M4 Pro VS. M1 Max Video Export Speed Test In DaVinci Resolve Studio - Mac Mini vs MacBook Pro

[@DavidHarry]

M4 Pro Mac Mini vs M1 Max MacBook Pro H.265 HEVC codec video encoding speed test in DaVinci Resolve Studio. This was a surprise. Video chapters: 00:00 Introduction 00:56 M4 Pro Resolve settings 02:51 M4 Pro export settings 06:35 M4 Pro export speed test 08:18 M1 Max Resolve settings 09:44 M1 Max export settings 10:55 M1 Max export speed test 12:06 End summary AMAZON ASSOCIATE DISCLOSURE: I am an Amazon Associate. My Amazon links are Amazon affiliate links. I earn money from qualifying purchases when you use my Amazon affiliate links. OTHER EARNINGS AND COMMISSIONS: I also earn money from other product links within my video description. For the sake of clarity and for the avoidance of any confusion, assume that I earn money from commissions from any and all links that I have within my video description. Amazon links to stuff used in this video: The gear that I use, have used or reviewed that I would recommend. Cameras: amzn.to/3PzC8mI Microphones: amzn.to/3tD6FaM Computers: amzn.to/3tuwHNr Phones: amzn.to/3ttqjWN iPads & Tablets: amzn.to/46t8wxG You can also send me a coffee donation via PayPal if you found my video super helpful: www.paypal.me/DavidHarry My Amazon pages with videos and product links: Amazon USA www.amazon.com/shop/davidharry Amazon UK www.amazon.co.uk/shop/davidharry If you would like to help my channel please use my global Amazon Affiliate links. I will be paid a small commission for anything you buy from Amazon when using these links. These commissions don't cost you any extra but really do help me to buy gear for my productions and to review: Amazon USA: geni.us/Amazon-USA Amazon UK: geni.us/Amazon-UK Amazon Deutschland: geni.us/Amazon-Deutschland Amazon France: geni.us/Amazon-France Amazon España: geni.us/Amazon-Espana Amazon Italia: geni.us/Amazon-Italia Amazon Canada: geni.us/Amazon-Canada Contact for product reviews: [e-mail] www.DavidHarry.com I’m David Harry. Thank you very much for watching this video, take care and goodbye now. Cheers, Dave.

[@asafblasbergvideographer]

Can you please test UHD 10-bit 4:2:0 H.265 source file exporting?

[@DavidHarry]

Will do in future videos 👍 Cheers Dave.

[@asafblasbergvideographer]

Hi Dave, I had a chance to test some UHD 10-bit 4:2:0 H.265 source files and exporting times in H.265. The M1 Max rendered at 172 frames per second with the UHD H.265 preset in Resolve; M4 Pro did it at 108 frames per second, and M1 Pro did it at 90 frames per second. So, while the Max variants have two encoders vs. one, it's not completely double the speed as you mentioned, it's about 1.59x faster. At the end of the day for me, I couldn't justify spending $4,000 on a M4 MAX laptop when the M4 Pro Mac mini was $1,399 for the base. For me I'd rather wait for the videos to finish longer and use that difference in the money in getting storage external hard drives and accessories. For reference, my PC build with NVIDIA RTX 4070 Ti did the test in 150 frames per second. The fact that a 3 year old Mac machine renders faster than an RTX 4070 Ti is amazing!

[@DavidHarry]

Hi. If you don't mind waiting then that's great 👍 However, there's no doubt that the 3 year old Pro and Max machines are still awesome machines. Don't forget, though, that something like the M4 Max is going to be way more powerful in the timeline compared to any M4 Pro variation. So if you are doing any serious processing and run into scenarios where the M4 Pro will lose real time ability, the export will be even slower. Doing the same thing with an M4 Max may still be in real-time in the timeline, or may lose real-time but will still be considerably faster than the M4 Pro. However, in either event, the final export will also be considerably than the M4 Pro. Saying that, if you don't push a system that hard, then other than the final export, you really don't have anything to gain spending more money beyond an M4 Pro. I think this is where the phrase "cut your cloth accordingly" works really well. There's absolutely no reason to spend on features that you simply do not need 👍 Cheers, Dave.

[@asafblasbergvideographer]

@@DavidHarry yes I agree with you 100%. Thank you for all your wonderful videos, I’m happy to be a subscriber of your channel!

[@DavidHarry]

@@asafblasbergvideographer I'm glad you like some of my videos and you are very welcome 👍 Thank you so much for the subscription, it really does help a lot 👍 Cheers, Dave.

[@danielealiani5484]

Amazing video! Hope to see a timeline scrubbing comparison!!

[@DavidHarry]

Hi. Yes, I’ll be doing some more video comparison stuff 👍 Cheers, Dave.

[@logesh5213]

@@DavidHarry That will be very helpful, also would like to see the see the frame drops when more effects are added, like noise removal, grains etc.

[@arkadiuszsyrowiec7710]

Yup, I'm actually using my M1 Max to edit 3/4 cameras video with NR applied to all of them + a few simple Fusion clips, sometimes combined both NR + Fusion and since I've started some more advanced projects my M1 Max doesn't playback the timeline in real time any more. I've started to wonder if upgrading to M4 would make any difference in timeline playback.

[@asafblasbergvideographer]

No question the MAX is the one to get for Resolve exports. :)

[@DavidHarry]

But it does depend upon what's going on in the timeline. Cheers, Dave.

[@brucekennedy5274]

I assume Resolve IS making use of the dual encoders on the M1Max.

[@DavidHarry]

That would make sense, however, after talking with other Resolve users on the Resolve forum the general consensus is that the encoders don’t double up on the final output.

[@50450720]

I too, am suprised at how poorly the mac mini performed compared to the m1 max, especially as you can now get a m3 max macbook from apple refurb for a similar price that the specked out mac mini cost you. Could part of the issue be that DR is not currently optimised for the M4 chip? And would a similar experiment be different in FCP? Thanks for showing this as ive been thinking of getting the m4 mini pro over a new macbook max. I think ill hang on and do more reserch, or even wait for the m4 mac studio. 👍

[@DavidHarry]

Hi. Yes, I was super surprised at the lack of difference. While the M4 Pro is more powerful as far as CPU power is concerned, the M1 Max was keeping up with it in other tests that I’ve done. I doubt very much that Resolve needs any specific optimisation for M4. The M4 Macs are running really well on it, the main difference here is just the differences between the Pro and Max chips. To be honest, it’s my own fault. I already knew from previous Mac’s what the difference between the Pros and Maxes were. However, I thought that because this was M1 and M4 I should have gotten a bigger difference in the basic core performance. I suppose you live and learn 🤣 I will be doing some FCP stuff as soon as I’ve caught up with more videos. I’ve just ordered the M4 Max MBP and will tryout FCP on that. BTW. If you are in the UK checkout www.KRCS.co.uk they have got 10% off all the M4 Macs, I’ve just saved £400 on my M4 Max MacBook Pro. Cheers, Dave.

[@Techno_Rama]

Devo my man, i tried to tell you last video, glad you figured it out, but there you go you checked it yourself...here is basic explanation...he Apple M1 Max and M4 Pro chips both feature dedicated media engines designed to accelerate video encoding and decoding tasks, enhancing performance for video editing, streaming, and playback. Here’s a comparison of their media engine capabilities: M1 Max Media Engine: • Video Encode Engines: Two dedicated engines for efficient video encoding. • ProRes Accelerators: Two accelerators specifically for Apple’s ProRes codec, enabling high-performance processing of ProRes video formats. M4 Pro Media Engine: • Video Encode Engines: One dedicated engine for video encoding. • ProRes Accelerators: One accelerator for ProRes codec processing. Key Differences: • The M1 Max offers double the number of video encode engines and ProRes accelerators compared to the M4 Pro. • This means the M1 Max can handle more simultaneous video streams and perform faster encoding and decoding, particularly beneficial for professional video editing workflows. Implications for Users: • Professional Video Editors: The M1 Max is more suitable due to its superior media engine capabilities, allowing for smoother handling of multiple high-resolution video streams and faster export times. • General Users: The M4 Pro provides ample performance for everyday video tasks, including streaming, casual editing, and playback, but may be less efficient for intensive professional workloads. In summary, while both chips are capable, the M1 Max’s enhanced media engine makes it a better choice for users with demanding video processing needs.

[@DavidHarry]

Alright, thanks for that info 👍 What I still don’t understand is that while the M1 Max has twice the encoders, Resolve is only using one engine to decode the ProRes file and one encoder for the encode to H.265 output. Having spoken to other Resolve users, the general understanding is that Resolve doesn’t utilise multiple media processors for any single task. Unlike Apple’s Compressor which shares the encode across two engines for an export on any Mac with multiple engines. Aside from the encoding/exporting, I’m not really impressed with the M4 Pro’s performance in the timeline either. Not that it’s bad, it’s really good but it just doesn’t feel like there’s any improvements over the M1 Max. I’m sorry I bought the thing now and I’ve got buyer’s remorse like you wouldn’t believe 🤣🤣🤣🤣 We’re All DEVO…. Cheers, Dave.

[@Techno_Rama]

@@DavidHarry Devo thats why i mentioned it to you, consider returning it and waiting for the M4 max/ultra studio if it will make you feel better(although i am bad at returning stuff). You are the ideal person to use m1 ultra (best cost effectiveness) or m2 ultra if want to maximize media engines as that has even double the max's(4x the pro). On the Resolve using all the decoders/encoders this is what chatgpt4o says with references and must be true simply from the times you got and my personal experience as well..."Yes, DaVinci Resolve is designed to fully utilize the media engines present in Apple’s M1 Max chip. The M1 Max features two video encode engines and two ProRes accelerators, which significantly enhance video processing tasks. DaVinci Resolve leverages these hardware components to accelerate encoding and decoding, particularly for ProRes and H.265 codecs, resulting in improved performance and efficiency during video editing and rendering. (Blackmagic Design Forum)" Stay cool, Devo and hope you enjoying the new PS Pro.

[@DavidHarry]

@@Techno_Rama Hi. Yes, I will be sending the Mac Mini back. I have already ordered the M4 Max MacBook Pro. I didn’t want another laptop, as I will use it 99% of the time in clamshell mode connected to a monitor, basically using it as a desktop. However, I can’t wait for the release of the Studio before I upgrade. Here in the UK the base M4 Max costs £4000. However, I found someone doing them on a Black Friday deal with 10% off, so £3600 for the base Max. This will help to reduce the loss I will incur when I come to sell it. The idea I had with Mac Mini was to sell it when the Studio come out and upgrade. So I will do the same with the M4 Max MacBook. I had the M1 Ultra Studio but did not keep it after I found out that it didn’t scale well, over the Max. There’s talk that the M4 Studio will be monolithic, if so, it will scale better than gluing two Maxes together 😆 As for Resolve. My understanding is that it does not use two encoders on a single stream encoder. Take care 👍 We’re All DEVO… Cheers, Dave.

[@asafblasbergvideographer]

Great video. Can you make another test but with "Optimize for Speed" on? It would be interesting to see. I'm sure the render times would decrease.

[@DavidHarry]

Thanks. Yes, I can do that in another video when I’m comparing encoding speeds. It’s about 30% faster with “Optimise for Speed”. Cheers, Dave.

[@asafblasbergvideographer]

@@DavidHarry Great! Thank you :)

[@Mr_ToR]

Very interesting. here is how I consider this and the apple silicon chipset. if we ignore the baseline chip, going from pro to max; apple kept the media engine count of 2 same but GPU cores and memory bandwidth is doubled. going from max to ultra its basically gpu cores, media engines and memory bandwith is all doubled. however cpu is kept same in all variants but we know cpu core count is not a big factor in media applications so we can disregard that. now eventhough the media engine count is same between pro and max, because the media engine uses the gpu and gpu cores are doubled going from pro to max, thats where the almost 2x speed difference comes for going from pro to max. Now let's try to compare m1 max to m4 max for media process: due to the 5nm process vs 3nm process difference (also some minimal architectural optimizations) there will be better thermals thus better efficiency thus better clock speeds so thats how the M4 is faster than the M1. In reality there isnt a big difference beween them. (disregarding the npu for our case) but apple also %20 increased gpu count to 40 gpu cores in m4 max compared to m1 max having 32 gpu cores. lets try to image the total speed difference between m1 max and m4 max. gpu is a bit faster and has more cores. from m1 pro to m4 pro you get %37 additional clock speed (3.2 GHz to 4.4 GHz), multiply that with additional speed from %20 additional gpu cores so you get like total %45 speed. add a bit of architectural optimizations maybe %50 However going from m1 pro to m1 max or from m4 pro to m4 max should be close to double due to the 2x gpu core count and 2x memory bandwidth. so: from m1 pro to m1 max you get close to 2x speed (%95 maybe?). So basically Im contemplating media applications where comparing to an imaginary baseline of M1 Pro benchmark, M1 Max should be 95% faster while M4 Pro should be only 50% faster.

[@DavidHarry]

Alright, Mr_ToR, I’m glad to see you again. Sorry I can’t be too long right now, I’m doing more testing. I totally agree with the logic you are applying with regard scaling, especially clock frequencies. And yes, the architecture isn’t a million miles different from M1 up to M4, regardless of iterations. As you say, most of these gains are probably due to the increased clocks as a result of the reduction in sizes and the efficiencies that brings and not least of all, more of the same in a smaller space. Altjough, at one point I did have an M1 Ultra and the scaling was quite poor, maybe due to the inefficiencies of fusing two separate SoCs? However, the issue I’m having in identifying certain processing tasks at any one particular time, isthat none of the monitoring software that I use is showing what you’d expect to see. That said, I find a lot of monitoring isn’t very accurate and some simply do not show anything, especially when it comes to encoding and decoding tasks. I did not show this in the video as it would have been more of a confusion than as help and it would have made the video longer than it was. BTW, I know these videos are very long for such easy tasks. I just like to be thorough in what I’m showing, so that people know exactly what is going on. I personally can’t stand seeing other KZbinrs just throwing up random graphs with numbers with no supporting evidence of how the data was collated. On the point of long videos. I’m just finishing off a video showing how to do HDMI capturing on Apple Silicon (specifically) using a UVC interface, OBS and Resolve free. The video goes right from the HDMI output of the gaming device to the upload file for KZbin or other streaming platforms. That video is well over an hour long. Anyway, back to this video. I’m still not sure that Resolve utilises all the available encoders when doing a single export task. Don’t forget, these parts of the SoC are multipurpose and can be used for a number of video related tasks in the timeline, where various types of codec manipulation is required. So yes, you definitely feel the benefit of the multiple video processors but I’m not sure that they are paralleled or serialised/doubled for any one export task. With regard my crash and corrupt media output issue. I am now talking with BMD via PM on their forum. They have now acknowledged that the corruption scenario I’d shown is definitely an issue, although, they say they can’t replicate the crash. What’s interesting about that is that they first questioned me about the crash without acknowledging the corruption issue. As in, they don’t see the crash but never admitted to the corruption either. As I knew that both scenarios stemmed from the same export setup, I pushed them on it. Because if they didn’t see the crash, the same export parameters would definitely have produced the corrupted media. It was only after pushing them they acknowledged the corruption fault, which they had to as I’d already proven it was reproducible and across different SoCs and others followed my workflow on the BMD forum and acknowledged the issue. To be honest, I found it very disingenuous of them to not have been straight with me immediately. I will also have to admit that I have bought the wrong thing. The M4 Pro definitely wasn’t/isn’t for me, which is my fault for jumping the gun. I have now done the thing that I should have done but didn’t want to and ordered the M4 Max MacBook. Either way I was going to need a stop gap with Thunderbolt 5 until the Studios are released. If the M4 Ultra is a more monolithic design and scales better linearly compared to the fused Maxes, I think I may break the bank and get something to last me a good few years. My main Mac is my daily workhorse for all my KZbin and video stuff plus I still dabble in music production. So buying once and having it last a long time is definitely something that I think’s a good idea, My M1 Max MBP has lasted nearly three years and has been an amazing purchase and well paid for itself. Finally. If you are in the UK and after a new M4 Mac, checkout www.KRCS.co.uk For some weird reason they are discounting all the M4 line by 10% for Black Friday, something I’ve not seen anyone do. Seriously, these Macs aren’t even a week old yet and there’s a discount. I just saved £400 on my MBP purchase. Take care 👍 Cheers, Dave.

[@TooMuch_TV]

So basically. Going from my 2019 MacBook Pro to a m4pro is a great move for me at this point. 😊.

[@EasyyokeFilms]

@@TooMuch_TVA very big move

[@TooMuch_TV]

@ I’m excited. It will be here soon!!!

[@Mr_ToR]

​@@DavidHarry Hi David, I want to make my argument about software vs hardware. I hope you have the time :-) So here I go; Back in the early 2000s at some point multi core was supported only with WindowsNT and since it was more stable everyone used it for professional stuff. While WinXP was less stable but faster without multi core support. Back then multi core was multiple physical CPUs on a single motherboard but essentially the same thing as what we have now. The important thing is that most of the motherboards did not support multi-core. However there was a software which utilised multi core on XP even while the OS was not supporting it. The software was Steinberg's Nuendo. This was a beautiful example displaying the significance of applications and the importance of how they were developed. I used to use this example back then to explain why software is the most important piece in the equation, not hardware. This still is very much valid, maybe even more applicable than ever. Let's look at what Apple’s been doing with their new Apple Silicon SoC and chipset technology since it came out. We’ll see that it's mostly driven by marketing with superficial updates to how it is set up or packaged. Let’s look at M1 to M2: CPU is exactly the same, GPU is essentially the same with more cores 8 to 10, clock is driven from 3.2GHZ to 3.5GHZ. Most important upgrade was the RAM: from LPDDR4X ram to more efficient LPDDR5 ram. These resulted in 50% memory bandwidth increase (from 68.25 GB/s to 100 GB/s) which resulted in better GPU and CPU performance. From M2 to M3: M3 was not a completed design. Apple decided to release it early mostly as a marketing strategy in their effort to be the first to manufacture a 3nm process chip. In May this year it was found by examining the M3 Max chip that it was missing the UltraFusion interconnect that connects two MAX chips to make an UltraChip. It was postulated then that there will not be an M3 Ultra variant. Now we found out this to be true since there is no Ultra variant for the M3. M3 essentially is something between M2 and M4. other than that, it's still the same story: CPU is exactly the same, clock is easily driven from 3.5GHZ to 4.06GHZ as a natural result of goingn from 5nm to 3nm. GPU is essentially the same but with higher max core count. Additional transistors in SOC for better neural engine, and now Media Engine supports AV1. There is no Memory Bandwidth increase from M2 to M3. However there should have been, since it was produced with the better 3nm process; which results in SOC with increased efficiency but probably Apple slowed it down somehow so somewhat benefiting from the increased clock but not fully. How can it still have the same memory bandwidth? Apple only lets people buy it for increased core count, better neural engine and AV1 but reserving benefits of 3nm for later offerings; well, it was after all as I earlier said, not a completed/final design. From M3 to M4: Now we get the real intended design and expected memory bandwidth coming with the 3nm process. Increasing from M2/M3 100 GB/s to M4 120 GB/s. This time no additional GPU core (happened in M3 already) so GPU is exactly the same but CPU setup is finally updated since M1 8(main)+4(efficiency) is now 10+6. However CPU and GPU are still the same thing. You see, the only real serious upgrade is perhaps the 3nm process SoC the rest is almost like what a PC enthusiast usually makes. Increasing cores, increasing clock, increasing ram, adjusting thermals and voltages, adding additional processors maybe? etc. During all these updates, developers are expected to convert their apps from intel to Apple silicon. But if you know the developers, similar to how Apple superficially updates their hardware, developers would rather do it as slow as possible too. Why? Because business… Apple has the Rosetta 2 translation layer. As long as the arm version performs faster than that, the developer is golden. Also, people constantly update their hardware for performance but never stop to wonder how much updating the software they use really needs. This is very hard to explain. But as a developer, considering only the software side, I can definitely say that, there is always room for performance increase in your software and programs are usually made in the fastest and most inefficient way possible. This is the way :-) Also, it is very easy for the developer to make the performance considerations an aspect of the hardware manufacturer and distance themselves as much as possible from any scrutiny for performance. So even with benchmarking a software system, the unknown part in the equation is how the software runs. When you use an open source software, be it a synthetic or real world process, you can then factor out what the software is doing and concentrate on the performance of the hardware. This is why benchmarking is quite difficult and comprehensive actually. But then again, how many souls do even care about the benchmark too anyways, it mostly always the shiniest, newest, most marketed, most talked, most expensive etc. Its even the same for many technical executives/managers too anyways. I hope I was able to make my point. Take care, ToR

[@paradoxicalcat7173]

M4 Pro has one media chip. M1 Max has TWO encoders, so half the time to export.

[@DavidHarry]

I know it has more, I say this in the video but I also say that there is no written evidence that Resolve doubles up on them, plus it's not half the speed.

[@garypranzo9334]

I am waiting for the M4 Studios to drop. I am a bit worried Apple will move away from the Ultra models though and maybe just have a new higher specced MAX

[@DavidHarry]

Yes, that's the best thing to do if you have the time to wait. If the M4 Max Studio had have been available now, I would have got it. Cheers, Dave.

[@Jason-Taylor]

Hi Dave, I’ve been enjoying your recent videos on your experiences with the new M4 Pro Mac Mini. I sense you’re finding that the mini is not necessarily worthy of upgrading to from your previous setup and that you may even be having some regrets. For someone like myself who is a non-professional currently using a 2019 iMac with an Intel i9 processor, 64GB of RAM and a 2 TB SSD to primarily record and mix music in Logic Pro and doing occasional video editing for my band, would the new Mac Mini M4 Pro be a viable machine for me to upgrade to in your opinion? I would imagine that moving to anything with an Apple Silicon processor would be a worthwhile upgrade in my case. I don’t run a business where I create content for a living so while I would definitely want a machine that is fast and efficient, it doesn’t have to be the fastest and most powerful since time isn’t money in my case. As long as it’s capable of efficiently running Logic Pro and Resolve, maybe After Effects etc. without constantly getting the beach ball or crashing like I occasionally get now, then is the new Mac Mini sufficient for my needs? I don’t mind upgrading it to be similar to your configuration with the Pro chip, extra cores and more unified memory options but wouldn’t want to spend the money for a M4 Max MacBook Pro or wait until there is an M4 Max Studio available as the cost of these is significantly more and would likely outweigh anything that I would ever get out of them. Or should I just consider buying an older M1 or M2 Max Studio instead which I can probably find for roughly the same price as a fully loaded M4 Pro Mac Mini?

[@DavidHarry]

Hi, Jason. You are very perceptive, yes I fu..ing hate this Mac mini 🤣 Nah, it's a great little machine but it's just not for me. In a previous life I used to be a music producer and still have a little dabble with it to keep my hand in. I use Cubase and have the Arturia V Collection and loads of VST FX plugins. The M1 Max has been insane for music production. Pair it up with a Focusrite 18:8:2 or similar, and you can do some serious damage. Even the original M1 Mac mini with 16GB is great for music and audio. Obviously not as powerful as the Max computers but it's great. Although, maybe a little underpowered for "serious" stuff. If I were you I'd look at the M1 Max machines, either the laptops if you need portability or the Studio if you don't need portability. There's going to be loads of them on the market now, with the M4 series being released. For music you don't need to worry about the GPU, so even the 24 GPU core Max Studio with 32GB RAM and 512GB SSD would be awesome. You can add a very fast Thunderbolt SSD for way cheaper than Mac storage, for your sample libraries, plugins, project/song data and audio. Use the search function on my video list and search for Thunderbolt, there will be a few guides on how to build the SSDs, they're super easy to put together. With an M1 Max, even a 24 GPU core one, you could easily do video work. If music is your main thing, I wouldn't recommend getting all bent out of shape over specifying the computer for video work, unless video is serious for you. Even then, my M1 Max only has 8 more GPU cores and it's been amazing. Don't listen to most "tech" KZbinrs, they're full of shit for the best part and most don't don't have professional experience in what they talk shit about and they have people buying into crap they need. You could go for an M2 Max but I'd only recommend that if the price is right. I live in the UK and have seen M1 Max Studios with 32GB RAM, 10 core CPU, 24 core GPU and 512GB storage for as low as £1000 on eBay. So for that price, it really is hard to recommend any new M4 series. In fact the MacBooks have now dropped a few hundred quid which is bad news for me as I've now ordered an M4 Max MacBook and now need to sell my M1 Max MacBook, talk about selling at the wrong time 🤣 Now, as far as the base M4 Mini is concerned. That may even be a thing. One thing I would say about the base M4 Mini is that it is insane for the money. I can say that without even trying one, as I know just how good the 16GB M1 Mini is and the M4 is defo better in a number of ways. If you're not tied to anything specific for video, you may want to consider Resolve, the free version is friggin' insane. Check out this video by Alex from MrAlexTech kzbin.info/www/bejne/fnzSfauFoMiZh5Ysi=IFRGokT8FmJmS0S7 BTW if you do think about going to Resolve, defo get into Alex's stuff he proper knows his shit and doesn't talk shit. If you live in the UK and are looking at any M4, there's 10% off on a Black Friday thing at www.KRCS.co.uk I've just saved £400 on my new M4 Max MBP with them. Anyway, I'm off to play some of the new season of Warzone Mobile, which has just been released a few hours ago. This is what iPhones were made for 🤣 Take care buddy. Cheers, Dave.

[@DavidHarry]

BTW, Jason. The only reason I'm buying into M4 is mostly for KZbin videos and because I've already committed to some of the manufacturers who want to send me some Thunderbolt 5 stuff to review. If I wasn't making KZbin "tech" videos, I would just stick with my M1 Max for another year or so. Cheers, Dave.

[@Jason-Taylor]

@ Thank you so much for your insightful reply, Dave. You’ve definitely armed me with some valuable information and provided me a new direction to look in. I will certainly be checking into some of older Max variants of the Studio as a possibility for my upgrade. Enjoy playing the new Warzone Mobile! Sorry to interrupt with another message but I just wanted to say, “Thank you”.

[@DavidHarry]

@@Jason-Taylor You are welcome, Jason. Good luck with your upgrade 👍 Cheers, Dave.

[@rohithmekala2608]

Like others said the Max chips have dual encoders. The M4 Pro is an upgrade over the M3/M2 Pros not the Max chips. They won’t be until the Pros get dual encoders

[@DavidHarry]

As I said in the video. Resolve doesn’t use the two encoders for a single encode task.

[@elysiumcore]

I've been debating if to upgrade my M1 Max 32 gig laptop for an M4 pro 14/20 Mac - I will definitely upgrade my M2 pro mini to M4 pro 14/20 - thanks for this test 😂

[@DavidHarry]

Hi. Seriously think twice about upgrading your M1 Max to the M4 Pro, I would recommend that you don’t. I’m really disappointed with the M4 Pro and I’m really regretting buying it 🤣 Cheers, Dave

[@elysiumcore]

​​​@@DavidHarryI will keep my M1 Max laptop - unless I went to M4 pro Max ( it's not an upgrade) - I will update the M2 pro Mini I have on desktop for the M4 pro 😂

[@DavidHarry]

@@elysiumcore 👍

[@tomsun3159]

2 encoders vs. 1 encoder, different memory bandwidth, higher number of GPU cores, that are probably fields where the M1 Max can shine. More interesting does the M1 Max have the same problem with the newest Mac Os and the newest Davinci Resove in corrupting files? Or is it specific for the silicon generation , and or specific for a specific model M4 Pro, or is it depending if binned or unbinned. or is it specific for a certain model (currently the question Mac Mini or MBP or is there a third model using the M4 Pro chip).

[@DavidHarry]

Resolve supposedly doesn’t double up on the encders for the exporters. Memory bandwidth won’t come into play as the speed of the encoders are way slower by comparison, effectively the encoders are the bottleneck. GPU cores don’t come into play as there is no processing going on. Yes, my M1 Max now has the same file corruption issue after updating to the latest OS and Resolve. I have also has BMD confirm that this is an issue.

[@tomsun3159]

@@DavidHarry So regarding file corruption, Old Resolve old OS works, new Resolve new OS corrupts, what is with new Resolve /old OS and old Resolve/new OS. This should figure out if it is the OS or Resolve who made a mistake. as i also heard of some games now crashing probably the OS. The speed difference in export is due to the higher number of encoders. For other works the GPU-core number can be the bottleneck, for other tasks the Bandwith of the RAM

[@DavidHarry]

​@@tomsun3159 The corruption issue happened after the macOS update to 15.1, suggesting it’s an Apple issue. As I’ve already said, Resolve supposedly doesn’t double up the encoders for the timeline export. If you know different where is the information saying that. Other tasks have nothing to do with anything to do with this test, it was very specific to a task that doesn’t tax the cpu, GPU or bandwidth. As explained, this was a test of the encoders not the general processing. Plus I also did make it clear that things would be different in other scenarios, such as timeline processing.

[@tomsun3159]

@@DavidHarry Its just my stupid logic, The M4 single core performance is significantly higher than the Singlecore performance of M1, if the M1 finishes much faster so it seems he can use provision the M4 not has. if you double the time from the M1 you get perhaps a performace equivalent to the differerence between M1 and M4 single core performance, it would explain the different values between your both devices for me.

[@DavidHarry]

@@tomsun3159 Sorry, I’m not understanding what you are trying to say.

[@devidasa9637]

I'm surprised you are surprised. Quite predictable if one understands the chips designs for its target users. And I see you're antagonizing people on the BM forums. Not cool. How about sharing logs with BM as was suggested to you in the forum??

[@DavidHarry]

How about you get specific instead being vague. Why would I share logs on an open when I'm dealing direct with BMD. So much for thinking that you know what you're talking about.

[@raw-viking]

This is comparing apples to oranges.. The M1 Max has dual media engines and 400 GB/s memory bandwidth. I’m expecting the M4 Max to beat it easily.

[@DavidHarry]

And you quite obviously don't know what you are talking about

[@raw-viking]

@@DavidHarryIt’s a valid comparison but it doesn’t relate all that much to M1 vs M4 and it certainly doesn’t show the latter to be worse. It’s more a comparison between Pro and Max (and yes, they have different media engines and memory bandwidth).

[@DavidHarry]

@@raw-viking It relates completely to M1 and M4 and the title is an exact description of what the video is, so I don't understand why someone would watch a video that they know they aren't going to like, or criticise, when they knew exactly what was going from the title. Plus, show me anywhere where it confirms that Resolve doubles up on the encoders for the main export. Plus, memory bandwidth has no bearing whatsoever in this scenario as the bandwidth of the encoders is way slower than the memory bandwidth and are the bottleneck. You're taking a maximum of 4K/60 RGB being sent to the encoders and then the encoders working way slower to encode and both of those processes are way slower than then memory bandwidth.

[@raw-viking]

⁠@@DavidHarry It seems clear that Resolve is using both engines. This was one of the reasons I switched to Resolve. And you can see it in the results you’re getting. While it was still possible to get the bandwidth usage I routinely measured 250-295 GB/s exporting video.

[@DavidHarry]

@@raw-viking You still haven’t given any evidence of the dual encoders, it’s just your opinion. Yes, it could be the case but where’s the proof. Now please explain what requires a minimum of 250GB/s bandwidth of any type when exporting any video 🤣 Seriously, you need to have proof of such things, not just make it up because it sounds good in your head 🤣

[@mrchrisbeaver]

I'm very tempted to upgrade from M1 Studio for DAW work, do you think the lower memory bandwidth would be a factor in audio work? Great video, thanks

[@DavidHarry]

DO NOT GO FROM ANY STUDIO TO AN M4 PRO. Wait for the Studios to be released.