HD Playback on Catalyst

[samsc]

I will try the CODEC that you mention above and see how that works for me. Speaking of different CODECs, what do you know about H.264? It looks like a very promising new HD CODEC and I would like to try this with Catalyst as well. Been hearing a great buzz about this.
Scott Riley

make sure you test the latest software build - m147 - there are some optimisations for higher definition codecs.

buzz yes - useful no.

h264 is not good.
try apple intermediate codec.
try different image sizes with this - and compare-

ignore the buzz - h264 is an internet codec- data rates are not contrained in a media server - what a media server needs is quick...

these are not the qualities one needs for a media server:

Ratified as part of the MPEG-4 standard (MPEG-4 Part 10), this ultra-efficient technology gives you excellent results across a broad range of bandwidths, from 3G for mobile devices to iChat AV for video conferencing to HD for broadcast and DVD.

all these advanced technologies mean its slow-

H.264 in QuickTime 7 for Tiger implements a set of advanced technologies and patent-pending techniques to create pristine video at low data rates. The H.264 encoder features:
Intelligent multi-pass encoding for the best possible results at the desired bit rate with the optimal number of compression passes.
Time-saving single-pass encoding for creating draft encodes, meeting impending deadlines and live encoding with QuickTime Broadcaster.
Peak-constrained VBR options for limited data-rate scenarios such as streaming and CD/DVD playback.
Advanced frame reordering (B-frame) support to more efficiently represent movie data.
4x4 integer transform for reducing traditional blocking and ringing artifacts, especially in areas of fine detail.
Improved intraframe prediction for more efficient compression of details and gradients in high-motion video.
Increased precision in motion estimation for crisp reconstruction of objects in motion.
Flexible block sizes in motion estimation for more efficient encoding of complicated motion in areas of fine detail.
Adaptively tuned in-loop deblocking filter for eliminating blocking artifacts, resulting in a smooth, clean image.

[OlliR]

I am quite impressed:

3 layers running 1920x1080p 25 fps with Apple Intermediate Codec!

That´s a really good result. Though Harddrives seem to have hard work - judging by the sound of them.

System setup: MacPro 8xCore 3G 4GB Ram internal 3x500GB SATA Raid Level 0

[samsc]

I am quite impressed:

3 layers running 1920x1080p 25 fps with Apple Intermediate Codec!

That´s a really good result. Though Harddrives seem to have hard work - judging by the sound of them.

System setup: MacPro 8xCore 3G 4GB Ram internal 3x500GB SATA Raid Level 0

try another 2 layers - you might be able to do it.

looking at your stats - you might get 5 layers.
will get a bit peaky and possibly drop a few frames.

[OlliR]

I did try a fourth layer - but that ran at 12 fps and layer 3 dropped to 18 fps.

For the next test I wanted to use reproduceable content for the test - so everyone can try them themselves. I encoded the 1920x1080 test movie that came with the latest Catalyst version (4 Frames Black / Colour Bars / Picture 1 / Picture 2) to Apple Intermediate Codec. So I have a new picture every frame and not big areas of the same colour.

It ran poorly (Layer 1 at 25 fps and Layer 2 dropping from 25 fps to 24 alot).
Then I wanted to make a more realistic test and copied the 4 frames up to a movie length of 30 second or 180 MB. I usually don´t have 4 frame movies to play back - in reality it would be 30 sceonds to some minutes.

I managed to run three layers with these longer test movies at 25 fps. Layer 4 would run at 22 fps - with layer 4 running layer 3 drops to 24 fps.
This works best if each layer runs its own file and not all layer accessing the data.

That gave me the conclusion that the effect that I saw with the short files is the access time of the SATA Raid level 0. The slow access time compared to fast SCSI disks reduces performance with short files (jumping from end to start alot) - on the other hand the bandwith of the SATA Raid has advantages when playing back long files.

Would be nice if someone could check the performance of the AI codec on an 8xCore with a Fast SCSI disk.

Apple Intermediate Codec seems to be a good way to go - quality of picture seems to be better than Photo-JPEG 50%.

Here are the results of the same file encoded to Photo-JPEG 75% and 50%:

Photo-JPEG 75% Layer 1 and Layer 2 25 fps - Layer 3 18 fps

Photo-JPEG 50% Layer 1 and Layer 2 25 fps - LAyer 3 20 fps

I would have thought there would be a bigger difference in performance than two frames...

I will go for Apple Intermediate Codec for HD. now its time for bed.

Cheers, Olli

[samsc]

I did try a fourth layer - but that ran at 12 fps and layer 3 dropped to 18 fps.
Cheers, Olli

you have the classic performance characteristic of slow sata discs.

they dont work well- doing 4 layers - is fine of almost anything.

more than that and they dont work so well

Please - suggest as a test - you UNRAID your discs, and just test with a single disc- and see what the difference is?

[samsc]

I am quite impressed:

3 layers running 1920x1080p 25 fps with Apple Intermediate Codec!

That´s a really good result. Though Harddrives seem to have hard work - judging by the sound of them.

System setup: MacPro 8xCore 3G 4GB Ram internal 3x500GB SATA Raid Level 0

i think - from my testing - you are wasting the power of your computer with your sata RAID.
I can get 7 layers with my MTRON SSD. same computer.
you do have x1900?
You need x1900 at these resolutions 7300 has very poor performance at hi-def sizes

[OlliR]

i think - from my testing - you are wasting the power of your computer with your sata RAID.
I can get 7 layers with my MTRON SSD. same computer.
you do have x1900?
You need x1900 at these resolutions 7300 has very poor performance at hi-def sizes

I have the X1900.

I did the test you suggested. Though not unraiding the disks but copying the test files folder to the system disk.

Same result 3 layers with 1920x1080 playback at 25fps.
Layer 4 plays back just below 25 fps.

So the RAID seems to be no advantage at all. I did not expect this. I believed the RAID should have a greater bandwidth and perform better with larger files.

For sure I will look into the MTRON SSDs. But 1500 Euros for 64 GB is quite a statement...

[samsc]

So the RAID seems to be no advantage at all. I did not expect this. I believed the RAID should have a greater bandwidth and perform better with larger files.

For sure I will look into the MTRON SSDs. But 1500 Euros for 64 GB is quite a statement...

raids do not work - and are pointless unless you are trying to achieve high data rates - with only 2 or 3 layers for uncompressed playback -
other than that they are pointless. I keep trying to say this-

the problem is that SATA discs have very poor random access data rates - sure they can do 150MB/s or more in a RAID config- but that is for 1 file - read sequentially.
catalyst does not read 1 file sequentially. It reads 1 file for each and every layer. In a random and unpredictable manner.
And SATA raids can actually perform worse with catalyst than a single disc.
A hard disc is a rotating device that typically rotates at 7200rpm - one revolution in 8.3ms - only a single read head - if you miss the data you need on one rotation - you have to wait up to another 8.33ms before it comes round again. 8.33ms is forever in computers.

The MTRON discs are worth every single cent - you get up to 14 layers of playback - with no dropped frames - on any layers - because there is no rotational latency - access times are 0.1ms -
Being able to almost guarantee this performance across all files - at all times - in the type of shows we do - is a life saver.
Performance is critical in almost all the shows i do.
As it is in most video applications.

---

Did you find a source for the MTRON discs in the EU?
for 6 1080p layers rather than 3 1080p layers - double the performance - i dont think the cost is much.

[jjrecort]

the problem is that SATA discs have very poor random access data rates - sure they can do 150MB/s or more in a RAID config- but that is for 1 file - read sequentially.
catalyst does not read 1 file sequentially. It reads 1 file for each and every layer. In a random and unpredictable manner.
And SATA raids can actually perform worse with catalyst than a single

But What about WD RAPTORS..are SATA but with the same specs of the SCSI.. 10.000 rpm and

- Read Seek Time 4.6 ms
- Write Seek Time 5.2 ms (average)
- Track-To-Track Seek Time 0.4 ms (average)

The problem is when we have to move HD files.. we need both.. high output and high access speed!

[samsc]

But What about WD RAPTORS..are SATA but with the same specs of the SCSI.. 10.000 rpm and

- Read Seek Time 4.6 ms
- Write Seek Time 5.2 ms (average)
- Track-To-Track Seek Time 0.4 ms (average)

The problem is when we have to move HD files.. we need both.. high output and high access speed!

raptors do not work as well as scsi.

Use MTRON SSD. works much better