So glad I came across your channel! It's refreshing see a TRULY pro approach to gear/software/video editing. No "influencer" hype or yelling, antics and BS. This is mature and respectful to the viewer. Thank you.
Thank-you. We appreciate that, we try to be as objective as possible and present the relevant information. Please let us know in the comments if you ever see anything we can improve on.
Thank-you so much. We're grateful to Audiio for helping support content production like this. Thanks to you too for watching and taking the time to leave a comment.
You all are simply the best channel there is for me. I am such a fan of your content and professionalism. Thank you so much for making everything so clear.
Thank you for the awesome video!!!! After 13 years with a 2012 Mac mini I upgraded to the base M4 Mini. It's an excellent machine. I am impressed with the render times I get on my videos using DaVinci Resolve- Worth every penny. The price was only $100 more than the Mac mini I purchased in 2012. I took the 1TB SSD out of the old Mac mini, formatted it and put it in an external drive housing and away we go. I expect this machine will last 10 years as well.
To whom it may concern. Benchmark A: AMD Threadripper Pro 12 Core / 128 GB Ram / 2 x RTX A4000 16GB / Windows 11 / 19.1 Studio First render attempt: 6:47 minutes / second: 6:25 minutes / third: 6:20 minutes All tests run from SATA SSD to (different channel) SATA SSD (onboard SATA 6G) Thanks for the free Benchmark 🙂
Thanks for this great video Benchmark A: Intel(R) Core(TM) i9-14900K / 64 GB Ram / 1 x Radeon RX 6900-XT SE 16GB / Windows 11 / 19.1 Studio First render attempt: 5:53 minutes / second: 6:06 minutes / third: 5:47 minutes.
Thanks for the benchmark. My results for Bench A A 671 - Mac Pro mid 2010 - 128GB DDR3 1333mhz - 2 x 3.46 GHz Intel Xeon 6 cores - AMD RX 6800 XT - 19.1.1 Studio. Not bad for a 14 years old computer.
Thank you, guys, for the valuable content on your channel. Keep up this great work. A 247 - M1 Ultra (20CPU 60GPU) - 19.1 STUDIO B 131 - M1 Ultra (20CPU 60GPU) - 19.1 STUDIO C 184 - M1 Ultra (20CPU 60GPU) - 19.1 STUDIO
Thank you for your helpful instructional videos. I am new to video editing and teaching myself your videos are a great help. Since I am just learning to use DaVinci Resolve I am still using the free version so here are my results from a new MAC Mini M4 Pro: B 294 - M4 Pro (12CPU 16GPU) - 19.1.2 build 3
Thanks Gary, great to have you here. Please check out our beginners tutorial for Resolve, we hope it's helpful for you. ua-cam.com/video/znBHzeXpsUw/v-deo.html
Thanks for great benchmark! Benchmark A: Ryzen 7 7700 / 32GB ram / RTX 3060 12GB / Win 11 / 19.1 Studio All 3 runs: 524, 553, 530 I guess it's slightly faster or on par with the new Mac Mini with M4 Pro base config (16 gpu). it's really interesting how 20 gpu version would perform!
As someone with an M1 Max who's new M4 Max is arriving on Monday, thank you so much for saving me the hassle of figuring out just how much faster Resolve is going to run!! (and also how resellable an M1 Max is 😆)
@@astromoosie looking for about £2000-2300 (about $2700). It was a spec’d to the max one tho. 8tb with 64gb ram so that will hopefully bump the price up a little
@ thank you for replying so quickly. This gives me a better idea of what I can get mine for. Still not sure if I even need to upgrade. It still handles everything extremely quickly.
Yeah, I think the name sneakily got changed to Advanced RISC Machine, but its heritage is Acorn! Loved my Acorn computers. Thanks for watching and commenting.
Thanks for putting this benchmark together, what a great idea! Here's my Desktop PC's benchmark results: A 325 - Ryzen 9 5900X, RTX 3090, 128GB DDR4 RAM - 19.1 STUDIO All 3 runs: 361, 325, 334
That's so awesome Kaur. Thanks you're on the scoreboard. Quite high too. I'd really love to see something beat the M4 Max. No one yet though :( I thought a 4090 could do it, but the one we've got on the board didn't make it.
Wow! Your video was recommended and I’m glad it did. Great content and helpful information and real world information, no useless benchmarks. Subscribed directly. Thank you
Just did a test real life test on the m4 vs m1 using different clips and codec and raw from sony and canon. And more importantly playback and editing not just render. I think it’s more useful seeing actual user experience than render time.
Hello! Yes, edit performance is more important. As mentioned in the video, render time will give an indication of edit page performance as they will be using the same internal engine. I.e. if it can render back in realtime that means it can play back in realtime. The reason we're using render time to benchmark, is because it's easier to extract statistics from a render. That's why render performance can be considered an indication of real-world performance. Hope that helps.
@@team2films I'm not sure actually that render time reflects application experience at all though. I have an 11700k and a RTX 4080 so my render times are fantastic, but timeline performance is horrendous. Picked up an M4 Pro to see if it makes timeline performance enjoyable for me, and I don't care about the render time at all.
@@OnceinaSixSide Hello, thanks for taking the time to comment, we appreciate that! Yes, of course there are other factors that affect edit page performance that can't be accounted for in the render benchmarks. For example drive latency will have less impact on rendering as it's more likely to be a continuous read. However it will affect random operations in the timeline where data is not read sequentially. As there's no simple method of programatically benchmarking edit page performance we decided to use render time as an approximation. Ideally, a suite of benchmarks could be designed to test and identify the various bottlenecks that might affect system performance. It can be really subjective describing edit page performance. 'This plays smoothly for me' is dependent on lots of different factors. Hopefully, the render times gives an objective analysis of performance in relation to other systems. For yourself it might be worth turning on caching. It sounds like you have a capable system, there could be a single component that is causing a bottleneck for you?
@@team2films Totally hear you on the subjective experience, and I wish there were a better way to benchmark timeline performance as well. I suspect something was off with my system as I just finished upgrading the mobo, CPU, and ram to x870, Ryzen 9 9900x, and 64gb DDR5 and the timeline performance is now similarly smooth to the M4 Pro I was testing. This is without any sort of Intel quicksync now as well and just pure CPU bruteforce transcode + whatever the 4080 does (which I dunno if its much in this regard). I'm really happy with this performance now and what's interesting is the M4 Pro definitely still had a smoother timeline experience. I tested some projects back and forth, and I was getting very smooth playback on some complex composites whereas the newly upgraded PC would hang every now and then. Renders though on the PC are significantly quicker though thanks to the 4080, as well as operations like Magic Mask which seem to average around 30fps versus the M4's 1-3fps.
@@OnceinaSixSide Thanks, we appreciate your balanced subjective. Hopefully we can develop a better timeline benchmark in the future. I think with timeline performance, latency might be having an impact. i.e. rendering highlights sustained performance well (once the computer is underway) but latency affects the responsiveness of the play button in the timeline. The M4 chipset architecture lends themselves towards being incredibly responsive. Not sure about the magickmask operations. That could be taking advantage of the 4080's better AI capabilities? If you have any future observations or ideas, please share them. It's great to have you here, thanks for taking the time to leave a comment.
@team2films i recently switched to mac and got the m2 pro with 16gigs. I am hitting the memory threshold all the time from multitasking, so I am looking at what direction to go. I have always been an AMD fanboy, but understand what apple brings. I find it so convenient on pc to upgrade as needed, but I really love the small footprint of what apple brings. It just sucks that I have to spend 1500 plus just to upgrade the ram at this point...
I ran a test in a second system, Benchmark A result. A 258 - Intel i7 13700K - 32GB DDR5 6000mhz - AMD 7900 XTX - 19.1 Studio Slightly faster than my Ryzen 9 5900x with the same video card, and a lot less memory in the Intel system. The 13700k has much better single core performance and higher boost clocks on P cores than the 5900x, along with faster ram shaves off 24 seconds vs the 5900x on ddr4.
A 1484 - M2 Air (8CPU 10GPU 16GB/512GB) - 19.1.2 Studio Thanks for the video and the benchmark project! I have been thinking about the base 16" Macbook so this is helpful to see what sort of improvements I might see. Thanks again!
Thanks for the value, this is called as useful content, with no hype and running for views. I saw in a video that a base M4 with more 64GB RAM outperforms M4 Pro with base 24GB RAM, is it true?
Scaling linearly with the number of gpu cores, is an amazing testament to the architecture of the chip… the scheduler apparently is really well-implemented, as it can divide the tasks so linearly among fewer and more gpu cores.
Thanks for the great share and video! Love your channel. I did try to benchmark my M4 ipad pro, with no luck. It had trouble decoding the RED files hence failed rendering. Tested my home workstation (win 11) and results: A 378 - Ryzen 9 5950X, 128Gb Ram, RTX 3090 - 19.1 Studio
Finally had a chance to benchmark my workstation: A231. i9 13900KS, 64Gb Ram, RTX 4090, Win 11 19.1 Studio. Full Watercooled loop incl GPU. Dissapointed it did not top the charts ;) With each successive render times went down. (ran 3 times) The good thing about watercooling is GPU never went above 44deg, barely raised a sweat! Could render this out all day..
Yeah, we’ve been disappointed with the 4090’s performance and impressed by the M4 Max. A laptop and top end GPU are basically equal in the test. There are some minor minor performance gains from different codecs… but nothing has touched the M2 Ultra. I guess Resolve benefits from the sheer number of cores or the unified memory. This is the great thing about real world tests. Synthetic benchmarks wouldn’t reveal that. You still have an amazing computer! Thanks for running the test and leaving a comment. Appreciate having you here.
Intel i9-13900KS / NVIDIA 4090 / 64GB / RESOLVE STUDIO 19.1.3 BENCHMARK A: 248 sec BENCHMARK B: 173 sec BENCHMARK C: 135 sec If my 5090 pre-order arrives anytime soon, I will test that as they have more encoders/decoders than the 4090 which I hear makes quite a difference with h264/5.
Thanks for this great video - I love the relaxed atmosphere! One thing to note is that for lighter workflows (less effects), the GPU has much less impact and the encoders are often the limiting factor. So for someone working with 4k at max and moderate amount off effects, a costly MAX might bring less benefit - in that case a base M4 might be as good or even better than a M1 MAX.
Yep, that's a great point. And it also underpins the fact that you don't need the best, you just need good enough. Thanks for watching, commenting and sharing your thoughts. It's appreciated.
im so glad i clicked on one of your videos contents are S-tier🔥 Subscribed! just a thing tho maybe its personal choice but i was bit sceptical about blue and white logo i think a vibrant logo would be better maybe like resolve's logo
My M1 is still busting through and ready for the upgrade to the m4 for my next big project. But with that said, I'm still very impressed with the M1 for basic projects and knowing how to use proxies and such
Hello! Thanks so much. Sorry, the M4 iPad wasn't featured in these tests as it runs a slightly different version of Resolve. It can't read all the codecs, can't process the same resolutions, etc.We really wanted it to be a high-end benchmark and include those codecs and high resolutions. We'll see if we can add a universal iPad benchmark at some point in the future.
Interesting result for the old workstation Xeon E5 1650v3, 256GB DDR4-2400, m.2 SSD with a newish Nvidia RTX3090: Benchmark A (Studio): 455s (7:35) Benchmark C (CPU): 680s (11:20)
@@team2films hi again, I did sent you an email on your site the same day but I never did get an answer , what version of Davinci are you covering in the course ? Thank you
Thanks! We added that little logo last minute to the video. I think we might do that with future videos too, put some info about what we are using to shoot it at the bottom
Yes, that’s true. Remember render times are a reflection of what edit page performance is like. Of course. More performance than you need is not helpful, so if you currently have more than you need then you are fine!
An amazing job, thank you for taking care of the community and looking out for the consumer! My data: Benchmark A: Mac Studio M2 Max / REFURBISHED / 12CPU 30 GPU 32 GB // - 19.1 Studio RENDER ATTEMPTS: 418 / 398 /396
Appreciate you sharing that. It's interesting that it gets faster with renders 2 and 3. I think that's a fan curve thing, where it is slow to turn the fan up and lets the chip thermal throttle. By 2 and 3 the fan is on constantly so it doesn't thermal throttle. Result added.
As a complete newbie trying to decide between a PC system and a Mac, I was wondering if you have any videos comparing the Mac Mini M4 Pro to any NVIDIA GPU?
One suggestion/caveat you might want to mention. For the folks that have those NVidia cards, 4080 and 4090 variants, they would probably see improvements over the Macs in your test and probably be at the top of the list (guessing) if they chose the AV1 dual encoding option. When you are dealing with 4k and higher, choosing mp4/AV1 will allow it to likely outperform even the mac ultra in this case....just throwing that out there. This test being h.264 and forcing a single encoder does seem to limit what the NVidia 4080/4090 series cards are capable of in the real world. If anyone owns a 4080/4090 and use 4k and higher, I can't imagine why you would not want to use AV1/Faster/High quality settings.
That's a good point, that some systems will perform better with particular loads. One challenge of making a universal benchmark is to use tasks that will run across all the systems that you might benchmark- h264 seems like a good candidate for that. However, the Benchmark is so computationally intensive, that the encoder doesn't have the opportunity to become a bottleneck. I ran Benchmark A on my computer rendering to h264 and then to an uncompressed avi. The results, h264: 4:10, uncompressed avi: 4:05. That margin is narrow enough to not be conclusive, as there is variance when running the same test. The point is, the encoder has a negligible effect on the render. We're not trying to cheat windows computers. Any modern editing computer should be able to render h264 effectively. At best you could say that a 4090 is better at encoding av1. But its computational performance in Resolve is still falling short of an M2 Ultra or at the very least, matching an M2 Ultra. Does that seem fair?
@@team2films I don't think you are trying to cheat windows computers, on the contrary it's still really impressive what Apple is doing with such low power usage. I just wanted to point out that if I owned a 4080/4090 or card with dual encoders and I was dealing with 4k and higher, I would definitely use AV1 and never use h.264, I do still use h265 for UA-cam uploads but I don't have the fancy 4080/4090. The dual encoders will reduce the time in this case, it's still encoding a fair amount of video and I think it would be neat to see both in their best form, but I understand there's really no way to produce some single benchmark for everyone/every system...that's impossible. This gives everyone some idea of what they can expect in given systems. I really appreciate your aggregating the list in the description, it has helped me....steer away from some upgrades. I am still curious to see how much it would lower the time with the AV1 dual encode and I am speculating, but from what I have seen in other tests from MrAlexTech (not my own) the dual encoder AV1 drastically reduced export times, but he was running tests on longer videos as examples, and probably not as complex of workflow as the bench here. Still the macs are crazy good considering what they are, a laptop that stays relatively cool and draws almost no power in comparisons is just...crazy that they even compete on any level with the fastest CPUs/GPUs on the market. Sadly, Apple knows this and their price reflects it. I'm sure the new Ultras will scale just the same like you mentioned, and destroy everything here, pending NVidias 50 series offerings.
Interesting! As eluded to, the codec format did improve render times on the 4090. Without (MOV) time was 231. Set Format to MP4, Type to Nvidia and render time dropped to 222 :) (on par with M2 Ultra)
@@TonkaFJ40 What is your time using AV1? the 4090 can dual encode a 4k or higher video with AV1 which is..quite nice. Love to hear if it improves, or if that 231 was dual encode.
Great video as usual. I currently am using the M1 Max 16" MBP, and it's just the best computer I have ever had. I am guessing I can probably get work to upgrade me to the m4 max, but I am not sure if this is the year or wait and see what the m5 has. One other really cool use case I have been exploring is getting a base spec m4 Mac Mini and using it for remote rendering. It won't render as fast, but I don't typically wait around while stuff is rendering, so it could be a pretty cost effective way to get more work time out of the laptop (because it is a bit annoying that you can't do other stuff while stuff is rendering), while I leave stuff rendering on the other computer. I think it will also make a killer home video server...just a thought. Keep up the great work! I hope you release some more courses soon!
Hi Max, thanks so much for your support. The M1 Max is amazing. If you're not sure about upgrading, it's probably worth waiting to see what the M5 brings. If it's a work expense though.... 🤣 Yeah, the M4 base Mini as a remote renderer is an awesome idea, I think we might experiment with that. $600?!?!? It's such insane value. And yes, we're talking about what courses we will make next.
@@team2films yeah exactly, at $600 it's just like begging for uses to be found. I think in some cases it is beneficial to scale wider not higher. anyway, can't wait to see what new courses you come out with. I hope you teach how to make such great educational videos...
@@MaximoJoshua Thanks Max, we have some fun stuff that we are planning, here's a preview of what's ahead (hopefully): Cine 12k, 35mm film, Storage, OLED
@@team2films those look like awesome topics, very interested in 35mm film (will this show scanning or working with scans?). Any topic you guys cover, you find a way to make it interesting and comprehensive. Thanks again for the benchmark, should prove very useful in the future.
Really clean presentation with good visuals! On my Mac Pro running Sequoia beta 3, the A Studio benchmark doesn't run quite clean and I've had two pretty hard crashes (which never happens). 2-3 clips seem to lock everything up (Relight and Depth map are the worst offenders). The B-version runs through as expected as such: B176-12 core Mac Pro 2019, two W6800X Duos, 240GB RAM. (anecdotally, if I turn off the effects Relight and Depth Map, but leave the clips as they are apart from that, the A-version renders in 177 seconds) Since this Mac Pro has a lot of GPU grunt, most "normal" projects run super smooth with typical effects like NR, Film simulation and so on.
Thanks so much for watching, we really appreciate that. And thanks for your kind compliments too. That's interesting. Sorry you're having problems. The Benchmark is running well across a variety of computers. Even old computers run it fine, but just take longer. It might have highlighted some issues with your system and Resolve installation.
this channel is new to me. looks like your Da Vinci Resolve benchmark is the one MaxTech were using in their M4 Pro vs M4 Max video, , yes it was. the #algorithm never sleeps. nice production style/values team (preferable to most of the Apple Mac + Video Production shouty "content" creators already!). NB Rendering time is the least of most editors concerns, and even for many compositors, it's more about what's achievable in the GUI without shuddering/down-sizing playback & previews/pre-rendering sub-comps/switching off layers that is where the time savings are.
Ah, that's so cool! We just went and checked out their video. It's a shame the test crashed on the M4 Pro as we've seen it work on the M4 Mini base spec. Hmmm. Yes, we agree that render time is less important. We're just using it a metric to evaluate edit performance though. If a video can render back at speeds higher than 1x playback, then the system will be able to playback without judder or issues. Quoting render times is less subjective than us saying 'it played back nice for us'. We're so happy you found us! Thanks for taking the time to leave a nice comment. It's always nice when people get what we are trying to do.
@@team2films ah, I'd assumed you sent them the benchmark for promotional reasons. (there's a Figma project they've been using for the last 20 odd years so you should be a lock for Da Vinci Resolve benchmarking!) MaxTech are all about volume and they've been putting out multiple videos per day since M4 Mac mini release, so that possibly explains the crashing in M4 Pro. I hear you on having objective testing, the kinds of things I meant are still objectively testable, and yeah, if it plays back it plays back. Problem is for heavier workloads compromises in workflow are always required. Knowing that in advance of purchase can help folks. Especially if RAM capacity is a relevant issue. Many of the kinds of stress testing I was referring to was done in a YT video you cited in this video did several specific tests of Da Vinci Resolve UI under specific loads (not just Noise reduction either!) [his video isn't showing in my history].
The M4 Max MacBook Pro can be upgraded to 128GB RAM, but it costs $2,000 more. Does DaVinci Resolve Studio take advantage of all that RAM? Does extra RAM have any impact on render times?
No, we don't think so. Extra RAM might help with heavy fusion comps, but we doubt it would impact general performance drastically. The internal SSDs are so fast that page file access doesn't have a big performance hit. That means that a Mac system will likely perform well even with lower amounts of ram (within reason).
the main difference with having SOOO much am is that you'll be able to render multiple apps at once. say ae, DaVinci and then also work on photoshop files while your rendering multiple videos at once.
It would be nice to see a video around what kind of PC build would be best to have. I am currently using a laptop for all my edits. Whenever I try to do something with higher production (more VFX/fusion comps), I have to strategize my renders in a way to remove as much stress on the GPU. Like figuring out a way to do something that would take 10 nodes down to 5 nodes. I also feel limited, since once I reach a threshold, Resolve crashes. I am definitely saving money for something, but not sure what is the best bang for my buck on Windows/PC side.
Hello, that's a great idea. We'd love to do something on that in the future. In the meantime, it sounds like you've got a great approach, keeping your comps as efficient as possible. It's great to have you here, thanks for watching.
M2 Max Studio 64/1TB 38 core GPU A:310 A:296: A:294 Great video. I downloaded your render to test my M3 MBP 18/512 14 core GPU and my M2 Max Studio 64/1TB 38 core GPU. The M3 MBP repeatedly crashed at 16% and the M2 Max knocked it out in 4:56 repeatedly first on DR Studio 19.1 and then after update to 19.1.1. Thanks for giving us a tool to play with.
@@team2films I installed DR Studio 19.1.1 this morning on the MBP 18/512 14 core GPU with the following results: (#1.Success) On Battery Powered A 804. 731 A 722 728 A 729 731
M3 MBP 18/512 14 core GPU on 19.1.2 scored 779 on power without crashing. Still disappointing as I really thought the M3 Pro would do better. I'm curious to see where the MBP Max loaded is performing.
Here is my contribution: A 231, 222, 220 - Mac Studio M2 ULTRA (24CPU 60GPU) - 19.1 STUDIO B 247, 278, 260 - Mac Studio M2 ULTRA (24CPU 60GPU) - 19.1 STUDIO C 151, 151, 151 - Mac Studio M2 ULTRA (24CPU 60GPU) - 19.1 STUDIO A CRASHED! - Macbook Pro M3 PRO (11CPU 14GPU) - 19.1 STUDIO B 440 - Macbook Pro M3 PRO (11CPU 14GPU) - 19.1 STUDIO C 330 - Macbook Pro M3 PRO (11CPU 14GPU) - 19.1 STUDIO Nice to see my Mac Studio flying, and each time faster than the previous one! Worried about seeing my macbook pro crash multiple times, with and without power. Thanks as always for the info, resources and community!
Thank-you, really appreciate you posting those results. We have added them to the description if that's ok! You are currently top of the leader board, but the M4 Max is really nipping at your heels :) Do you know what clip your M3 Pro crashed on? We've seen a crash in the M4 Pro too. It's weird though because the stock M4 can run the benchmark just fine.
@@team2films Sure! Glad to be first. haha. Looking forward to see what M4 Ultra will be able to. Mi M3Pro crashed repeatedly at 16%, finishing the 5th clip and starting the 6th, the train one. Interesting to see other crashes on M4Pro... Lets see if we can find out more... Thanks so much! :)
@@Miguel_Garcia_Iraburu Gotcha. It's speedwarp that's doing it. If you change even to speedwarp faster I believe the crash will no longer happen. Hmmm, will investigate this.
@@Miguel_Garcia_Iraburu That's good to know, really appreciate you confirming that. We'll send info to Blackmagic about the bug. In the meantime, the score is interesting... but it's not a fair comparison now as speedwarp better is less computationally intensive. Thank-you thank-you though. Super helpful.
Thanks for this great video, Very good solution to check performance between different machines, here I leave my performance in the project and waiting to receive my M4 MAX in a few days... Benchmark A: AMD Ryzen 7 7840HS / 32 GB Ram / 1 x NVIDIA RTX 4060 LAPTOP 8GB / Windows 11 / 19.1 Studio render attempt: 22:38 minutes (1350 seconds)
Hello. Most SSDs will be sufficient. That's because over a certain speed, the cpu and gpu will be your performance bottleneck, not the ssd. So any typical 1000MB and above SSD is going to be just fine. Does that help?
Any modern computer can run resolve or fusion, performance will depend upon the kind of footage you are working with and the effects that you apply. The stats in our video should hopefully help you understand the comparative performance you can expect from particular MacBook models.
Hey guys! Great video! I am torn, I am considering the Mac Mini spec'd out to M4 Pro chip with the 64Gb ram. Do you reckon this machine will easily handle 4K video editing. (I am a videographer / editor by trade) This will be my main machine as my custom windows laptop is starting to throw the towl in. I can't justify the crazy price increase to get the M4 Max MBP. Any guidance would be much appreciated!
Hello. Give or take.... The M4 Pro is twice the power of the base M4. The M4 Max is twice the power of the M4 Pro. That's a difficult question to answer. You can edit 4k or UHD video on a base spec M4. But it's also possible to edit 12k video on a base spec M4. Performance will decrease as you add effects and colour grades to your footage. It's possible to push an M4 Max to the point where it can't play back in real-time. And of course it's possible to overwhelm a base M4 too. We'd recommend downloading the benchmark. Check out the performance of your computer. Compare it against the specs for the M4 chips shown in the video. It will tell you how much better it will perform than your computer. I hope that makes sense! The short answer is, yes you'll love the M4 Pro chip. But it's hard to tell you what it will do without knowing more about the kind of editing you do! Thanks so much for watching, and please let us know if you have further questions.
I am pondering the M1 to M4 upgrade. The rendering tests are not that important to me. My videos are short and rendering time is less important to me than timeline performance. Does the M1 chip stall when you edit on your timeline? What happens if you apply a noise reduction node in the color panel? Does playback speed stay at 24/30fps? These are the important DR tasks that are of value to me and are harder to evaluate. FYI… I wanted to test the speed of the M4 Max Thunderbolt 5 USB ports at an Apple Store with various SSDs I work from. But Apple will not allow speed testing from store configured Macs 😢.
Hello, thanks for watching. Deliver page performance is representative of edit page performance. For example if a video can render above 1x speed then it will play at 1x speed in the edit or color page. Like you said edit performance is harder to evaluate. We avoided edit page tests, because it gets really subjective. For example, can a clip play with Denise applied? Yes.... but what's the resolution of the clip, what type of denoise is being used, how many other adjustments have been applied to the clip. Here's what we'd suggest. If you switch Resolve to 'Show All Video Frames' look at the playback FPS. Then you can use our benchmarks to work out the relative performance of the M4 and whether it will increase your performance above that real-time playback threshold. We're hoping to get hold of Thunderbolt 5 storage soon! Thanks for watching and commenting. Appreciate you!
Thanks for the benchmark. My results for Bench A A 282 - Ryzen 9 5900x - 128GB DDR4 3600mhz - AMD 7900 XTX - 19.1 Studio. Not too bad for an almost 4 year old CPU, the GPU is only about 2 years old though. I have a question if you have the time :). I don't care quite as much about render times like in this benchmark as I do timeline playback, or what I'm able to playback without being forced to generate render cache. With Benchmark A and my test system, I have to set render cache to "smart" in order to playback in the timeline reliably, without resolve crashing. If I don't generate render cache (and wait) It will playback ok for a bit in the timeline...but eventually it will crash if render cache is not generated first. I've been looking at upgrading my system (everything but GPU), then the M4 came out and I've been comparing results. How is the timeline playback/stability on the mac without generating render cache in Benchmark A? Or do you have to generate render cache on the mac as well for speed/stability in the timeline? My system never crashes during a render...only in timeline playback/scrubbing after a while with no render cache.
Hello, thanks so much for sharing your results. Sorry you are having problems with playback too. Your system should be able to playback without crashing (even if your system is old or not able to playback in realtime). We have older and slower systems that playback without crashing. Also, please note, we cannot play this timeline back in realtime, even on an M4 Max or M2 Ultra. I don’t think there are presently any systems that can play this timeline back. Basically you’d need to see a render time that’s lower than the length of the timeline. Render caching will always be an important part of workflows, as even though computers grow in speed, the complexity of the tools we use will likely grow too. So to answer your question, playback it’s stable for us but for speed, caching is required. Yes, we agree that timeline performance is more important than render speed. Our benchmark is designed to reveal timeline performance, we just use render speed as the metric for measuring that. The workload of rendering is almost identical to the timeline workflow with the exception of encoding. As h264 encoding happens on hardware for most people, then the impact of encoding on performance is very very low. Hope that all makes sense, please ask as many questions as you want. It’s good to have you here.
@@team2films Thank you for taking the time to answer! That makes sense about the render time vs. playback and now I think I understand where you are coming from with that as a test. I guess another "hack" I could use in edits would be to cut my video first, then add the layers of fusion/animations after all cutting is complete, and I could even disable/turn off fusion/color grading until I'm done "editing"...then turn on render cache toward the end of the edit to see it all before render. What I'm trying to do is prevent the annoying wait times of changing something/editing something, and having cache rebuild and having to wait in timeline on each/every edit. Thank you again for the tips/insight. I don't think it would really make much sense for me to upgrade much other than maybe CPU/RAM speed as that seems to be where my system is lacking. My GPU seems to keep up in your test. When I render your project, my GPU is almost always 80% or lower utilization which tells me that my CPU/RAM isn't feeding it data fast enough :). Something the Mac does well is the unified memory is shared between gpu/cpu, and it's generally faster than what you find on a PC unless you really buy high end memory and tweak it.
@@TheCarGuyOnline Yeah, that's a solid idea (compartmentalising stuff on different layers to make it easier to turn on and off). Don't forget the simple button above the viewer that turns off color page and fusion adjustments too. That's useful for when you want to focus on editorial changes so want the best possible timeline performance. Leaving render cache on as you go is a good idea too, because it only works in the background when you are not doing anything else.
I'm also having similar stability/quirky issues with my system running a 5950X with 7900XTX. T2F added my A time of 263 to the video description but it took me a few tries to get 3 full back to back renders without any freezing/crashing during rendering. I also sometimes have issues with timeline/playback ability and have to turn on Smart Cache. Sometimes even magic mask doesn't work properly. I have to save the project, close Davinci and reopen it again, then it will work. I don't know if its an issue with the AMD driver and the 7900XTX or what, but I find my system to be very inconsistent in Davinci Resolve where I too am thinking about getting an M4 or M4 Pro mini. It's a shame though because this 7900XTX renders this benchmark very well for what it is and its cost compared to something like the RTX 4090.
Just finished running it on my M4 Pro Mac Mini and I have a video posting tomorrow morning on my channel showing it and the results. Here is my report from that test. A 565 - M4 Pro (12CPU 16GPU, 24GB RAM) - 19.1 Studio
I’m rendering a 4k timeline on my M4 Max 16core CPU 40core GPU 48GB memory and I’m getting 4 -6 FPS and the render constantly crashes around 30 percent. I thought these laptops can handle anything we throw at them. I’m using NR Dehancer and another 8 nodes
Hello, thanks so much for watching. Sorry you are having problems. The M4 Max is incredibly capable. It’s hard to pass judgement on your particular circumstances without knowing more about the timeline you are processing. What I can say is that sadly no computer can handle ‘anything’ (or everything) you throw at them. The relative performance figures will give you an idea of how much slower your performance might be on another computer. The M4 Max is very close to the top of the score board.
Thanks for this video, it's perfect! I really thought my computer has held up well but holy moly these numbers are atrocious compared to everything else. I have a windows ryzen 9 3900x with 64gb DDR4 RAM and RTX 2080. It took my computer (drum roll please...) 29 mins. on the dot. 1740 secs. (for the A benchmark).
Oh no! Firstly, thanks for watching. Second, sorry about the score. We don’t know enough about the components you are using to say where the bottleneck is, but as Resolve relies primarily upon the GPU for rendering, we’d imagine it’s the GPU. Sorry though! Didn’t want to make anyone feel bad 😞. Here’s the important thing though… if your computer is doing what you need, then it’s good enough! Thanks for watching, it’s good to have you here.
@@team2films Update! I purchased an M4 MBP on Black Friday (I've been holding out on getting a laptop and can write it off for work anyways). 16" M4 Max 16c/40c, 48gb RAM. Here are the new numbers: Benchmark A- 4:12, 3:52, 3:50. And just because I was curious, I ran the same benchmark with the media stored on an external Samsung T5, and exported to the T5 as well. Interestingly, the numbers were about the same which is great! 4:03, 4:02
@@Megalukeification Whoop Whoop! That's an improvement! I love how the numbers speak for themselves. Yeah, we wouldn't expect to see a difference in score between the internal SSD and an external. There's fastest, then there's fast enough. In these tests, disk access speed is not the bottleneck. Most ssds will feed the footage faster than the GPU can process it, especially with those effects. Awesome stuff, thanks for sharing!
I am having using M1 Pro 14 INCHES MacBook Pro and I wanted to take another laptop as a secondary device which one do you suggest ? I am a wedding photography and I mainly use Final Cut Pro and davinci resolve what is your suggest value for money
The comparisons in the video will give you an idea of the performance you can expect from the new models. We would suggest comparing that against their cost and making a decision about what you can reasonably afford.
A223 - i9-13900KF 64gb (2 x 32gb) DDR5 6400mhz RTX 4090 - 19.1.2 Studio Wish I Bought Mac Book M2 Pro or M1 Max at time of building, needing portability now, and have worked on friends Mac systems, feel much better compatibility and less problems with applications like premiere, pro tools, and davinci in specific tasks, but 1 thing my build might have better render speeds but, Mac runs amazing with 4K footage in my experience without proxy, 😢
Glad the bios update fixed things. One of the advantages that Mac systems have is how prolific they are. Windows is very difficult to develop for because there are so many iterations of hardware. The most common system that DaVinci or Premiere runs on will be Mac silicon, so naturally it will always get more attention from devs. Just our thoughts! We love both windows and Mac though :)
I've tested your project on my machine: Intel(R) Core(TM) i7-14700K 3.40 GHz / NVIDIA GeForce RTX 4090, driver version: Nvidia Studio Driver 560.70 / G.Skill 64 GB DDR5-6000 / Windows 11 Pro 23H2 / Davinci Resolve Studio 19.0.3 build 5 Benchmark A: 238/231/224 But I also tried to use the voukoder utility to access custom settings for encoding, I've chosen Nvidia NVENC H264, the settings were: multipass=disabled preset=slow, profile=high qp=18 rc=constqp tune=hq, the results were almost the same: Benchmark A: 247/208/239 And finally I've tried to run it with Nvidia NVENC HEVC H265 encoder in voukoder, the settings were: multipass=disabled preset=slow profile=main10 qp=18 rc=constqp tier=high tune=hq, the results were much faster: Benchmark A: 215/183/188
That's cool to know that there are some performance gains with h265. I think it's really interesting that subsequent renders speed up too. I think it might be something to do with fan curves that allow thermal throttling before spinning up. I've added your h264 results to the scoreboard!
Ah nooo! Those final 2019 intel MacBook were not good. We know because we have one. The chipset has a lot of power, but it can't run under sustained load without thermal throttling. That's the problem.
So glad I came across your channel! It's refreshing see a TRULY pro approach to gear/software/video editing. No "influencer" hype or yelling, antics and BS. This is mature and respectful to the viewer. Thank you.
Thank-you. We appreciate that, we try to be as objective as possible and present the relevant information. Please let us know in the comments if you ever see anything we can improve on.
@@Moravia90s 💯
Yeah my ears so happy for this channel
Totally agree with @moravia90s. Thank you for the refreshing style.
The work put into creating and animating these videos is just mindblowing considering it's free content. Thank you for your work and patience!
Thank-you so much. We're grateful to Audiio for helping support content production like this. Thanks to you too for watching and taking the time to leave a comment.
That video was what we'd expect for real professional advices on real life performances. Thanks
Thanks, that's appreciated.
Not only do I enjoy your content, but I gotta say, your voice is so calming it feels like I’m listening to ASMR!
Aww thanks! What a nice compliment. Thank-you
@@team2films thanks to you 🙏
You all are simply the best channel there is for me. I am such a fan of your content and professionalism. Thank you so much for making everything so clear.
Thanks for being here. We appreciate everyone's support.
Thank you for the awesome video!!!! After 13 years with a 2012 Mac mini I upgraded to the base M4 Mini. It's an excellent machine. I am impressed with the render times I get on my videos using DaVinci Resolve- Worth every penny. The price was only $100 more than the Mac mini I purchased in 2012. I took the 1TB SSD out of the old Mac mini, formatted it and put it in an external drive housing and away we go. I expect this machine will last 10 years as well.
That’s awesome. Glad you’re happy with your purchase
To whom it may concern.
Benchmark A: AMD Threadripper Pro 12 Core / 128 GB Ram / 2 x RTX A4000 16GB / Windows 11 / 19.1 Studio
First render attempt: 6:47 minutes / second: 6:25 minutes / third: 6:20 minutes
All tests run from SATA SSD to (different channel) SATA SSD (onboard SATA 6G)
Thanks for the free Benchmark 🙂
Appreciate you. Added to the description. Thanks
Thanks for this great video
Benchmark A: Intel(R) Core(TM) i9-14900K / 64 GB Ram / 1 x Radeon RX 6900-XT SE 16GB / Windows 11 / 19.1 Studio First render attempt: 5:53 minutes / second: 6:06 minutes / third: 5:47 minutes.
Benchmark A Studio
AMD 5900x / 64GB Ram / rtx4090
1st run: 4:09 mins 2nd: 4:13 mins 3rd: 4:12 mins
Very helpful benchmark, thank you!
Got to love the am4 platform!
Thanks for the benchmark. My results for Bench A
A 671 - Mac Pro mid 2010 - 128GB DDR3 1333mhz - 2 x 3.46 GHz Intel Xeon 6 cores - AMD RX 6800 XT - 19.1.1 Studio. Not bad for a 14 years old computer.
We have to say, that is pretty excellent for a computer of that age. Well done, and thanks for sharing.
Thanks for great benchmark!
Intel 13900KS / RAM 32GB DDR5 / RTX 3090 / Win 10 / 19.1 Studio
All 3 runs: 251, 252, 250
That's a great score. Thanks for sharing. Added to the description.
Super professional video, what a great watch. No frills and very realistic
Thank-you! Frill-less is our aim ☺️
Thank you, guys, for the valuable content on your channel. Keep up this great work.
A 247 - M1 Ultra (20CPU 60GPU) - 19.1 STUDIO
B 131 - M1 Ultra (20CPU 60GPU) - 19.1 STUDIO
C 184 - M1 Ultra (20CPU 60GPU) - 19.1 STUDIO
These types of videos and this channel are a GODSEND. You all are amazing. Thank you and great job as always on the flow of these vids.
Thanks so much, we’re glad the videos are helpful. If there’s ever anything we can do to make future videos better please say.
I bought a second hand M1 Max 10 Core CPU, 64GB RAM and 32 Core GPU. Great value and definitely fast enough for my photo and video editing needs.
You two have a gift. Very helpful, always. Thank you both and Merry Christmas!
Thanks so much Norman! That's appreciated.
Thank you for your helpful instructional videos. I am new to video editing and teaching myself your videos are a great help. Since I am just learning to use DaVinci Resolve I am still using the free version so here are my results from a new MAC Mini M4 Pro:
B 294 - M4 Pro (12CPU 16GPU) - 19.1.2 build 3
Thanks Gary, great to have you here. Please check out our beginners tutorial for Resolve, we hope it's helpful for you. ua-cam.com/video/znBHzeXpsUw/v-deo.html
Thanks for great benchmark!
Benchmark A: Ryzen 7 7700 / 32GB ram / RTX 3060 12GB / Win 11 / 19.1 Studio
All 3 runs: 524, 553, 530
I guess it's slightly faster or on par with the new Mac Mini with M4 Pro base config (16 gpu). it's really interesting how 20 gpu version would perform!
Oooo😮 this is what I was looking for! I have a m2 pro with 16gigs. I am hitting the ram limit a lot and debating on which way to go.
Benchmark A Studio - 07:39 - M4 Pro Mini (12CPU 20GPU, 48GB RAM) 19.1 STUDIO
As someone with an M1 Max who's new M4 Max is arriving on Monday, thank you so much for saving me the hassle of figuring out just how much faster Resolve is going to run!!
(and also how resellable an M1 Max is 😆)
how much are you selling your m1 max for? i've got a m1 max studio with 2 tb of storage and 64 gb of ram
@@astromoosie looking for about £2000-2300 (about $2700). It was a spec’d to the max one tho. 8tb with 64gb ram so that will hopefully bump the price up a little
@ thank you for replying so quickly. This gives me a better idea of what I can get mine for. Still not sure if I even need to upgrade. It still handles everything extremely quickly.
Only one more day till it arrives! Let us know what you think when it comes 🤪
World class content! Smart and wise at the same time. Wish you guys the best.
That's really kind, thank-you.
Love the Homage to ACORN - no one ever has mentioned this, the respect is appreciated.
Yeah, I think the name sneakily got changed to Advanced RISC Machine, but its heritage is Acorn! Loved my Acorn computers. Thanks for watching and commenting.
The way that u presented was incredibly relaxed and easy to understand. Great job.
You are welcome. Thank-you.
Thanks for putting this benchmark together, what a great idea!
Here's my Desktop PC's benchmark results:
A 325 - Ryzen 9 5900X, RTX 3090, 128GB DDR4 RAM - 19.1 STUDIO
All 3 runs: 361, 325, 334
That's so awesome Kaur. Thanks you're on the scoreboard. Quite high too. I'd really love to see something beat the M4 Max. No one yet though :( I thought a 4090 could do it, but the one we've got on the board didn't make it.
@@team2films Yeah, it's starting to look like a no-brainer for colorists who don't need the best of the best (aka having a custom Desktop with 4 GPUs)
You are an exceptional presenter, very crisp, sort of soothing - well done, and wishing you best wishes in your career
Thanks Aaron!
Wow! Your video was recommended and I’m glad it did. Great content and helpful information and real world information, no useless benchmarks. Subscribed directly.
Thank you
We are glad you found us. Glad the real-world information is proving useful. Thanks for subscribing, it’s great to have you here.
Just did a test real life test on the m4 vs m1 using different clips and codec and raw from sony and canon. And more importantly playback and editing not just render. I think it’s more useful seeing actual user experience than render time.
Hello! Yes, edit performance is more important. As mentioned in the video, render time will give an indication of edit page performance as they will be using the same internal engine. I.e. if it can render back in realtime that means it can play back in realtime. The reason we're using render time to benchmark, is because it's easier to extract statistics from a render. That's why render performance can be considered an indication of real-world performance.
Hope that helps.
@@team2films I'm not sure actually that render time reflects application experience at all though. I have an 11700k and a RTX 4080 so my render times are fantastic, but timeline performance is horrendous. Picked up an M4 Pro to see if it makes timeline performance enjoyable for me, and I don't care about the render time at all.
@@OnceinaSixSide Hello, thanks for taking the time to comment, we appreciate that! Yes, of course there are other factors that affect edit page performance that can't be accounted for in the render benchmarks. For example drive latency will have less impact on rendering as it's more likely to be a continuous read. However it will affect random operations in the timeline where data is not read sequentially. As there's no simple method of programatically benchmarking edit page performance we decided to use render time as an approximation. Ideally, a suite of benchmarks could be designed to test and identify the various bottlenecks that might affect system performance.
It can be really subjective describing edit page performance. 'This plays smoothly for me' is dependent on lots of different factors. Hopefully, the render times gives an objective analysis of performance in relation to other systems.
For yourself it might be worth turning on caching. It sounds like you have a capable system, there could be a single component that is causing a bottleneck for you?
@@team2films Totally hear you on the subjective experience, and I wish there were a better way to benchmark timeline performance as well.
I suspect something was off with my system as I just finished upgrading the mobo, CPU, and ram to x870, Ryzen 9 9900x, and 64gb DDR5 and the timeline performance is now similarly smooth to the M4 Pro I was testing.
This is without any sort of Intel quicksync now as well and just pure CPU bruteforce transcode + whatever the 4080 does (which I dunno if its much in this regard).
I'm really happy with this performance now and what's interesting is the M4 Pro definitely still had a smoother timeline experience. I tested some projects back and forth, and I was getting very smooth playback on some complex composites whereas the newly upgraded PC would hang every now and then.
Renders though on the PC are significantly quicker though thanks to the 4080, as well as operations like Magic Mask which seem to average around 30fps versus the M4's 1-3fps.
@@OnceinaSixSide Thanks, we appreciate your balanced subjective. Hopefully we can develop a better timeline benchmark in the future. I think with timeline performance, latency might be having an impact. i.e. rendering highlights sustained performance well (once the computer is underway) but latency affects the responsiveness of the play button in the timeline. The M4 chipset architecture lends themselves towards being incredibly responsive.
Not sure about the magickmask operations. That could be taking advantage of the 4080's better AI capabilities?
If you have any future observations or ideas, please share them. It's great to have you here, thanks for taking the time to leave a comment.
The presentation skills are outstanding
Thanks. That’s kind of you to say.
Finally someone listen! Benching the same project! Congrats!
Hope its useful to everyone
@@team2films it's what I have been looking for for over a year!
@team2films i recently switched to mac and got the m2 pro with 16gigs. I am hitting the memory threshold all the time from multitasking, so I am looking at what direction to go. I have always been an AMD fanboy, but understand what apple brings. I find it so convenient on pc to upgrade as needed, but I really love the small footprint of what apple brings. It just sucks that I have to spend 1500 plus just to upgrade the ram at this point...
I ran a test in a second system, Benchmark A result.
A 258 - Intel i7 13700K - 32GB DDR5 6000mhz - AMD 7900 XTX - 19.1 Studio
Slightly faster than my Ryzen 9 5900x with the same video card, and a lot less memory in the Intel system. The 13700k has much better single core performance and higher boost clocks on P cores than the 5900x, along with faster ram shaves off 24 seconds vs the 5900x on ddr4.
Added to the scoreboard.
A 1484 - M2 Air (8CPU 10GPU 16GB/512GB) - 19.1.2 Studio
Thanks for the video and the benchmark project! I have been thinking about the base 16" Macbook so this is helpful to see what sort of improvements I might see. Thanks again!
So glad it was helpful.
Ahhh, the Acorn Atom 2k. What an exciting machine in its day!
Such fond memories. Had an Archimedes and a RISC PC too. Loved them
My laptop is M1Max，thanks for sharing! You are amazing :)
Glad it has been helpful :)
Thanks for the value, this is called as useful content, with no hype and running for views. I saw in a video that a base M4 with more 64GB RAM outperforms M4 Pro with base 24GB RAM, is it true?
We doubt it. The CPU and GPU have a bigger impact than the amount of ram. Thanks so much for watching.
Thanks for the clarification guys.
Scaling linearly with the number of gpu cores, is an amazing testament to the architecture of the chip… the scheduler apparently is really well-implemented, as it can divide the tasks so linearly among fewer and more gpu cores.
Yeah, that's true. Kudos to the BM dev team for making Resolve so well optimised to take advantage of those cores too. Thanks for watching.
Thanks for the great share and video! Love your channel. I did try to benchmark my M4 ipad pro, with no luck. It had trouble decoding the RED files hence failed rendering. Tested my home workstation (win 11) and results: A 378 - Ryzen 9 5950X, 128Gb Ram, RTX 3090 - 19.1 Studio
Finally had a chance to benchmark my workstation: A231. i9 13900KS, 64Gb Ram, RTX 4090, Win 11 19.1 Studio. Full Watercooled loop incl GPU. Dissapointed it did not top the charts ;) With each successive render times went down. (ran 3 times) The good thing about watercooling is GPU never went above 44deg, barely raised a sweat! Could render this out all day..
Yeah, we’ve been disappointed with the 4090’s performance and impressed by the M4 Max. A laptop and top end GPU are basically equal in the test. There are some minor minor performance gains from different codecs… but nothing has touched the M2 Ultra. I guess Resolve benefits from the sheer number of cores or the unified memory. This is the great thing about real world tests. Synthetic benchmarks wouldn’t reveal that.
You still have an amazing computer! Thanks for running the test and leaving a comment. Appreciate having you here.
Benchmark A 318 - Macpro 2019 3.3ghz 12core Xeon W, 2 x Vega II, 240GB, 1TB, 19.1, OS 14.7
Perfect video for me! I never upgraded the M1!
Awesome! It's a great computer.
m4 pro macbook pro 16'': render 1 :08.55, render 2 :08.43, Render 3 : 08.18
Davinci studio 19.1.1
That's awesome. Thanks for sharing.
Quite happy with my M1 Ultra studio… Will be upgrading when the M4 ultra comes to town. I want as much power as I can possibly get.
Awesome!
Intel i9-13900KS / NVIDIA 4090 / 64GB / RESOLVE STUDIO 19.1.3
BENCHMARK A: 248 sec
BENCHMARK B: 173 sec
BENCHMARK C: 135 sec
If my 5090 pre-order arrives anytime soon, I will test that as they have more encoders/decoders than the 4090 which I hear makes quite a difference with h264/5.
Thanks to this channel. Now, i'm going with mac mini m4 instead of going with 2nd hand m2/m1 macbooks.
Thanks for this great video - I love the relaxed atmosphere!
One thing to note is that for lighter workflows (less effects), the GPU has much less impact and the encoders are often the limiting factor. So for someone working with 4k at max and moderate amount off effects, a costly MAX might bring less benefit - in that case a base M4 might be as good or even better than a M1 MAX.
Yep, that's a great point. And it also underpins the fact that you don't need the best, you just need good enough. Thanks for watching, commenting and sharing your thoughts. It's appreciated.
im so glad i clicked on one of your videos contents are S-tier🔥 Subscribed!
just a thing tho maybe its personal choice but i was bit sceptical about blue and white logo i think a vibrant logo would be better maybe like resolve's logo
Thanks so much, we appreciate all feedback! Who knows, maybe we'll freshen it up at some point in the future. It's great to have you here.
A great video, and a useful benchmark.
Some results for a 9800X3D / 32GB / 7900 XTX on Benchmark A with DaVinci Resolve Studio: 256s, 252s, 255s
Results for Benchmark C: 252s, 243s, 247s
Thanks for sharing, that’s an excellent score.
My M1 is still busting through and ready for the upgrade to the m4 for my next big project. But with that said, I'm still very impressed with the M1 for basic projects and knowing how to use proxies and such
Yeah, the M1 is pretty capable.
Thank you for the insightful video and all the effort into pitting it together! It was very pleasing to watch.
We're happy it's going down well. Thanks for taking the time to leave a comment. We always appreciate that.
Benchmark A Studio - 04:36 - M3 MAX (16CPU 40GPU, 128GB RAM) 19.1 STUDIO
Benchmark C CPU - 03:38 - M3 MAX (16CPU 40GPU, 128GB RAM) 19.1 STUDIO
That's an awesome score. Thanks for sharing.
Nice comparisons. Perhaps I missed it, I would love to see the M4 Macbook and M4 iPad on Resolve performance comparisons.
Hello! Thanks so much. Sorry, the M4 iPad wasn't featured in these tests as it runs a slightly different version of Resolve. It can't read all the codecs, can't process the same resolutions, etc.We really wanted it to be a high-end benchmark and include those codecs and high resolutions. We'll see if we can add a universal iPad benchmark at some point in the future.
@@team2films thank you
Hey I just want to say that I am a big fan of your videos! They have been incredibly helpful.
Thanks Andy, we appreciate you taking the time to say that. Let us know if you have any ideas for future videos.
A 10min34sec (634): - HP Zbook Fury G10 13th Gen Intel i9-13950HX / 128GB ram / RTX 4000 Ada Laptop GPU 20GB / Win 11 / 19.1 Studio
Interesting result for the old workstation Xeon E5 1650v3, 256GB DDR4-2400, m.2 SSD with a newish Nvidia RTX3090:
Benchmark A (Studio): 455s (7:35)
Benchmark C (CPU): 680s (11:20)
He's that not too bad! Thanks for sharing.
Is there a video on what setting we should have in preference thanks 🙏
Hello. There’s no special settings to get extra performance. Just using the default stock settings. Thanks for watching
hahaahah don't even think about skipping to the conclusion , it's a trap :) well done guys ( I love the channel for sure )
Thanks Bernard, nice to have you here.
@@team2films hi again, I did sent you an email on your site the same day but I never did get an answer , what version of Davinci are you covering in the course ? Thank you
@@BernardBertrand Reply sent Bernard. Thanks for the reminder. Let us know if it doesn't come through.
my results:
A 445 - M1 Max 32GB (MacBook Pro 16") - 19.1 STUDIO - macOS 14.7.1
Loved the PYXIS footage... And the content as well !
Thanks! We added that little logo last minute to the video. I think we might do that with future videos too, put some info about what we are using to shoot it at the bottom
Thank you for all the effort to create this!
Our pleasure, we're glad people are enjoying it.
Render times are not issue for me. I want the editing to be as lag free as possible.
Yes, that’s true. Remember render times are a reflection of what edit page performance is like. Of course. More performance than you need is not helpful, so if you currently have more than you need then you are fine!
An amazing job, thank you for taking care of the community and looking out for the consumer! My data:
Benchmark A: Mac Studio M2 Max / REFURBISHED / 12CPU 30 GPU 32 GB // - 19.1 Studio
RENDER ATTEMPTS: 418 / 398 /396
Appreciate you sharing that. It's interesting that it gets faster with renders 2 and 3. I think that's a fan curve thing, where it is slow to turn the fan up and lets the chip thermal throttle. By 2 and 3 the fan is on constantly so it doesn't thermal throttle. Result added.
As a complete newbie trying to decide between a PC system and a Mac, I was wondering if you have any videos comparing the Mac Mini M4 Pro to any NVIDIA GPU?
No, sorry we don't have any videos like that. If you look in the description though you can see the scores of many PC systems compared to Macs
As always super solid stuff. Thank you.
Thanks George
A 399 - M4 MacMini Pro (14cpu/20gpu/64ram)
One suggestion/caveat you might want to mention. For the folks that have those NVidia cards, 4080 and 4090 variants, they would probably see improvements over the Macs in your test and probably be at the top of the list (guessing) if they chose the AV1 dual encoding option. When you are dealing with 4k and higher, choosing mp4/AV1 will allow it to likely outperform even the mac ultra in this case....just throwing that out there. This test being h.264 and forcing a single encoder does seem to limit what the NVidia 4080/4090 series cards are capable of in the real world. If anyone owns a 4080/4090 and use 4k and higher, I can't imagine why you would not want to use AV1/Faster/High quality settings.
That's a good point, that some systems will perform better with particular loads. One challenge of making a universal benchmark is to use tasks that will run across all the systems that you might benchmark- h264 seems like a good candidate for that. However, the Benchmark is so computationally intensive, that the encoder doesn't have the opportunity to become a bottleneck. I ran Benchmark A on my computer rendering to h264 and then to an uncompressed avi. The results, h264: 4:10, uncompressed avi: 4:05. That margin is narrow enough to not be conclusive, as there is variance when running the same test.
The point is, the encoder has a negligible effect on the render. We're not trying to cheat windows computers. Any modern editing computer should be able to render h264 effectively. At best you could say that a 4090 is better at encoding av1. But its computational performance in Resolve is still falling short of an M2 Ultra or at the very least, matching an M2 Ultra.
Does that seem fair?
@@team2films I don't think you are trying to cheat windows computers, on the contrary it's still really impressive what Apple is doing with such low power usage. I just wanted to point out that if I owned a 4080/4090 or card with dual encoders and I was dealing with 4k and higher, I would definitely use AV1 and never use h.264, I do still use h265 for UA-cam uploads but I don't have the fancy 4080/4090. The dual encoders will reduce the time in this case, it's still encoding a fair amount of video and I think it would be neat to see both in their best form, but I understand there's really no way to produce some single benchmark for everyone/every system...that's impossible. This gives everyone some idea of what they can expect in given systems. I really appreciate your aggregating the list in the description, it has helped me....steer away from some upgrades.
I am still curious to see how much it would lower the time with the AV1 dual encode and I am speculating, but from what I have seen in other tests from MrAlexTech (not my own) the dual encoder AV1 drastically reduced export times, but he was running tests on longer videos as examples, and probably not as complex of workflow as the bench here.
Still the macs are crazy good considering what they are, a laptop that stays relatively cool and draws almost no power in comparisons is just...crazy that they even compete on any level with the fastest CPUs/GPUs on the market. Sadly, Apple knows this and their price reflects it. I'm sure the new Ultras will scale just the same like you mentioned, and destroy everything here, pending NVidias 50 series offerings.
@@TheCarGuyOnline You're awesome. We're definitely on the same page. It's great to have you here. Thanks for the engaging conversation too.
Interesting! As eluded to, the codec format did improve render times on the 4090. Without (MOV) time was 231. Set Format to MP4, Type to Nvidia and render time dropped to 222 :) (on par with M2 Ultra)
@@TonkaFJ40 What is your time using AV1? the 4090 can dual encode a 4k or higher video with AV1 which is..quite nice. Love to hear if it improves, or if that 231 was dual encode.
Great video as usual. I currently am using the M1 Max 16" MBP, and it's just the best computer I have ever had. I am guessing I can probably get work to upgrade me to the m4 max, but I am not sure if this is the year or wait and see what the m5 has.
One other really cool use case I have been exploring is getting a base spec m4 Mac Mini and using it for remote rendering. It won't render as fast, but I don't typically wait around while stuff is rendering, so it could be a pretty cost effective way to get more work time out of the laptop (because it is a bit annoying that you can't do other stuff while stuff is rendering), while I leave stuff rendering on the other computer. I think it will also make a killer home video server...just a thought.
Keep up the great work! I hope you release some more courses soon!
Hi Max, thanks so much for your support. The M1 Max is amazing. If you're not sure about upgrading, it's probably worth waiting to see what the M5 brings. If it's a work expense though.... 🤣
Yeah, the M4 base Mini as a remote renderer is an awesome idea, I think we might experiment with that. $600?!?!? It's such insane value.
And yes, we're talking about what courses we will make next.
@@team2films yeah exactly, at $600 it's just like begging for uses to be found.
I think in some cases it is beneficial to scale wider not higher.
anyway, can't wait to see what new courses you come out with. I hope you teach how to make such great educational videos...
@@MaximoJoshua Thanks Max, we have some fun stuff that we are planning, here's a preview of what's ahead (hopefully): Cine 12k, 35mm film, Storage, OLED
@@team2films those look like awesome topics, very interested in 35mm film (will this show scanning or working with scans?). Any topic you guys cover, you find a way to make it interesting and comprehensive. Thanks again for the benchmark, should prove very useful in the future.
Turn the number into a score is very easy to understand, thanks 🙏🙏
I subscribe because of this.
That's so nice to hear. It makes us really happy when people pick up on the little details. Great to have you here.
I can confirm A 396 - M1 Max (10 CPU 32 GPU 64 GB RAM). I got the same rusult.
Awesome, glad to see it's consistent.
Really clean presentation with good visuals! On my Mac Pro running Sequoia beta 3, the A Studio benchmark doesn't run quite clean and I've had two pretty hard crashes (which never happens). 2-3 clips seem to lock everything up (Relight and Depth map are the worst offenders). The B-version runs through as expected as such:
B176-12 core Mac Pro 2019, two W6800X Duos, 240GB RAM.
(anecdotally, if I turn off the effects Relight and Depth Map, but leave the clips as they are apart from that, the A-version renders in 177 seconds)
Since this Mac Pro has a lot of GPU grunt, most "normal" projects run super smooth with typical effects like NR, Film simulation and so on.
Thanks so much for watching, we really appreciate that. And thanks for your kind compliments too.
That's interesting. Sorry you're having problems. The Benchmark is running well across a variety of computers. Even old computers run it fine, but just take longer. It might have highlighted some issues with your system and Resolve installation.
this channel is new to me. looks like your Da Vinci Resolve benchmark is the one MaxTech were using in their M4 Pro vs M4 Max video, , yes it was. the #algorithm never sleeps. nice production style/values team (preferable to most of the Apple Mac + Video Production shouty "content" creators already!).
NB Rendering time is the least of most editors concerns, and even for many compositors, it's more about what's achievable in the GUI without shuddering/down-sizing playback & previews/pre-rendering sub-comps/switching off layers that is where the time savings are.
Ah, that's so cool! We just went and checked out their video. It's a shame the test crashed on the M4 Pro as we've seen it work on the M4 Mini base spec. Hmmm.
Yes, we agree that render time is less important. We're just using it a metric to evaluate edit performance though. If a video can render back at speeds higher than 1x playback, then the system will be able to playback without judder or issues. Quoting render times is less subjective than us saying 'it played back nice for us'.
We're so happy you found us! Thanks for taking the time to leave a nice comment. It's always nice when people get what we are trying to do.
@@team2films ah, I'd assumed you sent them the benchmark for promotional reasons. (there's a Figma project they've been using for the last 20 odd years so you should be a lock for Da Vinci Resolve benchmarking!)
MaxTech are all about volume and they've been putting out multiple videos per day since M4 Mac mini release, so that possibly explains the crashing in M4 Pro.
I hear you on having objective testing, the kinds of things I meant are still objectively testable, and yeah, if it plays back it plays back. Problem is for heavier workloads compromises in workflow are always required. Knowing that in advance of purchase can help folks. Especially if RAM capacity is a relevant issue. Many of the kinds of stress testing I was referring to was done in a YT video you cited in this video did several specific tests of Da Vinci Resolve UI under specific loads (not just Noise reduction either!) [his video isn't showing in my history].
One thing to notice is the CPU and GPU are mostly used around 40% range, and only occasionaly hit 90%, even on a mid range system.
CPU and GPU performance metrics only offer a glimpse of the system utilisation. The actual performance utilisation will be more complex.
The M4 Max MacBook Pro can be upgraded to 128GB RAM, but it costs $2,000 more. Does DaVinci Resolve Studio take advantage of all that RAM? Does extra RAM have any impact on render times?
No, we don't think so. Extra RAM might help with heavy fusion comps, but we doubt it would impact general performance drastically. The internal SSDs are so fast that page file access doesn't have a big performance hit. That means that a Mac system will likely perform well even with lower amounts of ram (within reason).
the main difference with having SOOO much am is that you'll be able to render multiple apps at once. say ae, DaVinci and then also work on photoshop files while your rendering multiple videos at once.
Oh look, my top favorite UA-cam channel ever has dropped a new video! 😍
Hehehe! We appreciate you watching. Thank-you so much.
3rd run:
A 660 Ryzen 5 5600 (32GB RAM) RTX 3060 19.1 Studio
Awesomee, that's a good result!
A 1112 - M2 Pro (10CPU 16GPU) - 19.1 STUDIO
long have I watched a video that is as honest as this
Thanks so much. We appreciate you saying that.
It would be nice to see a video around what kind of PC build would be best to have. I am currently using a laptop for all my edits. Whenever I try to do something with higher production (more VFX/fusion comps), I have to strategize my renders in a way to remove as much stress on the GPU. Like figuring out a way to do something that would take 10 nodes down to 5 nodes. I also feel limited, since once I reach a threshold, Resolve crashes.
I am definitely saving money for something, but not sure what is the best bang for my buck on Windows/PC side.
Hello, that's a great idea. We'd love to do something on that in the future. In the meantime, it sounds like you've got a great approach, keeping your comps as efficient as possible. It's great to have you here, thanks for watching.
M2 Max Studio 64/1TB 38 core GPU
A:310
A:296:
A:294
Great video. I downloaded your render to test my M3 MBP 18/512 14 core GPU and my M2 Max Studio 64/1TB 38 core GPU. The M3 MBP repeatedly crashed at 16% and the M2 Max knocked it out in 4:56 repeatedly first on DR Studio 19.1 and then after update to 19.1.1. Thanks for giving us a tool to play with.
Glad you enjoyed the video and are finding the benchmark helpful. Can you try your M3 MPB with the latest release that came out yesterday?
@@team2films I installed DR Studio 19.1.1 this morning on the MBP 18/512 14 core GPU with the following results: (#1.Success)
On Battery Powered
A 804. 731
A 722 728
A 729 731
M3 MBP 18/512 14 core GPU on 19.1.2 scored 779 on power without crashing. Still disappointing as I really thought the M3 Pro would do better. I'm curious to see where the MBP Max loaded is performing.
this is such a professional documentary level of video
Thanks so much, what a nice compliment.
Great video! Love the attention to detail🔥
Thanks so much Michael. Great to have you here.
Here is my contribution:
A 231, 222, 220 - Mac Studio M2 ULTRA (24CPU 60GPU) - 19.1 STUDIO
B 247, 278, 260 - Mac Studio M2 ULTRA (24CPU 60GPU) - 19.1 STUDIO
C 151, 151, 151 - Mac Studio M2 ULTRA (24CPU 60GPU) - 19.1 STUDIO
A CRASHED! - Macbook Pro M3 PRO (11CPU 14GPU) - 19.1 STUDIO
B 440 - Macbook Pro M3 PRO (11CPU 14GPU) - 19.1 STUDIO
C 330 - Macbook Pro M3 PRO (11CPU 14GPU) - 19.1 STUDIO
Nice to see my Mac Studio flying, and each time faster than the previous one!
Worried about seeing my macbook pro crash multiple times, with and without power.
Thanks as always for the info, resources and community!
Thank-you, really appreciate you posting those results. We have added them to the description if that's ok! You are currently top of the leader board, but the M4 Max is really nipping at your heels :)
Do you know what clip your M3 Pro crashed on? We've seen a crash in the M4 Pro too. It's weird though because the stock M4 can run the benchmark just fine.
@@team2films Sure! Glad to be first. haha. Looking forward to see what M4 Ultra will be able to.
Mi M3Pro crashed repeatedly at 16%, finishing the 5th clip and starting the 6th, the train one. Interesting to see other crashes on M4Pro... Lets see if we can find out more...
Thanks so much! :)
@@Miguel_Garcia_Iraburu Gotcha. It's speedwarp that's doing it. If you change even to speedwarp faster I believe the crash will no longer happen. Hmmm, will investigate this.
Solved! With speedwarp faster it goes perfect! Here you have the results:
A 807 - Macbook Pro M3 PRO (11CPU 14GPU) - 19.1 STUDIO
@@Miguel_Garcia_Iraburu That's good to know, really appreciate you confirming that. We'll send info to Blackmagic about the bug. In the meantime, the score is interesting... but it's not a fair comparison now as speedwarp better is less computationally intensive. Thank-you thank-you though. Super helpful.
Thanks for this great video, Very good solution to check performance between different machines, here I leave my performance in the project and waiting to receive my M4 MAX in a few days...
Benchmark A: AMD Ryzen 7 7840HS / 32 GB Ram / 1 x NVIDIA RTX 4060 LAPTOP 8GB / Windows 11 / 19.1 Studio render attempt: 22:38 minutes (1350 seconds)
Thanks for sharing. How is your new M4 Max?
@team2films Thanks, it's the standard version with 48GB of memory. I'm really looking forward to it, as they say in my country, like water in May 🤣.
please make a video on best external ssd for video editing for mac.
Hello. Most SSDs will be sufficient. That's because over a certain speed, the cpu and gpu will be your performance bottleneck, not the ssd. So any typical 1000MB and above SSD is going to be just fine. Does that help?
I love the relaxed style presentation and suggestions. Congrats!
Thanks so much! Nice to have you here.
is a 24gb ram macbook pro 16inch 1tb laptop a good option, or will it not be powerful enough to use fusion and resolve
Any modern computer can run resolve or fusion, performance will depend upon the kind of footage you are working with and the effects that you apply. The stats in our video should hopefully help you understand the comparative performance you can expect from particular MacBook models.
Hey guys! Great video! I am torn, I am considering the Mac Mini spec'd out to M4 Pro chip with the 64Gb ram. Do you reckon this machine will easily handle 4K video editing. (I am a videographer / editor by trade) This will be my main machine as my custom windows laptop is starting to throw the towl in. I can't justify the crazy price increase to get the M4 Max MBP. Any guidance would be much appreciated!
Hello. Give or take.... The M4 Pro is twice the power of the base M4. The M4 Max is twice the power of the M4 Pro.
That's a difficult question to answer. You can edit 4k or UHD video on a base spec M4. But it's also possible to edit 12k video on a base spec M4. Performance will decrease as you add effects and colour grades to your footage. It's possible to push an M4 Max to the point where it can't play back in real-time. And of course it's possible to overwhelm a base M4 too.
We'd recommend downloading the benchmark. Check out the performance of your computer. Compare it against the specs for the M4 chips shown in the video. It will tell you how much better it will perform than your computer.
I hope that makes sense! The short answer is, yes you'll love the M4 Pro chip. But it's hard to tell you what it will do without knowing more about the kind of editing you do!
Thanks so much for watching, and please let us know if you have further questions.
everything is on point..thank you!
Thanks
I am pondering the M1 to M4 upgrade. The rendering tests are not that important to me. My videos are short and rendering time is less important to me than timeline performance. Does the M1 chip stall when you edit on your timeline? What happens if you apply a noise reduction node in the color panel? Does playback speed stay at 24/30fps? These are the important DR tasks that are of value to me and are harder to evaluate. FYI… I wanted to test the speed of the M4 Max Thunderbolt 5 USB ports at an Apple Store with various SSDs I work from. But Apple will not allow speed testing from store configured Macs 😢.
Hello, thanks for watching. Deliver page performance is representative of edit page performance. For example if a video can render above 1x speed then it will play at 1x speed in the edit or color page.
Like you said edit performance is harder to evaluate. We avoided edit page tests, because it gets really subjective. For example, can a clip play with Denise applied? Yes.... but what's the resolution of the clip, what type of denoise is being used, how many other adjustments have been applied to the clip.
Here's what we'd suggest. If you switch Resolve to 'Show All Video Frames' look at the playback FPS. Then you can use our benchmarks to work out the relative performance of the M4 and whether it will increase your performance above that real-time playback threshold.
We're hoping to get hold of Thunderbolt 5 storage soon!
Thanks for watching and commenting. Appreciate you!
if we speak about M4 base so it still will be struggling with noise reduction or any effects like this. Cuz GPU cores are almost the same
Thanks for the benchmark. My results for Bench A
A 282 - Ryzen 9 5900x - 128GB DDR4 3600mhz - AMD 7900 XTX - 19.1 Studio. Not too bad for an almost 4 year old CPU, the GPU is only about 2 years old though.
I have a question if you have the time :). I don't care quite as much about render times like in this benchmark as I do timeline playback, or what I'm able to playback without being forced to generate render cache. With Benchmark A and my test system, I have to set render cache to "smart" in order to playback in the timeline reliably, without resolve crashing. If I don't generate render cache (and wait) It will playback ok for a bit in the timeline...but eventually it will crash if render cache is not generated first. I've been looking at upgrading my system (everything but GPU), then the M4 came out and I've been comparing results.
How is the timeline playback/stability on the mac without generating render cache in Benchmark A? Or do you have to generate render cache on the mac as well for speed/stability in the timeline? My system never crashes during a render...only in timeline playback/scrubbing after a while with no render cache.
Hello, thanks so much for sharing your results. Sorry you are having problems with playback too. Your system should be able to playback without crashing (even if your system is old or not able to playback in realtime). We have older and slower systems that playback without crashing. Also, please note, we cannot play this timeline back in realtime, even on an M4 Max or M2 Ultra. I don’t think there are presently any systems that can play this timeline back. Basically you’d need to see a render time that’s lower than the length of the timeline. Render caching will always be an important part of workflows, as even though computers grow in speed, the complexity of the tools we use will likely grow too.
So to answer your question, playback it’s stable for us but for speed, caching is required. Yes, we agree that timeline performance is more important than render speed. Our benchmark is designed to reveal timeline performance, we just use render speed as the metric for measuring that. The workload of rendering is almost identical to the timeline workflow with the exception of encoding. As h264 encoding happens on hardware for most people, then the impact of encoding on performance is very very low. Hope that all makes sense, please ask as many questions as you want. It’s good to have you here.
@@team2films Thank you for taking the time to answer! That makes sense about the render time vs. playback and now I think I understand where you are coming from with that as a test. I guess another "hack" I could use in edits would be to cut my video first, then add the layers of fusion/animations after all cutting is complete, and I could even disable/turn off fusion/color grading until I'm done "editing"...then turn on render cache toward the end of the edit to see it all before render. What I'm trying to do is prevent the annoying wait times of changing something/editing something, and having cache rebuild and having to wait in timeline on each/every edit.
Thank you again for the tips/insight. I don't think it would really make much sense for me to upgrade much other than maybe CPU/RAM speed as that seems to be where my system is lacking. My GPU seems to keep up in your test. When I render your project, my GPU is almost always 80% or lower utilization which tells me that my CPU/RAM isn't feeding it data fast enough :). Something the Mac does well is the unified memory is shared between gpu/cpu, and it's generally faster than what you find on a PC unless you really buy high end memory and tweak it.
@@TheCarGuyOnline Yeah, that's a solid idea (compartmentalising stuff on different layers to make it easier to turn on and off). Don't forget the simple button above the viewer that turns off color page and fusion adjustments too. That's useful for when you want to focus on editorial changes so want the best possible timeline performance.
Leaving render cache on as you go is a good idea too, because it only works in the background when you are not doing anything else.
I'm also having similar stability/quirky issues with my system running a 5950X with 7900XTX. T2F added my A time of 263 to the video description but it took me a few tries to get 3 full back to back renders without any freezing/crashing during rendering. I also sometimes have issues with timeline/playback ability and have to turn on Smart Cache. Sometimes even magic mask doesn't work properly. I have to save the project, close Davinci and reopen it again, then it will work. I don't know if its an issue with the AMD driver and the 7900XTX or what, but I find my system to be very inconsistent in Davinci Resolve where I too am thinking about getting an M4 or M4 Pro mini. It's a shame though because this 7900XTX renders this benchmark very well for what it is and its cost compared to something like the RTX 4090.
Benchmark A_STUDIO - 10:17 - (Mac Mini M2 Pro - 32GB - 12 (8 performance and 4 efficiency) - Davinci Resolve Studio PAID - 19.1 Build 12
Benchmark_C_CPU - 05:28 - (Mac Mini M2 Pro - 32GB - 12 (8 performance and 4 efficiency) - Davinci Resolve Studio PAID - 19.1 Build 12
Hope that helps someone out.
M2 Pro with 12-core CPU and 19-core GPU
Thanks so much, that's awesome. Added to the description.
You are the best!
Just finished running it on my M4 Pro Mac Mini and I have a video posting tomorrow morning on my channel showing it and the results. Here is my report from that test.
A 565 - M4 Pro (12CPU 16GPU, 24GB RAM) - 19.1 Studio
I’m rendering a 4k timeline on my M4 Max 16core CPU 40core GPU 48GB memory and I’m getting 4 -6 FPS and the render constantly crashes around 30 percent. I thought these laptops can handle anything we throw at them. I’m using NR Dehancer and another 8 nodes
Did you run the other test for Free and cpu/gpu test?
Hello, thanks so much for watching. Sorry you are having problems. The M4 Max is incredibly capable. It’s hard to pass judgement on your particular circumstances without knowing more about the timeline you are processing. What I can say is that sadly no computer can handle ‘anything’ (or everything) you throw at them. The relative performance figures will give you an idea of how much slower your performance might be on another computer. The M4 Max is very close to the top of the score board.
Thanks for this video, it's perfect! I really thought my computer has held up well but holy moly these numbers are atrocious compared to everything else. I have a windows ryzen 9 3900x with 64gb DDR4 RAM and RTX 2080. It took my computer (drum roll please...) 29 mins. on the dot. 1740 secs. (for the A benchmark).
Oh no! Firstly, thanks for watching. Second, sorry about the score. We don’t know enough about the components you are using to say where the bottleneck is, but as Resolve relies primarily upon the GPU for rendering, we’d imagine it’s the GPU. Sorry though! Didn’t want to make anyone feel bad 😞. Here’s the important thing though… if your computer is doing what you need, then it’s good enough! Thanks for watching, it’s good to have you here.
@@team2films Update! I purchased an M4 MBP on Black Friday (I've been holding out on getting a laptop and can write it off for work anyways). 16" M4 Max 16c/40c, 48gb RAM. Here are the new numbers: Benchmark A- 4:12, 3:52, 3:50. And just because I was curious, I ran the same benchmark with the media stored on an external Samsung T5, and exported to the T5 as well. Interestingly, the numbers were about the same which is great! 4:03, 4:02
@@Megalukeification Whoop Whoop! That's an improvement! I love how the numbers speak for themselves. Yeah, we wouldn't expect to see a difference in score between the internal SSD and an external. There's fastest, then there's fast enough. In these tests, disk access speed is not the bottleneck. Most ssds will feed the footage faster than the GPU can process it, especially with those effects. Awesome stuff, thanks for sharing!
I am having using M1 Pro 14 INCHES MacBook Pro and I wanted to take another laptop as a secondary device which one do you suggest ? I am a wedding photography and I mainly use Final Cut Pro and davinci resolve what is your suggest value for money
The comparisons in the video will give you an idea of the performance you can expect from the new models. We would suggest comparing that against their cost and making a decision about what you can reasonably afford.
A223 - i9-13900KF 64gb (2 x 32gb) DDR5 6400mhz RTX 4090 - 19.1.2 Studio
Wish I Bought Mac Book M2 Pro or M1 Max at time of building, needing portability now, and have worked on friends Mac systems, feel much better compatibility and less problems with applications like premiere, pro tools, and davinci in specific tasks, but 1 thing my build might have better render speeds but, Mac runs amazing with 4K footage in my experience without proxy, 😢
Also had hell of a lot of problems until recently with intel cpu crashing and other things until intel recently fixed it with bios update
Glad the bios update fixed things.
One of the advantages that Mac systems have is how prolific they are. Windows is very difficult to develop for because there are so many iterations of hardware. The most common system that DaVinci or Premiere runs on will be Mac silicon, so naturally it will always get more attention from devs. Just our thoughts! We love both windows and Mac though :)
Unfortunately I cannot link all the clips in the project and some of them ar red 😔
I love his videos and his way of talking❤
Thanks!
I've tested your project on my machine:
Intel(R) Core(TM) i7-14700K 3.40 GHz / NVIDIA GeForce RTX 4090, driver version: Nvidia Studio Driver 560.70 / G.Skill 64 GB DDR5-6000 / Windows 11 Pro 23H2 / Davinci Resolve Studio 19.0.3 build 5
Benchmark A: 238/231/224
But I also tried to use the voukoder utility to access custom settings for encoding, I've chosen Nvidia NVENC H264, the settings were: multipass=disabled preset=slow, profile=high qp=18 rc=constqp tune=hq, the results were almost the same:
Benchmark A: 247/208/239
And finally I've tried to run it with Nvidia NVENC HEVC H265 encoder in voukoder, the settings were: multipass=disabled preset=slow profile=main10 qp=18 rc=constqp tier=high tune=hq, the results were much faster:
Benchmark A: 215/183/188
That's cool to know that there are some performance gains with h265. I think it's really interesting that subsequent renders speed up too. I think it might be something to do with fan curves that allow thermal throttling before spinning up. I've added your h264 results to the scoreboard!
Windows11 Pro: Core i9-14900KF, Asus Proart z790 w 64Mb ram, nVidia RTX 3090, all ssd drives: Benchmark A 4:50.
A 2569 - 2019 16inch MB Pro 2.3 GHz 8-Core Intel Core i9 // 32 GB RAM - 19.1 Studio
I need to upgrade.
Ah nooo! Those final 2019 intel MacBook were not good. We know because we have one. The chipset has a lot of power, but it can't run under sustained load without thermal throttling. That's the problem.