About this page

Q. Still not clear on the flow of video processing. A notional outline activity diagram can be found at EITS Video. However, some questions (noted by '?' on the diagram) remain.

Danelle Cline -> The EITS video page EITS Video has been updated.  Don't know the specific question this is referring to from the diagram, if you could post a more specific question that would be helpful.

Q. Lee needs to tell us how video and ancillary datDanelle Clinea will be delivered to shore.
Danelle Cline -> I think this is relevant. This is what it says in Lee's "Eye In The Sea - MARS Top-Level System Configuration" under Integration with VARS & AVED:

As a first thought...functions calls will be made by EITS software to VARS to log appropriate data...data stored into VARS by EITS will consist of a comma-delimted string, transmitted at a frequency of no greater than 1HZ...

[~erikar] The above quote is correct.  Lee will need an API from you to make calls and put data into the VARS system.
Brian Schlining From Feb 23rd 2007 conference call: Lee will deliver Video files ans CSV text anxillary data to the RAID drive on shore. We will write up a loader to stuff the anxillary data into EITS database tables.

Q. We intend to start processing video here at MBARI. Will the EITS team buy a computer cluster to use in Florida? (If processing is done at MBARI, we will need to back-up all video before we ship it to Florida.)

[~erikar] If all the initial AVED processing is done at MBARI, we will only need to view the processed video in FL. Therefore, our group will need to purchase a workstation in FL that can run VARS and AVED for post-processing purposes.

[~erikar] We still plan to keep all the original data from the mirrored RAID, as you mentioned.  It is my understanding that we will also need to have VARS and AVED loaded onto a computer here in Florida if we want to edit any AVED events; however, the computer in Florida will not require the same cluster arrangement as the AVED processing computer at MBARI.  Is this correct?  Our group will pay for the computer in Florida. Yes, you will not need a cluster. The project has allocated time to setup a workstation for deployment at ORCA which will have what you need to edit the events.  We also have travel money to come to Florida to help setup this computer and for any needed training. I imagine this computer will be a mirrored setup to what we have on the shore-side system here at MBARI (AVED, VARS, etc.), but we still need to work out those details.

Q. Where will we store all the data?

[~erikar] Our group would like to have back-ups of all video in FL and can store the data there; however, it might be easier to store the data at MBARI..........we will need to discuss this during our Friday meeting.
Brian Schlining From Feb 23rd 2007 conference call: The current plan is to host a replicated database both at MBARI and ORCA. MySQL 5.0 does support replication but we'll need to test this.
Brian Schlining From 2007-04-11. Setting up a server for integration testing. mbari649 (Mac OS X 10.4) is being setup with MySQL 5, Apache, and Darwin Streaming Server. Also setting up automount of RAID on nanomia.

Q. Is reprocessing of video in AVED a requirement? (Warn users that once data is edited in VARS it may not be practical to reprocess the corresponding video in AVED and load it back into VARS.)

[~erikar] Right now I don't see reprocessing video in AVED unless  major changes have been made to the code and earlier data needed to be reprocessed with the modified version.

[~erikar] Lee had a suggestion.  In the beginning we store 3 different video archives 1) all raw EITS data 2) AVED processed data and 3) edited AVED data.  VARS would have three fields, one for each archive.  The pointers in VARS would specify a directory path to the correct archive.  I can explain this better over the phone if you want to call me 772-321-5879.
Brian Schlining From Feb 23rd 2007 conference call: We will run though a demonstration of the entire data flow before we decide how to manage reprocessing.

Q. Where will the EITS computers live?

Initially at MBARI 

Q. What are the requirements for supporting time-code? Will we use time-code or run-time?
Danelle Cline -> The current plan on this is to use the AVED project cluster during the experiment, and setup a workstation for them to use to reprocess if they want. If the EITS team needs a compute cluster to reprocess video we can certainly help them setup, but it isn't in the proposed budget to buy a cluster. RE: ...we will need to back-up all video before we ship it to Florida.The current back-up plan is to create an mirrored image of the data. How/when this mirror is created is still TBD, but I imagine a mirrored RAID will be setup to handle this automatically.

[~erikar] Lee will provide you with a time-code from EITS
Brian Schlining From Feb 23rd 2007 conference call: Lee is going to investigate using a QuickTime container with MPEG-4 encoding. We'll send him sample code of how to write a TimeCode track. TimeCode will be written as GMT/UTC time.
Brian Schlining The current plan is for us to add the TimeCode track as part of our workflow. We have a script for doing this.

Q. We originally wanted to schedule a test-tank deployment in June; Craig says MARS the interface won't be ready. (i.e. no point in testing EITS). When will we reschedule?
[~erikar] forwarded the email about this:

Subject: RE: ROV and tank test questions
Date: Thu, 25 Jan 2007 16:53:39 -0800
From: Dawe, Craig <dacr@mbari.org>
To: Erika H. Raymond <erika@oceanrecon.org>

Hi Erika

As to test tank trials. I believe you are referring to a test using the Wet Node Simulator yes? If so, there are a couple of issues, first the simulator has to be built and that is dependent of some hardware from the MARS Electronics package. We do not know when that will arrive however..... The second issue is that the Wet Node Simulator is not scheduled to be completed until the middle of October. Hopefully, both of these issues will go away but that is the schedule for now.

Craig

Because of the test-tank delay, EITS delivery to MBARI may be pushed back to early April.

We will start our tank tests in Florida during April.