How would I go about creating a benchmark project for HitFilm 4 Express?


I'm not a programmer, and I don't understand programming. However, I really want a HitFilm 4 Express benchmark, so that we can easily  determine what hardware is bottlenecking for specific tasks and so on.

If anyone has any ideas, please share them here. I was thinking about a mouse recording software that would run through a script but no idea how to make it reliable so that nobody can manipulate the results...


  • Hey, I'm actually creating a Benchmark scene for HF4P, may take another week or so, gotta heal up first, due to a silly accident.

  • Are you going to use any scripts to verify results are reliable/accurate? Perhaps I'm asking for too much, I don't know how much goes into making something like this.

  • No, no scripts or code... just a few high poly 3D models with a few 4k textures.

    The only Benchmark scenes that I've tried out, are tested by rendering a single frame. You just take note of your computer specs and how long it took to render the single frame.

    There was a HitFilm Benchmark scene posted somewhere in the forums, a long time ago, not sure how it was supposed to work though, as in, do you render a single frame, or do you render an entire animation?

  • Triem23Triem23 Moderator

    @Spydurhank the previous benchmark scene was the old Harry Potter tutorial (so some grading, some lightning and some particles/forces/deflectors) . One was supposed to render the entire scene. 

    The Harry Potter scene and anything with models works for Pro, but not Express (no models or particles in Express). 

    The problem, Kevin, is a benchmark scene won't help find bottlenecks as such. They can just help compare relative overall system speeds. There are so-so many potential hardware configurations out there and so many things that can bottleneck a system that a benchmark won't narrow down the options. 

    Ok, so lets use my laptop as an example. I'm running an Intel i7-6700HQ @2.4 GHZ, nVidia 980m (8GB), 64GHZ of 2300MHZ DDR4 RAM with two M2 Sandisk 256GB SSDs for system, software and cache with a 1TB 7200RPM HDD as a work/library  drive. Right now I know a bottleneck is the HDD, but I also know my SSDs could have been faster if I had put in a single 512GB drive and let it have both M2 slots. What I DON'T know offhand is my mobo, and that could be another source of slowdown. My GPU is also overclocked. 

    Now, if we compare benchmark times what, exactly, accounts for time differences--mobo, CPU, RAM, GPU and/or drives could all be the difference. Even with an "identical" hardware configuration times might not match--perhaps the other system got a slightly better GPU off the line that overclocked a bit faster. Maybe my HDD was just defragged and the other isn't. 

    We haven't even gotten to software yet. My laptop is Windows 7. An identical hardware configuration might have a different speed on Win 8 ir 10 as both of those require more overhead. I don't have persistent Antivirus running. Someone with my identical hardware and OS migbt be running Avast or Norton which is sitting there using processor cycles and slowing things down. Maybe someone has Nero running or Adobe, without turning off Adobe download notifications. That's two more background processes eating up a small amount of CPU. 

    It gets complicated... 

  • I just put together my first PC build, it's actually a Workstation built for video editing, VFX and pretty much anything having to do with 3D software. 

    So, even though it has, and is set to use all of my 2 x 8 core processors with multi-threading enabled, 64gigs Ram and 3 x 4gig GPUs that can handle insanely heavy 3D scenes with multiple programs and services running in the background, I stop or disable any processes, services that I don't need. The only time that I have "Avast" active is if I go online.

    Other than that, I have a total of 24 processes running at any given time, not including HitFilm, Blender, Gimp, X-Normal and Foobar.

    HitFilm 4 Pro is the only program that I run set for automatic updates, and I manually check for updates on everything else. I turned off Windows update and indexing, as well as a bunch of "other stuff to long to list at the moment" and switched the theme to Basic so there are no fancy animations. This stuff helps.

    Also, my CPUs and GPUs are not overclocked, but optimizing your PC to run a bit faster is very easy and will give you a nice performance boost. Having your OS and software on an SSD helps a lot.

    Sure, you need to be careful when turning windows processes and services off but it's not as difficult as you would think. So optimize the crap out of your system and or upgrade if you have the budget. 

    My build would normally have been right around $7,500.00 which I did have the budget for, but my Fiancee did some looking around and found all the parts that I wanted for right about $2,100.00 which was pretty freakin' sweet. :)

    So yes there are a bunch of factors to consider but it doesn't need to be difficult because really... it's not. 

    edited August 2016

    The way I looked at it is that we could have seperate benchmarks, each would stress a different part of the system. Playback = I/O/CPU, and so on. Or is that still inaccurate? I don't think it would be!

  • That sounds really good to me, I think someone mentioned something similar on another thread, a separate BenchMark for 3D models, Particles, you know, stuff that will stress the GPU and another for CPU, not sure what, does what concerning the GPU and CPU in HF4P but I'm exited about this and can't wait to complete the BM to see everyone's system specs and render times.

  • edited August 2016

    Ha! Just found the thread where @Triem23 is the one that mentions the type of Benchmarks that we need.

    It's funny because it's my Thread, first post on the last page, you may need to scroll up? Scroll down and you'll see screens of my workstation... it's a Monster... a Monster I say! :)

  • Daaaaamn, what a beast workstation. My PC looks like a smartphone in comparison to yours, I'm jealous! :D

Sign in to comment