Vegas Movie Studio 14 and Pinnacle 20 GPU Render Tests
I just purchased a gtx1080 (actually a whole new Ryzen system) and thought that I would try out the new
Card vs my old AMD R9 290 to see if any improvements had been made for Rendering acceleration using GPU..
So anyway if you're interested check out this first movie which will just compare the two video cards.
I will give comparative render times throughout the video...spoiler alert...I end up being dissapointed.
I plan on doing a new movie to test the Ryzen chip next week.
Thanks for looking..
On the Dynamic Ram Topic - from the help file.
"Using dynamic RAM previews
In other words, you can playback a few seconds of video at full speed, if you have a very complex composite, CG overlays, or motion graphics. It has nothing to do with rendering video to file, or timeline acceleration.
By maxing your ram to dynamic preview, you cut down on Windows available ram which it reduced your disk cache. That meant your system was IOing more to compensate, which lead to slower render times. It is best to leave this at the default setting, unless you are using the feature as it is intended to be used. Some believe that a 0 setting gives better stability, but I believe that is due to other system instabilities.
SHIFT+M - builds a temp file that can be used to view the difficult section at full frame rate.
As far as the rendering tests go:
test the trial of Pro vs MS. I believe MS is gimped hence the lower price tag.
Source codecs matter and you say that your source is coming off your phone, and not an actual video device. There could be issues with that workflow that is slowing things down.
Not sure about MS, but all versions of Vegas have supported AVI, so the statement about MS not supporting AVI is odd.
Straight timeline encoding to MP4 will not be helped with GPU, since there is very little math there that OpenCL will help with. Another way to look at it is, the GPU is completing the math it is tasked with so fast, you are still waiting on the CPU to complete. Frame by frame, slow math kid vs fast math kid then combining the results. The fast math kid will be sitting waiting for the next problem, which looks like 0 GPU usage on an average meter system like most monitors are.
If you want to see how much GPU helps rendering to MP4, try 2 video layers, cookie cutter the top, and defocus the bottom track. Now try the same test in 32bit Floating Point mode. You should see a large improvement over CPU only. GPU acceleration is application specific. The uneducated marketing people have screwed most consumers here due to the complexity of the subject.
Rendering 60P means 2x the amount of frames to render, which means a longer render. The question have to ask yourself, is your material really worth or require 60P video?
GPU MP4 rendering was abandoned for a reason by most NLE and encoders. Instead favoring things like ASIC, or processor extensions like Quicksync. NLEs have not picked this up much, because of the dev time involved with coding for specific hardware. "Laying Up" a software only NLE is much easier, cheaper, and will run on all hardware. Just maybe not optimized. I have always found most low end NLE software to be a bit of a con job, you get what you pay for.
Thank you for your very detailed response.
I appreciate the time and have read everything twice.
I had previously assumed avi was not supported because my camera's (trail cams) making those files have never worked through various versions of movie studio. I have been informed both here and by others that only some AVI files are not supported.
I too have always wondered if Movie studio Render times vs Pro were clipped to push the professional version. Maybe that could be my next round of testing. ;)
You might consider changing your destination/render drive to anything other than your "C:\" drive. Then, re-evaluate the rendering times and post your results. I'm betting you'll get a faster performance.