[Jason Whetstone]

Yep, they are all mp4. sorry, should have specified that.

[Jason Whetstone]

Thanks Bryson, I appreciate your insight. I totally agree with you, and I’ve thought about this a lot. Let me give you some background. We have no project management tool in our department that will allow us to track project status, billing status, milestones, whether or not masters have been made, whether or not archiving has been done, who worked on what, etc. Those tools exist, but can be expensive, and they would rely on someone manually changing those statuses. Advanced integrations are usually custom and even more expensive. Also , you have different places where data is stored – the accounting database, the project management tool, the media manager… etc. When we used FCSvr, we were able to track a lot of this information using Productions, because FCSvr gave you the ability to customize metadata for Productions and attach automations and notifications to them. The productions encapsulated all the media for a project, so it seemed natural to use the same system to manage the information AND the media. It worked great (even though there was a LOT that FCSvr could not do). I could automate the creation of project folder hierarchies on our SAN, automate archiving of all the assets linked to a FCP project or Production (all of that done via scripting mind you). It really was great to have all the data in one place, and it was even better that the data about the project was in the same place as the data about the media.

So, FCSvr goes away, in comes CatDV. We all love it. It’s such a great media manager. But project management? Hmm. To have this kind of functionality with CatDV, I would need to be able to customize catalog metadata (not clip metadata) and attach Worker automations to it. I don’t know if the ability to add custom metadata fields to catalogs is something that is realistic to expect for the future of CatDV. The Production Blog is great, but it’s not exactly a great place to store structured information. So I came up with a way to do something similar with CatDV using a combination of a catalog of dummy clips and a production group. I have a catalog of “project clips” and a corresponding “Project Info” production group that contains the field definitions, views, and other settings that display the data the way I want. A “project clip” is really a dummy clip in this “Projects” catalog that is just a data placeholder. The fields are things like project code, client, project title, start/end dates, and checkboxes for project milestones (concepting, pre production, production, post, approval, audio, encoding, delivery, mastering, archive). There’s also a field for labor and materials submitted agains the project. These are filled in using the Worker CLI with queries into our SQL Server accounting database using a bash script and SQLWorkbench/J in between. I also have fields that reference the actual media (number of clips with the same project code, number of masters, are the clips archived, which tapes in the library contain clips for this project, etc). I fill these in using the CatDV Worker CLI. I had to write a pretty complicated bash script to tie it all together. It’s a little clunky because you have to switch into this group to be able to see the data you want to see. But it works pretty well, and I don’t have to be a front end developer to see all the data in one place.

We do have a department of dotNET developers that could easily do this, but unfortunately this is an internal initiative, meaning there’s no project that this development work can be billed to, and they are BUSY. So I’m on my own here. I can put a SQL query together with the rest of them, but a front end developer I am not (yet anyway) – I’m more of an integrator. An HTML report would be helpful, but I want to be able to sort, group, and filter the data. CatDV does that so well!

So that’s the background…

[Jason Whetstone]

Thanks Dennis. I hope the Squarebox folks think so too and choose to implement some kind of HTML rendering, even the basic web kit. Another idea along these lines would be a way to add HTML links so that clickable web addresses and mailto addresses could be added which would open in your default browser or mail client. Another idea yet would be a clickable catalog or clip link. You can do this with the Production Blog but it would be nice to also be able to do this in a field.

[Jason Whetstone]

Another great technique – try different booming positions! Don’t always assume that overhead or even under hand is the best way – try from the side from somewhere in front of the talent but off-camera, etc, whatever works. When dealing with a noisy environment, try identifying the source of the most noise and use the polar pattern of the mic to your advantage. For example, we shoot a lot in grocery stores in the produce and dairy aisles. In the produce section, if you’re always booming over the talent – what’s behind the talent? noisy produce cases! Boom from under? You’re aim is at the ceiling away from those noisy cases (but stay away from HVAC!). In the dairy aisle, once I actually boomed all the way over the entire shot and placed the mic IN the dairy shelf, off-camera, pointing out towards the talent. it worked pretty well. Also don’t discount the possibility of asking the director if you can shoot somewhere else, 90% of the time the answer will be no, but it doesn’t hurt to ask.

[Jason Whetstone]

Mariam, Keep in mind that every audio source is different as Peter said. Just as everyone’s voice sounds different and resonates at different frequencies, every recorded voice and piece of music will sound different as well. So, you can’t go by the clip meter to set everything the same. Most meters only measure peaks, not actual sound level over time. You want to set the “loudness” to the same level, not just the peaks. The best way to do this is to LISTEN! Levelator is a good solution if you don’t have an audio post engineer to help you out, it will add some dynamics compression and limiting to to achieve a better result than no processing at all. I also sometimes use an SPL meter to tell me how loud the source actually is (there are many free SPL meter apps available for iOS and Android). I add a limiter to the mix and push the source level until the SPL meter tells me that (over time) the level is what I’m going for. If you don’t use the limiter, you may find that you have clipping in some projects.

[Jason Whetstone]

I actually figured out a workaround for this. It took me awhile, but it seems to be working.

I had to duplicate all the migrated fields that I wanted to stay in the catalog as user-defined fields. In essence, I had to make what FCSvr considers “Custom Metadata” into what CatDV calls “User Metadata” since the migration tool does not make that distinction. All in all this amounted to around 15 fields. The rest of the fields were either not of value, they were inherent to the media files that could be reanalyzed by CatDV, or they were garbage from the FCSvr migration. I made note of the ID’s of these new User fields (User 11, User 12, User 13, etc).
I made a new view in CatDV that consisted of the fields Name, In, Out, File Name, and then all the new user fields.
I selected all (7748) clips and exported those as a tab-separated text file.
I opened the tab-separated file in OpenOffice.
I changed the headings to NAME, IN, OUT, MEDIA FILENAME, USER 11, USER 12, USER 13…USER 44 and saved this.
In CatDV, I imported the tab-separated text file into a blank catalog.
I selected all (7748) clips and copied them to the clipboard.
I selected all (7748) clips in the original catalog.
I selected “Paste Metadata” from the edit menu and selected only User fields, and selected Overwrite.
I reanalysed all clips.

There are a few things to note – making a new view with the fields I mentioned is really important, as CatDV needs them to identify the clips. Make sure the original catalog and new catalog you make from the tab-separated text file are unsorted so the clips stay in the same order. Make sure you read the help file on working with unsupported file types in the on-line CatDV help, it explains the user headings that need to be set for the imported fields.

It might be worthwhile to add a checkbox field to every migrated clip to indicate that it’s migrated from FCSvr, and save the original migrated catalog in a safe place. That way if you later decide you need additional data out of the original catalog that was migrated with the FCSvr migration tool, you can use this process again and will have any easy way to filter the original migrated clips after additional clips are entered.

Now I think we have what we need. I will have to do this to each catalog that has important metadata though. Luckily it only amounts to about 10 catalogs, although they are all large, consisting of several thousand clips each. FCSvr does have a clear distinction between Custom Metadata and other metadata in its postgres database – If Square Box support is willing to improve upon it, it may help FCSvr-deserters with their migrations!

[Jason Whetstone]

I have discovered something that may shed light on this, I have opened a catalog created with the CatDV FCSvr migration tool. I exported some clips from the catalog into a CatDV XML file and examined it. I see that all the fields from FCSvr show up between METADATA tags. It appears that Re-analyse media considers the values in these fields expendable. I created a new User-defined field, populated it with a value, and re-exported the same clips to another CatDV XML file. I see in this file that the new User-defined field is listed as USER1. The value in that field is preserved after the Re-analyse media is performed. So it would seem that i need the fields from FCSvr to show up as USER-DEFINED fields instead of general fields… Anyone else running into this?