Skip to main content

Do computers have a sound?

hey all. I've been bouncing this around around in the noggin. Take DAW of your choice, and picture a basic 24trk audio mix, with a plugin on each channel. This type of session would run on just about any computer hardware from the last ten years.

If you had the identical sessions, running on the exact same hardware, and same drivers/OS, on two separate computers. One as basic as it gets, but can handle this session without errors, even if it's pushed. Another, super high tech state of the art computer.

if you a/b the two systems, while listening live, and then auditioning the mix downs, (however you arrive at 2 track), would they produce identical results? In theory and in person?

I just got curious about this after starting to learn that there's different tolerance/quality levels of eletronic components, which seems to be one of the big selling points on higher end, and original vintage gear. If there was a MOBO created with 'boutique' caps and resistors, does this contribute to the daws final result? Is it a question more of longevity of the system instead? Is it just a limit on peak performance, or reliability?

Comments

pcrecord Thu, 05/05/2016 - 16:44

I'm a bit surprised by the question but let me say what I think
I think digital is a series of 0 and 1 and the sound you are recording or playing back isn't at all modified by what kind of electronic parts your MB CPU, Memory or HDD have and their quality.
It's true, any analog circuits your audio go through (like your audio inferface analog circuit prior to AD or after DA) will change the sound but in the digital domain it won't. At least the industry says so.

BUT, this is a big But... ;)..
We've spoken on tolerances of ressources of the computer. At some point with all the calculations, some bits could be droped or jitter could be introduced in the signal and that can influence the audio quality.
Thing is at this point in computer and software development I would expect that the creators of DAWs would be compensating in some way for the possibility that the CPU/Memore/HDD overuse could at some point degrade the sound
I wish that the sound would not be altered until it's too much and the audio engin start to fail or give clicks and pops to indicate it has too much to do..
But in reality, it is possible that the sofware makers have this big secret of signal tolenrance and the audio is different at a certain amount of ressources use.
This is a theory, no one dared to say, explain or test this. It's just me asking the question, is it possible it could happen?

Boswell Fri, 05/06/2016 - 02:17

Makzimia's correct - the outputs will sound identical, providing everything else that could make them sound different does not do so.

I helped a friend-of-a-friend recently who was transferring his mix setup from a desktop computer to a super new laptop. The software transfer had gone well, and his USB interface plugged into the laptop without problems, but he wasn't happy with the sound from his monitors compared with how it used to be. Before going there and hearing it, I had assumed this was going to be some sort of latency and/or buffer size problem, but no, on hearing it, the sound was "fluffy" and had an indiscernible background to it. I thought for a few moments, then did the simple thing of pulling the d.c. power connector out of the laptop while it was playing. Instantly, the sound recovered to how it had been from the desktop. I left the poor guy struggling on the internet to find a "professional" power supply for his laptop to replace the cheap, noisy POC it had been delivered with.

Beware!

pcrecord Fri, 05/06/2016 - 02:51

I know laptop's power supply are a pain. Specially if you use the internal sound card...
I had many DJ including myself complaining about that. I'm not doing DJ job anymore but I wonder how good those two products would perform
http://www.radialeng.com/jpc.php
http://www.radialeng.com/usbpro.php

Of course it doesn't apply to Bos friend who already had an external audio interface already.

audiokid Fri, 05/06/2016 - 11:25

pcrecord, post: 438368, member: 46460 wrote: I know laptop's power supply are a pain. Specially if you use the internal sound card...
I had many DJ including myself complaining about that. I'm not doing DJ job anymore but I wonder how good those two products would perform

(y)
Same: my experience, laptop PSU can be problematic / weak for pro audio.

kmetal Sat, 05/07/2016 - 13:02

Boswell, post: 438367, member: 29034 wrote: Makzimia's correct - the outputs will sound identical, providing everything else that could make them sound different does not do so.

I helped a friend-of-a-friend recently who was transferring his mix setup from a desktop computer to a super new laptop. The software transfer had gone well, and his USB interface plugged into the laptop without problems, but he wasn't happy with the sound from his monitors compared with how it used to be. Before going there and hearing it, I had assumed this was going to be some sort of latency and/or buffer size problem, but no, on hearing it, the sound was "fluffy" and had an indiscernible background to it. I thought for a few moments, then did the simple thing of pulling the d.c. power connector out of the laptop while it was playing. Instantly, the sound recovered to how it had been from the desktop. I left the poor guy struggling on the internet to find a "professional" power supply for his laptop to replace the cheap, noisy POC it had been delivered with.

Beware!

Is that saying basically the underlying audio/code Ect was the same, and it was the power supply 'masking ' the presentation of the audio/code?

When a power supply effects the sound, is it a matter of just proper supply voltage, or is it a build quality thing? Would in theory a generic and high end power supply 'sound the same' all other things equal? I'm sure it's not a simple thing and probably a question of overall build and design, like interfaces and everything else.

Is it an issue of headroom? That's what got me wondering if cheap small caps, and generic power supplies might be causing some headroom issues. I know Davedog had mentioned a while back that Mac minis were inadequate for recording due to shabby power supply. Is it wise to use more moderate levels on moderate computers, just like how a power starved audio interface like mine, tends to crap out sooner than better interfaces? Or is -18dbfs (or whatever each of you use for standard) the same across the board? Does it differ in ad vs da stages?

the reason I asked the original question, because I had remixed a song on my iPad in cubase, that was originally done on a Mac Pro and DP. I found the sound of the iPad mixes to be more solid sounding. The fact that a $450 tablet and app, on a set of mackies in the living room, could hang or outdo a 2k pro computer, multiple motu interfaces, every pluggin waves had, (I only used a few of my favs), and a set of Meyer HD 1 had me wondering. My experience was not scientific obviously, but it had me thinking about what is really effecting our sound.

All this talk about power supply came yesterday, while I was discovering my new computers psu uses a propriety 10pin plug for the motherboard. Apparently it's common for of the shelf computers to have this. It seems like the quality of psu is a fundamental element in the quality of a piece of gear.

Boswell Sat, 05/07/2016 - 15:45

Every case will be different. In the one I described above, the noisy PSU did not modify the underlying recorded sound in any way, it simply added different types of noise to it. As the power consumed by the computer and USB-powered interface varied with the amplitude envelope of the audio waveform, the added noise got modulated by the audio, hence the "fluffy" sound that Brien liked the name of.

kmetal Sat, 05/07/2016 - 21:53

Fascinating. It's amazing how many possibilities there are. Thanks for the story boz. I'm in a fascination phase with 'straight wire' or 'selective coloration' however u wanna put it. So I'm trying to learn how all the components interact, and how their individual designs and integrations differ, both hardware and software.

DonnyThompson Sun, 05/08/2016 - 03:28

I don't believe that computers have 'a sound", all things being equal in terms of specs and hardware, the only "sound" you would likely get would be a sonic degradation of some kind, like if the hardware had cheap converters, or as Bos mentioned, an insufficient power supply... those things can add bad things to audio, but that's not the same as a computer having a "sound" like certain pres, consoles or other OB gear.

That being said, I do believe that different DAW audio engines have different "sounds"... but it's not my intention to rehash that old debate. You either believe this to be true, or you don't.

kmetal Sun, 05/08/2016 - 09:06

Would say cheap/under spec caps, or an underpowered/inefficient power supply have any effect? If the computer or processor is lacks headroom in a psu sense, would that cause a starved sound, or does it simply not translate like that. What about error rate differences in drive type, like SSD vs hdd? Does less mechanical parts mean less errors? Or does that not matter, because it all goes down the same pipe ususingn standard protocol?

pcrecord Mon, 05/09/2016 - 03:06

I'm no computer hardware specialist. I think there will always be errors but if they are well bellow the tolerance, it won't affect sound.
Underspec caps will fail sooner but I'm guessing they will do their job until they do.
I do record big projects on ssd and small ones on hdd and I hear no difference in the sound quality.
I do use this very often to check my drives status : http://crystalmark.info/software/CrystalDiskInfo/index-e.html

The power supply is another story because many component need power to work properly
In Boswell story, what affected the sound was a lack of power in the usb port which directly relate to the audio interface performance.
I'm guessing (again) that if you'd deprive the cpu of some DC power, it would run slower or crash before you could hear the difference.

Boswell Mon, 05/09/2016 - 04:01

pcrecord, post: 438416, member: 46460 wrote: In Boswell story, what affected the sound was a lack of power in the usb port which directly relate to the audio interface performance. I'm guessing (again) that if you'd deprive the cpu of some DC power, it would run slower or crash before you could hear the difference.

I didn't take measurements or investigate in any great detail in this case, but my conclusions at the time was that the problem was due to circulating currents at the mains power supply's switching frequency.

You hear apocryphal stories about the design process of power supplies of this type being "value engineered" by chopping out components of the clearly over-designed prototypes until the unit fails to work as a d.c. supply, and then putting the last component back. I'm not a switching power supply designer, so my experiences in this matter mainly relate to the design of the audio part of the hardware, but I have on occasion been deeply disappointed by what can happen to a great-sounding prototype after it gets handed over to the Production Engineering department for manufacture. No names.

bouldersound Mon, 05/09/2016 - 07:11

Boswell, post: 438417, member: 29034 wrote: You hear apocryphal stories about the design process of power supplies of this type being "value engineered" by chopping out components of the clearly over-designed prototypes until the unit fails to work as a d.c. supply, and then putting the last component back.

That's paraphrased from a story about Colin Chapman and the Lotus 7. He is said to have removed frame tubes until it collapsed of its own weight and then put the last one back.

As far as digital errors, when I master a CD I export the finished audio, re-import it and do a null test. I haven't done a lot of them but they always null perfectly.

kmetal Mon, 05/09/2016 - 18:53

pcrecord, post: 438416, member: 46460 wrote: In Boswell story, what affected the sound was a lack of power in the usb port which directly relate to the audio interface performance.

That's what's making me wonder if other connection types would be immune to this. For instance if he was using the Ethernet port via Audiodante networked audio. I'm fairly sure Ethernet can also supply similar voltage and amperage as USB/FireWire can. Would the laptop still have been noisy, using a different connection.? If not, perhaps the power supply is only partly to blame.?

Would lack of power in a desktop cause a similar thing? For instance, my stock Lenovo psu is a gross 250w peak (and it's lovely proprietary 10pin) mobo plug. They installed a trash can graphics card stock, which is specs a min 300w psu. Does all the sudden my audio interface lose headroom now? Not just my case but in general. I'm thinking of just taking the card out and using the Intel onboard, to save power consumption, since there's not really a great need for a low end graphics card.

Obviously electrical headroom is good design, from the way it enter the house thru final use. I guess I'm trying to narrow down when other things in a circuit are interfering w audio, and if those things are necessary. In other words down to a mobo component level, how much could you remove from the 'signal path' so everything serves a purpose directly related to audio. As straigh wire as it gets, for a dedicated audio recorder.

I studied the replica/remake circuits and build, for the 1176, (I'll link it if I can find it), but the dude did high end replicas some had point to point wiring, Ect. But as a watched the evolution of the 1176s 3-4 revisions it seemed almost without fail, components were REMOVED, from the design, as the circuit was 'improved' or updated. It seems to hold true in a few of the other things of that nature I've been looking at. It seems like the circuits got consolidated, and smaller, with less stuff, over time.

Now it could be the corporate axe in most cases, like bozs. But even when Danny rebuilt/modded the trident channel modules, he mostly removed stuff, mainly caps, and replaced the ic chips, to gain 6db more headroom in the form of less noise. And also over specd some stuff in the psu and power related things in the strips themselves.

Boswell, post: 438417, member: 29034 wrote: I didn't take measurements or investigate in any great detail in this case, but my conclusions at the time was that the problem was due to circulating currents at the mains power supply's switching frequency.

What is a circulating current at the mains power supply? Is this something to beware of in all psu's, like a cheap import desktop switching power supply.?

bouldersound, post: 438418, member: 38959 wrote: As far as digital errors, when I master a CD I export the finished audio, re-import it and do a null test. I haven't done a lot of them but they always null perfectly.

That's pretty reassuring, at least if it's making mistakes, there in the same spot! I've got about 150 CDs to import, this calms my paranoia a bit.

DonnyThompson Tue, 05/10/2016 - 02:41

Boswell, post: 438417, member: 29034 wrote: but I have on occasion been deeply disappointed by what can happen to a great-sounding prototype after it gets handed over to the Production Engineering department for manufacture. No names.

I guess I always suspected that this happened, but I'm much more aware of it now since I met you, pal, and the stories you've told here on RO about your nice designs being cheapened up the ladder by bean-counting morons, often to save only a few cents; and I suppose I understand that from a business model, those few pennies can and do add up, but if I'm buying something where I know I'm going to pay for a certain quality, as a consumer, I expect that... I'd rather pay a few more dollars and get a well designed and good piece of gear, than save a couple bucks and get crap...

Boswell Tue, 05/10/2016 - 02:52

kmetal, post: 438424, member: 37533 wrote: What is a circulating current at the mains power supply? Is this something to beware of in all psu's, like a cheap import desktop switching power supply?

Most power supplies are commodity items and are designed as such. A computer power supply has to do two things: (a) deliver the correct d.c. Volts and Amps to the computer under varying mains input and d.c. output conditions, and (b) meet the applicable standards and regulations. Unfortunately, these standards and regulations do not include any mention of not causing unwanted effects in attached audio interfaces, so it's pot luck whether any particular power supply design causes audio troubles in such interfaces. The circulating current through the attached gear tends to be at the switching frequency of the power supply (several hundred KHz), and its amplitude and frequency will vary with load.

In my experience, if a better-quality supply does not eliminate burbles, fluffiness and whistles at the interface's analogue outputs, it may be a matter of trying a USB or FireWire optical isolator and powering the interface from a linear supply.

Boswell Tue, 05/10/2016 - 03:04

DonnyThompson, post: 438430, member: 46114 wrote: I guess I always suspected that this happened, but I'm much more aware of it now since I met you, pal, and the stories you've told here on RO about your nice designs being cheapened up the ladder by bean-counting morons, often to save only a few cents; and I suppose I understand that from a business model, those few pennies can and do add up, but if I'm buying something where I know I'm going to pay for a certain quality, as a consumer, I expect that... I'd rather pay a few more dollars and get a well designed and good piece of gear, than save a couple bucks and get crap...

I have often dreamed of getting the same identical paper design manufactured by two different audio manufacturers chosen from the high-end and the low-end of the commercial spectrum. I would then put the two products through a blind evalution process with a panel of audio pros drawn from across the industry. In particular, I would be very interested in whether it was possible to separate out what aspects of the performance were down to circuit design and what was production/manufacture. To the best of my knowledge, this has never been done in the audio field.

kmetal Tue, 05/10/2016 - 07:10

DonnyThompson, post: 438430, member: 46114 wrote: I guess I always suspected that this happened, but I'm much more aware of it now since I met you, pal, and the stories you've told here on RO about your nice designs being cheapened up the ladder by bean-counting morons, often to save only a few cents; and I suppose I understand that from a business model, those few pennies can and do add up, but if I'm buying something where I know I'm going to pay for a certain quality, as a consumer, I expect that... I'd rather pay a few more dollars and get a well designed and good piece of gear, than save a couple bucks and get crap...

Ya know Donny, that's exactly why I'm learning about this stuff. When I started getting serious about commercial gear, I naturally got sticker shock, realizing $100k doesn't really go as far as one might think, and in reality gets you about 24channels of pre/eq/comp at high end prices. This Doesn't count mics. Then you start hearing that 'older is better' and otften it's due do the design, and quality parts. But yet the prices are still top notch, even though the manufacturing is progressively cheaper. Then there's boutique handbuilders making copies of the gear, selling them for even MORE than the original brand companies are. So it told me two things, gear is possible to be reproduced by hand if your knowledgable and capable, and to do so costs in general 50-75% less (so far from my researching). The NOS parts are the difficult part. But even then, so you use a stock Jensen transformer instead of the custom wound Jensen BAE uses, you can still get a high performance piece of audio equipment. Or custom wind your own.

I'll bet the current 414 has been cheapened since I got an xls 8 years ago. They simultaneously came out w the 214, so we're obviously looking into how to cheapen a 414, if I were a corporate person I would also find it beneficial to cheapen the 414s design offing at the same time. Maybe using a cheaper ic chip, the one for the 214. Ect Ect. Next thing you know the 214 and the 414, sound "very similar" which is awesome for advertising blurbs.

Because I'm just getting into it I'm not completely sure, or educated, but from my limited knowledge I think I can assemble needs-based units Ala BAE style, for about $500 or less per channel.

Not surprisingly based on a lot of conversations around here, the power supply circuit diagrams etc. arent always the most accessible part of these things, so I think that's where a lot of The DIY ingenuity and research comes in, but again I'm a newbie to this so it could be that the power supplies are easily accessible for these types of things. There just doesn't seem to be another way to acquire a decent amount of high-end tracking year, as an average professional, especially one without a day job. It also doesn't mean that the home brew gear is "as good" as it's vintage or pro assembled counterparts, but a good design is a good design in a careful build is a careful build, and I think that you can get the effect you're looking for this way. Especially considering when talking to gear or vintage gear none of it sounds really the same anyway, I've had to mesa boogie triple rectifier's that sounded completely different. In other words you can get stuff that sounds "like a neve, or a pultec" in general tendency.

Fortunately, a lot of the best sounding older audio gear is some of the more simple stuff to work with, relative to others electronics designs. I realize it's not everyone's bag, but I'd patiently learn in patiently assemble, then go start a 30 year career at a bank, or live without the top end of outboard tracking year. Sitting around "wanting it" sounds just as bad to me as working at a bank to build up enough savings/credit to just buy the brand-name stuff. So this is a nice bridge, and has been opening up a whole new world of fascination with electronics to me. I'm looking at each little cat and resister as a tone effecting unit.

Boswell, post: 438431, member: 29034 wrote: Most power supplies are commodity items and are designed as such. A computer power supply has to do two things: (a) deliver the correct d.c. Volts and Amps to the computer under varying mains input and d.c. output conditions, and (b) meet the applicable standards and regulations. Unfortunately, these standards and regulations do not include any mention of not causing unwanted effects in attached audio interfaces, so it's pot luck whether any particular power supply design causes audio troubles in such interfaces. The circulating current through the attached gear tends to be at the switching frequency of the power supply (several hundred KHz), and its amplitude and frequency will vary with load.

In my experience, if a better-quality supply does not eliminate burbles, fluffiness and whistles at the interface's analogue outputs, it may be a matter of trying a USB or FireWire optical isolator and powering the interface from a linear supply.

Boz would/could similar issues arise in Ethernet, or pcie card based audio interfaces? Is this exclusively an issue with bus powered interfaces, and/or laptops in your experience?

DonnyThompson Tue, 05/10/2016 - 07:23

kmetal, post: 438435, member: 37533 wrote: Then there's boutique handbuilders making copies of the gear, selling them for even MORE. So it told me two things, gear is possible to be reproduced by hand if your knowledgable and capable, and to do so costs in general 50-75% less.

I'm envious of people like Bos who are capable, and intelligent, and experienced with doing this... I wish I had the knowledge ( and the tools, and the better eyesight, and the hands that don't shake, LOL) to do this myself ... I've often thought about how cool it would be to build a "Fairchild", or a "Pultec"... and, even cooler would be the talent to design my own "whatever" processor; something that was perhaps based upon a pre-existing model - yet with a twist - something I could have that no one else would... but alas, this old dog just doesn't have the knowledge, skill, talent or the experience to do that. I'm not saying it's easy - I'm saying exactly opposite - at least for someone like me, I recognize my own limitations in this regard.

And, maybe that's enough - to recognize and appreciate that the reason that so much of the "boutique" gear does cost substantial money, is because it takes talent and skill and know-how to actually design and build those pieces - talents and skills I don't have.

FWIW ;)

Boswell Tue, 05/10/2016 - 10:08

kmetal, post: 438435, member: 37533 wrote: Boz would/could similar issues arise in Ethernet, or pcie card based audio interfaces? Is this exclusively an issue with bus powered interfaces, and/or laptops in your experience?

Almost never in Ethernet connections as transformer-isolation is integral to the network jacks. PCI/PCIe cards can have their own separate problems, but the manufacturers go to some lengths to make them work acceptably in the hostile environment of a PC chassis. RME is a shining example of this.

As soon as you take digital connections such as USB or FireWire out of a computer to a separate chassis, you let yourself in for a different set of problems, especially if the external power is being provided down the bus. The main culprits are laptops, and particularly their power supplies. Generally speaking, desktop computers give less trouble in this respect,

kmetal Tue, 05/10/2016 - 11:25

Boswell, post: 438440, member: 29034 wrote: As soon as you take digital connections such as USB or FireWire out of a computer to a separate chassis, you let yourself in for a different set of problems, especially if the external power is being provided down the bus. The main culprits are laptops, and particularly their power supplies. Generally speaking, desktop computers give less trouble in this respect,

I must have missed that in the product description of the FW interface I got years back. So many things to consider!! I would complain, but I recorded with effects on a tablet, streaming to a tv in realtime, so technology sure is cool, the challeng is applying it appropriate.

kmetal Tue, 05/10/2016 - 21:51

Brien Holcombe, post: 438452, member: 48996 wrote: Could you explain that to me? Where is my name associated with a review for this product?

Simply because they used the word fluffy in the description brien. I just thought it was funny becasue you commented on boz using it to describe something, and then I see it used again in some random gear description I happened to be looking up a day later. Having not heard the word used much, especially for audio, it was strange to see it used twice in unrelated audio arenas. The reason I directed the comment to you was simply because you seemed amused when boz mentioned it. I was just messing around saying maybe you wrote the description. Perhaps misdirected, but you came to mind when I saw it. That is all I meant by it, just messing around.

Boswell Wed, 05/11/2016 - 02:29

I would never ascribe the term "fluffy" to the Neve sound. "Fluffy" is a word I use to describe the wooliness behind otherwise clean audio when corrupted by low-level amplitude-modulated noise.

I won't identify him, but after reading this thread, a member here sent me a message along the lines of: "At last! A description for what I can hear in my audio. Perhaps now I can fix it!"

Tags

x

Register