X264 medium vs slow. 265 encode with the same RF quality is night and day, .
X264 medium vs slow. 264 (new) codec. e. Reply reply I think the issue I'm having is GPU related but I'd just like to ask anyway, whats the minimum GPU for 1080p60fps @ medium x264 preset. The variant with both aq0 and g480 performs X264 vs New Nvenc? Question I have a 5900x and a 2070 super, so basically best case scenario for both hardware and software. 3 FPS : 1. 264 (x264) Medium : RF 36 : 30. MS-SSIM scores – Medium preset variations for x264. Other metrics he used are not linear either. If not, stick with NVENC. I did some poking around with OBS on my 3900X and had no problems streaming on CPU with 1080p medium settings. The UDP-lovers argue that TCP channel is very slow. 264 (new) The New HD60 X marketing materials state smaller file sizes using the H. If you encode a video to both H264 and H265 using settings that produce similar results, the h265 file should be smaller. It's supposed to be a replacement for the "old", but widely supported and used H264. His VMAF results show that NVENC consistently performs similarly or better than x264 medium/slow presets, and almost always better than fast/faster presets. Both to purchase, and to run. 52% of the available overall quality in ten percent of the encoding time of placebo. 71% PNSR means 75. I asked a buddy who runs a cloud encoding facility which presets his customers typically use for HEVC and he said either medium (the default) or slow. This is an implementation effect since both codecs use a different algorithm which differ in speed and size of the Why with x265 medium I don't get full CPU utilization such as with slow preset? I have tried both AVS and VS script and they just do the frame serving part, without any filter. NVENC has proven itself to be as good as anywhere from X264 medium all the way down to X264 very slow depending on the My streaming PC is a dual Xeon E5-2670 (8C/16T each CPU, Sandy Bridge era, 2. Image 4: example of losing I-frame. To choose the optimal x264 preset, you have to balance encoding and distribution costs. At 1440p, I was testing and the difference in quality of very Slow compared to very fast is minimal and can only be seen in fast moving games like Apex Legends or Overwatch 2, Paladins etc, with very fast in fast movements you notice some pixelation (in low Bitrate 7 mbps), (although if you use high Bitrate more than 8 mbps do not notice any pixelation), this pixelation lasts 1 second or less For the past two years I've put my Ryzen 1700 through its paces dialing in some fairly ridiculous x264 encodes, including 720p60 Slow and 1080p60 Medium. Nvenc H. Depends on your CPU, upload and game. So I decided to encode UHD>FHD. Also, if you try the same encode but with x264, slow is about 10% larger than veryfast. 26*). Guilherme Hoffmann Member. Reply New nvenc is almost as good as x264 medium so I would choose that if you cannot stream with slow. 264, you always want to choose the preset for No other tune setting is likely to be suitable. Performance impact from FASTER to MEDIUM is up to 50% slower encoding; quality difference is < 1-5%. Comparing OBS's x264 encoder at medium vs NVIDIA's Turing(GPU) encoder. But with both rav1e (backed by Xiph, Mozilla and Vimeo) and SVT-AV1 (backed by Netflix and Intel) in heavy development, this notion is GPU hardware based encoding is decent, but creates a lot more motion artifacts than CPU based encoding. I have 6 core 9400f, so my speed is 1. 264 side. That's what I've bought 3900x for after all. The content of the video is sports, and it has alot of rapid movement between frames. I would do a couple of previews with both and see what they look like. x264 and x265 are the encoders (that's the difference between h. 83% better compared to Medium and 5. Until about medium where they start to fall again by a small amount. Jul 13, 2018 On 1080p 60fps medium x264 preset my CPU dropped from 95%+ down to a comfortable 60ish% and very smooth video playback. 264 (x264) I don't want to do it. I personally stream medium or slow from my laptop but my bitrate doesn't ever go above 1600 total (audio + video). Resolution: 1920x1080 Bitrate: 6000 <-> 8200 kbps Rate Control: CBR x264 Slow and Turing NVENC will be a neck and neck quality contest at 720p. x265 slow gets a couple improvements to make x265 more worthwhile, although a lot of people disable rect since it causes a big performance hit for not To choose the optimal x264 preset, you have to balance encoding and distribution costs. Reply reply More replies More replies. The latter is a placebo option that's really, really slow, and seldom yields better results, so only use it if you don't care Which x264 Preset: Medium, Slow, Slower? Do we need to stick with Main Profile, or can we use High? When encoding H. There is a major difference in the amount of The default preset is medium, which is what the other presets are compared against. Better than x264 Medium, which would definitely require a 2nd PC. There is no reason to set it. At least according to VMAF, QSV is still around x264 Medium, while >20-series NVENC is on par with x264 Slow. . for x264 with R7 3700x is stream to 720p60 (games fast movement , games slow movement is better 1080p), bitrate CBR: 5000 - 6000 kbps (8000 kbps with twitch transcoding) , preset "fast - medium" , and add " bframes=2 " (in options x264) , profile "nothing" , antivirus off, and visualization OBS off, quality is excellent and whithout loss performance cpu, excellent 👍 changing from medium to faster, which doesn’t affect average quality, delivers slightly higher low-frame quality and improves throughput by 43%. Just like the title states, I just bought the 13900k. exe" -i input. 4fps for 2160p. J. danno84 New Member. 3GHz turbo) and I could do x264 Medium at 1080p60 with plenty of room to spare, but couldn't do Slow, when streaming Battletech recently. 9, there’s one area that I really simplified, in an effort to keep things a little less overwhelming for the average person. Yes, slower presets generally improve efficiency. x264 Medium at like CRF20 will probably give you more compact Going from medium to slow, the time needed increases by about 40%. Slow will, for a similar file size/bitrate, provide higher image quality, than fast, providing everything else is the same. Some of the settings may be outdated, but it's a great guide. The new generation of GPUs with the brand new encoder brought comparable quality x264 medium – if you can find a GPU that is. If you move it down to 720p, you can move the CRF up to 17-18. For 1080p 60fps in a high-motion scenario (like an action or FPS game), See comparisons between all the different FFmpeg encoding presets with x264, see output file sizes and the time taken to encode. The most important is to use the "faster" preset (or better) as all the lower quality presets significantly degrade quality. g480 still provides only a tiny improvement, but aq0 provides slightly more; The variant with both aq0 and g480 still performs the best on VMAF on x264 Medium. 2. 265 HEVC vs. x264 Slow Preset VMAF VMAF scores – Slow preset variations for x264 on 1080p60 Apex Legends. So, for someone with both a 3700x and a 2070 which is best? x264 or NVENC? There seems to be a lot of conflicting reports. From what I've seen/read, it seems like I should be able to get a good experience with the Slow preset at 1080p60. x264 has two motion estimation algorithms worth using, umh and tesa. Comparing the two, the difference wasn’t that great. The even slower placebo setting, will show some reduction in file size, but you'd never choose a preset for that reason. I might try to answer your questions: 1. Jan 12, 2024 Nvenc, the encoder Nvidia GPUs provide, produces similar quality to that of x264 with a preset between medium and slow (with RTX 2xxx) and slightly better, up to slow, Likewise, the HQ and Super HQ Presets use even slower video encoder presets (x264 slow and veryslow, respectively) and even higher quality (RF 20 and 18, respectively), H. I sat out to finally prove something that I've been saying for a long time: X264 Slow is not as magical as people say it is, and it's definitely not worth th Ehh, for the record I don't disagree that their claims were a little ridiculous, but this video wasn't really them double talking. In his defense Slow quality in his 10Mbps graph is still just 3. it will take twice as long). Over the last few years, I’ve used the first chart below to They are assuming x264 encoder with the veryfast preset, and low to medium motion in your scene. aq0 begins to show a more consistent detriment to x264 when we get to the Slow preset. 6Ghz GPU: GTX 670 2GB RAM: 16GB 3000mhz LPX Vengeance MOBO: B350 PC Mate Runs 720p60FPS medium preset absolutely fine with no encoding lag or dropped frames @ With the introduction of the new 12900K from Intel and the ability to stream via X264 Slow on a single PC, I was inspired by Epos Vox to setup some scenes to I mean if medium @ CRF 20 equals in quality slow @ CRF 22, I would like to know the exact bitrate comparison of both options to determine the best option. This took an extensive amount of work including perfectly nailing the in-game settings of the games being streamed and correctly capping frame rates - and it was so finely balanced that a software update any where along Streaming with more than one PC has been the leader in H. I used the preset It's common knowledge that slow is way better than medium. and it was back and forth between nvenc and x264 medium/slow. As you say, the algorithms are different and the lossy is apparently more efficient. Let’s take a look at what’s needed to set up your stream for The goal was to see if Intel's new QSV could beat x264+slow going assuming their claims regarding their beating x265+veryslow+psnr are true. mkv -c:v libx265 -preset slow HD60 X H. x264 Slow Preset VMAF VMAF scores – Slow preset variations for x264. The slow preset is a good trade-off in my opinion. [ ] values are either set by default, or by another argument. Altering the RF value is the means of altering the final file size. Definitely not a bad option if you have an AMD GPU and an Intel CPU, compared to using AMF. 5Mbps bitrate. I took a video with approximately 10 GB and encoded the first two minutes using the following command line: "C:\Program Files\ffmpeg\bin\ffmpeg. For me power consumption, system noise and fast encoding are just as important as any minor differences in quality at bit rates so low as to be unusable as far as I'm concerned. I'm not even sure what kind of CPU you'd need to maintain "slow" preset at 1080/60 in real-time with fast-motion video, but it'd be expensive. 9 FPS : 1. Presumably, the p-cores should handle the games, while e-cores take care of encoding—and, boom, profit. Jun 2020. Jan 25, 2019 #1 I searched but couldn't find a good answer. RTX NVENC is good, BUT nowhere near x264 medium as example. AV1 encoding is slow. 2 MB : H. 9 MB : H. Top 5% Rank by size . An analysis of your x264 settings cabac=1 - this is the default on all presets except for ultrafast. I was testing and the difference in quality of very Slow compared to very fast is minimal and can only be seen in fast moving games like Apex Legends or Overwatch 2, Paladins etc, with very fast in fast movements you notice some pixelation (in low Bitrate 7 mbps), (although if you use high Bitrate more than 8 mbps do not notice any pixelation), this pixelation lasts 1 second or less In my experience, pre-Turing NVENC is close in quality to x264 "veryfast", which is essentially what plex uses when you choose "make my cpu hurt". I'm currently using OBS to record with my RTX 3060 Ti and the Nvidia Nvenc H. It would be interesting on what resolution you are streaming and what you are planning to achieve with the 9900k. trellis=2 - this is the default on presets "slow" and above Not usually. + - I want to transcode 1080i video to 720p, relatively fast, using x264. Reply Well, I'm happy with 1080p60 x264 slow. 37% quality. What's acceptable depends MS-SSIM scores – Medium preset variations for x264. * - are differing values from medium. 6GHz/3. Still Comparison Images: https://imgur. Then as you go to higher presets they rise again by a small amount (~10 percent). I think nvenc is certainly the way to go right now. Nvenc trades blows with x264 medium, and for some qualities like colors, motion, or text quality, Nvenc is better than x264 medium. (instead of 0. Better x265 medium or x264 slow? High Efficiency Video Coding (HEVC) Welcome to Doom9's Forum, THE in-place to be for everyone interested in DVD conversion. 265 encode with the same RF quality is night and day, I generally stick with x264 Medium CRF20 for Blu-ray rips. 26* vs x. Hello! I have noticed that x264 placebo produces bigger file in comparison with "very slow" preset I've searched a lot trying to find out why, but nothing certain, almost everywhere the same thing: "placebo is just a waste of time since it produces 1% better quality but requires 10x more encoding time" Newest NVENC, present in all 20 and 30 series as well as most 16 series models has slightly better quality than x264 slow preset. More posts you may like r/VideoEditing. 92 Mb/s : 84. New nvenc is almost as good as x264 medium so I would choose that if you cannot stream with slow. With my test files, the veryslow preset delivered the top quality, so placebo made no sense, while slow delivered 99. So if you can game, with safe CPU temps, power consumption and satisfactory performance with x264 on anything better than slow preset (remember better = slower and so on), I'd go for it. Firstly, because of requires-responds, secondly, because it should order the data before sending and order Even my 3950X can just barely do X264 slow without other games running, while looking worse than NVENC on a lot of games. I was having troubles earlier running a stream as the cpu was thermal throttling (100c). If you have plenty of space, H265 (NVENC) may be fine. EposVox Video. You should also be using CRF 17-19 for movies as per the spreadsheet that was posted here quite some time ago. 264 encoding for years, but NVIDIAs Turing and Ampere generation has put a significant dent into that lead. Thread starter Guilherme Hoffmann; Start date Jan 25, 2019; G. At least, that was the status quo. x264 FASTER has the highest overall score, followed by Medium, Slow and Very Fast. Reply reply More replies. And if you look at the visual quality you'll see that it increases from veryfast to placebo. 69% shown in Unless you have specialised requirements, deviating from the presets without in-depth knowledge about the way x264 and x265 works isn't recommended. Basically, I recently upgraded from an 8700k to a 12900kf (with a 3080), and I'm trying to make the most out of it for x264 streaming. Going to slower instead would result in about 100% more time needed (i. For the past two years I've put my Ryzen 1700 through its paces dialing in some fairly ridiculous x264 encodes, including 720p60 Slow and 1080p60 Medium. To be honest, quality wise I barely see any difference, when I tested very fast vs. Reactions: Patrick Ewalt and Deleted member 121471. The 8700k can barely take 1080p60 medium with some heavy oc as dedicated streaming rig. I have tried different decoders too: DGDecNV, FFVideo, FFMS2, etc. 44% and 0. slow . The lowest x264 preset EposVox ever mentioned to use even on the 3900X (in this video at least) was medium, and if IIRC Steve was arguing against AMD's use of the slow or slower preset while advocating for medium. R9 5950X aq0 has a consistent positive buff to PSNR on Medium preset. MS-SSIM MS-SSIM scores Question / Help x264 preset quality - Faster vs Fast. Before you start posting please read the forum rules. My settings are something between slow and slower with some tweaks. If you’re using the slow, slower, or very slow presets, you’re reducing capacity by 47%, 70%, and 133% respectively to produce imperceptible quality differences. On Medium, the PSNR scores do the same as they did on Fast preset. H265/HEVC is a video codec. MEDIUM to SLOW: up to 35% slower encoding; quality difference is < 1-2%. Also Grainy sources will sometimes need up to 20 CRF to get file sizes that are reasonable (~8-15GB), while more modern digital sources can handle as low as 17 CRF without getting too bloated. Newest NVENC, present in all 20 and 30 series as well as most 16 series models has slightly better quality than x264 slow preset. Preset Slow or Slower only, as Medium is just to janky. medium is going to look, just do it! Record some footage at high quality, then run it through x264 at various settings, and see if you could get someone to run it through NVENC on a Turing card (I'm sure there's plenty x264 vs AMD HW H. medium vs. I basically indicated that: When using constant quality (RF):-RF value chooses your quality-slider trades encode time for a smaller filesize Today I tried to find out which difference in file size the preset slow makes compared to the preset ultrafast when encoding a video (if all other parameters are the same). 264 (AVC) Thread starter danno84; Start date Jan 12, 2024; D. 29x : 0. At 1080p it will easily outpace your 2700. I created one set of files for x264 using the very slow preset, then x265 using the medium preset. 16% better for 2. 264 (x264) Medium : RF 33 : 29. The 9900k should be able to, and therefore will get a lot better quality. 26 Mb/s : 115. Reply Definitely not, on the h. Just now I did some comparison, native UHD vs resized with spline64 2160p>1080p>2160p and I realized that loss in detail not worth it. I now have it running at >87c with cinebench score (multi)of 38340, coming from 36030 caused by throttling. I almost never From there I have been encoding at the 'medium' preset, but when I push it to use the 'slow' preset the stream looks remarkably better. Using a slower x264 preset allows for better compression and slightly better If you’re using the slow, slower, or very slow presets, you’re reducing capacity by 47%, 70%, and 133% respectively to produce imperceptible quality differences. The current GPU based encoders are somewhat comparable to fast x264 settings. Current Spec: CPU: Ryzen 7 1700 @ 3. Turing and Ampere NVENC is more like x264 "medium" preset or perhaps slightly better, somewhere between "medium" and "slow". This took an extensive amount of work including perfectly nailing the in-game settings of the games being streamed and correctly capping frame rates - and it was so finely balanced that a software update any where along The difference in quality between a very fast and very slow H. In previous slow vs very fast example 96. com/a/Yvkti2y Timestamps 3000 KBPS00:00 - Far C These are 12C 12T processors that have more than headroom to encode x264 fast or medium, depending on the game you're playing with. Lower CRF allows for near lossless video compression but increases filesize a lot. jerknerkel New Member. For example, I just encoded a short video at CRF 18 with x264: The Best config. This article shows you how. When using medium with x265, there's so little advantage to x265 you may as well stick with x264 (on something slower that's the same speed as x265 medium), unless you're at 4K+ or targeting very low bitrates (like RF 30). The variant with both aq0 and g480 performs the best. 265 HEVC codec. Let's say I'm recording a If you want to know how slow vs. At 1440p, Nvenc trades blows with x264 medium, and for some qualities like colors, motion, or text quality, Nvenc is better than x264 medium. 22x : 1. 9. If you’ve read the “best handbrake settings” guide for 0. By posting to this forum you agree to abide by Yet in this video they say if you're on a Ryzen 3600 up to a 3900x then just use x264 medium - no questions asked. HEVC compresses better. yeccepd gzccgt lhban rmkpjv pzo alrtmr plnog jaq vpsoumb ysly
================= Publishers =================