Stress test for CCE on your machine

[DDogg]

OK, I wish those of you having trouble with CCE and/or those who are overclocking would try this simple script below. This removes a lot of variables from the equation. This is the first place to test.

Just paste this into notepad and save as StressTest.avs. edit: Make sure you select all filetypes or it will get saves as stresstest.avs.txt. Check it by playing in media player before proceeding to encode. You should see color bars with time code at the bottom. Then drag the stresstest.avs into the CCE window.

#start of script
ColorBars (480,480)
ShowSMPTE(24)
ResampleAudio(44100)
#end of script

Make these settings in CCE by right clicking in CCE and selecting edit:

Uncheck audio (important or you will crash)
Set to 3 pass vbr
AVG 2250 Min 755 Max 2500
Quality 17 Anti-Noise 2
Video - check Add sequence end code, upper, Progressive and zigzag

Don't worry about the slow speed. My machine that normally does 1.3 to 1.8 gets about .37 on this test. Let us know what speed you get after about 30 minutes of the first vaf pass but let it run all night to completion if possible.

I have asked some folks in the avs forum to do something better. Ideally (I think) we need a fading from color to color in a loop. I don't know how to do that. See my post there if you do.

[DDogg]

Anybody try this?

[No Rice]

I am getting ready to do it right now. I will post first results.

[No Rice]

Well after about 45 mins, it is leveled off @0.219. It says about 4 hours to go. What would you like to see posted at the end of the test?

[DDogg]

Nothing more than seeing if it completes. You might want to monitor your cpu temp if you have software handy to do that. The RT number towards the end would be interesting. That is, I would be interested to see if it holds about the same all the way through.

I am hoping people with CCE troubles will run this test

This removes most variables like mpeg2dec, filters, etc. So, if people with problems can run this through to completion we can then start adding more complexity for them. Hopefully that will pin it down better than we have done.

The test is very crude and I hope to do something better later, but this should make sure the cpu and company are not the culprit (IMHO).

Also somebody with a monster fast machine could do me a favor and run this for an idea of the speed from the fastest side

[TheeeGod 3]

"cannot play back the file. the format is not supported." this is the message i get when i tried right now to play that file in media player.

[No Rice]

If you just hit cancel or something liek that, it will ignore and play it. I had the same error.

[Caliber]

After 20 minutes I am getting speed of .362. I can run the whole thing tonight after work, but is there any way of knowing the speed without watching it the whole time? Or maybe some way of timing how long it takes to finish?

Athlon XP 1800+
Soyo Dragon +
512MB (2x256) Crucial DDR CAS2.5

[chainsaw135]

"cannot play back the file. the format is not supported. this is the message i get when i tried right now to play that file in media player."

When you save the avs file in notepad, make sure after you select save as, you go down and change "save as type" from txt to "all files"



After 30 minutes I am getting speeds of .340

Athlon Thunderbird 1.4@1.5
512 megs (2x256) Samsung DDR2100 Cas 2.5
Windows XP Pro

I'll post back and let ya know if it finished or not.

edit:Also under this heavy load my CPU is at 51c which for my athlon that is normal under heavy load:)

[chainsaw135]

Over 300 views and we only have 4 or 5 people testing and posting back????? I know there is alot more then 4 or 5 people that are having problems with cce. So lets do our part and figure this out!

[TheeeGod 3]

i did as said last night and still get the same error while trying to play it in window media player.

[DDogg]

TheeeGod 3, ok, you mean you get an error when you try to play the script in mplayer 6.4? Or you get red text in Mplayer 6.4 on the screen with an avisynth error? Or, when you try to play the mpv file that has been created?

Do this:

Open notepad and make an avs script with just the word "version" in it. Save as test.avs and play in mplayer6.4. Tell us what happens exactly.

PS - Mplayer 6.4 is available via the "open with" when you right click on your AVS file. Choose "Choose program" and scroll down to where you see windows media player with the old icon.

If you know all that, sorry, but some reading this might not. ;)

[fgm]

Originally posted by chainsaw135
Over 300 views and we only have 4 or 5 people testing and posting back????? I know there is alot more then 4 or 5 people that are having problems with cce. So lets do our part and figure this out!

I hadn't tried it because I haven't had CCE problems and I'm not overclocking. Here it goes:
Speed: 0.233
Dual PIII 1GHz
512 MB PC133
Running 2 Seti clients at the same time ;)

I'll let you know if anything goes wrong

[DDogg]

I hadn't tried it because I haven't had CCE problems and I'm not overclocking

So, there is really no need for you to run it. Thanks for trying it anyway.

[jshumate]

I have a 1 GHz Athlon
Asus A7V motherboard
512 MB memory (2x256 MB, SDRAM, PC133, CL=2 from Crucial)
60 MB ATA/100 hard drive
I am not overclocking. I have had the CCE freeze issue. I ran the stress test for 34
minutes and it got to about .22. I had to use my PC, so I couldn't let it run longer, but
I might try to let it run overnight soon. Should I have it freeze up if I try to let it run
to completion or see anything else weird, I'll follow up on it.

[fgm]

Originally posted by DDogg

So, there is really no need for you to run it. Thanks for trying it anyway.

That's what I imagined. Anyway, it has been running without problems and at present reports:
File Size: (0.7 Mbps)
Speed: 0.212
Elapsed: 08:24:49
Time Left: 05:43:55

I have to kill the process now since I have to make an encode. Cheers.

[chainsaw135]

"I hadn't tried it because I haven't had CCE problems and I'm not overclocking. Here it goes:"
I wanted the users that overclocked or have cce issues to test, I was assuming that out of the 300+ views more then 4 people had either been overclocking or had cce crashes.



Ok I ran the test it went through without a problem, Took a little over 9 hours to run the test.

After 30 minutes I was at .340 then it dropped to .320 after a few hours, it stayed at that speed till the end.

I dont have cce freezes tho unless I set my swap file to small like under 256megs ect. I just did this test since I overclock:) any other testing dogg? That you'd like to see done let me know.

[DDogg]

Yeah, LOL, you can do about a dozen runs with the new beta to test quality between a two disk encode of CCE and TMPG. SOunds fun, yes? :D

[chainsaw135]

LOL thinking "what did i get myself in to":) Ok will get started on that tonight. I have 2 movies I can test it on..do you wanna me to post my results in this thread--->http://forum.doom9.org/showthread.php?s=&postid=87576#post87576

[DDogg]

I'll PM you. OHBoy, did you step in it now :) Better oil that machine up!

[chainsaw135]

.

[HeeD]

After about 10 minutes I get around .623 on this test.

I have never had CCE crash before so I aborted test.

System specs:

Windows XP Pro
Pentium 4 2200 Mhz
Asus P4B266 Mainboard
512MB DDR RAM
100GB Western Digital HD
80GB Maxtor
80GB Maxtor (external 1394 firewire)
Geforce 3 TI200 64MB
Toshiba SD-M1212 DVD-ROM
Plextor CD-R PX-W241040A

[pipo]

Hi,

When I play the test.avs in mediaplayers, I see a test image for about 75 minutes. Why not a shorter test (5 minutes?), if CCE stays below 0.1?

Pipo:confused:

[chainsaw135]

Well i'm sure he made the test for that long to stress your system, push it to the limit I know that it took me a little over 9 hours with a speed of .320 most of the time. I'm used to getting 1.5-1.6 this just showed me how hard this test pushed my system. So if you have problems with cce crashing/ or overclock you might wanna try this test also who ever does this test, see if you can monitor your temps also.

I do see your point tho wanting to make the test shorter because .1 would take forever, so maybe there will be a shorter test for those types of cases.

[xrv1138]

DDog
ill do some of these tests for u on both my machines
i was a perpetual cce crasher
tried everything on both machines
both are setup completely differently
(one amd one intel one o/clock one not etc)
have changed *everything* on both of them from bios to o/s settings
only thing that works for me is to use the bitrate gap of 2230 max ave and 2530 max
only had one failure out of 30 odd since ive stuck with that

but more than happy to do your tests
will have to wait for a couple of days
big queue of dvds :)

and also surprised that more ppl havent run this test for u given that so many were cce crashers as well :confused:
will post back in a couple of days...

[DDogg]

only thing that works for me is to use the bitrate gap of 2230 max ave and 2530 max ...only had one failure out of 30 odd since ive stuck with that

xrv1138, thanks. This test is only to detect those machines that are having hardware/memory glitches. Assuming somebody can run this test through to completion they should then try an encode with the settings you mentioned, or even a 300 gap. Then if they crash they should report back here and we can go to the next few steps.

[mkanar]

I shouldn't be posting; CCE crashes are far and in-between for me, but ohwell....

Dual P3 933, Intel 840 chipset and motherboard
512MB RAMBUS Dual-Channel PC800
WinXP

Presently at 38% of the .mpv and still going. Overall CPU usage is at 51-55%, Memory usage is at 255,444K, 255,460K now, 255,464K now..... 0.243fps

When I tried to load the .avs in Zoom Player 2.40, quartz.dll (a Microsoft library) crashed. I did however successfully load it into media player. I played the first 10 seconds of video, stopped, replayed. I was very fascinated that the first 10 seconds of video were much much faster the 2nd time around, but when I got to 'fresh' video, it slowed back down.....is this disk cache at work? I didn't expect that due to this access methodology.

Anyhow, thanks!
MKanar

[DDogg]

mkanar, something that might be fun for your dual setup is loading the avs in tmpg and running it. If you get around to it, let us know what the proc usage is. I don't know what mode to suggest. I would hazard CQ and 2pass.

[jasond]

anyone that has CCE freezes should kill all extra tasks going on and see if it still freezes overnight

I froze once during CCE last week and I endtasked my motherbaord monitor and it started fine again..

maybe its not CCE,DVD2SVCD,etc causing problems, but background processes using the memory at the wrong time,dlls are in use,etc..

Does CCE use any windows system dlls that may conflict with running processes?

I use wintasks2000 from http://www.liutilities.com

It works great at closing processes that run and you can raise/lower their priority in it.


maybe someone that has frequent CCE crashes/freezes can try this also to see if it makes any difference..
:)

[xrv1138]

grr
have to head away with work for a couple of weeks
will do this test on my machines when i get back :\

[jshumate]

I do have the CCE freezing problem. See an earlier post of mine to see my config. This
weekend I ran the CCE stress test successfully to completion. There's a post from
jasond worth looking at where he suggests a possible solution. That sounds reasonable
to me and I may try that. I also wonder if there might be a problem with the frame
serving aspect of DVD2SVCD and that CCE is OK, but the frame server is the real problem.
CCE stayed around .220 during the VAF pass of the stress test and stayed pretty steady around .200 on the 3 pass multipass step. I don't have a solution to the freezing problem, but maybe it's helpful to someone to know that I did run the stress test to completion.

[DDogg]

jshumate, exactly. So in your case we have now ruled out basic machine problems like overheating and such.

First thing for you to do is set your max 300 higher than the max average and run a test encode. Also, you might like to try setting your paging file min and max to the same size after you have defragged your drive. Make it big like 750 meg or 1 gig. If you have enough ram, you might try turning off virtual memory entirely for a test encode.

[sunfire]

0.450 +- 0-005 from start to 30min... is that god or bad??

My Computer is
P4 1.5
512Mb RAMBUS
60Gb UDMA100
using DVD2SVCD 1.0.7 build2

when i tried to ripp i taked 2h for the audio and 10h for the movie the movie i 1:32min with sub is that OK. or is anytinh wrong

[chainsaw135]

If the test completed thats good, if not then thats bad.:)

[Abend]

Just out of curiosity, I tested TMPG 2.52 and 2.02 on 32 CPU machine with 8GB RAM and Hitachi storage on 6 optical channels
1. 2.52 does not scale to more than 2 proc.
2. 2.02 did fine and achieved 96-100% CPU
3. 2 hour movie finished in around 40 min with 2.02 with 2pass 10bit DC and highest motion search precision
4. 2.52 - nothing to write home about
5. CCE does NOT scale 3-4% CPU
Same tests on 8 CPU with 2GB mem.
1. 2.52 does not scale
2. 2.02 around 3 hours

OS 2000 Datacenter

[chainsaw135]

Thanks for the info.

[Winroute]

My "DVD2SVCD-computer" have had some (alot of) problems, BUT now i have figured it out.....:-)))
Random crashes, lockups and so on specially when running CCE.....

My (built for) DVD2SVCD rig:

Abit BE6-II ver1.1 (flashed to ver 2.0)
Intel Celeron 850@1133 (running at 133 MHz FSB)
AOpen Slotket !!!!
1 x 256 MB Apacer CAS3 SDRAM (runned as CAS2)
1 x 128 MB Apacer CAS" SDRAM (-------"------)
2 x Seagate U6 20 GB drives in RAID 0
ATI XPert98 (with TV out)
Soundblaster 128 "something"
Creative DVD1241E (Revision: 0p44 (patched region free))

My problem dissapeared when i changed the AOpen slotket for a "new" ASUS-S370-DL slotket.
My DVD2SVCD-computer went from a "nervous wreck" to an ULTRA-stabil working-machine !
This means a rather fast, cheap unit doing a speed of 0,9+ in CCE made almost solely on second-hand parts !, and it is fast enough for me !

In other words: this little piece of #$%^ :devil: AOpen-slotket have made my
converting-attempts "miserable" the last couple of months.

"Nothing is stronger than the weakest link" !!!!!

Perhaps this little discovery could help some others with poblems?, so i desided to put this on the forum.

[marck77]

I tried the stresstest and it worked out in 15 hrs and 12 minutes, with a starting max speed of 0.212 and a final speed of 0.197. In this interval I never had more than 44°C of temperature in the CPU which is a standard value for my PC. I think that my system is very stable even if overclocked.

The configuration is Asus CUSL2 and Celeron2 566@850 Mhz, 384 Mb RAM, Windows XP professional, DVD2SVCD 1.0.7 build 1, CCE SP 2.50. This is what I saw while encoding PEARL HARBOR: CCE goes up to 100% of the first pass (at 0.603 on my slow CPU) and then it terminates instead of continuing doing the other 3 passes (I selected VBR multipass with 4 passes). I tried it twice with the same result. All the suggestions users gave here until now are verified. Anybody can help?

Marck

[DVDHack]

@marck77

The testing occuring here is largely aimed at determining if your system is fundamentaly stable with CCE.

The error that you seem to be getting is on pass changes in CCE on multipass encoding. There have been a lot of posts on this issue and you will find a number of suggestions on how to fix the problem in those posts.

Generally the problem is related to the gap between the average and maximum bitrates that you have set (or have been calculated). Try setting the average lower, I have successfully done Pearl Harbour and don't recall a specific problem but I did it on 3CDs which would mean a pretty low average with the standard max. If you are using 4 CDs you might like to manually set the average lower - I use 1900 when I hit this problem (can't be bothered doing multiple runs to find the ideal value).

If it still doesn't work you could do 1 pass VBR with TS and use 4 CDs.

[Winroute]

Now i have done the stresstest.
It used 14h. 24min. to finish.
Startspeed (VAF) 0,302
MPV-speed (after 3,5 hours of encoding) 0,278

It finished whithout errors.

Winroute :)

[DDogg]

A few people have had trouble understanding how to make the avs file so I will put a link here.

[jshumate]

I had a problem where CCE would freeze on my system. My PC passed the stress test and I found no memory problems. I finally decided to follow jasond's suggestion to go to http://www.liutilities.com and I bought
WinTasks2000 (the $27 version). I took a DVD that CCE would freeze on and I successfully did a multipass VBR encode on it twice. I'm inclined to think that my problem was that some program(s) on my system are competing for resources that CCE needs. I ran DVD2SVCD at RealTime priority in WinTaaks and CCE worked perfectly. Those of you who are having problems with CCE freezing might want to give this a try. I don't believe there is a demo version of WinTasks, but it's a good program and I think it's worth the money.

[SiliconSoul]

if this is correct this is the solution to many cce problems. have u used it to help anything else? cd burning and defragging or anything else?

[SiliconSoul]

anyone else test this program out?

im thinking of getting it because it will shutdown all the crappy programs that are running that are not needed.

i have over 49 processes that are runnign and im sure that many are not needed and probably cause problems with cce.

[chainsaw135]

It would make it easyer to use such a program, but the fact is you can shut down the processes manually and it will do the same thing and wont cost nothing.

Also just like jshumate, if any users end up using the program and your cce stops hanging let us know in here if possible, or if you shut down the processes manually and it fixed your cce also let us know.

[gerti67]

@SiliconSoul.

Geeeez! - 49 processes running? What the hell are you doing with your machine? - let me guess - 3 different firewalls - just to be on the safe side - and also 2 different anti virus scanners and so on? :D - Sorry couldn't resist. ;) - But really, that is very much programs running in the background.

I have 21 processes running after booting up my workhorse and there are some virtual print dispatchers and ISDN monitors I usually shut down before doing an encode - so I have only 18 processe running when encoding with CCE - never ever had a single crash with it - knocking on wood. ;)

You should really see what processes you can close - or better do a complete reinstall of your machine and only install the programs you really need. If you want to stay with it the program jshumate mentioned could safe you really some time with that many processes you are running.

Greetings,
Gerti

[SiliconSoul]

:p

ftp client, mirc, pop-up blocker, sb live driver, ftp server, Bestcrypt, 3com nic doctor, via drivers, virtual daemon, logitech mouse driver, clonecd with hidecd and amplyify weak sectors on ( :-) yeah baby! im in canada! ;) ) , powerstrip for videocard, volume, dyndns client , firewall, norton antivirus, trojan scanner, defrag.

:D

plus all the normal sys crap that runs

[chainsaw135]

I just checked my processes I have 23 running at 1 time, i prolly have 20 running during an encode tho as i shut a few things off myself.
Also every month or so I re do my whole system on both computers, they say 2 do that every 6 months but i like a fresh system at all times just seems 2 run better in my eye's.

[SiliconSoul]

trial version of wintasks 4 standard

http://www.lidownloads.com/demos/wintasksstdtrial.zip

pro

http://www.lidownloads.com/demos/wintasksprotrial.zip

both TRIALS! without any C r a cks or what not just to test it.

im not sure if its really any better than good ole windows task manager..

presets look nice but u have to set them up...

[gerti67]

Hi SiliconSoul,

if I were CCE and had been started on a machine like this I just would have crashed immediately! :D - Just kidding.

But really, I think that this many processes running on a machine at one time during a CCE encode is one reason for it to crash.

My aim is to run as few additional programs or drivers as possible on my machine as from my experince every additional program or driver just raises the risk of some incompabilities or even a crash.

Greetings,
Gerti