The following RAID solutions have been tested with with a range of HD and SD video modes and frame rates. Results have been obtained by the manufacture by self certification unless otherwise stated.

Tested Products

• ADTX ArrayMastor - L SCSI
• RORKE DATA MaxArray U320
• RORKE DATA Galaxy-i Series
  

For Standard Definition

Standard definition footage requires fewer resources than the High definition video.
The average transfer and data rate required is approximate 30 MB per second.
So for 1 minute of captured footage will require a storage array of 1800 MB or 1.8 GB, for 1 hour will require 180 GB of storage space. (See drive capacities section)

Minimum specifications
> Single channel SCSI or 8 port SATA
> 2 Ultra 160 drives at 7.2K, 10K RPM or 15K RPM
> Striped configuration software RAID

For performance (10 bit RGB, Multiformat playback and capture solutions)
> 4 or more Ultra 160/320 drives with dual channel card with even number of hard drives per channel
> Striped configuration in software RAID.

For High Definition

High definition 1920 by 1080 and 1280 by 720 capture and editing is very resource intensive and requires a considerably high transfer data rate than for standard definition editing.

The average transfer rate required for 1 minute of high definition editing without audio at the highest resolution is approximately 165MB per second. Even higher for 10 bit RGB file formats.
Therefore 1 minute of captured high definition footage will require a storage array of 9900 MB or 9 GB. 1 hour will require 594 GB of storage space.

Minimum specifications
> Dual channel Ultra 320 scsi card.
> Minimum 8 Ultra 320 drives at 10K RPM or 15K RPM
> Dual channel SCSI card with 4 drives per channel
> Striped configuration software RAID

For performance
> 10 or more Ultra 320 drives with dual channel card with 5 drives per SCSI channel
> Striped configuration software RAID

Performance of your storage solution can vary depending on the file format, bit depth, compression and resolution you use. RAID types and hard drives can also affect performance.

One of the most common methods to test your storage performance is to run an application that measures the data rate bandwidth from controller memory to the drive array. This test is always important because if your bandwidth is not greater than the format data rate, you will experience drop frames.

Once your bandwidth passes the first test you then need to run a series of capture tests to your array. Performance is greater when your array is empty compared to an array that is full and fragmented. The more full or fragmented the drive array, the more work the hard drives need to do to seek for space, that is where the performance of SCSI hard drives excel.

Disk benchmark tools do not fully test your array capacity, as the application will write a small file to the fastest part of the array, this measurement is mis-leading and inaccurate.

Bluefish444 approach method.

For Standard Definition testing

1. With Symmetry Application open, ensure that drop frame error reporting is enabled.
2. Capture to a range or clips with Symmetry starting with SD resolutions capturing to QuickTime and DPX files until the array is 30 percent full.
3. Capture to NTSC to a 1, 5, 10 , 20 minute clip in the QuickTime v210 file format
4. Do the same again with the DPX file format
5. Note any drop frame errors
6. Repeat step 2 again so the drive is approx 50 percent full
7. Repeat step 3 to 5 again, noting any drop frame errors.
8. Repeat step 2 finally filling the array up to 70 percent full
9. Repeat step 7 again.

For High Definition 1080 and 720p

1. Repeat steps 1 to 9 above but instead in step 3 capture to 1920 X 1080I 5994 or 1280X720P 60 and note any drop frame errors.

Please note the windows NTFS system will normally cause errors if the drive is near 90 percent capacity.