As a musician, I sometimes feel left out by tech channels

[The_Prycer]

Inb4 Cinebench

 

Look, I get that audio production maybe isn't the highest priority for LTT or other tech channels, there are content creators who focus on musical hardware and I guess they do a moderately okay job of it (while getting 1/10th of the views). However, one of the big things I feel is missing from reviews (particularly when Apple products are involved) is rendering times for high-quality audio files from moderately popular DAWs. Trying to find a baseline on how long a 4-8 track, 3 minute song takes to render at 320kbps in to a WAV or Mp3 format is nigh-impossible and when I get approached by other musicians on tech recommendations for their home studios I end up having to dig deep in to the Internet to try and find answers to some of their questions. 

 

I understand that certain VSTs are going to be more CPU heavy than RAM, if you make simple 4 track songs versus deeply complex, automated and EQ'd 32 track masterpieces your hardware requirements will differ greatly from one another (honestly you could probably dedicate an entire YouTube channel to that topic alone). I get that every DAW functions just a little bit differently and trying to cover them all would be a nightmare. However, writing the same simple 4-6 track 3 minute jingle in Garageband or Logic on every Apple product could provide a nice baseline for people who are looking at upgrading to a newer model for the sake of efficiency as Apple does some really neat things with their software in terms of short term CPU boosts for better renders than their cooling solutions and core count would appear to allow.

 

I'm not suggesting Linus & co. go out and buy an Ableton, FL Studio, Re:Noise, Cubase, Magix, and Reaper license to test all of this out. Just that maybe it's something to look in to because we are out here and being overlooked because apparently everyone is either a gamer, video producer, or digital artist now.

 

I guess what I'm trying to say is, Musical production is a big part of the "prosumer" scene and seeing it disregarded in favor of only video production or graphic design makes me a bit sad. 

[Crunchy Dragon]

Does LMG even have people experienced enough in musical production to make testing these applications worthwhile and profitable?

 

I agree, musical production is a big part of the professional scene, but I don't think it would be completely worth Linus' time to hire somebody who's proficient in that skill for the purpose of benchmarking different systems and applications specifically for that purpose.

[minibois]

It would be so cool if there was a benchmark tool that tests your system entirely in different program scenarios, including music productions.

 

It would only need to have a portion of:

- Ableton or another DAW to simulate a music track 'render' (probably not the correct term)

- a video editing program to simulate a video render

- a Blender render test..

- a Programming compile test (underrated benchmark test too!)..

- compression test (e.g. via 7zip)

- a Gimp image processing test

- etc. 

 

That could give a good grasp of how a specific PC/component will perform in such test.

Plus it would make it so much easier for reviewers to just run 1 program, that runs these applications.

Would be easier to retest equipment too, with improvement to these aforementioned programs.

 

This is way outside my boundary of programming, but I'm gonna check if this would be a viable solution to make.. some day.

[homeap5]

Graphics is very easy to benchmark - all you need is to run few games on few different graphics cards and you'll get results in real-time. Problem with music production is that type of tests have no sense since cpus are powerful enough to render music in realtime, so you no longer need to mixdown your project to hear what your song sounds. Final mix is needed once and sure - you can compare rendering speed, but that is just small fraction of time you spend on whole project. The same way you don't testing which text editor saves faster. Or even how fast final project is saved in Photoshop as flatten tiff. It's just not necessary. I don't believe that anyone will choose DAW only for it's final mix speed.

 

Sure, I remember days when more channels may produce delays, stuttering or whatever, but that was years ago. Now I can't imagine how many plugins you should add to your tracks to made any programs reach 100%.

 

This is, imo, main reason why nobody tests and compares daw rendering speeds.

[The_Prycer]

12 hours ago, Crunchy Dragon said:

Does LMG even have people experienced enough in musical production to make testing these applications worthwhile and profitable?

 

I agree, musical production is a big part of the professional scene, but I don't think it would be completely worth Linus' time to hire somebody who's proficient in that skill for the purpose of benchmarking different systems and applications specifically for that purpose.

For the expansive PC and even Linux DAWs I don't think there's enough time unless they only chose to focus on something like Cubase or FL Studio exclusively. I will say that I know FL comes with artist's songs already laid out for study so whoever was benchmarking simple export times to various audio qualities wouldn't even need to know anything about audio production, just record how long their test bench took to export the file at whatever quality and file format was chosen.

 

As for Apple (given their seeming lack of upgrades to most products) I'd suggest having someone just go in and make noise. It doesn't even have to sound good. Just have a certain length, track count, and file size upon exporting.  But if Linus really wants one, I'll write him a song he can use to benchmark Apple products entirely in Garageband.

 

As for tests I'd like to see done, they don't really have to be familiar with audio production, mostly just conducting stress tests in the form of adding more audio tracks until stuttering or crashing occurs, seeing how responsive the program responds under different hardware conditions, and export times of commonly sized audio files at high bit rates.

 

While these tests could certainly go deeper, having simple benchmarks of "this product exported the file at X% faster/slower than a competitor" or "this product didn't stutter after we recorded Y number of tracks in to it" would do it for me.

 

11 hours ago, homeap5 said:

Graphics is very easy to benchmark - all you need is to run few games on few different graphics cards and you'll get results in real-time. Problem with music production is that type of tests have no sense since cpus are powerful enough to render music in realtime, so you no longer need to mixdown your project to hear what your song sounds. Final mix is needed once and sure - you can compare rendering speed, but that is just small fraction of time you spend on whole project. The same way you don't testing which text editor saves faster. Or even how fast final project is saved in Photoshop as flatten tiff. It's just not necessary. I don't believe that anyone will choose DAW only for it's final mix speed.

 

Sure, I remember days when more channels may produce delays, stuttering or whatever, but that was years ago. Now I can't imagine how many plugins you should add to your tracks to made any programs reach 100%.

 

This is, imo, main reason why nobody tests and compares daw rendering speeds.

I admit it's a trickier thing to benchmark, but having someone at least make some noise with a few different plugins, maybe some vocal layering or guitar tracks, and just try to make a DAW go slowly could provide some neat metrics as to what seems to be the more important parts of your audio workstation.

 

Will the 6 core 2.2Ghz processor hold the project back versus a 3.7Ghz 4 core? Are hyper threaded CPUs able to produce the same project faster? Do your really need 16GB of RAM to make trap beats, etc.

 

But yes, I also remember when a 6 track recording would cause an entire project to halt upon playback because the system just couldn't handle it.

[Flying Sausages]

Do you make trap music? Do you have a mixtape?

[.Apex.]

50 minutes ago, Crunchy Dragon said:

Does LMG even have people experienced enough in musical production to make testing these applications worthwhile and profitable?

I think a one off would be adequate, they do this kind of thing all the time, once you come up with a real world standardized test you can reuse that for a long time, so they don't have to hire anyone full time

Though this definitely doesn't fall on LMG, I'm just saying it's simpler than it seems

54 minutes ago, The_Prycer said:

However, one of the big things I feel is missing from reviews (particularly when Apple products are involved) is rendering times for high-quality audio files from moderately popular DAWs.

It's not going to make a difference between a minute or two of rendering, so that test doesn't make much sense honestly, no one upgrades to improve audio rendering times

29 minutes ago, homeap5 said:

Sure, I remember days when more channels may produce delays, stuttering or whatever, but that was years ago. Now I can't imagine how many plugins you should add to your tracks to made any programs reach 100%. 

That's not true though, it's a lot easier than you think to stress a CPU to 100% with VST's

[Fasauceome]

Basically they need a cinebench for music

[The_Prycer]

24 minutes ago, OlympicAssEater said:

Do you make trap music? Do you have a mixtape?

I write Industrial and Doom Metal mostly with some DnB influences, but I do have a mixtape and it is straight fire

[Flying Sausages]

1 minute ago, The_Prycer said:

I write Industrial and Doom Metal mostly with some DnB influences, but I do have a mixtape and it is straight fire 

Link to your mixtape?

[homeap5]

1 hour ago, _Syn_ said:

That's not true though, it's a lot easier than you think to stress a CPU to 100% with VST's

I'm not "thinking", I'm using it. Before on i5 3470, now using i7 8700k. Maybe I don't use as heavy vst as yours, but I'm also not trying to optimizing by using lighter ones. And I knew person who uses even worse computer than my old one.

 

But I can imagine that someone uses old 32 bit software that do not utilize more than one cpu core and have problem.

[LogicalDrm]

Order to this be a thing, they would first need someone who understands audio to begin with. They did ask for such person in last recruitment ad, so hoping that someone got it with that background. But until then, let your ears bleed as much as my head aches when they even mention GPS, mapping or 3D point clouds.

[homeap5]

@The_Prycer A little ot - I'm interesting what you're doing. Can you PM me? Only small percentage of users create something so being old freak (and musician too, sort of) I want to hear your work.

[The_Prycer]

5 hours ago, LogicalDrm said:

Order to this be a thing, they would first need someone who understands audio to begin with. They did ask for such person in last recruitment ad, so hoping that someone got it with that background. But until then, let your ears bleed as much as my head aches when they even mention GPS, mapping or 3D point clouds.

And here I could have been applying to work for Linus had I joined the forums earlier.

 

Maybe they still need a "full builds for $400" or "how to do basic msconfig to speed up your system" or even a "720p gaming" person

 

My ears bleed whenever I see someone using earbuds for mastering. At least have the decency to go do the car test if you can't afford studio monitors.

[LogicalDrm]

2 hours ago, The_Prycer said:

And here I could have been applying to work for Linus had I joined the forums earlier.

 

Recruitment ads were on all of their social as well as YT and WAN Show.

 

 

2 hours ago, The_Prycer said:

Maybe they still need a "full builds for $400" or "how to do basic msconfig to speed up your system" or even a "720p gaming" person

 

My ears bleed whenever I see someone using earbuds for mastering. At least have the decency to go do the car test if you can't afford studio monitors.

 

[homeap5]

5 hours ago, VegetableStu said:

car test? as in play back the file in a car? o_o

Well, probably if it's mindless techno, it will be good idea.

 

[The_Prycer]

5 hours ago, VegetableStu said:

car test? as in play back the file in a car? o_o

Yes.

 

The car test is to put your mix on either a CD, a USB stick, or even a phone and then go listen to it through a car's stereo system. It's not the absolute best way to identify flaws in your mix, but if you don't have studio monitors it can be a valuable tool in determining if what sounds great through a set of headphones falls flat or is overpowering through a more robust stereo system that includes a subwoofer.

 

Sort of the same idea as using a lower hertz tv monitor to test your gaming PC's performance. It isn't going to be as pretty as if you hooked it up to a high refresh 1080, 1440, or even 4k monitor, but it'll at least let you identify if something is obviously flawed.

[RasmusDC]

just have to ask, does it matter? 

 

if a PC is fast at a multicore cinebench element, does that not corrolate into it being fast at encoding audio, unless we are talking hardware supported codecs? or am i missing something..

 

i´m just curious, not a .. ."angry thread", but synthetics are most applicable in my case to things like rendering and encoding, this is where they are most usable, since it normally is a CPU intensive load, and multi core in some areas, are more value than single core speed.

 

so if you have a reference point on an old macbook,, and see the new one being 40% faster in cinebench, would it not be around there in encoding?

 

i guess the same mindset can be used on games, test something GPU heavy, something CPU single core enhanced something multicore enhanced when talking CPU´s and stop there..

 

so i might be contradicting myself.

[The_Prycer]

4 hours ago, RasmusDC said:

just have to ask, does it matter? 

 

if a PC is fast at a multicore cinebench element, does that not corrolate into it being fast at encoding audio, unless we are talking hardware supported codecs? or am i missing something..

 

i´m just curious, not a .. ."angry thread", but synthetics are most applicable in my case to things like rendering and encoding, this is where they are most usable, since it normally is a CPU intensive load, and multi core in some areas, are more value than single core speed.

 

so if you have a reference point on an old macbook,, and see the new one being 40% faster in cinebench, would it not be around there in encoding?

 

i guess the same mindset can be used on games, test something GPU heavy, something CPU single core enhanced something multicore enhanced when talking CPU´s and stop there..

 

so i might be contradicting myself.

For the most part this is true, but this was more me saying "I keep seeing cinebench and Adobe premiere times, but where are my Logic Pro times?"

 

It annoys me that when I see LTT review a Mac product they never delve in to it's performance in Garageband or Logic despite that proprietary software being a major reason for purchase for a lot of people, but they'll happily talk about render times, scrubbing, and all manner of other video production related topics that I barely care about.

[RasmusDC]

28 minutes ago, The_Prycer said:

For the most part this is true, but this was more me saying "I keep seeing cinebench and Adobe premiere times, but where are my Logic Pro times?"

 

It annoys me that when I see LTT review a Mac product they never delve in to it's performance in Garageband or Logic despite that proprietary software being a major reason for purchase for a lot of people, but they'll happily talk about render times, scrubbing, and all manner of other video production related topics that I barely care about.

i do agree, but the hate is strong, now i am not a mac fan, i loved my thinkpad, i did however buy a unibody early 2010 Macbook, and it still lives, and is a nice laptop to work on with a SSD and a system update.. so my son has it.

 

they can see value in RGB but not paying for design, it is a bit "fun" but i think you need to look elsewhere., this is not a mac friendly channel.