LibTiff Mailing List
TIFF and LibTiff Mailing List Archive
July 2007
Previous Thread
Next Thread
Previous by Thread
Next by Thread
Previous by Date
Next by Date
Contact
The TIFF Mailing List Homepage
This list is run by Frank Warmerdam
Archive maintained by AWare Systems
Thread
2007.07.03 14:23 "Re: Big TIFF Compression", by Andy Cave
Hi Frank, Thanks. That's interesting info. >From what you say, with the 16 bpp or float data types, I presume either LZW or Flate is used (if any). It sounds like people currently don't write 1m x 1m pixels (and clearly probably don't since they'd exceed classic TIFFs file size), but might start trying to once BigTIFF is out. In comparison, pre-press separations at their largest are typically 100K - 150K (on each side) or thereabouts. Security printing does go up to 10,000 dpi, but then the page sizes are smaller, so overall dimensions not much different. This is unlikely to change (no real need for higher resolution - and even if you do, the halftone dots get bigger, so compressed data size is not that much different). It will be interesting to track peoples experience as they do so, as I think a whole bunch of new issues will turn up (performance of reading/writing large files, OS limitations, corruption, copying these files around(!), ftp'ing them, etc...). Regards, Andy. ----- Original Message ----- From: "Frank Warmerdam" <warmerdam@pobox.com> To: "Andy Cave" <andy.cave@hamillroad.com> Cc: <tiff@lists.maptools.org> Sent: Tuesday, July 03, 2007 2:55 PM Subject: Re: [Tiff] Re: Big TIFF Compression > Andy Cave wrote: >> I presume that georeferencing and medical imaging are all 8 bits per >> pixel, multi-channel for the former (and possibly latter). > > Andy, > > Speaking for geospatial images, they are not all 8bit per pixel. Many > are 16bits per pixel (the IKONOS sensor for instance), and scientific > data sometimes uses other data types (floating point for instance). > > Multi-channel (ie. multi-sample) is common but there are also a lot > greyscale datasets (one sample per pixel). > >> I'm interested in hearing real examples as to why BigTIFF is useful for >> georeferencing and medical imaging. Are the images really that big? After >> compression or only without? Are they big 'cause they are multi-channel, >> or multi-page, or uncompressed, or, ...? > > Geospatial images are often mosaiced into seamless images for large > regions. As you can imagine an image with 1m x 1m pixels for the > entire united states gets very large. Previously other formats were > used for this, or supertiling of many medium sized TIFF files. > > Best regards, > -- > ---------------------------------------+-------------------------------------- > I set the clouds in motion - turn up | Frank Warmerdam, > warmerdam@pobox.com > light and sound - activate the windows | http://pobox.com/~warmerdam > and watch the world go round - Rush | President OSGeo, http://osgeo.org >