Soldato
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
NVIDIA 4000 Series
- Thread starter MrClippy
- Start date
Man of Honour
Nice one cheers, just updated and ran the same file through to compare but this time set a constant bitrate. I forgot that bluray rips and downloaded movies etc are all in the 15Mbps range, and knew that AV1 can output the same (if not better) quality than H265 at half the bitrate, so put the AV1 encode to 8Mbps. From a 7GB source file at HEVC 105Mbps (unnecessary bitrate I know) it came down to under 600MB using AV1 and between the two, there is zero visual differenceIf you download the nightly build from github:
Also noted that the nightly version of Handbrake didn't use any of my CPU cores during NVEnc encoding using any codec, whereas it did in the public release version, so even better still. All 4 codecs encoded at ~200fps average.
Game changer!
Constant bitrate isn't good for efficiency. If you encode with NVEnc HEVC at the same bitrate can you see any difference in quality between AV1 and HEVC?Nice one cheers, just updated and ran the same file through to compare but this time set a constant bitrate. I forgot that bluray rips and downloaded movies etc are all in the 15Mbps range, and knew that AV1 can output the same (if not better) quality than H265 at half the bitrate, so put the AV1 encode to 8Mbps. From a 7GB source file at HEVC 105Mbps (unnecessary bitrate I know) it came down to under 600MB using AV1 and between the two, there is zero visual difference
Also noted that the nightly version of Handbrake didn't use any of my CPU cores during NVEnc encoding using any codec, whereas it did in the public release version, so even better still. All 4 codecs encoded at ~200fps average.
Game changer!
Your source file is AVC by the way not HEVC.
Last edited:
Man of Honour
What would be the best settings to save in this case? Normally I've just left it on defaults after making sure the resolution and encoder was set accordingly. I'm not seeing a difference visually between the two if they are both set at the same bitrate.
Edit*
Sorry this is what I meant by the settings I am using, Average bitrate, not constant quality mode, and the framerate is set to constant framerate (was before just toggling it in this screenshot).
Edit*
Sorry this is what I meant by the settings I am using, Average bitrate, not constant quality mode, and the framerate is set to constant framerate (was before just toggling it in this screenshot).
Last edited:
Sure the quality isn't great but how it is created is cool - you just type in "bear playing guitar" and it creates that video. Software like this already exists for photos and people have been using it to make fake photos - a fake photo just this week won a famous global art competition. Soon we'll be able to make high quality fake video
You should probably use AV1 10 bit instead of AV1. Constant quality is best for efficiency, but the CQ values aren't equal between HEVC and AV1 as AV1 results in a larger file size with the same CQ as HEVC. On encoder preset you should use slowest for the best efficiency.What would be the best settings to save in this case? Normally I've just left it on defaults after making sure the resolution and encoder was set accordingly. I'm not seeing a difference visually between the two if they are both set at the same bitrate.
Edit*
Sorry this is what I meant by the settings I am using, Average bitrate, not constant quality mode, and the framerate is set to constant framerate (was before just toggling it in this screenshot).
Last edited:
There's no Founders Edition cards that interest me this time around. I think because it was never Nvidia's goal for these to be sold in high numbers they are purely for marketing purposes, and for Nvidia to get better reviews on product launches.
If they could do a RTX 4070 TI FE for £650, I can see that selling - therefore, I doubt they would consider it. Maybe if they did that alongside intended price drops to the AIB cards, it could work.
If they could do a RTX 4070 TI FE for £650, I can see that selling - therefore, I doubt they would consider it. Maybe if they did that alongside intended price drops to the AIB cards, it could work.
Last edited:
Man of Honour
Found the Firestorm app for the 40 series it's linked through the Zotac product page not their downloads section lol. Set the LED now to a dim white just to outline the logo which is how the 3080 Ti FE was, just how it should be 
The organisers of that photo competition knew it was an AI piece because he told them beforehand: https://www.vice.com/en/article/dy3vxy/sony-world-photography-awards-ai-generated
Declining the prize and saying it was fake is a forced publicity stunt.
People already have workflows for AI created videos using Midjourney, ChatGPT and so on. Sure a bit more manual and hands on required.
Last edited:
It's not quite that magical I'm afraid - in my tests, AV1 will completely blur out film grain if you go too aggressive with it - I'm quite the film snob and my 4k AV1 encodings ended up 20GB or so smaller than their HEVC counterparts at what I'd describe as 'visually' lossless. Still nifty though - the only problem I have at the moment is that AV1 isn't supported by the usual suspects when it comes to streaming boxes. I was planning on getting an Apple TV 4k but I'm going to have to wait for now.I forgot that bluray rips and downloaded movies etc are all in the 15Mbps range, and knew that AV1 can output the same (if not better) quality than H265 at half the bitrate, so put the AV1 encode to 8Mbps. From a 7GB source file at HEVC 105Mbps (unnecessary bitrate I know) it came down to under 600MB using AV1 and between the two, there is zero visual difference
Caporegime
Intel IGPs support AV1 decode AFAIK?
I was referring to things like Nvidia's Shield Pro, Roku + Apple TV (plus Kodi boxes etc.) - you can get random Chinese Android boxes that supposedly support AV1 decode but I'd rather not risk my network security on an Alibaba streaming box (I need something that does 4k HDR AV1 hardware decode).
Man of Honour
I'm still happy with H265 though tbh, 15Mbps vs the AVC game recording from shadowplay shaves off massive amounts of disk space and visually looks the same, whilst ~22 CQ appears to look identical in AV1 whilst still being slightly smaller.
I only intend to upload these exports to youtube which supports AV1, and then internally the platform presents as VP9/AV1 whatever is supported. It's mainly so upload times are faster I guess because you're halving the file size, so it's useful for that, and streaming of course. I don't really archive movies any more because download speeds are so fast I just watch on teh fly, if a movie needs to be rented/downloaded then a 4k HDR flick takes several minutes to download.
Ewww!
Blu-ray > 4k streaming. I prefer to get the movies I love on 4k UHD and rip them - massive file sizes though hence my interest in AV1. I agree that the convenience of streaming is great though even if the quality isn't (still, AV1 should help with that too).
Man of Honour
Yeah that's what I mean, HDR 4K BR rips, usually around 30-40GB, Gigabit internet has changed the way I consume movies 
I took a random 1 minute segment of one of my Shadowplay recordings and tested NVEnc HEVC and AV1 on it with a different CQ value to get a very similar file size.
NVEnc HEVC - CQ 28 slowest preset. File size 235,636 KB.
NVEnc AV1 - CQ 36 slowest preset. File size 235,635 KB.
Overall bitrate is 31.9 Mb/s in both.
Then I took a screenshot of a random frame and did a 1920x1080 crop so you can more easily see the differences in quality. Can you tell which one is which?
Soldato
Some expert pixel peepers on here will find this a breeze!