[PROJECT HAS ENDED] The "Crowds" Library

Contribute or start a crowdsource project here.
Post Reply

Would you like to contribute to the "Crowds" Library?

Yes
99
100%
No
0
No votes
 
Total votes: 99

User avatar
samueljustice
Critical Listener
Posts: 24
Joined: August 21st, 2013, 2:04 pm
Contact:

Re: The "Crowds" Library

Post by samueljustice » August 6th, 2014, 6:08 am

I'm sure others do, but I have a copy of injector lying around somewhere. I can create a tab from text file and reinject the old files with the new metadata and chuck it up on GDrive if that also is of interest to anyone.

User avatar
stevexavier
Intern
Posts: 3
Joined: September 30th, 2013, 5:20 pm
Location: Southern California

Re: The "Crowds" Library

Post by stevexavier » August 6th, 2014, 9:25 am

MikeQuell wrote: I'm taking a guess here and assuming a decent amount of contributors wont have a file searching database equivalent to Soundminer's capabilities. Without the abbreviations in the file names, people can do simple searches in a finder/explorer window for things like "joshua" and get results for "_joshua_tree_".
Mike, I know a standard Finder search will include some metadata in the scope. To what extent, I've never tested, but some libraries pop up in standard finder searches that only match metadata.

Maybe we can run some tests to find out how to do that so that any extra keywords included in the description field will be included in the finder search.
I'll get with Alex on this.

User avatar
axk
Critical Listener
Posts: 22
Joined: September 30th, 2013, 4:23 pm
Location: Los Angeles, CA
Occupation: Sound Editor
Location: Los Angeles, CA
Contact:

Re: The "Crowds" Library

Post by axk » August 6th, 2014, 11:42 am

stevexavier wrote:Mike, I know a standard Finder search will include some metadata in the scope. To what extent, I've never tested, but some libraries pop up in standard finder searches that only match metadata.

Maybe we can run some tests to find out how to do that so that any extra keywords included in the description field will be included in the finder search.
I'll get with Alex on this.
Spotlight comments are stored separately from the actual files (unlike BWAV files, where metadata is stored in the file header) in an invisible .DL_Store file. If you've ever transferred files from a Mac-formatted hard drive to a HFS+ (Windows) formatted drive and seen those show up, that's what they are.

Soundminer does not primarily handle Finder comments when looking at each sound file that it imports/exports. Long story short, the only way to put info into Finder comments is through a work-around that requires some Automator/AppleScript batch processing. I'll see what I can come up with, but I'm not sure it's going to be the best solution. Plus, quite frankly, anybody that's going to be using these files should be doing so through some kind of database software. AudioFinder is a great solution and it's $70. MiniMiner, while far from a complete product, is $200. Basehead is awesome, although more expensive at $300, but people don't have to be using Soundminer v4.5 Pro to work with these files.

The game audio workflow is definitely a different workflow in comparison to post, and I imagine there is a much heavier reliance on filenames when working in a game engine than there is when working with a DAW for film sound. However, I certainly hope that anyone using these files is going to rely solely on Finder comments and filenames to search through them.

Mike, are we overcomplicating this, or is this worth looking into further?

User avatar
samueljustice
Critical Listener
Posts: 24
Joined: August 21st, 2013, 2:04 pm
Contact:

Re: The "Crowds" Library

Post by samueljustice » August 6th, 2014, 12:52 pm

The game audio workflow is definitely a different workflow in comparison to post, and I imagine there is a much heavier reliance on filenames when working in a game engine than there is when working with a DAW for film sound. However, I certainly hope that anyone using these files is going to rely solely on Finder comments and filenames to search through them.
You'll also find in the games industry a much heavier leaning towards Windows as opposed to Apple, primarily due to game engine toolsets only being developed for the single platform. Couple with tight deadlines, fast turnaround, crunching - speed is of the essence, so filenames (and filenaming convention) is massively important. Especially when you're dealing with 10-20k+ files.

User avatar
axk
Critical Listener
Posts: 22
Joined: September 30th, 2013, 4:23 pm
Location: Los Angeles, CA
Occupation: Sound Editor
Location: Los Angeles, CA
Contact:

Re: The "Crowds" Library

Post by axk » August 6th, 2014, 2:56 pm

samueljustice wrote:You'll also find in the games industry a much heavier leaning towards Windows as opposed to Apple, primarily due to game engine toolsets only being developed for the single platform. Couple with tight deadlines, fast turnaround, crunching - speed is of the essence, so filenames (and filenaming convention) is massively important. Especially when you're dealing with 10-20k+ files.
Yep! That's a great point. All the more reason that I don't think it'll be useful to use an approach that doesn't embed metadata in the files. I grew up on Windows and love it, but it does a lot of things that are drastically different from Apple's workflow...so we should be taking into account that whatever we do with metadata needs to translate across platforms.

User avatar
MikeQuell
Site Admin
Posts: 412
Joined: June 6th, 2013, 4:15 pm
Location: Los Angeles, CA
Occupation: Sound Designer
Contact:

Re: The "Crowds" Library

Post by MikeQuell » August 6th, 2014, 10:10 pm

Agreed. I was just suggesting we don't abbreviate any potentially useful words used in the filenames. I'm not saying we should embed more info in file header/spotlight comments on top of the BWAV metadata (soundminer tagging).

My Joshua Tree example was me running a search with the Windows' "search" bar for "Joshua" and not finding "_JshTree_" but it would find "_Joshua_" (even with the underscores on either side). So if someone wants to search for "Canada" they should be able to find results (instead of us using an abbreviation like "CNDA"). It's more about taking care of the hobbyists and people trying to get into the industry that don't have the same tools as us. I think we can find a happy middle ground between creating a professional quality library and something that's less intimidating for newcomers to be a part of and use.

-Mike

DanCostello
Critical Listener
Posts: 13
Joined: April 4th, 2014, 3:23 pm
Location: Boston, MA
Occupation: Sound Designer

Re: The "Crowds" Library

Post by DanCostello » August 8th, 2014, 11:10 am

Keep in mind that Windows also imposes tighter limits on filename lengths than OSX does. I forget what the limit is, but I know I've gone over it when unzipping on my PC commercial libraries that were developed on a Mac.

-Dan.

User avatar
samueljustice
Critical Listener
Posts: 24
Joined: August 21st, 2013, 2:04 pm
Contact:

Re: The "Crowds" Library

Post by samueljustice » August 8th, 2014, 12:43 pm

DanCostello wrote:Keep in mind that Windows also imposes tighter limits on filename lengths than OSX does. I forget what the limit is, but I know I've gone over it when unzipping on my PC commercial libraries that were developed on a Mac.

-Dan.
The most annoying part about that is that Windows also uses the path as part of the filename length as well. Very frustrating when trying to deliver a session that sits fine inside your computer, but in a larger dropbox path, half the audio files will not transfer. I think keeping filename size under 20 chars is best working practice.

argyllrb
Intern
Posts: 3
Joined: October 17th, 2013, 4:59 pm
Location: Edinburgh
Occupation: Sound Designer

Re: The "Crowds" Library

Post by argyllrb » August 18th, 2014, 6:59 am

Hi all,

Can you offer any guidance on the amount of editing that should (or shouldn't) be done to the submitted sounds? i.e. would you prefer a totally raw recording or is a little EQ ok if required (probably just a high-pass)?
Some of the recordings I've made have obvious disturbances/distractions in (sudden loud voices nearby, people passing by the mics, etc). Is it acceptable to cut down the recordings and fade the useful crowd/walla together?

On the subject of embedding metadata, I work under linux so I'd appreciate a file with comma/tab separated values or an embedded iXML chunk.

Thanks for your work on this.

Many thanks,
Richard

User avatar
axk
Critical Listener
Posts: 22
Joined: September 30th, 2013, 4:23 pm
Location: Los Angeles, CA
Occupation: Sound Editor
Location: Los Angeles, CA
Contact:

Re: The "Crowds" Library

Post by axk » August 18th, 2014, 4:08 pm

argyllrb wrote: Can you offer any guidance on the amount of editing that should (or shouldn't) be done to the submitted sounds? i.e. would you prefer a totally raw recording or is a little EQ ok if required (probably just a high-pass)?
Some of the recordings I've made have obvious disturbances/distractions in (sudden loud voices nearby, people passing by the mics, etc). Is it acceptable to cut down the recordings and fade the useful crowd/walla together?
This has been a huge issue with larger sound libraries and longer recordings. There are a few different schools of thought on how much processing work should be done when creating library sound effects, but at the end of the day, I personally will prioritize usability above all else (aside from high sound quality, of course).

I will always remove anomalous sounds from my recordings to keep them as easy to use as possible. If I'm using an ambience with some odd disturbance in it somewhere, I don't want to have to cut out the same sound over and over again. I'd rather do it once while I'm mastering my recorded files, then toss it into my library and not have to worry about it ever again.

Also, as far as noise reduction/EQ/etc...obviously you want to do everything you can to record things properly in the first place, so anything you can do in the field to avoid unwanted noise is helpful. However, it's extremely important to avoid too much processing when making a library. Everyone has their own different tastes when it comes to what a file should sound like to begin with, so it becomes potentially dangerous to do more than just some basic broadband noise reduction or a little roll-off to avoid low-end wind bumps and the like.

At Universal, I'll usually try to eliminate any unwanted wind noise or hum that isn't useful/intentional in the file I'm mastering (ie I don't want wind noise or birds in my car bys, or distant voices in my quiet forest ambience...but maybe I want that fluorescent light hum in my room tone) but I keep any non-organic editorial/cleanup to a minimum to avoid artifacts.
argyllrb wrote: On the subject of embedding metadata, I work under linux so I'd appreciate a file with comma/tab separated values or an embedded iXML chunk.
As far as metadata goes, check out the Ambient World Metadata Thread for a download link to a tab-delimited text file of the previous library's information. Let me know if you have any questions or comments!

Post Reply