NEW Rendertest-HDV.veg

Comments

neilslade wrote on 7/10/2008, 1:33 AM
Finally, got my new system together-- the less expensive online purchases all worked first time out-- whew!!!

Anyway, did the render test-- no overclocking, just stock Q9450 with 4GB G.Skill memory, the Gigabyte GA-P3d-S3G motherboard

118 seconds

Now, for fun, overclocking... results--

3.25 GHz= 98 seconds

and that's about as fast as this Q9450 will go as far as I can tell. I'm just using the Gigabyte EasyTune utility, and beyond that is beyond my ability.

Can't say that I'll run this machine overclocked to save a few seconds, and as I've observed previous--- render while doing chores or sleeping, and a P4 will do just fine, and this was all mostly for jollies as far as I can gather.

Glad I figured how to save a few bucks in the process.

If there are some tweaks to optimize in Vegas for this quad, I've missed it-- anyone clue me in if there's something else I should have done... :-)

Cheers.
boggaf05 wrote on 7/10/2008, 6:52 PM
2:13 on Q6600
Aegis Kleais wrote on 7/11/2008, 10:21 PM
2:14

Q6600 2.4GHz
4GB PC2-8500
800+GB of 7200RPM 16MB SATAII HDD
Windows XP Pro SP2
rs170a wrote on 7/12/2008, 7:34 AM
What is the "standard" output being used for comparison?

From a post by John Cline much earlier in this thread:
...it is intended to be rendered to an MPEG2 HDV file using the default "HDV 1080-60i" template.

Mike
Joe Balsamo|LVX wrote on 7/12/2008, 8:14 AM
My time is 4:26.

I'm running a Sony Vaio laptop with a T9300 (2.5GHz) Duo Core, 4GB of RAM and 32 bit Vista.

Joe
JoeMess wrote on 7/12/2008, 8:28 AM
New HP dv9700 notebook with Turion TL-60 dual core running at 2ghz, 2 gigs of RAM and NVIDIA 7150M.

8:44

Joe
alltheseworlds wrote on 7/14/2008, 9:30 PM
9450 quad under XP with 4gb ram & everything on default settings: 118 secs (exactly the same as an earlier post by another 9450 owner)
JulianCH wrote on 7/15/2008, 7:22 PM
nice test
2':10 on custom built machine
ASUS P5E
Intel Quad Q6600 @ 2.4GHz
2Gb RAm
Windows XP PRO SP3
neilslade wrote on 7/16/2008, 2:38 PM
I made a few changes to my machine- so, of note for anyone building a new system.

I'm sticking with XP Pro, incidentally...

Primarily, I switched from IDE to SATA drives-- Microcenter (and I would expect others) had these for $99 for the Green Energy SATA 500GB Western Digital (my fave) dual speed drives....

No change in rendering times, even though my render files are going to a 2nd hard disc (as before)-- apparently, doesn't matter.

Please note: If you are changing from IDE to SATA, I had to stop using Ghost 2003 as my backup-- happily worked for years, but will not work easily, if at all, with SATA drives. I got Acronis True Image 11-- and not only did it work flawlessly with my new SATA drives, but the user interface is MUCH MUCH smarter than Norton Ghost. Bye bye to Ghost forever.

2) I experimented with OCing with the Q9450, and it's not really enthusiastic about overclocking on the Gigabyte board (which seems more than capable), and 3Ghz is about all it will handle-- I've gotten faster speeds on one day, and limited at under 3.1GHz on another. Doesn't seem worth it to save 5% on time and perhaps risk or shorten the life of any components for a teeny bit more speed.

Don't think it's a temperature thing, because everything is running very cool with five fans, the new HR-01 Plus Thermalright cooler with a big 120mm SIlenX fan running full blast, as are the case fans (4 of them)


All in all, happy with the upgrade, wish I knew then (before building) what I know now. :-)

Aegis Kleais wrote on 7/16/2008, 2:58 PM
OT : Your CPU may become unstable if it's not getting enough voltage to support the OC. Temps being low is a good sign, but if you bump the FSB (I think the Q9450 Multi is locked at 8x) then you need to make sure it gets the proper incremental boost to voltage sometimes.
Michael Daul wrote on 7/28/2008, 8:28 PM
Just an the render test on my new system...

2:02 (122 seconds)

Q9300 @1.175 V
8gb Ram
Asus P5Q Pro
Ati 4850
1 x WD 320gb system drive
2 x WD 640gb (raid 0) render drive
Vegas 8.0b
Windows XP x64
neilslade wrote on 8/11/2008, 2:24 PM
Some things definitely are a bit slower on Vegas 8-- I captured some new footage, and it sure seems to be a bit pokey on some capture operations-- nothing deadly horrible, but I noticed it right away.

Vegas 8 is supposed to be using a different codec, which is VERY welcome as far as I'm concerned.... recently I started looking at my projects with a High Def television monitor, and it alerted me to some vertical banding in saturated red areas next to dark areas or contrasting color areas like red next to green.

I also spotted banding with saturated colors against a very dark background.

At fist I thought it was my LCD TV, but then saw it on another LCD TV, and then went back to my original AVI captures--- THERE IT WAS.

It was hard to pin down discussion of this, but eventually I did.

Yep, its a real problem, and you won't notice on a standard monitor or TV-- but use a high def monitor-- and voila-- there she blows- and boy, does it BLOW.


Earlier versions of Vegas most definitely has this problem, and V 8 is using a Cinepack codec, which is supposed to eliminate this compression banding problem. I'll test this week and report back. Another cure for earlier versions of Vegas is to just replace the codec and use another like cidoceda.

Note- less codec compression will result in larger file size, and this may be the reason for the slow down while capturing.

Any thoughts anyone?

Thanks

Neil
http://www.BrainRadar.com

http://www.EasyPaintYourCar.com
http://www.InkJetHelper.com
http://www.MyOwnPublishing.com
http://www.neilslade.com/Papers/Painting.html
John_Cline wrote on 8/11/2008, 4:07 PM
The is nothing wrong with Vegas. The high-contrast, saturated color vertical banding is a function of DV's 4:1:1 color sampling, although the Vegas DV codec handles it somewhat better than other DV codecs. It's not a Vegas-specific problem at all.

The "damage" is already done in the camcorder when you shoot scenes which trigger the problem. Using the Cineform (not "Cinepack") codec will help avoid the problem when using generated media, graphics or text, but it will do nothing on footage that already contains the problem.

Magic Bullet has a "de-artifactor" plugin that will reduce or eliminate the problem and the free video processing program, Virtual Dub, has a "chroma smoother" plugin that works almost as well.

If you want some "worst case" sample files to test the various codecs, I have some that I posted on my web site when this subject has come up before:

www.johncline.com/colorspace.zip

Also, some LCD panels have banding issues themselves, but they manifest the problem with kind of a stepped contouring effect, not the vertical banding that you described.
JJKizak wrote on 8/11/2008, 4:44 PM
My experience with banding suggests that the TV stations are to blame for about 98% of it. OTA watching the same channel on a Sony CRT analog and a Sony XBR LCD 1080P showed identical banding with analog/SD digital channels. HD was very rare. I haven't seen any with the Sony Z1 yet. If there is any I can't see it with the Z1.
JJK
neilslade wrote on 8/16/2008, 7:32 PM
Reply to John per Vegas Banding--

Yes, I know its DV color-- just didn't know if this was a VEGAS issue or not-- on the web, some have reported this as a Vegas issue in particular. I've nothing to compare it to-- Of course I'm a big Vegas fan, and I'm not going to stop using it


Thanks for your input-- I didn't get around to testing V8 versus V6 to see if things improved-- apparently not.

I've documented the banding observation on my web page per today's test
www.neilslade.com/Papers/vegasbanding.html

The vertical banding seen in the captured video footage is NOT from
camera shake. These photos reveal exactly what I've seen in areas of high
color saturation, especially against a contrasting background. The same
vertical banding is present in both Sony and Panasonic captured footage.

I realize this is a problem with DV capture-- and the only cure I've found is adding
a chroma blur effect in Vegas, that greatly minimizes the problem post production.
I thought Vegas 8 would be better-- good grief, it looks worse than the same footage captured using Vegas 6 OW!

Ah well--- so goes life with MiniDV


I haven't been able to figure out of the Cineform codec is automatically used or not, or if I'm supposed to specify it somewhere-- if anyone knows--- clue me in.
Thanks



7) Replacing the original Vegas codec with the cedocida codec didn't help
Terry Esslinger wrote on 8/16/2008, 8:08 PM
Another time, if anyone is still interested:
7:46
Intel PentD duo core 2.66 MHz 2G mem
John_Cline wrote on 8/16/2008, 8:52 PM
Neil,

Yep, that's the 4:1:1 DV color sampling that's causing the vertical stripes on the "T." The vertical striping artifact has already occured in the camera as it got compressed to DV just before it was written to tape. Also, there should be no difference in the capture between any version of Vegas as it's just a straight file copy between the camcorder and the computer, there is no codec involved. As you figured out, a little chroma blur helps. If you happen to be generating high-contrast graphics, like the red T, then adding a bit of noise to the image will help if it needs to get compressed to DV when you render.

I'm not sure why V8 should look any worse than V6, I don't think the DV codec has changed in quite a few versions. By the way, the Vegas DV codec is the absolute best DV codec available (and I've tested them all.)

John
neilslade wrote on 8/16/2008, 11:35 PM
Thanks very much-- I feel better now :-)

At first this drove me crazy-- I noticed as soon as I got my HD LCD TV- and thought it was the TV at first, and eventually realized it was on the avi capture.

I had a belly dance recital that looked REALLY bad-- bright costumes against a black curtain--- good grief-- a Panasonic DVX100-- and it was AWFUL. Glad to hear you've done what I won't have to do, test all the codecs.

SO-- guess this means I'm due to upgrade my camera to keep up.

I am surprised in general how well the DVX footage looks, and mine is the first one, not even the 100A or 100B version, and the chroma blur NEARLY makes all this banding go away. I also found that burning to a 8.5GB dual layer disc is an improvement over the single layer 4.5 discs, if I need the extra sharpness.

Alas, really sharp will require an HD camera I suppose--, although obviously, I can't fit much HD on a single or dual layer disc. Will an HD camera solve this color sampling problem? Obviously not, if the camera is still using miniDV for tape---

any thoughts?

I can see banding in the V6, its just that it looks blurred in the frame I picked-- akin to adding chroma blur-- I dunno. I saw similar banded frames in the V6 capture as well. Just the one I picked was a little better.

Incidentally, this is off topic but-- I switched from Ghost to Acronis for backup after I put my new machine together and changed from IDE to SATA drives. Ghost just couldn't do it-- and if it could, I no longer had the patience to figure it out. Acronis is just SO MUCH better, and worked first time, no hitches at all.
John_Cline wrote on 8/17/2008, 12:35 AM
DV and HDV use completely different video encoding methods. The ONLY thing that they have in common is the tape itself. DV is 4:1:1 color sampling, HDV is 4:2:0, as are DVDs. Between the two, HDV's 4:2:0 is much better at handling high-contrast, highly saturated video.

Also, if you shoot 4:1:1 DV and then convert it to 4:2:0 MPEG2 to put on a DVD, then the color sampling you end up with on the disc is 4:1:0, which contains only 12.5% of the original color resolution. If you start with 4:2:0 HDV and keep it 4:2:0 through the editing process and then encode and burn to DVD, it keeps all the original color resolution it had to begin with.

Shooting HDV and then downsizing it to a standard definition DVD looks better than shooting DV.
neilslade wrote on 8/17/2008, 12:07 PM
cool, thanks
neilslade wrote on 9/11/2008, 10:32 AM
Hey John Cline,

Any problems rendering HD footage with V8 (done that?)

I'm reading a few horror stories on Creative Cow-- thinking that my next big project
I might just stick to V7... knowing that if I start it on V8 I won't be able to open it with
V7 if my quad core starts throwing fits in the middle of the thing

So far, on small regular non-HD projects I've had no issues.

See down the page

http://forums.creativecow.net/thread/24/882908
John_Cline wrote on 9/11/2008, 11:58 AM
Neil,

Obviously, there are people on the forum that have reported real issues with Vegas, I just don't happen to be one of them. I haven't had any problems whatsoever that could be attributed to Vegas v8.0b. I have it installed on three machines with completely different hardware configurations and I regularly do long-form HD projects for both Blu-ray and broadcast.

However, the game is about to change with the release of Vegas Pro v8.0c and the 64-bit v8.1 next week. We'll see what happens then.

John
Christian de Godzinsky wrote on 9/11/2008, 1:36 PM
Hi,

Recenly upgraded (or downgraded - according to some people) from XP x64 to Vista x64. I haven't made up my mind yet if that was a step backward or forwad ;) Yes - I know -Vegas 64 is supported only on Vista 64. Why do you think I switched!? ;)

My (and still the forum's? best) "new rendertest" record time worsened, from 82 seconds to 84. This is - without doing any tuning on either system (by removing unnecessary system processes). Both setups have exactly the same software installed.

Probably I will not bother to do some Vista tuning at this time, if the 64 bit 8.1 version is just around the corner.

Do you John (or anyone else) know if the 64 bit version already uses SS4 instructions? Is the promised/upcoming speedup when editing/rendering due to this fact, of just because of generally imporved code when now re-written as 64 bit?

Christian

WIN10 Pro 64-bit | Version 1903 | OS build 18362.535 | Studio 16.1.2 | Vegas Pro 17 b387
CPU i9-7940C 14-core @4.4GHz | 64GB DDR4@XMP3600 | ASUS X299M1
GPU 2 x GTX1080Ti (2x11G GBDDR) | 442.19 nVidia driver | Intensity Pro 4K (BlackMagic)
4x Spyder calibrated monitors (1x4K, 1xUHD, 2xHD)
SSD 500GB system | 2x1TB HD | Internal 4x1TB HD's @RAID10 | Raid1 HDD array via 1Gb ethernet
Steinberg UR2 USB audio Interface (24bit/192kHz)
ShuttlePro2 controller

jrazz wrote on 9/11/2008, 4:46 PM
1 minute 2 seconds with 8.1 pro default

1 minute 1 second with 8.1 pro with internal prefs set to 8 threads.

When I changed the settings to good from best it dropped to 31 seconds.

8 cores (2.50ghz each) and 8gigs of ram.

j razz