Pages: 1 [2]   Go Down

Author Topic: Good vs. not-so-good output profiles  (Read 14236 times)

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Good vs. not-so-good output profiles
« Reply #20 on: February 05, 2008, 09:23:39 am »

Quote
As to the accuracy of the device and its program, I did some round trips and found it to be very accurate that way, not with a reference spectrometer, but with known Lab values, and what the spectro read and what Photoshop reported.

Well there's the issue. If it's off but repeatedly off, that's not all that good.

And as we've heard just here (let alone the endless posts on DP Review), its a hit or miss process. Not good. Not necessarily a good bargain either.

Quote
There was a review of several profilers (I think it was on Luminous Landscape), including the PrintFixPro, and it did surprisingly well. 

Again, I'm not saying it can't build good profiles. I'm saying it can also build crap profiles and why, no one from CV seems to know or say. Flip a coin. My suspicion is the assumptions made by the hardware. I suspect if you could somehow feed it good spectral data, it would be far more consistent. But then the unit would cost more.

Doesn't CV have to update the software from time to time to different papers or inks? I don't know. What I do know is it gets mixed reviews in the field. I don't hear that from other products. And to get back to my original point, putting sliders into a profile package seems like an excuse to 'fix' what a good profile should produce out of the box.
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

AaronPhotog

  • Full Member
  • ***
  • Offline Offline
  • Posts: 128
    • Dygart Photography
Good vs. not-so-good output profiles
« Reply #21 on: February 07, 2008, 12:15:02 am »

Quote
Well there's the issue. If it's off but repeatedly off, that's not all that good.

And as we've heard just here (let alone the endless posts on DP Review), its a hit or miss process. Not good. Not necessarily a good bargain either.
Again, I'm not saying it can't build good profiles. I'm saying it can also build crap profiles and why, no one from CV seems to know or say. Flip a coin. My suspicion is the assumptions made by the hardware. I suspect if you could somehow feed it good spectral data, it would be far more consistent. But then the unit would cost more.

Doesn't CV have to update the software from time to time to different papers or inks? I don't know. What I do know is it gets mixed reviews in the field. I don't hear that from other products. And to get back to my original point, putting sliders into a profile package seems like an excuse to 'fix' what a good profile should produce out of the box.
[a href=\"index.php?act=findpost&pid=172449\"][{POST_SNAPBACK}][/a]

Andrew,

Where did I say it was off, or repeatedly off?  It's repeatedly accurate for me.  The package is quite good for the price.

Why would a proflier have to be upgraded for a paper if it's able to read the reflective densities and colors?  I've been able to spot paper inadequacies to be sure.  Some of them produce nice smooth curves, some don't, but the data is accurate, and repeatable, warts in the paper curve and all.  The DataColor program has been recently updated, but I think it's mostly a new interface.  It works just as well as before with a few new wrinkles.

If you read all the way to the end of the comparison article in the links I posted, you'll find that it did quite well, unless you are trying to make a perfect print of a grainger rainbow, an artificial construct.  In my case, I have no evidence of the kind of banding you can get with that in real images.  Sure, I'd prefer one of the expensive high end packages, but, hey... it works!!  Can there be operator error, or can other mistakes or bad choices be made in the profiling process?  You betcha!  But that's true of any package from the inexpensive to the pricy ones.  This one, because it is operated by hand rather than machine, may be the subject of mixed reviews for that reason (I haven't seen them, but I'll take your word for it).  I've seen some posts that sounded like other people besides myself were having issues with overinking that they couldn't recognize and didn't know how to fix.  You also have to be able to recognize good results from bad, no matter what program you use, and know how to take appropriate action (which can result in starting over in a different way).  I'm sure you use considerable care and quality testing to get the kinds of profiles you are making for Epson.

Your own book says that several of the better profiler packages, such as Eye-one XT, Fuji's Colourkit, Monaco Proflier Platinum (w/sliders), etc. have profile editing and adjusting capabilities.  Is this a sign of inadequacy or of flexibility?  It's not a matter of "fixing" but of having options and the tools.  Do I edit the profiles? No, most of the time I do not.  But, if I think that something needs adjusting, I expect the program to have the capability to do that.  This one has that capability, just like the "big boys."  Yes, I have used those sliders, but rarely, and not drastically in any case.

Aloha,
Aaron
Logged
Aaron Dygart,
Honolulu

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Good vs. not-so-good output profiles
« Reply #22 on: February 07, 2008, 09:29:05 am »

Quote
Where did I say it was off, or repeatedly off?  It's repeatedly accurate for me.  The package is quite good for the price.

There's really only one way to define accuracy here. One is to measure color patches with the product and compare that to a known, higher grade reference device.

The other thing you can do is repeat the measurements with the same unit and compare the differences, that tells you nothing about the accuracy of the data, only the repeatability.

And lastly, unless you measure the same target with two different instruments and build two profiles using the host software, all you can say is, you're happy with the profiles from the one product (you can't discuss accuracy).

Quote
Why would a proflier have to be upgraded for a paper if it's able to read the reflective densities and colors?

Here's what I suspect is going on, its only a guess based on what I've read about this product, what happens when people get poor results and when they get newer software to adjust/compensate/recognize this paper. I suspect this device isn't really anything much more than a colorimeter, perhaps with more filters but not a true Spectrophotometer (its not reading spectral data) and thus, it has to make a bunch of assumptions about what its measuring. If the assumptions based on the software are correct, you get a decent profile, albeit with tweaking and printing using the provided sliders (another reason I question what's going on under the hood). If the assumptions are off, you get a crap profiles.

There's no question in my mind that, based on those users reporting both good and poor results, there's something not totally well defined in this process. This is NOT my experience with other profiling packages using true Spectrophotometers (Moanco, GretagMacbeth, Praxisoft, Agfa, Fuji, etc).

Quote
If you read all the way to the end of the comparison article in the links I posted, you'll find that it did quite well, unless you are trying to make a perfect print of a grainger rainbow, an artificial construct.

Not at all. As I've said, its hit or miss. The links you provide do not address all papers and printers and I can if I had to, find plenty more posts where users reported problems.

Quote
Your own book says that several of the better profiler packages, such as Eye-one XT, Fuji's Colourkit, Monaco Proflier Platinum (w/sliders), etc. have profile editing and adjusting capabilities.  Is this a sign of inadequacy or of flexibility?

I also say they are and should rarely be used, that if you start with a poor profile, don't even start to think about editing it. If you want to do cross rendering, its useful. If you want to do very minor tweaks, long after printing lots of images, fine. This isn't how the CV editing tools are presented to the user. 99% of the time, you should never need to touch any such sliders if the profile is of good quality out the gate. And you'll notice that ALL the products you name provide a secondary editing process, not something you'd ever address while building the profile itself.
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".
Pages: 1 [2]   Go Up