Movie Edit Pro 2015 64 bit Performance

sirsteve wrote on 9/16/2014, 8:00 PM

I was just one week into evaluating the 2014 Premium edition and the 2015 edition was announced. I was excited to hear about 64bit support. So I timed the rendering of the same movie clips using identical settings and in each case the 2014 version was faster than the 64bit 2015 version. For a 1080p WMV export, 2014 took 58 seconds and 2015 took 62 seconds. For a 1080p MP4 export, the 2014 version took 38 seconds and the 2015 version took 57 seconds (67% longer). Also, regarding the file size, the 2014 MP4 was 24.5mb and the 2015 MP4 was 59mb (more than twice the size). This is far from scientific testing at this point but i was expecting the reverse results. Both products were installed with default options.

Has anyone else noticed any performance hits and/or gains with the MEP 2015 version? Are there default settings that may need to be changed to optimize performance?

I'm running Win 8.1 with an i7-3770 CPU @ 3.4GHz with 32GB memory and SSD drives

Comments

terrypin wrote on 9/17/2014, 2:00 AM

Hi,

If borne out by wider based tests that would be especially disappointing given the potential 64 v 32-bit potential.

But I can't say I'm too surprised. I had the same sort of result to my tests of MEP 2014 rendering speed versus MEP 2013 a year ago, contrary to the advertsising, as reported here:

http://www.magix.info/uk/2014-mp4-export-slower-than-2013.forum.1028732.html

I returned to 2013 for a while but its other strengths (particularly proxy files) soon had me back using MEP 2014. Plus the factor that my free one year's technical support on 2013 ended.

On a few occasions since then I've made similar comparisons (typically exporting 1920x1080 MP4s) with the same sort of edge for the older version.

And although I can't immediately find the thread, I recall a similar conclusion about 2013 versus MX.

I've learned to take Magix hype about both feature and performance improvements with a large pinch of salt!

--
Terry, East Grinstead, UK

Scenestealer wrote on 9/17/2014, 5:47 PM

Hi sirsteve

Are there default settings that may need to be changed to optimize performance?

For your exported file to be twice the size there would have to be some noticeable differences in the export template's Advanced Encoder Settings and the files subsequent bitrate / compression. You would need to compare all advanced settings in both versions prior to export.

Edit: - Also be aware that the default Mpeg4 encoder in 2015 is an Intel. Up until now MEP always used a Mainconcept encoder which you now access after an additional free download.

You can also create variances by opening a project created by an earlier version in a later one so each version should create a new project then add the same material.

Also a 1 minute project is rather a small sample.

johnebaker wrote on 9/18/2014, 2:45 PM

Here are some comparative speed tests I have made between MEP 2014 and 2015 exporting the same HD project composed of AVCHD and mp4 video, still images, crossfade transitions and Speed effect applied to one clip.

The same settings were used for export ie 1920* 1080 at 25fps using QuickSync (Hardware acceleration) and Calculate video effects on GPU in the export dialog turned off

Format	MEP2014	MEP2015
MXV	1m 38s	1m 24s
MP4	2m 03s	1m 33s

As you can see there is a definite increase in speed compared to MEP2014 for MXV and a significant increase in speed for MP4.

In the above tests when I turned on the Calculate video effects on GPU option in the export dialog there was a definite slowing down, the exports taking about 20% longer.

Project loading times show very little difference ~ 6% faster in MEP2015

John

Scenestealer wrote on 9/18/2014, 4:55 PM

Hi John

Was that using the Intel encoder or MC?

How many minutes was the timeline and what were the resulting file sizes?

Thanks

Peter

johnebaker wrote on 9/19/2014, 3:28 PM

Hi Peter

The tests were carried out using the Intel encoder (Quick Sync) in both MEP 2014 MC codec and 2015 Intel codec

The section encoded was an exported range of 2m 30 s from the same project.

Files sizes are:

	MEP 2014	MEP 2015
MXV	998 MB	925 MB
MP4	200MB	375 MB

The difference in the mp4 file sizes is a result of MEP 2015 using a higher bitrate for the video 10529 kbps - 19855 kbps which gives better quality and a lower bitrate for audio 158 kbps - 37 kbps.

The Intel encoder is set for higher quality video and a longer GOP size then the Main Concept encoder.

Audio sample frequencies are different 48000 vs 44100 khz (2014 vs 2015)

Repeating the tests for the MP4 export, setting the Main Concept encoder (2014) to the same bitrates for video and audio, GOP length and profile as the 2015 settings made no difference to the times I previously posted. The file sizes are comparable - the mp4 MC encoded with the new settings jumps to 369 MB

John

Scenestealer wrote on 9/19/2014, 10:40 PM

Hi John

Thanks for the additional tests and info.

Interesting, I was not aware because I do not have a current CPU, that the use of Quicksync in 2014 used a different codec to MC, but it stands to reason. Had you noticed any "improvements" in quality before 2015?

I am a little surprised that changing several parameters in the encoder settings didn't change the encoding times, but I guess one change might have cancelled out another.

Not really sure what is going on with the audio BR and freq.? I assume that is AAC but 37kbps seems very low. What does it sound like?

I assume if you choose a DVD template for output it will still use 48000 hz.

Peter