Any Minolta/Sony users using UFRaw and GIMP?

Discussion in 'Digital Cameras' started by Jeffery Small, Apr 5, 2014.

  1. Jeffery Small

    Eric Stevens Guest

    No. You are saying it. On checking I find that that is your statement.
    The first use of these words appears in Message-ID:

    "Classic troll diversion. You have yet to tell us how it is
    *evident* that:

    1. It has improved his workflow
    2. His workflow was previously ineffective"

    Later on, in Message-ID: <>
    Tony Woods attempted to mollify you by writing:

    "OK...just for shits and giggles I'll concede this point and revise
    my statements:

    1. The use of LR has not improved nospam's workflow at all. It's
    as creaky and useless as it was before.

    2. His workflow was not ineffective before, it was simply the best
    that could be done under the GIGO rule.

    I'll let nospam correct whichever one of us that he wants to
    correct."

    I'm not sure whether or not nospam accepted the invitation of the last
    sentence. In any case you now appear to be arguing with words created
    by yourself.
    If you think I'm wrong you should falsify my statement. Simple denial
    will not do.





    Message-ID: <>
    Classic troll diversion. You have yet to tell us how it is *evident*
    that:

    1. It has improved his workflow
    2. His workflow was previously ineffective



    Message-ID: <>
    OK...just for shits and giggles I'll concede this point and revise my
    statements:

    1. The use of LR has not improved nospam's workflow at all. It's
    as creaky and useless as it was before.

    2. His workflow was not ineffective before, it was simply the best
    that could be done under the GIGO rule.

    I'll let nospam correct whichever one of us that he wants to correct.
     
    Eric Stevens, Apr 21, 2014
    1. Advertisements

  2. Jeffery Small

    Guest Guest

    the random image was one i shot a month ago at a dance performance and
    it's not for public consumption. one performer in the image was holding
    a guitar and that is where the visible change was. there was also some
    minor differences on her clothing.

    the key is that anyone who claims converting to lab is lossless is
    mistaken.
    open image, duplicate (image/duplicate), convert to lab and back to rgb
    in the second image (image/mode) then subtract (image/calculations).
    look at the histogram to see the extent of the changes.
     
    Guest, Apr 21, 2014
    1. Advertisements

  3. Jeffery Small

    Guest Guest

    it does, but what it doesn't do is convert on the fly. it *uses* lab as
    a reference. read what andrew wrote.
    that is incorrect, as explained by andrew rodney in what i quoted and
    easily demonstrated.

    converting to/from lab mode is a lossy operation. whether that matters
    to the user is up to the user to decide but anyone claiming it's
    lossless (namely dan marguilis) is mistaken.
     
    Guest, Apr 21, 2014
  4. Jeffery Small

    Guest Guest

    there's a faint line where the stitching is.

    however, to really see what changed, check the histogram (command-l).

    that stuff on the left means there's a difference and looks like it
    extends to 27 out of 255 levels. slide the white (rightmost) slider to
    the left and you can maximize just what was actually lost and it's more
    than just the stitching.

    now create a new image filled with pure black and look at its histogram
    in case you need to see what zero really means.

    aka: you are wrong.

    after that, go ask anyone on the photoshop team if a lab conversion is
    destructive. they'll set you straight. i know you don't believe me but
    maybe you'll believe one of them.
     
    Guest, Apr 21, 2014
  5. Jeffery Small

    Guest Guest

    there's a difference there too.

    i did:
    image/duplicate, append -converted to the name.
    image/mode/lab
    image/mode/rgb
    image/calculations, original on top, converted below, subtract mode

    optionally, just before the calculation step, switch between the two
    images while looking at the histogram palette. there's a difference.
    then do the calculation step.
     
    Guest, Apr 21, 2014
  6. Jeffery Small

    Guest Guest

    read what andrew rodney wrote. it uses lab internally for a reference.
    it does *not* convert from lab on the fly, as that would be slower.
     
    Guest, Apr 21, 2014
  7. Jeffery Small

    Guest Guest

    i ignored it because there is no gigo rule. tony is trolling, as usual.

    my workflow at the time was the best it could be given what tools were
    available. when new tools were released, my workflow improved. simple
    concept.

    in the future, there will be even better tools available, with a
    further improvement in workflow.
     
    Guest, Apr 21, 2014
  8. Jeffery Small

    Sandman Guest

    This is the statement:

    Eric Stevens
    Re: Any Minolta/Sony users using UFRaw and GIMP?
    04/19/2014 <>

    "Well of course he didn't actually say that. The particular
    meaning congealed from with a cloud of diffuse verbiage."

    Just how do I go about falsifying that statement, Eric? My inability to
    falsify a vague statement based on an unknown number of vague sources does
    not mean that the statement is correct.

    You made the claim, thus the burden of proof lies upon your shoulders.
    Claims doesn't become facts just because others can't disprove them.
    Indeed. I'm waiting.
     
    Sandman, Apr 21, 2014
  9. Jeffery Small

    Eric Stevens Guest

    I don't really understand what he is saying. He seems to be thinking
    of using a colour engine which works differently from what is
    described in most of the books I have read. What he says is also
    contradictory. For example:

    "Some users are under the impression that Photoshop does all its
    conversions to and from LAB, converting on-the-fly. this is untrue
    as it would greatly slow down performance. Instead, Photoshop uses
    LAB as a reference when conducting many operations. Photoshop is
    not actually converting pixel data between color spaces unless you,
    the user, actually ask for this."

    Photoshop must be converting pixel data unless the image source, your
    monitor all have identical profiles.

    You can get access to this 2005 publication at
    http://tinyurl.com/m3u98fb . His description of working spaces doesn't
    sound at all right and certainly doesn't include device independent
    colour spaces.

    On page 54 he says:

    "The RGB working spaces provided by Photoshop are synthetic
    mathematical constructions. In fact, for a time, I used to call
    these RGB working spaces 'Quasi-devide-independent' until my
    technical editor, a true color scientist asked me to refrain from
    the practice."

    Rodney's criticism of the burden imposed on the CPU by the conversion
    to Lab space has probably been rendered irrelevant by the advent of
    fast multicore processors and the GPU processors which can help with
    intensive general processing.

    But that was back in 2005 or earlier. It may well be that at that time
    Adobe was using their own synthetic RGB working space.

    The best clue I can get as to what Adobe might be doing now is on
    pages 15 and 16 of http://tinyurl.com/jwskdc6 (Color Image Processing,
    Methods and Applications - Edited by Rastislav Lukac, Konstantinos N.
    Plataniotis - 2007) where the general discussion is of the Adobe Color
    Space and the file tags required to convert images to Lab. But 2007 is
    still some time in the past and I would like to see something more
    uptodate.
    -
     
    Eric Stevens, Apr 21, 2014
  10. Jeffery Small

    Eric Stevens Guest

    I've never said it was not destructive.
     
    Eric Stevens, Apr 21, 2014
  11. Jeffery Small

    Eric Stevens Guest

    I agree with you. He was poking you with a stick.
    I think everyone understands that.
     
    Eric Stevens, Apr 21, 2014
  12. Jeffery Small

    Eric Stevens Guest

    Your ability to falsify that statement is equalled only by your my
    ability to point you to a single source.
     
    Eric Stevens, Apr 21, 2014
  13. Jeffery Small

    Sandman Guest

    I haven't asked you to point to a single source. I have asked you to
    support your statement, if it requires multiple sources, then so be it.

    I'm waiting.
     
    Sandman, Apr 21, 2014
  14. Jeffery Small

    Guest Guest

    i don't know what books you've read but andrew rodney is more likely to
    be right than another book, where the authors don't know the photoshop
    team personally.
    how is that contradictory? photoshop uses multiple buffers and could
    easily have a second copy in lab in addition to rgb, or it could
    convert to lab for a specific operation.
    actually you can't. that link comes up blank here, but even if someone
    does search for a phrase or sentence to get google books (which is what
    i suspect you did), it's a session based link that is for just that
    person and will time out eventually.
    lab is device independent.
    it's certainly smaller but it's non-zero. there's no need to convert
    when a conversion is not needed.
    that book looks like it's more about algorithms rather than the
    internals of photoshop.
     
    Guest, Apr 21, 2014
  15. Jeffery Small

    Guest Guest

    you did ask why a conversion to and from lab would matter.
     
    Guest, Apr 21, 2014
  16. Jeffery Small

    Alan Browne Guest

    So find a photo in your vast photographic repertoire that you can share
    without embarassment and post the results.

    You're really in "walk the walk" territory now.
    Fine. I did exactly that with 2 other images and posted them.

    No visible difference in either case.

    So now it's up to you to show us a difference.

    Walk the walk nospam.
     
    Alan Browne, Apr 21, 2014
  17. Jeffery Small

    PeterN Guest

    Then use a MacBeth color chart.
     
    PeterN, Apr 21, 2014
  18. Jeffery Small

    PeterN Guest

    Don't hod your breath. nospam will produce even less then Bruce.
     
    PeterN, Apr 21, 2014
  19. Jeffery Small

    Alan Browne Guest

    Not on my screen. Even @ max brightness. Which difference image are
    you looking at anyway? Is it the one marked "-D" in the file name?
    Nope. The only thing that shows is am uncorrected (not healed) dust
    mote on one image (Lab or RGB) that was not corrected on the other (RGB
    or Lab). Funny that you don't mention that at all - 'cause that's the
    only thing that does show in the histogram when the sliders are moved.

    But do post what you're seeing.
    So it's up to you to post what you have.
     
    Alan Browne, Apr 21, 2014
  20. Jeffery Small

    Eric Stevens Guest

    I don't think I asked that. I have asked why converting to Lab would
    require more conversions than to any other working color space. You
    still haven't told me.
     
    Eric Stevens, Apr 21, 2014
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.