a slightly irrational situation on the face of it worthy of further discussion and investigation/explanation. I'd really like someone technical to explain why. Albeit, your results do generalise to demonstrate significant gains in file size reduction using lossless LZW or ZIP, confirming the advantages from their use . Thanks for the sound work and trouble with this, I'm going through the day now going......"hmmmmm...there's a thing"! | ||
| Kevin Elliott New member Username: Kevgermany Post Number: 138 Registered: 09-2004 |
James, size increases have been a feature of compression algorithms since the beginning. They work by substituting codes for repeating data values. If there's little repetition the replacement will then incerase the size. Binary files such as executrables and image files are hard to compress becasue of the lack of repetition. Jpegs work by throwing a lot of data away - in other words increasing the amount of repetition, so a jpeg converted to tiff contains a lot of repeated data - instaed fo the original detailed version. Thus it will compress, lossly, but the losses are already there... There's a lot on the web about data compression, but it's, of necessity, rather technical. | ||
| James Winthorpe New member Username: Cobaltgreen Post Number: 8 Registered: 03-2005 |
Thanx Kev for the explan..... "size increases have been a feature of compression algorithms since the beginning"....well search me..this is the strangest life I have ever known, back on the planet Janet, compression usually has the effect of less mass. Well there y'go, yer lern summat every day...!! | ||
| Bill Ross Unregistered guest |
"Binary files such as executrables and image files are hard to compress becasue of the lack of repetition." Just being binary does not rule out repetition - note the favorable ratio that the original poster reports from TIFF LZW in Photoshop. You can try compressing a single-color image to see how this is possible. Fine random detail on the other hand will likely result in expansion of the file. (A 'tar file' of the basic linux executables in /bin compresses by 50%.) | ||
| Kevin Elliott New member Username: Kevgermany Post Number: 139 Registered: 09-2004 |
Bill, you're right, bad wording on my part - low/er level of repetition would have been much better, sorry. There are also special compressoin algorithms designed to handle binary data, LZW was designed more for text where there is lots of repetition. | ||
| Vincent Ehrstrom New member Username: Vincente Post Number: 5 Registered: 06-2005 |
Why would one want to keep 16bit TIFFs as an archive? My workflow is RAW (with adjustements) -> TIFF 16 for further work & adjustments -> TIFF8. IMHO, the benefits of full colour depth are when you need to tweak the image exposure. Once this is done, then I have found TIFF 8b to be quite OK. I haven't noticed any difference in tone quality from prints done out of 16 bits rather than 8 bits. So I believe a good solution is to convert to TIFF 8 bits when one is satisfied with the exposure and curves and also keep the RAW as a "digital negative", since RAW files do enjoy a rather good compression for their level of colour richness. But maybe I'm not as thorough as some of you in examining quality... Don't we try sometimes to reassure ourselves with technology, thinking that the more the better, just in case? In this example, thinking the more data in our files the safer we are? I can't help to think that although digital photography has opened new possibilities, it has also introduced vasts mounts of questions, dilemmas and sometime even anxieties that are jsut not founded. I have B&W prints made 25 years ago on which I worked a full day to get them just the way I wanted them. I have a single copy of them and there is no way I could redo them exactly the same (I know that professional printers write everything to be able to reproduce that, but I certainly never did this). I still have the negative so I could re-print them. But it would be a new creation and the result would never be identical. I don't lose sleep over this and actually, there is something real special about that. Think of it: this is a unique piece! A bit like a painting that cannot be done over. That's why I think that having my negative (the RAW) + an archive of my image in a format that enables me to print the same picture as many times as I want is quite fine for me. IMHO, I believe that our obsession to control everything in quest of perfection, although a main driver of this industry, is defocusing us from a lot of the actual rewards and pleasures of this form of expression. Whao! I didn't set off to end up there. Sorry about that digression guys... This is a technical forum. | ||
| John Lodge Unregistered guest |
There have been long discourses elsewhere about the removal of support for RAW formats as a particular camera ages. This goes together with the general issue of the archival of Information Technology data. Indeed even TIFF format (no not the current TIFF-6). But if you were to look at TIFF formatted data all the way back to TIFF-1 you will find so many variations that not a single software product can decode it all, as well the operating environments (complete context) on which the old software ran is becoming unknown and unknowable. Even Adobe DNG is not a solution it is a wrapper for the RAW; but depends on the DLL understanding the individual RAW formats which may not be the case over time. At the present the best we probably have it TIFF-6. There are simply so many different software DLL's that can read it that it is safe. (Not I did say DLL most software uses a much smaller group of dll's for this purpose). But nothing in the digital world will have the archival safety of negative film. We can still take 100 year old negatives or more and read them. Try projecting any of the digital technologies forward 100 years. A gut reaction is worthless here but there are quite a lot of good information on the extreme issues NASA has had in going back to old data. The "solution above" of keeping the RAW and 8bit processed TIFF is not the archival equivalent of a film negative. It is more or less what I do (and hope for the best). In the same vein the files must then be stored in a media that is very common and by a storage method that is standard. |
