DO NOT USE THESE SSDs (one 7200rpm recommended) at DVinfo.net

Go Back   DV Info Net > High Definition Video Acquisition > HD and UHD ( 2K+ ) Digital Cinema > Silicon Imaging SI-2K

Silicon Imaging SI-2K
2/3" 1080p IT-integrated 10-bit digital cinema w/direct-to-disk recording.


Reply
 
Thread Tools Search this Thread
Old October 31st, 2010, 01:29 AM   #1
Major Player
 
Join Date: Feb 2009
Location: toronto, canada
Posts: 212
DO NOT USE THESE SSDs (one 7200rpm recommended)

ADATA - 128gb 500 series s596 (has a built in USB connection) - using the built in USB, the drive randomly buffers out on my m750 toshiba laptop.

mediasonic hd30su3-bk smart drive usb 3.0 external enclosure
WITH
Kingston SSDNow V-Series G2 128gb
Randomly buffering out during capture using the above external enclosure on usb 2.0 interface.
after approx 6mins and above.

Was having success with OCZ 128gb Solid series (with USB connection built in). Though, after 2 years, the drive is now having problems. The image will freeze for less than a second on the DVR monitor once every 5-10 seconds. Captured file is fine, but sometimes the image freezes for longer in which case the system buffers out.

May work with other configs, ie esata.. but I have had no luck using USB with the above drives.

Capturing successfully with 7200RPM Elephant drives so long as they aren't being bumped around. Captured over an hour in 2k filmscan 2 mode (highest quality cineform compression) 23.98.

Also capturing successfully to my internal 128gb SSD on the toshiba m750.
__________________
Justin Lovell
cinematographer//8mm.16mm>HD.transfers . www.framediscreet.com
Justin Lovell is offline   Reply With Quote
Old October 31st, 2010, 03:20 AM   #2
Inner Circle
 
Join Date: Dec 2003
Location: PERTH. W.A. AUSTRALIA.
Posts: 4,356
Justin.


Sometimes, if the ram buffer bar is climbing its way inexorably but persistently towards the top in something like a rock concert or musical performance, you may sometimes be able to extend or save your take if it is possible for you to zoom in close and reduce the fine detail in the image to broader shapes.

The ram buffer may even fall slowly again and allow you to go wide again for a while. The zoom-in and zoom-back has to be fairly quick.
Bob Hart is offline   Reply With Quote
Old October 31st, 2010, 03:32 AM   #3
Major Player
 
Join Date: Feb 2009
Location: toronto, canada
Posts: 212
you mean zooming with the lens, or tapping the zoom in hot spot?
im mostly on primes, but that is an interesting idea.
__________________
Justin Lovell
cinematographer//8mm.16mm>HD.transfers . www.framediscreet.com
Justin Lovell is offline   Reply With Quote
Old October 31st, 2010, 11:27 AM   #4
Inner Circle
 
Join Date: Dec 2003
Location: PERTH. W.A. AUSTRALIA.
Posts: 4,356
Justin.


For some of the following I may deserve a slap over the ears for my observations and assumptions. I would hope that those with more competence than I add their own corrections to any bum steer I head you on..


On the night I was using a Nikon stills zoom lens.

SI2K LOW LIGHT QUICKTIME GRADING TEST. By Bob Hart On ExposureRoom

This was the location, Sonic Sessions at the Fremantle Arts Centre hosted by Lucky Oceans. I could not post any of the actual show because of copyright issues. That became pretty much moot because a lot of people had their handycams at the five sessions anyway.

There was an interesting mix of cameras, Steve Rice's SI2K, Two RED Ones, two JVC GH-HD100s plus an EX1 on two nights for front of stage and front of audience roving. All cams and operators were supplied and operated on a sort-of prob-bono/deferrment arrangement.

They were good gigs, run as live shows with interview interludes by Lucky Oceans between sets.

Performers included Abby May and they managed to snag the legendary Hank Marvin. Sadly the production entity was unable to sell the concept to the chosen broadcasters and the project was shelved last I knew.

This was way back when we were still working out how to operate the camera and had not woken up to the "adaptive mode".

I was the operator on this night, had to set up the two JVCs and shepherd one of the operators, a student with no hours on the camera type, set up the sound feed to the SI2K and operate as well and on top of that, I did not have much of a clue. I had operated one of the two JVC GY-HD100s the two weeks before and had good nights with it.

My image did not look right. I was not then yet onto the false-colour exposure system or spot metering. I went across to the two guys operating the two REDs and studied their histograms, picked the one which was holding a similar frame to mine. I thought "yeah I can do this" and set my gain as low as I could and set the lens iris to make my histogram look like the RED's.

Mmmh-yeah wrong move. Major epic fail and underexposed to the max, almost black. This was when I discovered just how much the SI2K can give back to you when you stuff up. I discovered later the monitor needed to be set correctly, all part of the steep learning curve at the time.


The SI2K behaves the opposite way to long GOP cams like the Sony Z1 and EX1.

With the long GOP cams, you attempt to minimise movement or heavy changes of the image between individual frames to preserve fine detail, unless you are actually following violent moves, then it doesn't matter as motion blur and the action takes care of it. The codec throws away detail in favour of maintaining the frame rate.

So a zoomed-in hold on a gyrating performer will load the codec as there are considerable differences between each frame. Go wide and keep the camera moves to a minimum and your image is cleaner.


What I found with the SI2K was that it was the actual amount of complex detail in the wide frame which loaded the codec and drove the ram indicator bar to the top in the quality setting we were using.

I found that when zoomed in and holding on the performer, although there was movement by the performer, the overall complexity in the image was less with larger areas not containing fine detail differences and the buffer would slowly drop back down again. There was a point about two-thirds into the zoom range where the buffer would float at where it had been.

Now I am a bit older and wiser.

Last edited by Bob Hart; October 31st, 2010 at 11:45 AM. Reason: error
Bob Hart is offline   Reply With Quote
Old October 31st, 2010, 07:52 PM   #5
Trustee
 
Join Date: Sep 2003
Posts: 1,435
Bob wrote:

>> For some of the following I may deserve a slap over the ears for my observations and assumptions.

Slappety slap!!

>> you may sometimes be able to extend or save your take if it is possible for you to zoom in close
>> and reduce the fine detail in the image to broader shapes.

I'm not so sure about this. In my own observations, content of the frame did apparently affect bandwidth. But I thought it had something to do with the brightness of the scene. For instance, panning across the room would actually tax the SI2K system more on the parts when pitch-black was taking significant part of the frame.

Then again, this was not a scientific observation. Ari and the gang can probably shed the light on this issue.

My own take on this is as follows... if contents of your frame are affecting your ability to secure the take, then you have a weak system that needs to be upgraded.
Alex Raskin is offline   Reply With Quote
Old October 31st, 2010, 11:59 PM   #6
Inner Circle
 
Join Date: Dec 2003
Location: PERTH. W.A. AUSTRALIA.
Posts: 4,356
That was not a slap. That was a real hillbilly mom walking forward, "doubling" her retreating kid over the ears with both hands. Never mind. I like to think I am tough enough to take it ( he says as he runs howling from the house, chucks rocks at the roosters in passing, turns and screams "you'll be sorry an' I'm never coming back." ).

Levity aside - Agreed on the ability of the system to pass data. Using framing to save the take is a desperation measure but it is nice to know that it will sometimes work when the buffer is on the point of tipping.

Your comment on the blacks is interesting. Within the wide frame, there was a lot of dark area because the stage and the performers were more strongly lit. Subtle movements of people in the crowd in that darker area might also have been enough to add more to the system's workload.

I must drag the DVR2.0 demo into the system again to see what that does. On initial impressions, it seems to run smoother on the recorder unit/camera combination. Affordabiity is the big issue, so it is a luxury I cannot indulge right now.

Last edited by Bob Hart; November 1st, 2010 at 12:18 AM. Reason: error
Bob Hart is offline   Reply With Quote
Old November 1st, 2010, 01:14 AM   #7
Major Player
 
Join Date: Feb 2009
Location: toronto, canada
Posts: 212
i've also been told that a black frame will tax the system more. IE. if you put the lens cap on the camera and have buffer issues, you'll see it climb as the system has to work harder.
__________________
Justin Lovell
cinematographer//8mm.16mm>HD.transfers . www.framediscreet.com
Justin Lovell is offline   Reply
Reply

DV Info Net refers all where-to-buy and where-to-rent questions exclusively to these trusted full line dealers and rental houses...

Professional Video
(800) 833-4801
Portland, OR

B&H Photo Video
(866) 521-7381
New York, NY

Z.G.C.
(973) 335-4460
Mountain Lakes, NJ

Abel Cine Tech
(888) 700-4416
N.Y. NY & L.A. CA

Precision Camera
(800) 677-1023
Austin, TX

DV Info Net also encourages you to support local businesses and buy from an authorized dealer in your neighborhood.
  You are here: DV Info Net > High Definition Video Acquisition > HD and UHD ( 2K+ ) Digital Cinema > Silicon Imaging SI-2K

Thread Tools Search this Thread
Search this Thread:

Advanced Search

 



Google
 

All times are GMT -6. The time now is 06:39 PM.


DV Info Net -- Real Names, Real People, Real Info!
1998-2017 The Digital Video Information Network