VideoHelp.com Video Benchmarking

[rallynavvie]

There were two video files originally so you had to do a batch encode of the two. Then BJ_M streamlined it to one file to make it easier for folks to participate. Apparently only mods are interested in this endeavor though (and coalman).

[thecoalman]

It was one file. I didn't see two there. Just one file in two parts, after making the single file (one was an exe, the other a rar) I encoded it in 1:09. If there is a 2nd file I missed it. I might be able to shave a couple of seconds by disabling a few things like the itunes stuff, AV, and random crap but its not worth it.

My system is very tight right now though. Recently defraged, SP2 and everything. Very smooth. Now in 6 months... who knows?

I had the first set of files which is why I asked if it was both files, I went back and got the single file for my second batch of tests.

Regardless this still doesn't explain the difference between BJ_M's results and mine. If they were close it would be one thing but they are too far apart. It's the same processor and like mine is a system that has not been overclocked, except his is running at nearly a full .35ghz lower than mine, 300MHz less bus speed, he has 512mb less of ram and probably lesser HD's and that pisses me off (no offense BJ_M).

I don't think the HD's have anything to do with it since encoding to the same or seperate HD produces the same results anyway. Matter of fact my HD speed is great, I always judged by how fast I enter a game when it restarts and I'm always one of the first.

Could the Motherboard have anything to do with this? If I remember correctly I was reading a review about the chipset on the one I have and they weren't exactly overwhelming.... of course this was after I purchased it.

Apparently only mods are interested in this endeavor though (and coalman).

Yes I'm very interested since my system seems to be running sub-par.

mine runs about 38degree idle i seem to recall -- but that is when i boot it and leave it on bios to check -- so i really dont know how realistic that is .. maybe diff is in diff. version of tmpgenc ?

i also never use HT ..it just slows most things down ..

Mine is 30-32 from bios, the other temps I posted came from a utility that came with the MB. I'm using version 2.524 which was just downloaded and installed. All the boxes were checked under CPU except "enable 3-d now" which is greyed out. The only setting I changed was to assign it high priority under task priority. Disabling HT produced a 2 second faster time.

Why is the multiplier and voltage different on yours? And does that make a difference? I'm guessing it shouldn't since your mhz is lower anyway. I'm a complete newb when it comes to processor settings and the like so bear with me.

[Faustus]

Motherboard can have ALOT to do with this. Especially the chipset.

Also what ram are you running?

[thecoalman]

Motherboard can have ALOT to do with this. Especially the chipset.

Also what ram are you running?

Same chipset, I just looked....

Corsair pc3200 xms, it is running in dual channel, motherboard fully supports it..... http://www.corsairmicro.com/corsair/products/specs/twinx512-3200c2.pdf

I have nearly the latest bios too.

[Faustus]

Motherboard can have ALOT to do with this. Especially the chipset.

Also what ram are you running?

Same chipset, I just looked....

Corsair pc3200 xms, it is running in dual channel, motherboard fully supports it..... http://www.corsairmicro.com/corsair/products/specs/twinx512-3200c2.pdf

I have nearly the latest bios too.

How long ago did you install windows?

[thecoalman]

About 9 months, latest updates. I'm not going through that 2 day ordeal..... Besides I had a bitch of a time getting one old but important app installed correctly which I don't want to go through again.

[offline]

I have a P4 2.4 with ASUS board. Batch file
run on same drive as video clip - details
see above computer profile and the CPUZ
thread started by Cobra.

1.22 in standard mode (HT enabled)
1.09 auto mode (HT enabled)
0.59 overclocked (HT disabled) which is how
I run this box 24/7
Only installed XP twice since purchase. The
1st time was thanks to ATI's crappy drivers

HT slowed things down by ave. 2 seconds
(HT was only disabled from Tmpgenc environment, not bios)

[BJ_M]

OK WTF am I doing wrong, I could see if there was a marginal difference but comsidering all my shit is faster than yours. Is that for both videos? Am I missing something here, I'm not too familiar with the numbers and exactly whjat they represent.

It was one file. I didn't see two there. Just one file in two parts, after making the single file (one was an exe, the other a rar) I encoded it in 1:09. If there is a 2nd file I missed it. I might be able to shave a couple of seconds by disabling a few things like the itunes stuff, AV, and random crap but its not worth it.

My system is very tight right now though. Recently defraged, SP2 and everything. Very smooth. Now in 6 months... who knows?

no -- just one file there

[BJ_M]

I have a P4 2.4 with ASUS board. Batch file
run on same drive as video clip - details
see above computer profile and the CPUZ
thread started by Cobra.

1.22 in standard mode (HT enabled)
1.09 auto mode (HT enabled)
0.59 overclocked (HT disabled) which is how
I run this box 24/7
Only installed XP twice since purchase. The
1st time was thanks to ATI's crappy drivers

HT slowed things down by ave. 2 seconds
(HT was only disabled from Tmpgenc environment, not bios)

sounds about the same as me -- if you turn off HT in bios , may shave a few more points off ..


coalman .. i am running a combo of ide and scsi drives in this machine .. 7 drives ... the boot (c) drive is a Cheetah 15K RPM 146GB Ultra320 SCSI ...

the file source drive is a IDE 133 drive (200GB) - as that is more the bottle neck , i dont think having the scsi drive would speed anything up ..

also -- turn off any antivirus .. i noticed that that will add 2-3 seconds just having it on .. though it may because i dont have much memory in this machine .. i should really increase it -- but i have my other monster machine here at home (and several others on the network here i can use for distributed rendered) ..

i bet the motherboard is a big factor .. the XP on here was installed a year ago , sp1 only .... no real major windows tuneup done except running only old style windows clasic look and turned off most services not required ..

[offline]

BJ_M i disabled HT in bios and in tmpgenc and
gained 4 seconds on average??????

[BJ_M]

sounds right .. HT really only works for very few things effectivly .. and I gave up on it .. good for a few things though i understand ... not sure what though .. a dual cpu is much better way to go (if you can afford one) - though even dual cpu systems have issues with some things (slower on many games for example)

[offline]

I had a dual Xeon 1.2 and was not impressed with
the cost versus performance. Back in those days
only tmpgenc seemed to make use of it & many
apps were slower. Better to invest in SCSI drives
IMHO

My next system will most likely be a single AMD
and maybe a DSP board for video encoding.

[rallynavvie]

This is unexpected but I like where this is going. This thread is not only about getting some sample benchmarks but to get other members' computers working up to their full potential. I like that, seems a little more useful than just getting encoding results

HT is pretty much useless on my system. I enabled it and did some testing with it when I first put the processors on the board but haven't had it enabled since. SMP systems generally aren't best for price/performance if you're just looking at numbers like this. However if you need all the advantages that a workstation like that can provide then it's worth every penny. I wouldn't be able to make do with an AMD or P4 solution, even if I had several of them. Batch encoding/rendering is the best thing ever since I can just set up the queue and let it go while I'm at my day job or sleeping. I won't see twice the performance of a single 3.0 P4 but the 50-60% over one is still a big deal. Besides, I still have a dual 1900 MP rig I can farm out smaller encoding jobs to, in fact that one still sees a lot of MPEG encoding use now that the big rig is doing rendering and assembly. GbE is a godsend for video work

Anywhoo, I'm gonna free that one up today and get my results posted up. I'm guessing I should be seeing about 45 seconds to do the encoding, maybe 40 if I play with it.

[rallynavvie]

42 seconds the first try, 41 after closing a few more processes. Read from 10k SATA boot drive and written to 7200rpm SATA storage drive. HT is disabled, of course.

System specs from the txt output of CPU-Z:

Number of CPUs 2

CPU #1
APIC ID 0
Name Intel Xeon
Code name Prestonia
Specification Intel(R) Xeon(TM) CPU 3.06GHz
Family/Model/Stepping F29
Extended Family/Model 0/0
Brand ID 11
Package mPGA-604
Core Stepping D1
Technology 0.13µ
Instructions Sets MMX, SSE, SSE2
Clock Speed 3066.8 MHz
Clock multiplier x23.0
Front Side Bus Frequency 133.3 MHz
Bus Speed 533.4 MHz
L1 Data Cache 8 KBytes, 4-way set associative, 64 Bytes line size
L1 Trace Cache 12 Kµops, 8-way set associative
L2 Cache 512 KBytes, 8-way set associative, 64 Bytes line size
L2 Speed 3066.8 MHz (Full)
L2 Location On Chip
L2 Data Prefetch Logic yes
L2 Bus Width 256 bits

CPU #2
APIC ID 6
Name Intel Xeon
Code name Prestonia
Specification Intel(R) Xeon(TM) CPU 3.06GHz
Family/Model/Stepping F29
Extended Family/Model 0/0
Brand ID 11
Package mPGA-604
Core Stepping D1
Technology 0.13µ
Instructions Sets MMX, SSE, SSE2
Clock Speed 3066.8 MHz
Clock multiplier x23.0
Front Side Bus Frequency 133.3 MHz
Bus Speed 533.4 MHz
L1 Data Cache 8 KBytes, 4-way set associative, 64 Bytes line size
L1 Trace Cache 12 Kµops, 8-way set associative
L2 Cache 512 KBytes, 8-way set associative, 64 Bytes line size
L2 Speed 3066.8 MHz (Full)
L2 Location On Chip
L2 Data Prefetch Logic yes
L2 Bus Width 256 bits


Northbridge Intel i875P rev. A2
Southbridge Intel ID25A1 rev. 2
AGP Revision 3.0
AGP Transfert Rate 8x
AGP SBA supported, enabled
AGP Aperture 128 MBytes
Memory Type DDR
Memory Size 2048 MBytes
Memory Frequency 133.3 MHz (1:1)
CAS# 2.0
RAS# to CAS# 2
RAS# Precharge 2
Cycle Time (tRAS) 6
Performance Mode enabled

[offline]

No people It too 4 seconds longer to
encode with HT disabled in the bios!

[BJ_M]

gain ussually means better

:P

[vhelp]

Here's my results ...



-vhelp 2991

[offline]

gain ussually means better

:P

The Q I have is why?

[thecoalman]

Here's my results ...



-vhelp 2991

I don't think you used the file settings... they can be found in the first post by Rally.

[Cobra]

vhelp,

You need to use the exact settings rally gave you. You also need to go into CPU-Z to the last tab and give us a machine report so we know what your exact specifications are.

[Heywould3]

need to upgrade i guess.. but not bad for an OCed old as dirt computer BTW i have a few things running in BG might get a bit faster but not much.. good thing i dont use this for much encoding
1:37 OC
1:47 stock @1400mhz (1600+)


CPU-Z Report
(Over CLocked)
--------------------------------------------------------------------------------

CPU-Z version 1.26.


CPU(s)
Number of CPUs 1

Name AMD Athlon MP
Code Name Palomino
Specification AMD Athlon(tm) XP 1900+
Family / Model / Stepping 6 6 2
Extended Family / Model 7 6
Package Socket A
Core Stepping A5
Technology 0.18 µ
Supported Instructions Sets MMX, Extended MMX, 3DNow!, Extended 3DNow!, SSE
CPU Clock Speed 1575.3 MHz
Clock multiplier x 10.5
Front Side Bus Frequency 150.0 MHz
Bus Speed 300.1 MHz
L1 Data Cache 64 KBytes, 2-way set associative, 64 Bytes line size
L1 Instruction Cache 64 KBytes, 2-way set associative, 64 Bytes line size
L2 Cache 256 KBytes, 16-way set associative, 64 Bytes line size
L2 Speed 1575.3 MHz (Full)
L2 Location On Chip
L2 Data Prefetch Logic yes
L2 Bus Width 64 bits



Mainboard and chipset
Motherboard manufacturer MSI
Motherboard model MS-6380, 2.0
BIOS vendor American Megatrends Inc.
BIOS revision 07.00T
BIOS release date 01/31/2002
Chipset VIA KT266/333 rev. 0
Southbridge VIA VT8233 rev. 0
Sensor chip Winbond W83627HF

Graphic Interface AGP
AGP Status enabled, rev. 2.0
AGP Data Transfert Rate 4x
AGP Max Rate 4x
AGP Side Band Addressing supported, enabled
AGP Aperture Size 128 MBytes



Memory
DRAM Type DDR-SDRAM
DRAM Size 512 MBytes
DRAM Frequency 150.0 MHz
FSBRAM 1:1
DRAM Interleave none
CAS# Latency 2.0 clocks
RAS# to CAS# 2 clocks
RAS# Precharge 2 clocks
Cycle Time (TRAS) 5 clocks
# of memory modules 1
Module 0 Apacer Technology DDR-SDRAM PC2100 - 512 MBytes



Software
Windows version Microsoft Windows XP Professional Service Pack 2 (Build 2600)
DirectX version 9.0c

CPU-Z Report
(Stock)

--------------------------------------------------------------------------------

CPU-Z version 1.26.


CPU(s)
Number of CPUs 1

Name AMD Athlon MP
Code Name Palomino
Specification AMD Athlon(tm) XP 1600+
Family / Model / Stepping 6 6 2
Extended Family / Model 7 6
Package Socket A
Core Stepping A5
Technology 0.18 µ
Supported Instructions Sets MMX, Extended MMX, 3DNow!, Extended 3DNow!, SSE
CPU Clock Speed 1399.3 MHz
Clock multiplier x 10.5
Front Side Bus Frequency 133.3 MHz
Bus Speed 266.5 MHz
L1 Data Cache 64 KBytes, 2-way set associative, 64 Bytes line size
L1 Instruction Cache 64 KBytes, 2-way set associative, 64 Bytes line size
L2 Cache 256 KBytes, 16-way set associative, 64 Bytes line size
L2 Speed 1399.3 MHz (Full)
L2 Location On Chip
L2 Data Prefetch Logic yes
L2 Bus Width 64 bits



Mainboard and chipset
Motherboard manufacturer MSI
Motherboard model MS-6380, 2.0
BIOS vendor American Megatrends Inc.
BIOS revision 07.00T
BIOS release date 01/31/2002
Chipset VIA KT266/333 rev. 0
Southbridge VIA VT8233 rev. 0
Sensor chip Winbond W83627HF

Graphic Interface AGP
AGP Status enabled, rev. 2.0
AGP Data Transfert Rate 4x
AGP Max Rate 4x
AGP Side Band Addressing supported, enabled
AGP Aperture Size 128 MBytes



Memory
DRAM Type DDR-SDRAM
DRAM Size 512 MBytes
DRAM Frequency 133.3 MHz
FSBRAM 1:1
DRAM Interleave 4-way
CAS# Latency 2.5 clocks
RAS# to CAS# 3 clocks
RAS# Precharge 3 clocks
Cycle Time (TRAS) 6 clocks
# of memory modules 1
Module 0 Apacer Technology DDR-SDRAM PC2100 - 512 MBytes



Software
Windows version Microsoft Windows XP Professional Service Pack 2 (Build 2600)
DirectX version 9.0c

[vhelp]

hmm...

Whenever I File\Open Project\.. I get an Illegal stream format. error
when I try to open this file:

* mpeg1.mcf

I even renamed it to *.tpr (cause i'm assuming that this is actually a project
type file) anyways, there were no provisions for opening *.mcf file types
anywheres inside TMPGenc.

.. You know.., I wasn't sure why I needed a settings file for this test.
.. Basically, you just have to encode a 214 frame divX file to MPEG-1 (I did
.. mpeg-2) but it doesn't matter. And, weather you choose VBR or CBR doesn't
.. matter either. And that's what I did.., unless there is something that is
.. CPU related that needs to be checked or something, I don't know.
.. But I encoded the source as such:
.
.. * CBR 7000 bitrate
.. * 720 x 480
.. * mpeg-1 and mpeg-2 (w/ same results as my pic above)

Anyways.


..and..


The pic above is from one of my lastest (last) sample clips from last years
posting:

* VHELP's Samples..

I found (through a few tests) that different types of AVI formats encode
faster. It seemed that the encode times were slower for a divX source file
vs. my DV avi test file, which was approx 5 sec's faster (given the same specs
above @ 9000 bitrate, 5 seconds slower, than if I had encode at 7000 bitrate
(was quicker) @ 29 sec)

I leave you w/ my last pic results.. (I seem to not be able to perform the test
consistantly and w/ the same setup, ) so.. I guess I'm all washed out here.

Cheers to all
-vhelp 3000

[rallynavvie]

OK I'm keeping records of overclocked and stock results seperate so for those that are overclocking if you can note your stock specs like processor speed and bus. I'm also keeping a tab of overclockers' achievements as well just for fun.

offline I need your memory timings and chipset.

I'm gonna dust off my Athlon MP machine and give it a go as well

[rallynavvie]

VBR/CBR makes a huge difference as it prescans the content once so I've generally seen at least a 50% increase in encoding times when VBR is set. Wierd that you can't import the MCF file when everyone else was able to. Which version of TMPGEnc were you running?

[Cobra]

I'll clock my machine back and retest it for you when I get a chance, rally.

How about editing your original post to make the results clearly visible? As in, add it to your first post as people produce results - maybe rank them in order of speed (as long as I'm in the top five).

[rallynavvie]

Waiting to get a few more responses and then posting graphs and the like, something easily visible.

[thecoalman]

hmm...

Whenever I File\Open Project\.. I get an Illegal stream format. error
when I try to open this file:

* mpeg1.mcf

Open with the load button down in the bottom right hand corner...

[BJ_M]

hmm...

Whenever I File\Open Project\.. I get an Illegal stream format. error
when I try to open this file:

* mpeg1.mcf

Open with the load button down in the bottom right hand corner...

i was going to say the same thing

[Delerious]

Number of encode frame = 214
Length of GOP = Average 35.67 field(s), MAX 40 field(s)
Average bitrate of video = 6995.19 kbps
VBV buffer size = 304 KB

Finished encoding video

-------------------------------------------------------------

48 seconds - HT on processes not disabled






J

[Heywould3]

vhelp.. you are prolly not saving the template correctly.. i ahd a problem, the first try.. just open the file select all then open a new text doc .. paste and save as XXXX.mcf you prolly have the html header in the file.. you can also just edit that out if the one you have.. BTW.. you want to goto the lower right and hit "Load" to load the mcf file.. not open project.

P.S. you have the same computer specs as me pretty much so youll get close to same times as me.. (looked under profile)