It's not silly, but the 8K support they called out is specifically about using an AMD VCN2.0 GPU chip to import 8K files.

That doesn't mean PD can't work with 6/8K files without the chip, there just isn't any hardware support. There are many people who've been using PD (including older versions) to work with and produce videos above 4K, and this recent post is links to a 5.7K example.

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°

It would be nice if Cyberlink could create the possibility in a future update in PD to be able to manually enter the output size in pixels in addition to the given resolution so that you don't have to edit the INI file!

@optodata "AMD VCN2.0 GPU" I haven't found anything on google now - do you have a tip which AMD graphics card would be suitable here? Liebe Grüße / Regards
Oliver

Quote @optodata "AMD VCN2.0 GPU" I haven't found anything on google now - do you have a tip which AMD graphics card would be suitable here?

Sorry, I don't have a clue beyond what CL posted in their promotional materials and the English PDM FAQ

Quote It's not silly, but the 8K support they called out is specifically about using an AMD VCN2.0 GPU chip to import 8K files.

That doesn't mean PD can't work with 6/8K files without the chip, there just isn't any hardware support. There are many people who've been using PD (including older versions) to work with and produce videos above 4K, and this recent post is links to a 5.7K example.

Unclear to me what CL implied with that reference. One can import 8K (some containers) with PD16, 17, 18, 19 and produce 8K with proper hardware and profile, yes a user workaround as optodata eluded to vs UI standard. Maybe finally users with AMD VCN2.0 cards can now do it too. I have one VCN2.0 GPU in a box, never got excited about trying with PD again as usually too many issues.

Jeff

See this article for vcn 2.0 in Navi10, Navi14 gpu implementation: https://en.wikipedia.org/wiki/Video_Core_Next . These are the amd cards at the bottom of the list that have the AMD Navi 10 GPU sold today: https://www.techpowerup.com/gpu-specs/amd-navi-10.g861 . They are not expensive at all.

I believe that one can import, edit, and produce 8k videos with a custom profile in the previous version PD14. See this post: https://forum.cyberlink.com/forum/posts/list/45694.page . YouTube can be a source of 8K vp9 videos that can be imported and edited in PD as a test without a powerful gpu. Works for me.

Edit: Jeff answered first. I don't have one of the vcn 2.0 cards on the list.

Thanks for the links!
I.e. an AMD Radeon RX 5700 would be such a graphics card with the chip mentioned? Hm - as far as I can see, a GeForce Rtx 2070 is available for the same price and it is faster ...
Does that also apply to video editing?
I am less concerned with producing (rendering) the finished video - that is now fast enough with my current GTX960.
It is mainly about the editing (preview) itself!
I now have a GoPro Hero 9 that can handle up to 5K resolution. Editing such a video is a pain! The preview is excruciatingly slow / faltering / jerky ... Of course, I can have everything "pre-render" - whatever takes forever ... (maybe a higher-performance graphics card would bring the desired speed boost ...), but for the moment editing 4 and 5K videos is not much fun!
So if I upgrade my PC, it should also help with video editing - I'm not a gamer ... Liebe Grüße / Regards
Oliver

Quote Thanks for the links!
I.e. an AMD Radeon RX 5700 would be such a graphics card with the chip mentioned? Hm - as far as I can see, a GeForce Rtx 2070 is available for the same price and it is faster ...
Does that also apply to video editing?

Yes, indeed the AMD Radeon Pro 5700 is at the top of that list in the link.

As an experiment I decided to checkout 8k video editing on PD14 as it was done 6 years ago in the link supplied. The vp9 encoded 8k 60p and 30p YT videos have audio only in preview. Converted them with HandBrake to h.265 at both 60 fps and at 30 fps at the low bitrate of 21.6 Mbps which is the same as the source. To my surprise these 7680 x 4320 resolution videos can be edited easily in both of my pc with i7 cpu’s without shadow files.

My GoPro hero6 4k 60p h.265 68 Mbps encoded videos can hang in the preview while editing and requires shadow files to be enabled for a good editing experience.

8k Vp9 videos can be edited natively in the later versions of PD like the converted h.265 ones. So why can I edit these 8k YT videos with Nvidia hardware decoding unchecked and without an AMD Nav10 gpu?? … whereas my 4k GoPro hevc 60p videos require shadow files to be enabled. The answer I believe is in the low bitrate. You can perform your own tests and post what you find.

Quote Yes, indeed the AMD Radeon Pro 5700 is at the top of that list in the link.

As an experiment I decided to checkout 8k video editing on PD14 as it was done 6 years ago in the link supplied. The vp9 encoded 8k 60p and 30p YT videos have audio only in preview. Converted them with HandBrake to h.265 at both 60 fps and at 30 fps at the low bitrate of 21.6 Mbps which is the same as the source. To my surprise these 7680 x 4320 resolution videos can be edited easily in both of my pc with i7 cpu’s without shadow files.

My GoPro hero6 4k 60p h.265 68 Mbps encoded videos can hang in the preview while editing and requires shadow files to be enabled for a good editing experience.

8k Vp9 videos can be edited natively in the later versions of PD like the converted h.265 ones. So why can I edit these 8k YT videos with Nvidia hardware decoding unchecked and without an AMD Nav10 gpu?? … whereas my 4k GoPro hevc 60p videos require shadow files to be enabled. The answer I believe is in the low bitrate. You can perform your own tests and post what you find.

Not sure I see anything unusual with your noted behavior.

But yes, bitrate and codec complexity is what pulls PD timeline fluidity down. Keep in mind resolution is contained in bitrate. The basic math was shown here https://forum.cyberlink.com/forum/posts/list/18251.page#post_box_93985 In my experience I'd say PD timeline fluidity starts to have some issues even on high end equipment at above say 40Mbps or so for a single track of advanced codec content, so it does not matter if 2K, 4K, 6K, or 8K for the most part as that's already taken into account in the 40Mbps bitrate. But for similar quality playback, higher resolution needs higher bitrate if truly playing back on appropriate equipment so not really proper to scale to same Mbps as a lower resolution.

My guess, you simply can't produce (full encode) to a 8K user defined profile unless you have the hardware to do it, PD CPU encoding won't cut it nor many iGPU's or say a 900 series Nvidia or below or I'm guessing here, below VCN2.0 for AMD.

PD really does not edit a 8K VP9 video natively, PD versions can simply import it but the timeline display is never 8K so you are always editing a proxy display, not native. VP8/9 decode part of Win10 current release.

Jeff

The posts in the forum does stir up interest at times. 5.7 K videos have been edited and produced with earlier versions of PD. One user produced a 8k video. Suffice to leave it at that as I don’t have a 5.7k or an 8k camera to use. Of course Full HD and 4k UHD are the limits to preview quality.

Does that mean that the bit rate is crucial for the fluid display in the preview?
So a video with 5k and 50000bits / s would be displayed more smoothly than a video with 4k and 100000bits / s?
Isn't the bit rate decisive for the compression of the video and the associated artifacts? Liebe Grüße / Regards
Oliver

Quote Does that mean that the bit rate is crucial for the fluid display in the preview?
So a video with 5k and 50000bits / s would be displayed more smoothly than a video with 4k and 100000bits / s?
Isn't the bit rate decisive for the compression of the video and the associated artifacts?

It appears to be so from my previous tests. There are exceptions to this. There are users who edit videos with relatively uncompressed formats like the lossless ones at several thousand Kb/sec and the editing is easier than any other formats whether highly compressed or not. These users may be able to answer your question better.

I think the fluency and bitrate JL_JL referenced is with normal, encoded clips. HEVC clips will take more processing to decode than AVC clips, and with lossless clips (using MagicYUV or other "intermediate" codecs) the issue becomes more about data transfer speeds.

Even with a 10 year old PC, with the MagicYUV codec installed you would be able to work with this 5.7K 1.3Gbps clip in Full HD preview mode if the spinning HDD is fast enough to keep up:



The trade-off is the gigantic file size, but with a more modern PC and SSDs you can get full HD fluency with multiple MagicYUV clips.

In this post I showed a maxxed out SSD transfer rate from running several MagicYUV clips at once on the timeline. That was with a SATA connection, and you can get almost 10x that performance when using an M.2 socket.

Back in 2015 I was a huge proponent of this and you can check out my slightly over-the top video showing the free helper app I created:
.

Quote I think the fluency and bitrate JL_JL referenced is with normal, encoded clips. HEVC clips will take more processing to decode than AVC clips,.....

optodata, that's not really true. I said 40Mbps as more of a generic threshold that one often starts experiencing many time playback fluidity issues, not necessarily a specific number for H.264 or H.265, as I wrote, "single track of advanced codec content". But more when bitrate is around 40Mbps, user can often expect some fluidity playback issues, H.264 and/or H.265 with PD.

Since appropriate hardware is very good with H.265 encoding as you were enlightened to here: https://forum.cyberlink.com/forum/posts/list/77898.page#post_box_319522 and the real reason why one buys hardware for specific needs when properly supported by software. You can see the same effect with hardware decoding timeline playback in PD as shown in the attached pic. Here I made a H.265 file with very similar specs as your H.264 skateboard file. Notice the lower timeline playback GPU avg load, 19% vs 33%, for the H.265 section. Even though H.265 is more complex vs H.264, specific hardware can make it more palatable. I used basic HD preview and results are from a RTX2070. You won't however see this same effect with CPU decoding of the timeline.

Jeff

Thanks for your contributions!
I experimented a little. In fact, with the same resolution and HEVC codec at a lower bit rate (60000 instead of 100000), the preview is smoother (even without a shadow file).