Whoever told you Ryzen is geared for gaming was dead wrong. It actually isn't as good at gaming as the i7 and i9 chips, but kills the intel chips (not counting i9) at workstation use. Ryzen's quad and 8 core performance is far better (especially 8 cores) than i7. That makes it a very good chip for video production, CG/motion graphics, or any other type of production that utilizes multithreading extensively. I am using a Ryzen 7 1800x and have been blown away by the performance gains over my FX8350, at least 2-4x faster at any given task. Pair that with some fast DDR4, a solid GPU (I prefer NVidia, but not the best for vegas) and a Samsung Evo Pro M.2 hard drive for your system/apps, and you've got a respectable editing system.

I am yet to see proof that the Ryzen is better than a HEDT i7.

I figure the needs for gaming & video/photo processing are similar so Ryzen would help my needs.

I am sure many of the "Gurus" in this forum will disagree with that statement, gaming and video editing are two very different operations.

I am yet to see proof that the Ryzen is better than a HEDT i7.

Single threads, intel is still king (in most cases)... once you get past 4 cores being used at once, that's when Ryzen pulls ahead. It's also worth noting the Ryzen 7 1800x is $400 cheaper than the 6900k.

Cinebench 15 Singlethreaded
Ryzen 7 1800X -161 points
Intel Core i7-6900K - 155 points

Cinebench 15 Multithreaded
Ryzen 7 1800X - 1,628 points
Intel Core i7-6900K - 1,477 points

PCMark 8 Work (non-OpenCL)
Ryzen 7 1800X - 3,620 points
Intel Core i7-6900K - 3,171 points

These are benchmark figures and not many relate to what is going on inside Vegas. The only way to really test it, is comparing render times, like it was done in the past. The SCS Benchmark project is still the best way to tell and it can be scaled up to 4K.

FYI, as far as I know from this forum, many of the Vegas FX are single threaded.

I never said it was better at Vegas, only that in general, as a workstation chip for people who do video production and motion graphics, it's the best option. Ryzen isn't a slouch in the single threads as well, as you can see it beat the i7 at the c4d benchmark even single threaded. Then with the price factored in, I'd say it's the smarter buy as you could invest more money into eliminating bottlenecks elsewhere in the system.

I never said it was better at Vegas, only that in general, as a workstation chip for people who do video production and motion graphics, it's the best option. Ryzen isn't a slouch in the single threads as well, as you can see it beat the i7 at the c4d benchmark even single threaded. Then with the price factored in, I'd say it's the smarter buy as you could invest more money into eliminating bottlenecks elsewhere in the system.

Yes and no. The Ryzen is sure no slouch, no doubt and for the price maybe the better value. BUT, you also have to factor in that you get by far less PCIe lanes, 24max! The 5820 has 28 and the 6900K has 40. With 24 lanes, you are left with 8 lanes once you plug in your graphic card, which eats up 16 lanes. So there isn't much left for things like a RAID controller or additional sound card, capture card and so on.

Anyways, I took the time and converted the SCS Benchmark project to 4K. These are the steps:

1) Save as a new project file, we don't want to overwrite the original file

2) Change project resolution to 3840x2160, 29.97p, Best render quality.

3) Select all media and run the "Resize Generated Media" script; it's a script supplied with Vegas.

4) Render the 3 video files to XAVC-I 3840x2160, 29.97p

5) Use "Replace Media" and replace the 1080 files with the new 4K files.

6) Upscale the two PNG files to 3840x2160 and replace the project PNG files with the new ones

7) Disable Resample on all media.

8) Save the project.

Now you converted the SCS Benchmark project to 4K.

Here are some render times (mm:ss) on my aging 3930K@4.3GHz, 32GB RAM and a Fury X and GTX580:

XAVC-I 2160 29.97p: 3:11

XDCAM 1080 29.97p: 1:03

MC AVC 1080 29.97p (CPU): 2:05

MC AVC 1080 29.97p (CUDA): 1:04

MC AVC 2160p 29.97 (CPU): 6:22

MC AVC 2160p 29.97 (CUDA): 1:37

MC MPEG2 DVDA 720x480i (video only): 1:04

I have set the Fury X as my main acceleration card but I can still use CUDA for MC AVC due to the GTX580 that is also in my system. My system has 40 PCIe lanes and can run both cards in PCIe x16.

I was actually surprised to see that the rather old MC AVC CUDA encoder could handle 4K resolution, sweet 😊.

Now, maybe someone with a RYZEN system can post their render times for the same project too. I am interested to see how the RYZEN compares. I would even say that if it can come close, a second or two off my times that it is a very good CPU for Vegas. Maybe it's even faster, that is also welcome!

PS: All render test where done in SVP13 but I also checked two renders in VP14 and the times are identical.

Edit: My 3930K scores 1083 in Cinebench 15 🙁

I don't need raid controllers, my system drive is faster than any SATA striped raid configuration can get and I clone my external drives often so the redundancy aspect of raid setups isn't needed either. I have my blackmagic decklink 4K card and room to add a nice sound card (which I would just use a USB 3.0 audio interface if I needed that), so 24 lanes is more than enough for me. There are also some motherboards that add an additional 8 lanes.

If you'll post that project file (I can't find it), I'll run some tests.

I don't need raid controllers, my system drive is faster than any SATA striped raid configuration

Well, I doubt that too. M.2 RAIDS are very very fast and when you do 4K multicam editing you will see what I mean. There is only so much storage a single drive can provide and with the upcoming trend to PCIe drives, you will run out of lanes pretty fast on a RYZEN system. Anyways, for the majority of users, 24 lanes will be sufficient. I personally stay away from anything Vegas uses system wide and is connected via USB, eSATA or any "hotplug" connection.

I found my old thread on this topic but the link to the files isn't working anymore. This link may work.

I said SATA striped raid, not M.2. If you want to start raiding M.2 you're talking about a lot of additional cost to your build, and I edit 4K multicam good enough as it is using my SATA video drives. I don't put video on my system M.2 drive. I'll post benchmarks tomorrow. Downloading sloooooooow.

EVGA GeForce GTX 1050 Ti, 4GB GDDR5, DX12 OSD Support

Poor choice for Vegas

Anyways, I took the time and converted the SCS Benchmark project to 4K.

Thanks for that OldSmoke. I will steer people to these instructions in the future to try and standardize a benchmark.

On those benchmarks, are you rendering the regions or the entire timeline?

The entire timeline.

I didn't disable resample on each individual clip, but in project settings instead. Is that good enough? I also had chrome running in the background (this page), so not exactly a clean benchmark, but as you can see by the specs I provided, at no point did any encoder ever max out my system in any way. MC 4K came the closest of them all. CPU averages were usually in the middle of the range I provided.

For all 7 regions rendered at once, my times are:

(Specs: Ryzen 7 1800x - 8 cores, 16 threads, 3.6ghz with 20mb cache, 32GB DDR4 3000mhz, NVidia GTX 970 (unsupported by Vegas, so not used), Samsung Evo 850 M.2 system drive (media was sourced from this drive), 4TB SATA External drive (file was rendered to this drive)

XAVC-I 29.97p: 21:52 - CPU Usage: 18-39% Disk Usage: 1-2% Memory Usage (32GB): 16%

XDCAM 1080p: 2:28 - CPU Usage: 52-94%, Memory Usage: 19% Disk Usage: 2%

MC AVC 1080: 3:14 - CPU Usage: 64-99%, Memory Usage: 20%, Disk usage: 2-10%

My GPU is not recognized by Vegas, so no CUDA :(

MC AVC 4K: 8:17 -CPU Usage: 35-99%. Memory Usage: 21%, Disk Usage: 2%

DVD 480i: 2:27 - CPU Usage: 63-96%, Memory Usage: 20%, Disk Usage: 2-5% with occasional jumps as high as 41%

This is with no GPU acceleration at all, as vegas won't even see my GPU.

No idea why the first one took so much longer than the rest...

Fr0sty,

looking at your CPU only numbers and mine, I wonder how much difference is contributed by the fact that my CPU is running at an almost 20% higher clock speed. Can the Ryzen 1800 be overclocked? I heard in the beginning there where issues with overclocking on that CPU, it overheated fast and wasn't stable beyond 4GHz.

Anyways, a 5820K will outrun it when it comes to Vegas as it can be overclocked without any troubles to 4.3-4.5GHz.

Note that the new RX580 isn't doing well with Vegas and there is no guarantee the new ones will either. I would buy a used R9 290X or 390X with 8GB from eBay.

Fr0sty,

looking at your CPU only numbers and mine, I wonder how much difference is contributed by the fact that my CPU is running at an almost 20% higher clock speed. Can the Ryzen 1800 be overclocked? I heard in the beginning there where issues with overclocking on that CPU, it overheated fast and wasn't stable beyond 4GHz.

Anyways, a 5820K will outrun it when it comes to Vegas as it can be overclocked without any troubles to 4.3-4.5GHz.

Yes, it can be overclocked, but because I'm still using the heatsink from my fx8350, I'm not going to go there with it yet. That said, CPU utilization was pretty low on most of those encodes, so it's hard to say if that would be the issue or not. I think it just has to do with optimization. I just don't think Vegas is tuned to take advantage of 16 threads while rendering. Look at that first XAVC render and you'll see why I think that. There is no way that raw clock speed, thread count, etc can cause that big of a difference in render times. The raw specs aren't that far apart from each other, and other than clock speed and single core strength, work in my CPU's favor if anything.

"Can't win. New cards that were $300 are now $500. Used prices are all up, too. I'll wait it out as I'm not pressed to upgrade now."

Honestly, I'd hold off until 15 to get a new GPU. I just can't justify sinking money into old hardware (especially if it is getting more expensive) just so a NLE with GPU support from the stone age can use them, I'd rather wait until it supports the latest and greatest.