Posts tagged with "metadata"
Tuesday, 18. September 2007, 09:38:32
Leo Stein, Gertrude Stein, metadata, Adobe
...
After asking Adobe why their new CS3 software strips the metadata from images, I've now got two responses from the software giant. My main question, "Why does APCS3 strip metadata?" was answered yesterday afternoon, ten days after I posted the question (they claim they respond within 24 working hours). Here's the reply:
You reference to your issue, this is a 'feature request', not an issue with the software, can I check that the software that you have purchased from Adobe works as designed, if you enter metadata in the Bridge for an image, is is retained when viewing inside of Photoshop?
My first reaction is that this is the corporate equivalent of sticking fingers in both ears and going 'LA-LA-LA-LA' when someone tells you something you don't want to know.
So my 'feature request' (they kindly provided a URL ending in 'Wish List'!) would be on the lines of 'Wouldn't it be great if Photoshop CS3 DIDN'T strip all the metadata from images?'
This is not a desirable feature. It is a mistake in the software.
You have to smile at the corporate mentality which is so convinced of its own perfection that it is incapable of acknowledging that it may possibly have made an error. There were a couple of small boys at school like that; I know they sincerely and genuinely believed they were perfect, and any failure or mistake on their part was met with tears of impotent and baffled rage. They're probably running Adobe now.
I'm reminded of a description I read of Leo Stein, Gertrude's elder brother:
"There seemed to be something in him which took it for granted that anything said by anybody except himself needed immediate denial or at least substantial modification."
The other reply to my more facetious question about 'Why is CS3 Design Premium $1,799 in the USA and $2,858 in the UK?' was, in full, from Adobe:
Taxes, US Paid-For Support, etc.
The prices I quoted were pre-tax from the Adobe Online Stores in the US and the UK. The answer isn't really satisfactory.
I don't choose to be a cynic, but that dismissive "Taxes, US Paid-For Support, etc." is enough to convert even Pollyanna into an embittered (and impoverished) old misanthrope.
And here I am.
Thursday, 6. September 2007, 12:20:25
metadata, Adobe, Photoshop, bug
...
As a picture library / stock agency it's vital for us to protect our members' copyrights. At fotoLibra our members add their captions and keywords, and we copyright the images in their names, adding the Copyright Status, Notice and Info URL.
All this metadata is neatly stripped out for anyone who uses Adobe Photoshop Creative Suite 3. Adobe Support charges us £30 if we want to speak to them (the CS3 suite costs a mere $3,200 in the UK) and meanwhile unscrupulous people can get hold of images with no clue as to their origin.
How do we get Adobe to listen to us? How do we contact them? How soon can they fix this bug? Sure, WE can use workarounds, but we have no idea what software our customers are using. And we are getting complaints from both our members and our customers that we're selling images with no metadata attached.
Of course we're not, but it's hard for them to believe it's the fault of giant Adobe rather than tiny fotoLibra.
But I do believe it is.
Friday, 24. August 2007, 12:17:10
stock agency, metadata, keywords, picture library
No apologies for going back to one of my favourite subjects. We're rewriting and rebuilding the fotoLibra site from the ground up for v4, and one of my tasks is to check all the advice pages on the site and tighten up the writing. I always end up adding more than I take away, so I'm splitting the pages into two Sections: SUMMARY — what you HAVE to do, and REASONS — why you must do it. It follows that SUMMARY is pithy and short, while REASONS allows me to go on at length. You don't have to read it, but it helps.
So here's an example of the new format:
Caption and Keywords: The Secret to Success
SUMMARY
Use dictionary headwords only — Nouns, Active Verbs, Adjectives, Adverbs.
Separate them with commas or semi-colons. No sentences.
REASONS
Without relevant descriptive keywords, images can’t be found, and if they aren’t found, they won’t be bought. However wonderful an image may be in composition, subject matter, lighting or rarity, a researcher can only track it down by the caption and keywords associated with it.
The CAPTION is a brief image description, restricted to 36 characters – “Camel Face On” is ideal, “You lookin’ at me?” is not. You’re not captioning a cartoon, you’re helping a buyer find your picture.
KEYWORDS must describe the image in detail – what it shows, where it is, the emotions it evokes, the time of year – using single words or very simple phrases, separated by commas and spaces. They must be accurate, apposite, concise, correctly spelt, descriptive, precise and relevant.
An essay on the preferred prey, terrain and mating habits of the Siberian tiger is irrelevant if the image is a close up of the tiger’s head.
fotoLibra runs an automatic translation facility on keywords for non-English language picture researchers. So
“Lion,aggression, hunter, tawny”
is translated automatically into Russian as
Лев, агрессия, охотник, смуглавый
and back into English as
“Leo, aggression, the hunter, tawny”
which is close enough.
Whereas
“The saying that the female of the species is deadlier than the male is particularly true in the case of lions, where the adult females usually hunt down the prey while the males stay at home drinking vodka”
is translated automatically into Russian as
Высказывание что женщина вида чем мужчина определенно поистине в случае львы, где женщины взрослого обычно охотятся вниз с пока мужчины остаются дома выпивая водочкой
and back into English as
“Statement that woman it saw than man definitely truly in the case the lions, where the women of adult usually okhotyatsya downward from thus far the man tyuey ostayutsya house drinking by vodka”
which isn’t really helpful.
Imagine you don’t speak Hittite. You are looking at a picture on a Hittite picture library web site with Hittite keywords. You have a Hittite dictionary. Bingo! (or !Ognib, as they say in Hittatia). You look up the keywords and there is the complete, accurate description of the picture.
Write your keywords as if your market were monoglot Hittites.
Saturday, 4. November 2006, 14:13:04
IPTC, metadata, CEPIC, Photoshop
...
OK, sorry about yesterday. There was more work needed on the Submission Guidelines than we forecast, but they've now been revised and the new version is online and ready for you to download. You can get it from
http://www.fotolibra.com/about/metadata.php along with downloads of the BAPLA / Pic4Press panel and the BAPLA Digital Guidelines.
It's worth reading the Metadata page before you upload images to fotoLibra. It's full of good advice.
I know. I wrote it!
Thursday, 2. November 2006, 17:40:00
metadata, IPTC, BAPLA, CEPIC
...
At last it's been released. Of course this means we now have to rewrite the Submission Guidelines which we only posted on the site last week. Still, we moaned about the pace of change in the document itself.
I'll put it up on the fotoLibra site and post a link to it from this blog tomorrow. It only works in Photoshop 8 and above, but we are working on a cunning plan to allow you to adapt and correct your metadata live on the fotoLibra site.
More information tomorrow!
Tuesday, 17. October 2006, 12:08:46
CEPIC, stock agency, picture library, Photoshop
...
I cannot emphasise enough how important metadata is. Without accurate and accessible metadata, picture sales become a mere lottery. No matter how good your picture, if the buyers can’t find it they won’t buy it. It is that simple.
And despite the (deliberately, I suspect) complex terminology, it’s basically quite easy to deal with.
Metadata really just means an extended description. It is a set of data that describes other data, in fotoLibra’s case an image. All our images are digitised, so the image becomes a digital object. In order to give it a value and to market it, this digital object needs to be identified and described. Without the proper metadata, it is valueless. The label has fallen off the print.
We need to know what the picture is of, how big it will print, who owns it, where it can be bought, what rights are available and so on. If the image is born digital (taken originally on a digital camera) then the camera software will input shooting data such as the exposure, whether the flash fired or not, the date and time of photograph and loads more. Rather scary really, but we don’t have much of an alternative. They already know everything else we do, and now they know I prefer wide apertures.
We are going to make it as easy as possible for fotoLibra members to supply metadata for their images. As soon as the BAPLA/Pic4Press XMP panel guidelines are established, we will make the panel available as a download from our site.
Until then, please caption and keyword your images using the Description panel in Photoshop which maps to the correct IPTC data. When you upload to fotoLibra using FTP or drag and drop, the metadata is retained and used to describe the image.
Thursday, 28. September 2006, 19:20:58
BAPLA, metadata, picture libraries, picture library
...
The first field in the Adobe Photoshop XMP Description Panel is "Document Title".
At fotoLibra we interpret that as "Caption", and images which have metadata added to this field and are uploaded via FTP or fotoLibra DND will have the information preserved and placed in the field on the site labelled "Caption."
Not for much longer. Take for example the image captioned "The Kick". This is referenced by the fotoLibra database as Image ID #149815. The image IDs are fine for internal use, and for invoicing and such like, but it's not a title or a caption.
The new BAPLA / pic4press proposals want this field to be renamed "Image Ref ID", and for it to contain the picture library's internal reference number. So instead of a Document Title or Caption of 'The Kick', it will now be called 149815.
Somewhat to my surprise, I was the only dissenting voice. The overwhelming majority of picture libraries favoured using an internal reference number in this field rather than a named caption.
So if this proposal is finally accepted and goes ahead, we will have to change and reorder our database to comply and be compatible. More upheaval, more expense; irritating but necessary.
This means members will have to learn to leave this field blank, because it will be overwritten on upload by our automated numbering system. You will have plenty of warning.
Additionally it was suggested that all picture libraries have a two- or three-letter suffix to distinguish individual image ref IDs. So 'The Kick' will become FL149815 or FOT149815.
Hmmm. I'm not convinced.
But we'll go with the flow.
Tuesday, 26. September 2006, 15:11:55
metadata, XMP, BAPLA, picture libraries
...
If you open a picture in Photoshop and then go to File> File info, you get an XMP panel, which by default will open in Description. You will see the following fields:
Document Title:
Author:
Description:
Description Writer:
Keywords:
Copyright Status:
Copyright Notice:
Copyright Info URL:
Fairly straightforward, you might think, and you’d be right. But as in every business the devil is in the detail. We could argue for hours (and in the BAPLA Metadata committee, we did) about every nuance of meaning in each descriptor.
At this moment — and it may well change — the BAPLA / pic4press proposed XMP Panel has the following fields:
Image Reference ID: (maps to Document Title in Photoshop’s ‘Description’ panel)
Description: (maps to Caption in the ‘Description’ panel)
Credit: (maps to Credit in Photoshop’s ‘Origin’ panel)
Licensing Contact: (maps to Copyright Info URL in the ‘Description’ panel)
Creator: (maps to Author in the ‘Description’ panel)
Date Created: (maps to Date Created in the ‘Origin’ panel)
Copyright Notice: (maps to Copyright Notice in the ‘Description’ panel)
Rights & Restrictions: (new field)
Headline: (maps to Headline in the ‘Origin’ panel)
Job Reference: (maps to Transmission Reference in the ‘Origin’ panel)
Clearly this will mean a change in the way we all work. But if we’re all working to the same standards, it should make life easier in the long term.
PACA, the American equivalent of BAPLA, has proposed a vastly more complex metadata array that includes such fields as ‘Ethnicity of Model’, which we felt belonged among the keywords.
Hang about, the sharper-eyed among you will say. What’s happened to the keywords?
Good question. I’ll find out and let you know by the end of the week.
Monday, 25. September 2006, 10:36:46
metadata, XMP, BAPLA, picture libraries
...
I went to a meeting of the BAPLA Metadata Committee on a blazing hot day last week. We are trying to set common standards for metadata fields which can be used by both picture libraries and publishers. We are working jointly with pic4press.
As fotoLibra members do their captioning and keywording themselves, I'm going to share the information with you so the images you upload are as close to the industry standard as possible.
Not quite the same excitement as pressing that shutter at the critical moment or post-processing the digital image, but every bit as essential to the sale and protection of your images.
The meeting was patiently chaired by Sarah Saunders of Electric Lane. What was shown was a new custom file info panel to slot into Adobe Photoshop. What was discussed was the nomenclature to be used on that panel.
My blog postings this week are largely going to be about this XMP panel. Use Trumalia or Google to find out more about XMP.
Photoshop CS2 comes with the ability to load additional panels to the basic Description / Origin / Advanced. When this BAPLA / pic4press file info panel is approved and released we will post it on the fotoLibra site so everyone can download and use it.
So what do you, as a fotoLibra member, need to do about this? If you use Adobe Photoshop, add metadata to every image you upload to fotoLibra by going to File> File Info> Description. If you don't have Photoshop, search Trumalia for alternative ways of inputting metadata.
My next four blogs will discuss each proposed field of the BAPLA / pic4press custom file info panel and show where these fields map to in other XMP data files.
WIDGETIZE