Longtime PD user looking for preferred upgrade options maybe multiple SSD options?

Shakeyleg

Newbie Joined: Sep 17, 2019 20:56 Messages: 23 Offline

Solved by

May 26, 2021 17:02

I am increasingly upgrading my video cameras but with the current GPU supply shortage I am trying to make due with my current PC. I have the following setup:

AMD Ryzen 7 2700X Eight-Core Processor 3700 MHZ
32 Gigs of Ram
B450 Tomahawk (MS-7C02) board
Multiple SSDs including a 1gb Sabrent M.2 Rocket and a couple of Samsung 1TB SSD 860 drives
Radeon RX 570 GPU

I film concerts and can have up to 10 4K video cameras running but usually cull that down to about 6 during the editing process. So I am getting lag during the editing process and lately as I use more effects and layer in lots of cameras my 45 minute videos can take all night to produce. It has gotten to the point that almost every night I just produce whatever I have finished becuase it takes so long.

I do usually try to finish my projects one at a time so my workflow is to load all the camera and audio footage to the M.2 Rocket and just work from that. (My PD software is not housed on that drive) WHen I look at the task manager the CPU is hit 100% a lot and the GPU is usually around 20%. My disk drive never shows more than 2 or 3 percent so it seems the disk isn't the bottleneck. But I was wondering if I split the videos up on a couple different SSD drives would that matter? I have one UHD camera (3840X2160 4K) and I have noticed if I disable that on the timeline it seems to help. I have shadow files turned on and I always use almost the lowest preview resolution as possible. This does sometimes make it difficult for me to determine the best footage I have available as it makes everything look like VHS footage from the 80's during preview. I have updated the drivers as well.

My current project is 429GB of footage but as I upgrade the UHD cameras that number will climb.

Any suggestions would be helpful. I could try to upgrade pieces if you think that would help.

Thanks! Maybe Optodata has a magic wand he could wave at me? lol

I am a huge PD fan and don't want to entertain any other software at this time.

JL_JL

Senior Contributor Location: Arizona, USA Joined: Oct 01, 2006 20:01 Messages: 6091 Offline

May 26, 2021 17:30

Quote But I was wondering if I split the videos up on a couple different SSD drives would that matter?

Assuming 10 simultaneous source streams of like 60Mb/sec it will make no editing or producing improvements if source video uses a compressed video codec like H.264.

Jeff

Shakeyleg

Newbie Joined: Sep 17, 2019 20:56 Messages: 23 Offline

May 26, 2021 18:02

Quote

Assuming 10 simultaneous source streams of like 60Mb/sec it will make no editing or producing improvements if source video uses a compressed video codec like H.264.

Jeff

Thanks for the input! I did run another test and if I break it up into two parts and produce them separately I can combine them later into one and the GPU runs at almost 100% and the CPU runs at about 20% and it produces much faster.

JL_JL

Senior Contributor Location: Arizona, USA Joined: Oct 01, 2006 20:01 Messages: 6091 Offline

May 26, 2021 18:05

Quote Thanks for the input! I did run another test and if I break it up into two parts and produce them separately I can combine them later into one and the GPU runs at almost 100% and the CPU runs at about 20% and it produces much faster.

Well of course, even better in that case is to use SVRT if it works for your produced format, combine in a fraction of what reencoding will do.

Jeff

optodata

Senior Contributor Location: California, USA Joined: Sep 16, 2011 16:04 Messages: 8630 Offline

May 26, 2021 18:11

You're working your system hard with all those 4K clips, and unfortunately I don't have a magic wand to make everything run 10x faster

Like Jeff said, you're not really limited by drive transfer speed unless you'd converted all your clips to an intermediate codec like MagicYUV. While that would let you edit the full resolution clips with ease, your storage requirements would go up by at least 10x and you would have to worry about having too many streams at once due to the very high bitrate.

The only hardware change I'd suggest looking into is a motherboard and more powerful CPU. Your Ryzen 7 2700X has a good Passmark score of 17.6k, but something like the Ryzen 9 5950X has double the number of cores and is 2.6x faster. That's also the cheapest high-end CPU by far and that would make the biggest difference in editing (and possibly producing, depending on the specific kinds of edits/corrections you're making).

The only other changes you can make are to workflow.

Like you just saw by breaking up a video into two sections, you may find it quicker and easier overall to work with one track at a time, even before you worry about which sections will actually be in your final video.

If you have lots of processing just to get the clips visually in tune, make those changes to the entire clip/set of clips from a single camera in a separate project then produce them all at once (maybe overnight) using Batch Produce.

While that will take a significant amout of time upfront, you'll end up with clean new clips ready to put into your master project and you won't have to deal with anywhere near the sluggishness brought on by PD having to render all the timeline edits in real time while you preview and edit.

Depending on your project, you might be able to skip using shadow files entirely and still preview them at Full HD resolution. That's my minimum requirement and I'll switch to non-real time mode if I need to see everything in one section clearly.

YouTube/optodata

DS365 | Win11 Pro | Ryzen 9 3950X | RTX 4070 Ti | 32GB RAM | 10TB SSDs | 5K+4K HDR monitors

Canon Vixia GX10 (4K 60p) | HF G30 (HD 60p) | Yi Action+ 4K | 360Fly 4K 360°

Shakeyleg

Newbie Joined: Sep 17, 2019 20:56 Messages: 23 Offline

May 27, 2021 12:56

Quote You're working your system hard with all those 4K clips, and unfortunately I don't have a magic wand to make everything run 10x faster

Like Jeff said, you're not really limited by drive transfer speed unless you'd converted all your clips to an intermediate codec like MagicYUV. While that would let you edit the full resolution clips with ease, your storage requirements would go up by at least 10x and you would have to worry about having too many streams at once due to the very high bitrate.

The only hardware change I'd suggest looking into is a motherboard and more powerful CPU. Your Ryzen 7 2700X has a good Passmark score of 17.6k, but something like the Ryzen 9 5950X has double the number of cores and is 2.6x faster. That's also the cheapest high-end CPU by far and that would make the biggest difference in editing (and possibly producing, depending on the specific kinds of edits/corrections you're making).

The only other changes you can make are to workflow.

Like you just saw by breaking up a video into two sections, you may find it quicker and easier overall to work with one track at a time, even before you worry about which sections will actually be in your final video.

If you have lots of processing just to get the clips visually in tune, make those changes to the entire clip/set of clips from a single camera in a separate project then produce them all at once (maybe overnight) using Batch Produce.

While that will take a significant amout of time upfront, you'll end up with clean new clips ready to put into your master project and you won't have to deal with anywhere near the sluggishness brought on by PD having to render all the timeline edits in real time while you preview and edit.

Depending on your project, you might be able to skip using shadow files entirely and still preview them at Full HD resolution. That's my minimum requirement and I'll switch to non-real time mode if I need to see everything in one section clearly.

So very very helpful as always! I have never tried to batch produce and it looks like SharperTurtle has a few tutorials on that so I will research. So in your opinion if I am working on a full concert and I batch produce the clips at the beginning of the editing process I will not see a significant reduction in video quality if I then use the newly produced files to edit with? I should have posted here years ago before I learned bad workflow habits. Always refreshing to get real solutions here sir. Thank you

optodata

Senior Contributor Location: California, USA Joined: Sep 16, 2011 16:04 Messages: 8630 Offline

May 27, 2021 14:01

Quote So in your opinion if I am working on a full concert and I batch produce the clips at the beginning of the editing process I will not see a significant reduction in video quality if I then use the newly produced files to edit with?

That's correct, as long as you're producing to the proper format and have a high enough bitrate.

An easy way to do that is to place a 4K clip on the timeline then go to the Produce page and click on Profile Analyzer. You should be presented with a Best Matched Format and possibly also an Intelligent SVRT profile. If you produce your edited 4K to that profile, PD should be able to produce your final project using SVRT which does not require re-rendering for most sections.

With a mix of 4K and HD clips in the project, SVRT may not be available or it may actually be slower than hardware producing if large sections need to be rerendered. That's because SVRT will use the CPU for all required rendering which is usually slower than using the GPU. You may want to experiment with a mix of clips to find the best producing option.

In any case, the I think you should plan ahead and find the optimized output profiles now. Then you can (pre)produce the individual clips knowing that your final project will be easier to edit and faster to produce at full quality 👍

Shakeyleg

Newbie Joined: Sep 17, 2019 20:56 Messages: 23 Offline

May 27, 2021 18:49

Quote

That's correct, as long as you're producing to the proper format and have a high enough bitrate.

An easy way to do that is to place a 4K clip on the timeline then go to the Produce page and click on Profile Analyzer. You should be presented with a Best Matched Format and possibly also an Intelligent SVRT profile. If you produce your edited 4K to that profile, PD should be able to produce your final project using SVRT which does not require re-rendering for most sections.

With a mix of 4K and HD clips in the project, SVRT may not be available or it may actually be slower than hardware producing if large sections need to be rerendered. That's because SVRT will use the CPU for all required rendering which is usually slower than using the GPU. You may want to experiment with a mix of clips to find the best producing option.

In any case, the I think you should plan ahead and find the optimized output profiles now. Then you can (pre)produce the individual clips knowing that your final project will be easier to edit and faster to produce at full quality 👍

Incredible. Thank you sir. I have another concert to film in 3 weeks so I will get this figured out and be much better prepared.

This message was edited 1 time. Last update was at May 27. 2021 18:53