Comments

ahshing wrote on 8/31/2017, 2:07 AM

Did you try to use the GPU that is not selected for rendering as the display acceleration GPU? e.g. set the Intel GPU for display acceleration and render with NVENC.


If I select Intel Graphic from Preferences>Video, then select NVENC in render, it still uses QSV to render in final, seems the priority of selecting "GPU acceleration for processing" will override the setting in Render menu

Peter_P wrote on 8/31/2017, 2:28 AM

Strange, do you have both options (NVENC & QSV ) in the render templates?

ahshing wrote on 8/31/2017, 2:34 AM

Strange, do you have both options (NVENC & QSV ) in the render templates?


Yes, the result of NVENC is slower than QSV, I'm sure that the Nvidia GPU works when using NVENC, but not sure is it normal to use "copy back" instead of "3D" engine.

Wolfgang S. wrote on 8/31/2017, 2:44 AM

Strange, do you have both options (NVENC & QSV ) in the render templates?


Why not? If it is a Workstation with both Quick Sync and if there is an nvidia Card in the System both should be available.

Desktop: PC AMD 3960X, 24x3,8 Mhz * RTX 3080 Ti (12 GB)* Blackmagic Extreme 4K 12G * QNAP Max8 10 Gb Lan * Resolve Studio 18 * Edius X* Blackmagic Pocket 6K/6K Pro, EVA1, FS7

Laptop: ProArt Studiobook 16 OLED * internal HDR preview * i9 12900H with i-GPU Iris XE * 32 GB Ram) * Geforce RTX 3070 TI 8GB * internal HDR preview on the laptop monitor * Blackmagic Ultrastudio 4K mini

HDR monitor: ProArt Monitor PA32 UCG-K 1600 nits, Atomos Sumo

Others: Edius NX (Canopus NX)-card in an old XP-System. Edius 4.6 and other systems

Peter_P wrote on 8/31/2017, 3:15 AM

Strange, do you have both options (NVENC & QSV ) in the render templates?


Yes, the result of NVENC is slower than QSV,

That's why I prefer to use the Intel GPU 😀

There are also other indications in that direction ....

May be it is different with other NVIDIA cards.

NickHope wrote on 8/31/2017, 3:48 AM

Strange, do you have both options (NVENC & QSV ) in the render templates?


Why not? If it is a Workstation with both Quick Sync and if there is an nvidia Card in the System both should be available.

The strange thing is that QSV is getting used for render when NVENC is selected as the encode mode.

bitman wrote on 8/31/2017, 1:12 PM

My experience is the reversed, the Magix nvec render is a bit faster than the magix intel render. I use the video HW acc setting (in the preference tab) to match it with the render setting. the Nvidea render is probably faster since my videocard gtx titan x (maxwell) is more recent than my intel i7 4790k

It could also be because my test project has a lot of newblue plugins and newblue text standalone, and I always got the impression/gut feeling that newblue plugins are (and were) very well supported by nvidea HW.

I get the feeling that people are focusing a lot on the Vegas performance with regard to AMD vs Nvidea discussion, but correct me if I am wrong, but the plugins in your project may also have their own HW preferences to render out in their engine (as some can even be used as stand alone).

Last changed by bitman on 8/31/2017, 1:22 PM, changed a total of 1 times.

APPS: VIDEO: VP 365 suite (VP 22 build 194) VP 21 build 315, VP 365 20, VP 19 post (latest build -651), (uninstalled VP 12,13,14,15,16 Suite,17, VP18 post), Vegasaur, a lot of NEWBLUE plugins, Mercalli 6.0, Respeedr, Vasco Da Gamma 17 HDpro XXL, Boris Continuum 2025, Davinci Resolve Studio 18, SOUND: RX 10 advanced Audio Editor, Sound Forge Pro 18, Spectral Layers Pro 10, Audacity, FOTO: Zoner studio X, DXO photolab (8), Luminar, Topaz...

  • OS: Windows 11 Pro 64, version 24H2 (since October 2024)
  • CPU: i9-13900K (upgraded my former CPU i9-12900K),
  • Air Cooler: Noctua NH-D15 G2 HBC (September 2024 upgrade from Noctua NH-D15s)
  • RAM: DDR5 Corsair 64GB (5600-40 Vengeance)
  • Graphics card: ASUS GeForce RTX 3090 TUF OC GAMING (24GB) 
  • Monitor: LG 38 inch ultra-wide (21x9) - Resolution: 3840x1600
  • C-drive: Corsair MP600 PRO XT NVMe SSD 4TB (PCIe Gen. 4)
  • Video drives: Samsung NVMe SSD 2TB (980 pro and 970 EVO plus) each 2TB
  • Mass Data storage & Backup: WD gold 6TB + WD Yellow 4TB
  • MOBO: Gigabyte Z690 AORUS MASTER
  • PSU: Corsair HX1500i, Case: Fractal Design Define 7 (PCGH edition)
  • Misc.: Logitech G915, Evoluent Vertical Mouse, shuttlePROv2

 

 

bipwave wrote on 8/31/2017, 2:22 PM

I test a clip video with GTX970 43s and with GTX1080 39s, two test with MAGIX NVEC render.

The same render in Vegas 14 2m 12s. It is also true, the visualization is faster in Vegas 14 than in Vegas 15.
This should improve.

 

Peter_P wrote on 9/1/2017, 12:56 AM

My experience is the reversed, the Magix nvec render is a bit faster than the magix intel render. I use the video HW acc setting (in the preference tab) to match it with the render setting. the Nvidea render is probably faster since my videocard gtx titan x (maxwell) is more recent than my intel i7 4790k

That's not surprising with a 4790k CPU. The newer, that also provide support for HEVC encoding are much stronger.

ricky-k wrote on 9/5/2017, 3:26 AM

Can anybody tell me why my gtx 1070 needs 36 min to render 4:30 min video and audio with some fx. I am exporting AVC 720p mp4 with nvenc?

Peter_P wrote on 9/5/2017, 11:33 AM

Did you disable so4compoundplug.dll ?

ricky-k wrote on 9/6/2017, 4:05 AM

I tried, same results. Is it because of effects and timestrech?

Peter_P wrote on 9/6/2017, 4:14 AM

May be, but a 7,5 * realtime render on AVC 720p is really very high. What CPU are you using?

Marco. wrote on 9/6/2017, 4:25 AM

It's high but "… with some fx" can quickly let render times explode.

ricky-k wrote on 9/6/2017, 1:09 PM

May be, but a 7,5 * realtime render on AVC 720p is really very high. What CPU are you using?

It's high but "… with some fx" can quickly let render times explode.

It's Intel i3 2100, when I try to render 6:30 min video without effects or timestrech it renders for 3 min

monoparadox wrote on 9/6/2017, 5:30 PM

Just shot some h.265 1080p @60fp with my Phantom 4 Pro. Vastly improved playback and improved rendering. Flying on a i7 4790 24/gig ram with R9 200. If they enhance rendering on the Radeon anytime soon I'll be shooting everything in h.265 and rendering to h.264. It looks great. sharp as a tack, smooth as silk with good coloration.

-- tom

dtran02 wrote on 9/8/2017, 8:27 AM

I test a clip video with GTX970 43s and with GTX1080 39s, two test with MAGIX NVEC render.

The same render in Vegas 14 2m 12s. It is also true, the visualization is faster in Vegas 14 than in Vegas 15.
This should improve.

 

 

What driver version are you using with GTX 970? My computer uses the CPU 100% most of the time.

Former user wrote on 9/8/2017, 10:17 AM

Just another piece of info on render speeds in VP15. I've been doing similar material recently, SD 720 x 576 to .mp4, from mpeg2.

It was typically taking about 90/95% of clip time length to render.

I'm finding thats come down to around 60/66% render time, nice improvement. Thats using the new Magix MC encoder without QSV or Nvenc render enabled. The gtx 1080 is enabled for HW acceleration as previously. Same settings and data rate set in render template as was with older MC encoder.

So maybe either the HW Acceleration using OpenCL is improved or the new encoder is improved for speed, probably the former.

OldSmoke wrote on 9/16/2017, 12:04 PM

I bought at GTX 1080Ti to test with VP15 and the NVENC and I must say I am not impressed. The red car project in it's original form, not upscaled to 4K, rendered with NVENC in 27sec using the MAGIX AVC 1080 NVENC template. That is the exact time as my old GTX 580 and the old MC AVC encoder with CUDA. Considering that the GTX 1080 Ti has a much higher spec and NVENC, I am rather disappointed. I don't mean that MAGIX has done a poor job but rather that NVIDIA has. But, maybe that is all just related to my system. Needless to say, the GTX 1080Ti is going back. Let's see how the implementation with AMD's VCE goes if it ever comes.

Proud owner of Sony Vegas Pro 7, 8, 9, 10, 11, 12 & 13 and now Magix VP15&16.

System Spec.:
Motherboard: ASUS X299 Prime-A

Ram: G.Skill 4x8GB DDR4 2666 XMP

CPU: i7-9800x @ 4.6GHz (custom water cooling system)
GPU: 1x AMD Vega Pro Frontier Edition (water cooled)
Hard drives: System Samsung 970Pro NVME, AV-Projects 1TB (4x Intel P7600 512GB VROC), 4x 2.5" Hotswap bays, 1x 3.5" Hotswap Bay, 1x LG BluRay Burner

PSU: Corsair 1200W
Monitor: 2x Dell Ultrasharp U2713HM (2560x1440)

fr0sty wrote on 9/16/2017, 3:47 PM

Nvidia claims the 10 series has 2x the NVENC speed if I remember right. I wonder if it is on Vegas' end? If not, that is disappointing. I do remember another user here stating something similar about not having any gains with the 10 series cards in VP15.

Last changed by fr0sty on 9/16/2017, 3:47 PM, changed a total of 1 times.

Systems:

Desktop

AMD Ryzen 7 1800x 8 core 16 thread at stock speed

64GB 3000mhz DDR4

Geforce RTX 3090

Windows 10

Laptop:

ASUS Zenbook Pro Duo 32GB (9980HK CPU, RTX 2060 GPU, dual 4K touch screens, main one OLED HDR)

Former user wrote on 9/16/2017, 5:19 PM

Oldsmoke, some results from the Red Car test without mods.

Red car test ...

i7-4790K + GTX 580 + 16gb ram.

MC ... 1:40s

i7-4790K + GTX 1080 + 16gb ram.

MC ... 2:15s 

i7-4790K + GTX 1080 + 16gb ram.

Magix MC Nvenc ... 0:27s

So, systems vary I guess but I saw a substantial improvement over my existing system with the GTX 1080 and also the previously used GTX 580.

 

OldSmoke wrote on 9/16/2017, 5:21 PM

Oldsmoke, some results from the Red Car test without mods.

Red car test ...

i7-4790K + GTX 580 + 16gb ram.

MC ... 1:40s

i7-4790K + GTX 1080 + 16gb ram.

MC ... 2:15s 

i7-4790K + GTX 1080 + 16gb ram.

Magix MC Nvenc ... 0:27s

So, systems vary I guess but I saw a substantial improvement over my existing system with the GTX 1080 and also the previously used GTX 580.

 

Did you use CUDA with the MC AVC encoder and the GTX580?

Your 27sec for NVENC are spot on, I am surprised the GTX580 isn't.

Last changed by OldSmoke on 9/16/2017, 5:22 PM, changed a total of 1 times.

Proud owner of Sony Vegas Pro 7, 8, 9, 10, 11, 12 & 13 and now Magix VP15&16.

System Spec.:
Motherboard: ASUS X299 Prime-A

Ram: G.Skill 4x8GB DDR4 2666 XMP

CPU: i7-9800x @ 4.6GHz (custom water cooling system)
GPU: 1x AMD Vega Pro Frontier Edition (water cooled)
Hard drives: System Samsung 970Pro NVME, AV-Projects 1TB (4x Intel P7600 512GB VROC), 4x 2.5" Hotswap bays, 1x 3.5" Hotswap Bay, 1x LG BluRay Burner

PSU: Corsair 1200W
Monitor: 2x Dell Ultrasharp U2713HM (2560x1440)

Former user wrote on 9/16/2017, 5:31 PM

At this remove, it was a while ago, I'm pretty sure I did ...

I've extracting the above subset of results from these following notes I kept around the time I did the testing ...

Red Car test ...

 

i7-4790K + GTX 580 + 16gb ram.

 

XDCam 0:33

MC 1:40

 

Laptop ...

XDCam 3:07

MC 2:57

 

Redid Red Car test on PC, previously forgot to include GPU on (off) values as other posters have.

 

GPU on (off in brackets)

i7-4790K + GTX 1080 + 16gb ram.

 

The fps was at max 29.97, preview setting was at best-full.

XDCAM .. 0:31s (2:58s)

MC         .. 2:15s (3:11s) [1:31 28-March-2017, cpu OC'D + mem boost.]

OldSmoke wrote on 9/16/2017, 5:57 PM

I get 1:37 with MC AVC and CPU only. I think you didn't enable CUDA, for that template the default is CPU only.

Last changed by OldSmoke on 9/16/2017, 5:58 PM, changed a total of 1 times.

Proud owner of Sony Vegas Pro 7, 8, 9, 10, 11, 12 & 13 and now Magix VP15&16.

System Spec.:
Motherboard: ASUS X299 Prime-A

Ram: G.Skill 4x8GB DDR4 2666 XMP

CPU: i7-9800x @ 4.6GHz (custom water cooling system)
GPU: 1x AMD Vega Pro Frontier Edition (water cooled)
Hard drives: System Samsung 970Pro NVME, AV-Projects 1TB (4x Intel P7600 512GB VROC), 4x 2.5" Hotswap bays, 1x 3.5" Hotswap Bay, 1x LG BluRay Burner

PSU: Corsair 1200W
Monitor: 2x Dell Ultrasharp U2713HM (2560x1440)