@terrypin

Hi Terry

. . . .I'm curious about your version of Windows OS, which your sig says has the format '19041.450 (2004)' . . . .

The format the version as written in @CubeAce sig is not the standard format used by Microsoft - Windows 2004 (OS Build 19041.450) which is the same version as I am on except mine is the Home variant. Windows variants are like for example Music Maker - the base Windows is the Home edition, Pro and Enterprise add additional features and capabilities for the environments they are designed for.

@pmikep

Hi
. . . . I presume that my bottleneck here is my CPU and if I had more cores, then the bottleneck would move to the nVidia . . . .

That depends on which tests you are relying on to come to the above conclusion see below.

. . . . I will try the "Let windows decide" setting. It's the only thing that I haven't tried. . . . .

Have you carried out comparison tests using your own projects for the different combinations of settings?

If not, I would suggest you do to give you a bigger picture of your PC's capabilities with the type of projects you create.

I am not sure of the relevance of the handbrake test with respect to MEP. AFAICS all it shows is that Handbrake uses NVENC and is very fast in encoding as it should be, I get the same when rendering with Blender - it is fast using the NVidia.

MEP does not use the NVidia for encoding. On my PC it is now being used for decoding after the reset I had to do, encoding is still done between the iGPU and CPU depending on what effects etc are being used.

I have yet to revisit the Neat Video plugin to see if it uses, and there is any improvement with, the NVidia.

John EB

@johnebaker

I have not tried my other tests yet. (Usually 1080p material with minimal editing.) I will plan to do those after I try "Let Windows Decide" with Ray's benchmark. (Or after I remove MEP from Windows' Graphics Settings.)

The relevance of the handbrake test was simply to show what might be possible if a program made use of all the hardware available. Say, if Magix were to do similar.

@pmikep@pmikep

Hi Mike.

I don't think an editor can do the same as a program designed to do just one job if it's doing more.

Even a simple thing like putting a few clips together with no fades and a straight conversion requires re-rendering at the joins.

I haven't tried a straight conversion with MEP yet. I'll give that a go tomorrow. and take some screen shots.

For now I'm turning in.

Ray.

True. But programs like VirtualDub(2) and my competing Video Editor use a smart rendering protocol, where they pass frames through to the output where there are no changes. I routinely trim TV movies using VD2. It takes about 5 seconds to process a 1.5 hour movie. (Although you did say with conversion. So nothing would pass straight through.)

@pmikep

Hi Mike.

True, but MEP on research hasn't been the fastest editor out there for for some time. It doesn't even currently make sixth place. Some of the other editors have other problems though if you look through the user forums so it's swings and roundabouts. On average I have chosen programs on how easy I find them to use along with functionality. I chose MEP over other software through a mixture of factors. If it hadn't been MEP it probably would have been PowerDirector but they have gone over to an annual or monthly plan. Filmora is the same even with the one time payoff. After that they get progressively more expensive or need add on packs for more options. Davinci resolve is not that stable on my machine. I haven't tried the cut down Adobe offering or Corel or others so can't comment.

They all seem to have their positive and negative points though so I suppose it just comes down to individual needs.

Personally I've got used to how MEP and now VPX works and I'm happy with them and the workflow patterns.

MEP so far also hasn't fallen over as a project has got more complex.

I also like the exported files quality wise. Working at 1080p, exports for me are now close to real time playback speeds so I'm OK with that.

In fact I've been doing just that. Taking the two movie clips within the Test project and placing them side by side and exporting the result.

The resulting video is 4K UHD and 2 minutes 13 second and 19 frames.combine file size is 1558.79 MBs.

Export time for MP4 at either Best setting or Balanced is roughly the same at 5: 53s. File size is 1.27GB

Exporting to MP4 at 1080p best quality takes 2 : 4s. File size is 423MBs

Exporting at 4K HEVC Best setting takes 7: 19s. File size is 242MBs

So there are a few more comparisons for those that are interested along with a bit of info on how MEP changed its allocation of resources on my machine for each.

But there still could be other variables involved. You didn't say what your motherboard ram speed or type was.

Does your motherboard still use a north and south bridge? Do we have different versions of the audio chip?

I don't know if any of that makes a difference or not but few two systems are ever even close to each other. It's difficult to say what influences results.

Ray.

@CubeAce

Yeah, I did. A few posts back. Also in my Sig. 2400 MHz, CL 15. My mobo is a Dell. So no telling exactly what they've done. It does have a N and S bridge, although I don't have control over their freqs.

I agree with you that the reason I tried MEP again was because it's more ergonomic than Power Directory. (BTW, one can still by the perpetual version of PD. They just bury it on their website. I bought PD18 on a smokin' sale on Amazon.) But PD sure is fast. Although it has a bug where it won't use my GTX when an Intel iGPU is present.

But when it comes to editing, even simple things like crossfades are a pain with PD. Once you figure out how to crossfade video, the audio doesn't crossfade! If Magix could get near the speed of PD, it would be an amazing product.

@pmikep

Hi Mike.

I could have sworn I'd looked and didn't see it. So not a lot of difference there but my Motherboard doesn't have a North & South bridge. The new replacement for that is supposed to be a lot faster. Again I'm not technical enough to know how but my son says it's something to do with less bottlenecks with data packets.

The lifetime version of Power Director does not come with any upgrades. That is still a separate subscription and I like their version of a library but I have enough monthly bills as it is.

To me, workflow is as important if not more important than export speed. Time lost in editing makes up for the export time where I can be doing something else. I personally don't have that many deadlines to meet for video. On the other hand, photos are often needed before I've taken them. 😁.

My son who was using filmora found audio mixing and sound sculpting a challenge in it and his export times are much slower than mine and we use the same cameras, but again he is using my old motherboard and ram with a newer i7. I did like the quality of his outputted files though. Since his move to Adobe they have improved but the cost! and he has crashes more than I do.

I have never owned a Dell pc but from what I hear they are not bad but seem to have their own way of configuring Windows. I don't know if that is true or not.

I have always run the Pro or Enterprise versions of Windows. Just habit mostly after our first and only prebuilt computer (MESH) came with whatever was the highest version of Windows back then. It was a nightmare with configurations that we eventually had to change. Good spec though.Still have the Super VGA 17" Sony monitor sitting at the back of the garage.

Still. Off topic. 😇

Ray.

I didn't realize that there was another new update to the nVidia Studio Driver. (Less than a month apart.)

I updated my driver and am now down to 18:37 on Ray's benchmark.

Within spitting distance of my previous best with MEP 2020.

@pmikep @johnebaker @browj2

Glad to hear that Mike.

This version of the program seems very picky on its running requirements.

Which nvidia driver are you now using? I'm on 26.21.14.4292.

Ray.

@CubeAce

Hi Ray

. . . . Which nvidia driver are you now using? I'm on 26.21.14.4292. . . . .

There is a newer version available - I have the following installed:

• NVidia Studio Driver - 452.06 (27.21.14.5206 in Device Manager) 18 Aug (latest)
   
• Intel UHD 630 - 27.20.100.8587 - 5 Aug (latest not including the new beta version)

John EB

@johnebaker

Hi John.

I hadn't noticed but have now updated the driver for the nvidia.

Ran my 'project' tests over again.

There has been no appreciable gain in the performance of MEP beyond clips seem to be importing slightly faster but that could be my imagination. Export times for various file formats remain the same as do the settings needed.

So I'm not sure what the 'improvements' are between those drivers but makes little difference on my system with MEP.

Ray.

Just thought that I would confirm changing the location of the Videocontent during installation.

Since I didn't have MEP2020, there was no folder for Movie Edit Pro Plus under: C:\Users\Public\Documents\MAGIX\Common\Videocontent

I first installed MEP Plus 2021 (to C) with the above as the default folder for VideoContent, uninstalled everything 2021 including the content, deleted the folder that was created under C:\Users\Public\Documents\MAGIX\Common\Videocontent, and reinstalled. This time, I created similar folders on another drive, which is actually where my documents go to by default. I installed the program again to C but indicated the new location during installation for extra additional content (7.86 GB) and everything installed to the new folder structure under:

B:\Documents\MAGIX\Videocontent\Movie Edit Pro Plus

So far, I have had no problems with the additional content in MEP2021. Everything shows up. Store purchases also go to the ShopManager folder under the above folder structure. Of course, they will likely have to be downloaded again.

For anyone with limited C drive capacity, I recommend using another drive for the extra content.

Note that if you have already installed a previous version with the VideoContent set to the default location on C, then when you install the new version, the same folder for additional content will show in the default settings and you will not be able to change it. The program detects that the folder and its content exists and migrates the previous content with any new content. You would first have to uninstall all extra content, copy somewhere any effects and other modifications that you have may have added, and then delete the folder entirely. Then, when you install, you will be able to modify the extra content location.

Has anyone else done this but had a bad experience?

John CB

@browj2

Hi John.

Does that mean your content resides on the same drive as your source files?

Ray.

@CubeAce

Hi Ray,

I have several drives. My photos and early videos are on my drive B, as are my project MVP files. Later videos are on 2 other drives, as is anything special that I put in a project. I have project folders for this. I don't copy photos and videos to a project folder; I leave them in their original locations. I export to an external SSD drive.

So, I can say yes to your question. I don't see that this makes any difference; content is content. The program itself is on C, everything else is on other drives.

John CB

@browj2

Hi John.

I'm just interested and wonder how having files in use within MEP (or VPX) influences how the program works.

I know I that so far I have got my smoothest playback and fastest render times on my system with the program running on the C: drive. My project folder and source files on a second drive and exporting to a third drive. I have found that rendering a file back to the drive containing the project folder, even if to a different folder location on the same drive slows down the export considerably. So in effect I have found so far that the use of three drives to be optimum for me.

I was just wondering if you had four drives in action during exporting. I can see the relevant transfer of data during exporting or playback but yet to understand the impact as the drive read / write times during such action gets nowhere near their maximum transfer rates regardless on how it's set up and yet the differences remain.

As it's impossible for me to see how that works internally I'm wondering if putting additional resources like the programs' content on yet another drive would be beneficial or not, or if indeed placed along with the source files.

It could of course make no difference at all where they reside short of making more storage space to the C: drive.

As I say, just curious. Not something I've tried personally and you did ask if anyone had had a bad experience so I too am interested in that answer.

Ray.

@CubeAce

Hi Ray

. . . . I can see the relevant transfer of data during exporting or playback but yet to understand the impact as the drive read / write times during such action gets nowhere near their maximum transfer rates regardless on how it's set up and yet the differences remain . . .

Here are the results of a quick test - 4K UHD export, 4 mins project, project and source files on W: exported to SSD, third HDD and the same drive (W:)

SSD export time 3m 47s, write speed 0 - 11 MB/s, active time 7%, response time 19 - 42ms

3rd HD export time 3m 48s, write speed 0 - 11 MB/s, active time 5%, response time 0 - 30ms

same drive (W:) export time 3m 52s, write speed 0 - 11 MB/s, active time 6%, response time 5 - 25ms

From these results a quick conclusion would be writing to a third drive would be better than to the same drive, however if you express the difference as percentage it comes to 2.5% which is not a significant difference.

John EB

@johnebaker

Hi John.

Was this a straight transfer or did it include effects etc?

My tests have been with a working project of mixed graphic, cross-fade types, and video clips.

Unfortunately at present I don't have a working SSD to compare to but you seem to be getting just over twice the write speed I'm getting writing to a WD Black with the aforesaid project with exports topping out at 4.27MB/s but around halves that write time when exporting to the same drive.

Ray.

@CubeAce

Hi Ray

Cuts only, no effects, for maximum transfer speeds on writing - this is where in theory the SSD should out perform the hard drives.

I would expect effects etc would slow down the write speeds due to the encoding being slower, however I would not expect much difference between drive types as the write phases are in bursts when a chunk of data is ready.

The 2 hard drives have 256MB caches, AFAICS your 2 WD Blacks have a 64 MB cache this may have a small effect on the write times.

John EB

@johnebaker

So many variations 😩🙁.

I use VirtualDub2 for a quick end trim of TV movies. Using Stream Export from one SSD to another, I see a 1.5 hour video process in a few seconds. I presume that is all hard drive "latency," since the CPU isn't doing much of anything on a stream export.

@CubeAce @johnebaker

An SSD is not a fixed transfer rate. It can be using SATA or PCIe technology.

SATA is 3 or 6 Gbps wheras PCIe can transfer up to 31,5 Gbps. Usually the SSD otself also have a max read/write speed (at least for PCIe) that May be slower than the interface.

It would be useful to actually specify interface and SSD specs when reporting on these findings.

True. But I didn't think my specs really mattered at this point. That is, even if a Samsung EVO was faster (and it is), I wouldn't buy one just to save one second off a few second transfer.

As for my particular specs, they're in my signature.