[codec] dBpoweramp Bench Mark Tester

[Spoon]

The purpose of this bench mark suite, is to test a computers performance (speed), using up to 16 CPU cores, encoding audio is a good test of a computers speed (with audio encoding it is possible to 100% 8 cores of a system). Results are presented as encoding speed (combined from all CPU cores) and time taken to encode. Encoding speed is x(times) real time, for example an audio file which is 60 seconds long takes 6 seconds to encode, the encoding speed would be x10. If 2 cores of a CPU were used, it is likely that the encoding speed would be x20. A higher encoding speed = faster computer, a lower time to encode = faster computer.

Installation:

Install dBpoweramp Music Converter,
Install this test suite

Download:

http://www.dbpoweramp.com/beta/dBpoweramp-Codec-BenchMarkTest.exe

1.4MB

[bhoar]

Re: dBpoweramp Bench Mark Tester

Spoon - have you thought about creating something similar for benchmarking burst rip speed?

-brendan

[Spoon]

Re: dBpoweramp Bench Mark Tester

Many drives limit the speed depending on disc quality, so I dont think it would be conclusive.

[bhoar]

Re: dBpoweramp Bench Mark Tester

Many drives limit the speed depending on disc quality, so I dont think it would be conclusive.

True, but for benchmarking, you'd use known good discs on known good drives - mostly, I'd like something like this to look for any problems in the IO chain (interfaces, bridges, windows IO subsystem, dma handling, interrupt handling, etc.) and/or any differences in IO chains.

-brendan

[Stuart.Truman]

Re: dBpoweramp Bench Mark Tester

Are you collecting results?

Using the defaults on a twin X5460 Xeon sytem with 4GB Ram and XP 64 I got the following:

238x Combined Realtime Encoding Speed
Time to Encode: 33.5 seconds

I'm fairly happy with that. Roll on a pair of W5580's :D

[Spoon]

Re: dBpoweramp Bench Mark Tester

Might as well post results.

[Snowknight26]

Re: dBpoweramp Bench Mark Tester

Guess you run this from the command line or do you do it from inside of dBpoweramp?

Also, from the command line, is there any way to specify the location of the benchmark file? Reason I ask is that the HDDs on my machine aren't what I usually use to convert files - why use one slow HDD when its connected to a really fast RAID over Gbit LAN?

[Spoon]

Re: dBpoweramp Bench Mark Tester

All explained in the BenchMark help file.

[Snowknight26]

Re: dBpoweramp Bench Mark Tester

Makes sense. However, it doesn't give an answer to my second question.

[Spoon]

Re: dBpoweramp Bench Mark Tester

Internally it uses very little HDD as it generates the sound its self.

[neslekkim]

Re: dBpoweramp Bench Mark Tester

If collecting result should be of any interrest, I guess we should tell which codec teste also
On my 2xXeon E5440 (8 cores total, 16GB RAM, Vista 64), I get the following:

Encoding To: mp3 (Lame)
183x Combined Realtime Encoding Speed
Time to Encode: 43,6 seconds

Encoding To: flac
404x Combined Realtime Encoding Speed
Time to Encode: 19,8 seconds

Kinda ok results..

[LtData]

Re: dBpoweramp Bench Mark Tester

My results:
Intel Q9650 (quad-core 3.0GHz), 8GB RAM, Vista 64

mp3 (LAME)
109x Combined Realtime Encoding Speed
Time to Encode: 1 minute 13 seconds

FLAC:
242x Combined Realtime Encoding Speed
Time to Encode: 33 seconds

[ChrisNYC75]

Re: dBpoweramp Bench Mark Tester

LAME:
209x Combined Realtime Encoding Speed
Time to Encode: 38.2 seconds

FLAC:
449x Combined Realtime Encoding Speed
Time to Encode: 17.8 seconds

(i7 940 @ 3.91Ghz)


Interesting program

[Radish]

Re: dBpoweramp Bench Mark Tester

And what about encoding options? I can't find any information about options for encoding used by this benchmark.

[Spoon]

Re: dBpoweramp Bench Mark Tester

The standard settings are used for the respective encoders.