@johnny-s When I'm home later I'll share that clip I've been showing, but I've tried other media & as I wrote I converted that clip & render was still slow.

I thank & appreciate all who've helped, I was just trying to get an overall consensus as to whether there was a problem generally or if it was just me. edoardo-l & I have super slow, Yelandkeil is slower but not so much but yours is faster.. I wonder why that is.. Whereas the Intel crowd are getting super fast speeds..

Google -

Does the 7950x3d have integrated graphics?

This processor features the Radeon Graphics integrated graphics solution.

There's no integrated graphics on mine, edoardo-l nor Yelandkeil 's 🤷‍♂️

@Gid

That's a good way to go. Hopefully if several users can test your test clip (using VP22 and VP23) then the issue can be narrowed down. Maybe its an Amd only issue.

Do also include a screenshot perhaps of your render template for consistency.

I only used Nvidia for encoding and decoding with both of my laptop and PC tests.

My 6900xt is not in a system at the moment but I do have a 5700xt in an old xeon/128gb system and just tried it with vp22 and 23. Rendered a single song trimmed out of a big project shot with three 4k 10bit hevc cameras plus 1 additional camera in 4k avc 8bit. Project is lightly graded with cgp with a png watermark applied. I see a problem with Amd black-bar generated media... the black color comes up white and the fadeout turns into a burnout. Other than that, the render I did with vp23 with the Vce HD 12/24 default template was a few seconds faster than vp22:

vp22 5700xt vce hd default: 2:07.80
vp23 5700xt vce hd default: 2:05.57

I did notice that the default 4k templates in vp23 are not the same as they used to be before vp22. So if you use those to compare, make sure bit-rates are the same.

Same test with an Nvidia 5080 in an 11900k machine was about 1 minute quicker; vp22: 1:10.59 and vp23: 0:52.98

Thought I'd try the test on an 9900k/32gb ram machine with a Radeon VII and it didn't end well for vp23: instant crash. Vp 22 completed it in 2:29.52.

@johnny-s Your results with an Nvidia gpu and an Amd cpu suggests it's more of an issue with Amd gpus rather than Ryzens.

@Howard-Vigorita But I don't have an AMD GPU ..,?

@Gid here are 2 render tests. The files are from Panasonic, 10bit 422 MOV UHD 4K 25P. The render settings were AMD VCE and the same bitrate in both. VP22 utilized more of the GPU and produced a faster render

First is VP22

VP23

@Reyfox Thanks , so ... slower..

VP22 - 2.41mins - 24.96fps

VP23 - 3.12mins - 21.05fps.

@Gid no problem! I haven't tried using 8bit 420 4K yet. But most of the footage I shoot is 10bit 422 4K.

I will re-run the test and report back. So much on my "plate" coming back from vacation.

@Gid a second run with VP23.

@Reyfox Thanks, so like I say slower than VP22.

I noticed this in Preferences, I don't know what these mean but they have worked on Vegas's engine, have they prioritised Intel, have they done something to improve one at the expense of the other, did they know about this before release, is this AMD vs Intel or am I barking up the wrong tree , , so many questions ...

@Gid while I think the ability to customize to a high degree, I would like to know what those settings do or don't do.

I'm pretty sure OpenCl doesn't exist anymore with the move to DirectX and the direct GPU path is now the default in the new video engine. If anything the devs have prioritized NVIDIA GPUs so I'm surprised the CPU makes any difference (perhaps it's something else with the overall AMD architecture that VEGAS isn't optimized for).

Thanks, so that's 4 people with AMD CPU's without IGPU's that get slower render times.

It looks like I'll be sticking with VP22, & seeing as I'm on the Trial I'll prob not know if this has been acknowledged or addressed unless there's a public notice in future update postings.

@RogerS As there does seem to be a link would you change the title of this post to -

VP22 vs VP23, AMD vs Intel rendering speed, VP23 ½ slower than VP22...?

Thanks.

Be helpful to hear from more Ryzen users with Nvidia gpus to see if they all have the issue. Based on my test with a Xeon, which also lacks an igpu, I doubt that lack of igpu is a significant factor. In fact, I routinely disable them in bios or at the driver level in my Intel desktops because they're not that great with vp21 b300 and later, vp22 and vp23.

@Howard-Vigorita 👍 AMD IGPU is the only link I can see but yep it'd be nice to hear more from all users AMD & Intel, Nvidia & AMD, an influx of Intel users saying everything's great won't be that helpful tho, so agreed we need more AMD users 👍

For me the render speed is important, my vids are roughly 20mins long, at the end of a long night or two of editing I can render that project in VP22 in about 10-15mins, I can then watch it, re-edit if needed & render again before uploading to YT... then bed.

Over the last few yrs with VP21 (300) - VP22 I've appreciated the big reduction in render times but double the time to render in VP23 = 40mins, that's out of the question & a massive step backwards for me.

I know there's Benchmarks but this is a fairly simple test, if users can just render a short 2-3min unedited clip with no fxs in VP22, then try it in VP23 with the same render settings & post whether they see if it's slower or faster it would be helpful.

Here is my render of 3m14s 10 4K XAVCS clips:

V23: H264 NVENC 3m54s (RTX4080 Super)

V23: H264 CPU 5m30s (Intel i913900K)

V22: H264 NVENC 1m16s (RTX4080 Super)

V22: H264 CPU 4m17s (Intel i913900K)

Same template Magix H264

Gid yes V23 slower render than V22

In V23 (trial), I have made release video memory True (false is default) and disabled Extended Color Format Conversion.

@andyrpsmith Thanks, that's interesting because you have an Intel but are getting similarly ½ speed slower render times as I do 🤔😒👍

----------

Maybe not related but I wonder if the Trial affects render times as it does have an overlay (a simple overlay shouldn't slow renders this much), I'm on the Trial too , I'd have to go back through this post & @edoardo-l's post to see if he's on the paid version. @Reyfox before I asked you to do this test I read in another post you'd 'downloaded the files'.. are you on the paid version or the Trial?

Here's a simple X-AVC S test between VP 23 and 22. 8-bit UHD to MagixAVC (NVENC) UHD.
VP 23.278 with color format conversion on: 17:15
VP 22.250 26:03

i5-13600K & RTX 5070