Sound Designers

>I obviously am not explaining the differences well enough. the
>Workspace is an iterative search engine which is why it works the way
>Advanced Search and Search in Field works. You will notice the
>workspace is considerably slower because of this design. It isn't
>treating it as separtor - it is just a character in front or after
>the other characters. This is the nature of iterative search methods
>and why they are slow. There's nothing specific to French users -
>this the WAY iterative engines work. They are considerably slower
>but do not distinguish between terms so the apostrophe is never an
>issue - it is just another character and it doesn't care what is
>before or after. A boolean engine, all boolean engines work the way
>I described unless we programme something differently. Instead of
>hitting cmd-F you hit cmd-shift F and you get the iterative engine.
>You have two ways of searching in SM not just one like in the
>workspace. This is the way it is for english persons looking for
>something with no metadata and only a filename "123mysong.wav". If
>you search 'my song' in the boolean engine it would not find it. It
>won't find it because it does not see a term for 'my' and 'song' but
>a term beginning with 123. But if you hit command shift F and typed
>'my song' it would find it. As an english user I can add that to my
>thesaurus or just add metadata to find it based on the terms 'my AND
>song". There is plus and minus to both way of working and as a
>working editor I always make it a point of knowing which search
>method is best for what I am doing. It isn't restrictive and it is
>not limiting to french users. This is search engine design.
>
>I'll ask what more we can do.

Thanks a lot for your anser Steve, it's very appreciated.
Even if i'll finaly get rid of all the apostrophes in my descriptions because of this (until you find something before september?).

Have you tried scanning sounds with ' in them into 4.1.8(into a new
database?), if the metadata doesn't contain Keywords then it should
split off the apostraphe

so d'ait becomes air in the keywords, that way a search for d'air will
find it in the filename(or description) and a search for air will be
found in keywords.

Try creating a new db and scanning some sounds in and viewing the
Kewords field. 4.1.8 should add keywords to the fields that get
searched for new databases.

Let us know what you find!

nx

Justin

Mix7 a écrit :Je viens de constater qu'il n'incrémente plus les noms de fichiers !

Je sélectionne une première partie du son "auto, skid, long", tout est ok !
Mais lorsque je veux utiliser une autre partie de ce même son (avec le même nom, donc), c'est la guerre.
Que faire, c'est une option dans Soundminer ou ProTools 7.4 ?

Dorian
Firstly, make sure you can rescan your files so the new routine can be
ingested by the scanner and potential keywords identified. It is
important to note that this in our estimation has more to do with
knowing the differences between the boolean and iterative search
routines in v4. It is NOT french specific, but may be more persistent
given common use of apostrophes in french..... but let's use Dorian's
example to illustrate:

Robust advanced search engines use boolean parsing and full index
searching in order to get speed optimizations. Instead of looking
through every character of every field, they look at a select number
of fields and create a new field called a 'full index' which contains
all the 'terms' found that would be useful for searching. When you
see a search engine alert you that it is indexing, it is doing this
kind of update to its full index in most cases. v4 users will
recognize the Modify Search index menu in the DATABASE menu - this
tells the boolean search engine which fields to build a full index
from. Terms are searched using AND/OR/NOT modifiers and they are
searched from the BEGINNING of the term. So this method is NOT best
if you are searching for a 'contains' set of characters. Searching
for 'Fond AND air' will not bring up 'Fond D'air' if the 'term' in the
full index is 'd'air' not 'air'. This new version if rescanning your
files should however break the apostrophe apart so that 'air' gets
added to your keywords field automatically and thus would be
returned. So while the engine cannot distinguish the apostrophe as a
separator the scanning process does and that will automatically create
a keyword addition 'air' to your keyword field. I have tested it and
it does work in the examples I have used...You CAN always add to your
thesaurus that 'd'air' and 'air' are synonyms or you can manually add
the term 'air' to your keywords or description field so it always
comes up as well. We have tried to give you the most flexible options
here and it holds true in english as well. In other words, it is the
same rules.

Now the other thing to note is that if you use SEARCH IN FIELD or
Advanced SEARCH (cmd-opt-F) you are not using the boolean parser but a
standard iterative engine. This looks for any set of characters
before,middle or end. It is a slower way of searching but is not
looking for 'terms' . This is the way Itunes or the workspace in PT
works. If you have a file called '123mybestfx.wav' and no other
metadata in the file, then these routines are your better option. The
main find box will look for a 'term' and none exist in a simple
filename all strewn together. Hit cmd-shift-F and you can use
'mybestfx' and find the file. Similarly, if searching for 'Fond
D'air' and you have no metadata and no keywords were created in the
scanning process, you can find this file by using the Search in field
or Advanced SEARCH windows.

Obviously we'll continue to see if any other refinements can be made.

steve