Im having trouble reading tiff images rendered by houdini
on other software … like aftereffects or finalcut
Ive tried all the avialable format options.
Im making 16 bits per RGB pixel images.
Has anyone else noticed this …
is there a fix?

Try changing the compression type to something else. When you save, click on the arrow next to the file type. This expands additional image options.

Tiff is usually the best option when you start to get into more than 8 bits, and the target software doesn't support OEXR, but I must admit to wondering about this whole tiff compression issue. Back in the day, when the initial threats of lawsuits over the lzw compression were raised, the sky was falling and suppliers started pulling it from their export options. I do recall SESI keeping it(I thought), and there being a slight issue over “Adobe Deflate” compression. My understanding was that this was essentially lzw compression with Adobe doing the typical corporate thing and sticking that phrase into the magic portion of the file, but otherwise it was just plain ol' lzw. I have noticed for some time that a number of apps, Adobe as mentioned and some Mac apps won't read SESI's compressed tiffs. At least that's the whine I hear from the Mac camp. Can anyone speak clearly on this issue? It's been pretty frustrating for us having to use uncompressed tiffs.

It is worth mentioning, however, that much of the time I've actually spoken to those mac users, I find out they're using apps that only support 8 bit. Clearing the chaff from that wheat has been a pain.

Cheers,

J.C.

AdobeDeflate is actually a gzip compression and is the same as “Deflate” except with a different compression type id. It's true though that Houdini avoided LZW compression due to patent issues.

I asked about the difference between AdobeDeflate and Deflate on the libtiff mailing list a while back. According to Chris Cox of Adobe (which I suspect is from the Photoshop side), this is “Because someone outside Adobe defined a flate compression value and ignored the value defined by Adobe as part of the TIFF standard.” He never replied to my comment that After Effects didn't seem to load AdobeDeflate files though. I would have though that After Effects would support AdobeDeflate by now though.

For some reason, libtiff, the library by which what most people use to read and write tiffs, only supported Deflate initially. Houdini up until version 5 or so wrote out tiffs using Deflate by default. At that point, some customers complained that they couldn't use the tiffs output by Houdini with some Adobe products and so we changed the default to AdobeDeflate. This of course then broke other apps which used old versions of libtiff.

The short of it is to set your tiff compression type to the one that you need if the default doesn't work for you. I believe iconvert can be used to to change the compression type on existing tiff files when given the right options.

Excellent - that clears up a lot of questions, Edward…

Cheers

J.C.

good info edward.. thanks..


willa:

AFAIK AE won't support 16 bit tif's.. In fact they claim only 24 and 32 bit files are supported along with some 16bpc quicktime codecs. seems really stupid considering you can work in 16bpc.. as a side note, i tested out by just creating a 16 bit tif and psd out of photoshop.. tif was a no go in different compression methods but the psd was.. go figure..

another option that you might want to consider is working with openexr and eLin…

http://redgiantsoftware.com/elin.html [redgiantsoftware.com]

One really nice thing about using Houdini tifs with photoshop (at least 8bit ones) is that they have an “assiociated” Alpha channel, i think that's the expression. Anyway what it means is that photoshop reads the alpha channel correctly and converts it to it's own transparency type. If you don't do this and just load the alpha channel as a selection after opening a image and then delete the background it leaves black in the foreground portion that shouldn't be there. Not sure it After Effects does the same thing. But basically if you want transparency in an Adobe product probably best to start with tif.

AE will just read the image and if it sees an alpha, it will ask you what you want to do with it upon importing (ignore, premultiplied, straight, invert).. it's then brought in to your bin where you can go back at any time to change those settings..

We had trouble in windoz aftereffects but got it going by
gettting the free supertiff plugins at the site below
http://www.fnordware.com/ [fnordware.com]
there are dowloads for MAC as well

for your info
uncompr … 16bit per channel uncompressed
packbits … 16bit per channel compression=packbits
adeflate … 16 bit per channel compression=Adobe Deflate
sdeflate … 16 bit per channel compression=Standard Deflate

the compressions are all lossless but the deflates makes the smallest files
90% of the time.

in WINDOZ:
aftereffects 6.5
-adeflate
-packbits

quicktime 6.5.2 pro
-uncompr

photoshop CS 8.0
-uncompr
-sdeflate
-packbits
-adeflate



Pete M. at axyz(Toronto) made this nice summary for MAC software

Final Cut Pro v4.5
-uncompr

Photoshop CS
-uncompr
-sdeflate
-packbits
-adeflate

Preview v2.1.0
-packbits
-uncompr(but messed up)

Shake v3.5
-uncompr
-sdeflate
-packbits
-adeflate

After Effects v6.5
None. Parsing error This surprised me
(he had not yet acquired the supertiff plugin from
http://www.fnordware.com/ [fnordware.com]
)

Illustrator v10.0.3
None. Error saying can't open files with that bit depth.

iPhoto v5.0.2
-uncompr

QuickTime 6.5.2
-uncompr
Haven't loaded QuickTime 7.0 yet. But apparently it is now 32 bit. Also Final Cut Pro 5.0 will be out in a couple of days and it is 32 bit as well.

Keynote v2.0.1
-packbits
-uncompr(but messed up)

Pages v1.0.1
-packbits
-uncompr(but messed up)

Summary

It seems that uncompr is the most compatible and second place would be packbits. Although Shake and Photoishop CS were fine with all four.

nice find WillA… thanks for the info..

I wonder what compression types in 8-bit tiff files would a stock After Effects handle then.

No compression and packbits were a go.. Adobe and Standard Deflate were not.. stupid after effects..

Thanks Will and Dave … It's good to know what to say next time someone asks about this.

I don't know if this has been fixed in v8 but in Houdini 7 (up to and including 7.0.412 on WinXP), TIFF compression will revert to the Adobe Deflate scheme even if you've chosen Uncompressed and saved out the .hip file…

It took me a while to trace the bug - rather annoying all those unreadable file type messages in Nuke!

——–
Roy Stelzer
Stelzer Productions
Switzerland

Hmm … as far as I can tell, that compression type dialog is purely UI and not actually a parameter so isn't saved anywhere at all other than in the current session. You should send in something to support.

rxs3817
I don't know if this has been fixed in v8 but in Houdini 7 (up to and including 7.0.412 on WinXP), TIFF compression will revert to the Adobe Deflate scheme even if you've chosen Uncompressed and saved out the .hip file…

It took me a while to trace the bug - rather annoying all those unreadable file type messages in Nuke!

——–
Roy Stelzer
Stelzer Productions
Switzerland

Is there a way to resolve this issue ?

You can edit the $HFS/config/FBoptions file, and change the compression type to whichever one you are using. The option is ‘compression’, the tags are ‘none’, ‘packbits’, ‘deflate’ and ‘lzw’ (see the FBoptions file for examples of how to do this).

It would certainly be nice if there was an easier way to set the default than editing this file. I'm not sure why the options aren't being saved in the hip files, as they are saved into a string parameter as option/value pairs. Perhaps they are, but aren't being restored correctly.