First shoot with nano at DVinfo.net

Bob Griffiths · March 19th, 2010, 05:02 PM

Hi all,

Just finished my first shoot with the nano...19 days on location. Recorded just over one terabyte of info (Long GOP 100, 1080 30p). Used the nanoFlash Public Beta 1.5.31. Nothing fancy... just straight ahead recording. The nano performed pretty much flawlessly. Twice, it didn't want to go into Record. We were using HD-SDI embedded TC as the trigger. We simply stopped tape, re-booted the nano, and we were back in business. Used a remote tally light that George Griswold made for me. Worked great, George!

The picture quality is wonderful! I haven't played with anything higher than 100 Mbps so I don't know if it is the "sweet spot"... but it certainly looks sweet! And it's really efficient. There were days we shot 6 long interviews (nearly an hour each) and we still managed to cram it all onto 4 64Gb cards. Used the SanDisk Extreme Pro cards... flawless.

Did on-site back-ups to a Nexto 2500... nightly back-ups to a CalDigit VR Mini (RAID 1)... it took 3 1Tb drives... and a 4Tb CalDigit VR (RAID 0). So we walked away with a primary/back-up on one drive and totally separate copy on another drive. The Producer will take the VR home with her as a fail safe back-up and I will take the Mini's home with me for editing. Again, everything went flawlessly.

We ran tape as a further back-up. Anybody want to buy 50 DVCProHD, 33 minute, "one-pass wonders"? ;-)

Thanks to everyone in this forum and everyone at CD for helping me make this transition to tapeless acquisition!

Be well...

Dave Sperling · March 21st, 2010, 07:53 AM

Good to know.
On the average, how many takes a day? (19 days of shooting provides enough data for a rough 'relaiability percentage' ratio).
I'm assuming it was all tripod mount. did you try to mount your tally near the camera VF, or keep it on the Nano?

Bob Griffiths · March 21st, 2010, 10:16 AM

Well... hard to estimate by take numbers. Of course, the answer is... it depends. The most was on interview days where we did 6 formal interviews a day. They were about 45-50 minutes each. So figure that was about 300 recorded minutes per day. Lighter "b-roll" days could be as little as 90 mins. per day. My 4 SanDisk 64Gb cards will hold a total of 335 minutes at 100Mbps.

Yes, most was on a tripod. We snaked the cable for the tally up the camera body and gaffer taped it near the VF. Worked great. The nano itself also has a status light next to each card slot. Whichever slot is recording at the time, the status light blinks red, acting as a tally. The way I have the nano mounted, this flashing light can be seen from the operators side of the camera. You can see it here: http://www.dvinfo.net/forum/converge...-solution.html

HTH,

Dave Sperling · March 21st, 2010, 10:48 AM

Hi Bob,
Thinking more in terms of # of takes. (Normally, if I've had a problem, it's been at the start of recording for a take.) Just mathematically, if you averaged 11 takes a day for 19 days, that would mean a 99% reliability rate (2 roll-start failures out of ~200 takes). My first shoot with the Nano we were using it as a relatively unattended 'video assist' recorder - we had it in a different room (with the clients) at the end of a LONG bnc (that had been looped through the director's monitor,) so on occasions when a problem occured we weren't immediately aware, and there could have been any of many causes for signal problems along the way. On that shoot we averaged about 35 setups a day times 3-4 takes per setup, but didn't get an accurate problem rate (since the Nano was just for client viewing/ playback convenience), though it seemed like a couple of times a day (~98.4% reliable) The last shoot with it we used it as the primary record deck, kept the cable runs short and went straight from the camera to the Nano, and had a Zero failure rate.
Sometimes having an estimated predicted failure rate - ie Reliability Rate - can help (like MTBF on hard drives) make clients more understanding of the need to spend a few seconds on a reboot when a problem occurs.
Best,
dave S

Dan Keaton · March 21st, 2010, 01:51 PM

Dear Bob and Dave,

We, of course, are trying for 100 percent reliability.

The Public Beta that Bob was using had a feature, designed to allow takes to be reviewed in camera, which was enabled whenever one used Trigger On Incrementing timecode.

Since this was enabled, if a "high" timecode value was received by the nanoFlash, followed by a "lower" timecode value, the nanoFlash, by design, would not go into record.

We have decided to make this feature optional.

And, the default would be to have the feature off. When the feature is off and Trigger on Incremented Timecode is on, any change in timecode triggers recording.

So, as I read Bob's post, it wasn't that a recording was bad, it was that it did not go into record (which, of course is just as bad). This points to this feature being the problem, especially when the nanoFlash was power cycled, it worked fine. (By design, we do not remember the "High" timecode value across a power cycle.)

Having this as an option allows those who want to review in camera turn on this option, and those who do, or do not want to take a risk in the nanoFlash getting an invalid "High" timecode value, to have this feature off.

This is already programmed in our next Public Beta which we hope to release soon. It is undergoing final testing.

Bob Griffiths · March 29th, 2010, 08:35 AM

Quote:

Originally Posted by Dave Sperling

Sometimes having an estimated predicted failure rate - ie Reliability Rate - can help (like MTBF on hard drives) make clients more understanding of the need to spend a few seconds on a reboot when a problem occurs.

Hey Dave,

Sorry, I've been in internet hell for the last week and out of touch.

I just looked at my recorded files and I was averaging about 70 "takes" a day... a take being defined as a record "start/stop"... including files that the nano automatically creates on long record times. I had 16 actual shoot days. So about 1100 takes. Divide that into 2 and you get a failure rate of 0.00181818... statistically insignificant. Of course, if it happens on the 1 take to REALLY need, it still sucks.

HTH

March 19th, 2010, 05:02 PM	#1
Bob Griffiths Regular Crew Join Date: Sep 2009 Location: Carbondale, Colorado Posts: 117	First shoot with nano Hi all, Just finished my first shoot with the nano...19 days on location. Recorded just over one terabyte of info (Long GOP 100, 1080 30p). Used the nanoFlash Public Beta 1.5.31. Nothing fancy... just straight ahead recording. The nano performed pretty much flawlessly. Twice, it didn't want to go into Record. We were using HD-SDI embedded TC as the trigger. We simply stopped tape, re-booted the nano, and we were back in business. Used a remote tally light that George Griswold made for me. Worked great, George! The picture quality is wonderful! I haven't played with anything higher than 100 Mbps so I don't know if it is the "sweet spot"... but it certainly looks sweet! And it's really efficient. There were days we shot 6 long interviews (nearly an hour each) and we still managed to cram it all onto 4 64Gb cards. Used the SanDisk Extreme Pro cards... flawless. Did on-site back-ups to a Nexto 2500... nightly back-ups to a CalDigit VR Mini (RAID 1)... it took 3 1Tb drives... and a 4Tb CalDigit VR (RAID 0). So we walked away with a primary/back-up on one drive and totally separate copy on another drive. The Producer will take the VR home with her as a fail safe back-up and I will take the Mini's home with me for editing. Again, everything went flawlessly. We ran tape as a further back-up. Anybody want to buy 50 DVCProHD, 33 minute, "one-pass wonders"? ;-) Thanks to everyone in this forum and everyone at CD for helping me make this transition to tapeless acquisition! Be well... __________________ Bob FireDancer Productions, Inc.

March 21st, 2010, 10:16 AM	#3
Bob Griffiths Regular Crew Join Date: Sep 2009 Location: Carbondale, Colorado Posts: 117	Well... hard to estimate by take numbers. Of course, the answer is... it depends. The most was on interview days where we did 6 formal interviews a day. They were about 45-50 minutes each. So figure that was about 300 recorded minutes per day. Lighter "b-roll" days could be as little as 90 mins. per day. My 4 SanDisk 64Gb cards will hold a total of 335 minutes at 100Mbps. Yes, most was on a tripod. We snaked the cable for the tally up the camera body and gaffer taped it near the VF. Worked great. The nano itself also has a status light next to each card slot. Whichever slot is recording at the time, the status light blinks red, acting as a tally. The way I have the nano mounted, this flashing light can be seen from the operators side of the camera. You can see it here: http://www.dvinfo.net/forum/converge...-solution.html HTH, __________________ Bob FireDancer Productions, Inc.

March 21st, 2010, 01:51 PM	#5
Dan Keaton Inner Circle Join Date: Dec 2002 Location: Augusta Georgia Posts: 5,421	Dear Bob and Dave, We, of course, are trying for 100 percent reliability. The Public Beta that Bob was using had a feature, designed to allow takes to be reviewed in camera, which was enabled whenever one used Trigger On Incrementing timecode. Since this was enabled, if a "high" timecode value was received by the nanoFlash, followed by a "lower" timecode value, the nanoFlash, by design, would not go into record. We have decided to make this feature optional. And, the default would be to have the feature off. When the feature is off and Trigger on Incremented Timecode is on, any change in timecode triggers recording. So, as I read Bob's post, it wasn't that a recording was bad, it was that it did not go into record (which, of course is just as bad). This points to this feature being the problem, especially when the nanoFlash was power cycled, it worked fine. (By design, we do not remember the "High" timecode value across a power cycle.) Having this as an option allows those who want to review in camera turn on this option, and those who do, or do not want to take a risk in the nanoFlash getting an invalid "High" timecode value, to have this feature off. This is already programmed in our next Public Beta which we hope to release soon. It is undergoing final testing. __________________ Dan Keaton Augusta Georgia

March 21st, 2010, 07:53 AM	#2
Dave Sperling Major Player Join Date: Jan 2010 Location: Tinton Falls, NJ Posts: 780	Good to know. On the average, how many takes a day? (19 days of shooting provides enough data for a rough 'relaiability percentage' ratio). I'm assuming it was all tripod mount. did you try to mount your tally near the camera VF, or keep it on the Nano?

March 21st, 2010, 10:48 AM	#4
Dave Sperling Major Player Join Date: Jan 2010 Location: Tinton Falls, NJ Posts: 780	Hi Bob, Thinking more in terms of # of takes. (Normally, if I've had a problem, it's been at the start of recording for a take.) Just mathematically, if you averaged 11 takes a day for 19 days, that would mean a 99% reliability rate (2 roll-start failures out of ~200 takes). My first shoot with the Nano we were using it as a relatively unattended 'video assist' recorder - we had it in a different room (with the clients) at the end of a LONG bnc (that had been looped through the director's monitor,) so on occasions when a problem occured we weren't immediately aware, and there could have been any of many causes for signal problems along the way. On that shoot we averaged about 35 setups a day times 3-4 takes per setup, but didn't get an accurate problem rate (since the Nano was just for client viewing/ playback convenience), though it seemed like a couple of times a day (~98.4% reliable) The last shoot with it we used it as the primary record deck, kept the cable runs short and went straight from the camera to the Nano, and had a Zero failure rate. Sometimes having an estimated predicted failure rate - ie Reliability Rate - can help (like MTBF on hard drives) make clients more understanding of the need to spend a few seconds on a reboot when a problem occurs. Best, dave S