Pentax PPL strictly 8bit. what about others?

[jeffkrol]

after a bit of internet research I have come to the conclusion that Pentax photolab is useless for 16bit Tiff's (thanks Blende et. al.) and possibly dysfunctional due to the way the bits are incorrectly packed.
see
Here is what I found out: Pentax SLR Talk Forum: Digital Photography Review

Now it is most likely (circumstantial and others) that it also only processes in 8 bit as well........ NOT GOOD... how else can I say it.
Now this poses an interesting question regarding other processing software and their methods.. Are we getting cheated by many of them? Which ones?
Anyone have "for sure" insight into the processing levels of any other Pentax RAW compatible software? ASSuming seems to be a no no. I find it rude to export at least 10bit RAW and have them truncated to 8 bits immediately......
TO BE HONEST this may not be true but I'd like some research (best yet insight from others who already "invented the wheel")..
EDIT: Good thing it's free but I guess I know why now....
SECOND EDIT: Anyone think 8bit processing is not a bad thing, and reasons why.
EDIT:a good discussion on the 8 vs 16bit w/ LAB nr and other lab tricks thrown is as well....
http://www.dgrin.com/showthread.php?t=84751&page=3

[glasbak]

I have no reason to suspect that PPL does not take advantage of the 4 extra bits of raw.
I can not tell if additional filter functions are done with 8 or 16 bits.

Anyway, every program that shows histograms from 0..255, cant be using 16 bits

George

[jeffkrol]

Originally Posted by glasbak

I have no reason to suspect that PPL does not take advantage of the 4 extra bits of raw.
I can not tell if additional filter functions are done with 8 or 16 bits.

Anyway, every program that shows histograms from 0..255, cant be using 16 bits

George

Not sure where, or what software this was good for but I do remember a discussion that, though the display is 8bit, in effect it is functioning in 16bit for math... As a side note Adobe 16bit is really only 15bit....... w/ one bit used for something else. this is easily proved.
Clarkvision: Raw Converter Shadow Detail
Photoshop does 15-bit processing and I believe that is causing some issues with the "16-bit" processed data. I've been experimenting doing more and more processing in ImagesPlus as it does scientifically correct math (64-bit processing, but I don't believe that 64 or even 32 bits are really warranted-it's more the math).
...............Photoshop does 15-bit math to speed operations, and it appears it is limiting the noise and dynamic range of current DSLR cameras. .................
More on this p51
http://books.google.com/books?id=5WN...esult&resnum=8
http://mike.russell-home.net/tmp/DanMargulis16Bit/

[Maffer]

All right. This might explain something and then again, maybe not. Anyway, if I just load RAW files directly into photomatix to make an HDR image the color gradients are generally smooth after I'm done with tonemapping. This, however, produces softer result than creating the HDR from TIFF-files generated with PPL from the RAW-files (no curve adjustements, everything as neutral as can be). So the results after tonemapping are:

1) Sharper output than with RAW files converted by photomatix
2) Somewhat more crude color gradients with skies and weirdo halo effects here and there

Also, pulling back slightly overexposed details in PPL often results in crude banding versus pulling details back in some other major converter. See this thread: I feel bad for saying this... but who else is waiting for the k-7's DXO mark? - Ishpuini pulled off some major recovery with lightroom. I cannot see that happening with PPL.

[thazooo]

SECOND EDIT: Anyone think 8bit processing is not a bad thing, and reasons why.



8 bit processing works fine for me. I work in RGB and 8 bit fits within the color space.
I've tried using the 16 bit processing and saw no difference in prints. I agree that 16 bit will give smoother colors when you do the math, but on most folks systems it's 8 bit for display and printer. I imagine if you've got a 16 bit printer and switch to prophoto rgb or wide gamut rgb , you can pick up a few more colors. Problem being
would you see these on your display as you make your adjustments or would they just show up once printed ?

[jeffkrol]

Originally Posted by thazooo

SECOND EDIT: Anyone think 8bit processing is not a bad thing, and reasons why.



8 bit processing works fine for me. I work in RGB and 8 bit fits within the color space.
I've tried using the 16 bit processing and saw no difference in prints. I agree that 16 bit will give smoother colors when you do the math, but on most folks systems it's 8 bit for display and printer. I imagine if you've got a 16 bit printer and switch to prophoto rgb or wide gamut rgb , you can pick up a few more colors. Problem being
would you see these on your display as you make your adjustments or would they just show up once printed ?

Actually this issue remined me about the old Dan Margulis challenge re: 8 vs 16 bit processing:
This 4 page paper sums it up quite well (this discussion is apparently over 8 years old):
Part 1 of 4



to take one minor quote of a contrarian to Dan:
Bruce Fraser - 1:07pm Sep 19, 05 PST
I'd also suggest that you need to weigh the consequences of working in 8-
bit and finding you need 16 versus working in 16-bit and finding you only
need 8. My lack of clairvoyance is one of the factors that leads me to work in
16-bit....
Something that sticks in my mind is as sensors and circuits get better the gap between 8 and 16 bit processing (if you believe there is one) could increase...
The 8 vs 16 bit debate is based mostly on "after conversion". I think few would argue that during RAW conversion you should most likely work in at least 12 bit integer math... some state anything short of floating point math is destructive (though slow on many systems and therefore not socially acceptable)
Very few would argue that the RAW data has less than 9 fully useable bits (some seem to believe that 10 is easy and 11 is becoming standard today. Of course marketing is pushing 14 bits) so again 8 in RAW IS destructive.
Now I'm only speculating based on what I read and the use of a few RAW converters and photo editing software... and I'm miffed at the 16bit TIFF lie.

[thazooo]

to take one minor quote of a contrarian to Dan:
Bruce Fraser - 1:07pm Sep 19, 05 PST
I'd also suggest that you need to weigh the consequences of working in 8-
bit and finding you need 16 versus working in 16-bit and finding you only
need 8. My lack of clairvoyance is one of the factors that leads me to work in
16-bit....

Great ! Plan ahead :-)) Thinking like this was the reason I started using 16 bit
processing, I did 16 bit for several years and after a talk with a friend who also
did 16 bit, we decided to try the 8 bit mode. Neither of us saw a difference in output, I do print more and didn't see it there either. Since I've upgraded my programs and work flow from what I used at that time, it may be time to revisit
16 bit processing.


'Something that sticks in my mind is as sensors and circuits get better the gap between 8 and 16 bit processing (if you believe there is one) could increase...'

This is the area to really watch, I suspect in the near future that all equipment
will move to a 16 bit friendly environment. From some of the charts I've seen of camera color space vs. RGB color space, Brand X already surpasses RGB foot print.
I'm sure this will lead to all brands shortly. Once this is main stream, we should be able to see the difference in print. Right now the only time I run into problems is when I shrink the image and make a JPG. Banding shows up.



'The 8 vs 16 bit debate is based mostly on "after conversion". I think few would argue that during RAW conversion you should most likely work in at least 12 bit integer math... some state anything short of floating point math is destructive (though slow on many systems and therefore not socially acceptable)
Very few would argue that the RAW data has less than 9 fully useable bits (some seem to believe that 10 is easy and 11 is becoming standard today. Of course marketing is pushing 14 bits) so again 8 in RAW IS destructive.'

You just jumped WAY beyond me :-) My programming experience and what goes on in the machine dates from 1970.......lol.

'Now I'm only speculating based on what I read and the use of a few RAW converters and photo editing software... and I'm miffed at the 16bit TIFF lie '

The program I'm waiting for is the one where you process the RAW file Completely
in the converter with out having to use another program for tweeking.

[Mountain Vision]

Hmm, I have used PPL for many images, and I can tell you this, the conversions were better than ACR. I now use Lightroom, but occasionally still process a shot or two on my wifes laptop using PPL and it is absolutely fine. And yeah, I print my photos. So if for some reason those 16bit tiffs I output are inferior, I have no idea that they are, and would love to see proof.

BTW, the real Silkypix would be a nice addition to the K7, I used a trial of it for a few weeks, and it was awesome.

[silvershoes]

Originally Posted by Mountain Vision

Hmm, I have used PPL for many images, and I can tell you this, the conversions were better than ACR.

I agree. I've tried several times to use ACR, I even bought Bruce Fraser's book, and always came back to PPL.

It's possible ACR works better for other camera brands, although I've seen similar discussions by Nikon users and many prefer the Nikon software even though it's admittedly kludgy.

As for the whole 16-bit thing, a lot of people (including Fraser) point to the histogram in PS to "prove" how "destructive" various processing methods are. Histograms are useful to a point, but they're programs like any other and a lot of their value depends on what's going on in the calculations. It's my understanding that some features of PS haven't changed since being introduced - new versions introduce new features but aren't a complete rewrite.

If it makes you feel better to process in 16 bit, by all means do so. I think the real test of processing is in looking at the output with an educated eye and not worrying so much about theoretical numbers. If you print an 8 bit file and a 16 bit file and can't see a difference, the difference isn't enough to worry about.

[jeffkrol]

Originally Posted by silvershoes

I agree. I've tried several times to use ACR, I even bought Bruce Fraser's book, and always came back to PPL.

It's possible ACR works better for other camera brands, although I've seen similar discussions by Nikon users and many prefer the Nikon software even though it's admittedly kludgy.

As for the whole 16-bit thing, a lot of people (including Fraser) point to the histogram in PS to "prove" how "destructive" various processing methods are. Histograms are useful to a point, but they're programs like any other and a lot of their value depends on what's going on in the calculations. It's my understanding that some features of PS haven't changed since being introduced - new versions introduce new features but aren't a complete rewrite.

If it makes you feel better to process in 16 bit, by all means do so. I think the real test of processing is in looking at the output with an educated eye and not worrying so much about theoretical numbers. If you print an 8 bit file and a 16 bit file and can't see a difference, the difference isn't enough to worry about.

Easy to say when you don't have a hard drive full of "fake" 16bit TIFF's. Good thing I still have the RAW... Actually what bothers me is not 16 vs 8 and does it matter but 16bit is just a lie. I recently been trying GIMP and it at least warns me "we can only process this in 8 bit, file will be truncated" or some such error message..I really do expect to at least know the basics of what my software can and can't do. Honestly though, this is a lot harder then most imagine.
As to Silkypcs retail I would not pay $150 for an an 8 bit only program especially if it indeed processes RAW in 8 bit (which is quite primitive if you ask me). THIS IS RUMOR, I have currently no way to confirm or deny this
Again, to be honest for years I've been doing major post conversion tweaks in Corel Photopaint (which I buy because it can, wait for it, do some 16bit manipulations). I've always wondered why my histograms got so spiky when doing some work on 16 TIFF's from PPL . Though, since the Corel histograms, color pallettes are all listed in 8 bit anyways, it's hard to completely fault the TIFF's.
Yes this is an annoying "property" of Corel but it is quite visible...
ACR and the associated CS whatever is 1)too bloody expensive 2)from what I gathered it has it's fair share of processing problems in regards to a few key areas (personal opinions mostly).
There are always work arounds and quality differences are hard to see when you are viewing on low gamut monitors, printing w/ low gamut printers and maybe here the 8 bit is an advantage because it has already reduced your files and truncated your color pallet....
IS it really too much to ask to know what your software can or can't do??????????
AS to output.. what's good enough today may not be good enough tomorrow.. besides the final image is usually truncated to 8bit jpg for printing but it may not always have to be..
I ADMIT that I may be a tad blowing this whole issue out of proportion but then again others may want to know what their buying as well. Hard to fault PPL since it's free though and my MAJOR issue is no longer with them per se.. it's different when I shell out cash, convert something to supposedly 16bit and someone tells me it's just a stuffed 8 bit file...... ohhh how special is that..
On the "good news" front I see PSP now (and actually has been for awhile )
supports 16bit and CMYK (what about LAB format???) so my options appear to be growing... Not sure about their RAW editor..
Oh well enough rambling........

[silvershoes]

Originally Posted by jeffkrol

Easy to say when you don't have a hard drive full of "fake" 16bit TIFF's. Good thing I still have the RAW... Actually what bothers me is not 16 vs 8 and does it matter but 16bit is just a lie. ..I really do expect to at least know the basics of what my software can and can't do. Honestly though, this is a lot harder then most imagine.

Two thoughts:

(1) I never have a hard drive full of anything. You need to investigate a good redundant backup scenario.

(2) The web is full of rumors and "expert" opinions and claims, some of which are simply a sales pitch. A lot of scanners that claim 16 bit have been shown to be 12, for example; the scans are still lovely. If it really matters to you, why not go to the source (the people who make your software package) and ask customer support?

You can download trials of just about every RAW processor on the market and do your own head-to-head comparison with your own images. The software's only half the equation; your own processing preferences and skill level matter just as much as the number of bits.

Good luck!

[jeffkrol]

Originally Posted by silvershoes

Two thoughts:

(1) I never have a hard drive full of anything. You need to investigate a good redundant backup scenario.

Never worked in a small biz where they just never upgrade much... anyways it was a bit of exaggeration.

Originally Posted by silvershoes

(2) The web is full of rumors and "expert" opinions and claims, some of which are simply a sales pitch. A lot of scanners that claim 16 bit have been shown to be 12, for example; the scans are still lovely. If it really matters to you, why not go to the source (the people who make your software package) and ask customer support?

You can download trials of just about every RAW processor on the market and do your own head-to-head comparison with your own images. The software's only half the equation; your own processing preferences and skill level matter just as much as the number of bits.

Good luck!

Yes but even with this hard line exporting FAKE 16bit files is just dumb or insulting. Whichever you prefer..
I'm looking into lightznoe right now (quite cheap currently) and yes there is a lot more to photography then processing (but it's fun to state the obvious)
A corollary to this is 8 bit color LCD's (x million colors or "UP TO" x million colors )or are they 6 bit interpolated. And did they change panels even keeping the same model numbers???
There are a LOT of windmills out there... I have mine...

No confrontation intended, just conversing.....

[Dom]

UFRaw saves in 16-bit TIFF.
UFRaw - Home

CinePaint you can work in 32-bit IEEE float if you want
CinePaint

Edit: If you want CinePaint for Windows, bad luck it's still in beta.

[philbaum]

I use Lighroom exclusively for processing, but whenever i transfer to Adobe Elements 7 for the edit module, it stops me from doing anything and asks if i want to convert the Lighroom TIFFs to 8 bit.

Photoshop may have 16 bit, i don't know, but Elements sure seems like an 8 bit operation.

Phil

[jeffkrol]

Originally Posted by philbaum

I use Lighroom exclusively for processing, but whenever i transfer to Adobe Elements 7 for the edit module, it stops me from doing anything and asks if i want to convert the Lighroom TIFFs to 8 bit.

Photoshop may have 16 bit, i don't know, but Elements sure seems like an 8 bit operation.

Phil

Pretty sure you are right regarding Elements.and Photoshop definitely has 16bit (and at times 32/64 bit) capability.. You may be interested in this though:
Elements+ unlocks hidden features in Photoshop Elements
No guarantees on my part .............
BTW: I'm leaning towards Bibble as of now.. not that it really matters