seems you are quite a shit disturber John! howdy hehe

its all fun,

Yes to to each their own of course. I thought this exact thing for years and espoused the benefits of analog summing on various forums. The more I tested and tried to prove my opinion the more I realized I couldn't. Where I've ended up is believing that digital summing doesn't have a sound and that great analog gear sounds great. I'm not asking you to believe that. I've never been able to prove digital summing even comes close to tampering with audio. Every time I think it is it turns out to be something else. How have you proven to yourself that it is digital summing has a sound?

Would you like fries with that?(I mean this in all humor and not condicention)

Kurt Foster, post: 409436 wrote: let's try to not make this personal ...

i am not advocating any expenditure. i'm not saying go buy this or that. i am only saying that summing analog sounds better than summing digitally. if you think it doesn't or that digital summing sounds better or there is absolutely no difference based on a test you performed on your computer fine. if you trust what a computer shows you over what your ears tell you, good for you. i hear a difference and it's not subtle.

if you don't trust you own ears, you should be flipping burgers, not mixing audio ....

@ audiokid:

LMAO!

I totally get what your saying and it is a good line of thinking. Your line of thinking begins with "you can't trust a computer" and " it can't possibly make all the calculations it needs to and get it right". My line of thinking is "I haven't been able to prove a computer is miscalculating or making errors" and "I am choosing to trust that computers are accurate in areas I haven't tested yet because they have been were I could test". Neither of these is more valid then the other IMO. Based on these two views our decisions each make sense.

We agree that something is lost everytime you ADDA. Certainly not as much as back in the day of tape or even earlier converter days but a difference none the less.

I spent a full year mixing OTB using only stock plugs. I round tripped sometimes to get my 2-bus mix. Other times I recorded to my ADAT HD that I was using for converters anyway. I monitored from my D&R console. The ADAT was hooked up via adat cable so it doesn't meet the separation requirement you have but I was not doing a round trip. At the time I did an experiment where went Daw -> interface/converters-> console -> monitor analog->record in adatHD and record in round trip to Daw. I then imported my adat file into the daw and nulled it against the Daw file perfect null. This isn't perfectly conclusive but it was good enough for me at the time.

For my workflow I found being able to use plugins after analog was a benefit. The additional automation and keeping analog gear quiet automating AFTER the analog stage was a massive help. Instant recall from recording my tracks back in that had analog gear on them was the biggest help. For me all this was a bigger benefit then the potential problems I couldn't prove or hear about digital summing(which was what I was trying to prove at the time).

This is a workflow personal taste with built in trade offs. Your workflow will have trade offs as well that impact your ability to mix and ultimately effect the sound of your mixes.

It is possible that I hear the phase you distain and quite like it. I am perfectly happy with the state of pop mixing and think most pro recordings sound killer(the same ratio that always sounded good). I think mixing has progressed in the last 10 years. Sure there are cheap sounding recordings like the glee/auto tune soundtrack that sounds very amature to me. Then there are the black keys stuff that sounds big, bold and vintage mixed ITB by Tchad Blake(or so I recall).

I dunno but then again I trust computers.(sort of. I hate them to)

audiokid, post: 409438 wrote: Indeed. not to mention.... the time it takes to return back to the session.

The concept that the DAW will correct this latency to exact is also something I doubt is that accurate. But hey, everyone thinks its rocking solid.

So lets put this together...

Why would you want your analog hardware ( any hardware) to remain constant in the first place? Does any of this make sense to you? Not to mention, why would you want to redigitize it two more times before it is even summed with the final mix? You are forcing two extra SR ( ADDA) and time shifting it as well. You get this right? I bet you are using plug-in on that round trip too?

Summery:

You mix ITB, DAC a track OTB to an analog product ( that you expect would add some glue) then return it BACK to the same session in hopes it not only improved the SOUND of that track, but even more ironic, re sampled it back to the original session SR in hope it made a difference.

And now, you are going to tell us that it will null or something.
I'm currious as to why you would want your lush analog track to sound similar to what it was before it left? I mean, you are returning it back to the same session. Do you ever think that the analog is in better shape before it returns back to that same session?

I'm sorry but its no wonder you are having trouble with this.

John. It is all good. I don't feel anyone here has gotten too far out of hand or denigrated to name calling. Has been a stimulating discussion.

Paul999, post: 409451 wrote: Where I've ended up is believing that digital summing doesn't have a sound and that great analog gear sounds great.

EXACTLY!!! add to that the more you ask of the DAW the worse it sounds.

Paul999, post: 409451 wrote: ........ I've never been able to prove digital summing even comes close to tampering with audio. Every time I think it is it turns out to be something else. How have you proven to yourself that it is digital summing has a sound?

take a multi track file and run it out of the box to a mixer (mackie / beheringer anything) so each track has it's own channel on the mixer (this is how the big boys do it). use a pre fader send on the DAW. now put up a simple no efx or eq mix of the file on both the DAW and on then mixer trying to keep the levels the same as best you can (remember louder always sounds better).

now tell me which one sounds better? i already know the answer.

kmetal, post: 409446 wrote: well i could. cuz apparently pro tools m-powered doesn't even have plugin delay compensation. as far as the liquid mix, my best guess is the quality of the dsp chip design, the buffer size (2048) and the bandwidth of the firewire bus.

i'm not a super tech, so i can't explain other than what i hear, which is similar to the characteristics of something that introduces phase cancellation (thinning, depth collapse), and a buildup of nasty upper mid range frequencies, and grainy sound, which i have heard in badly designed conversion/interfaces, like the digi 002.

again my real point in all this is overuse of plugins. if i had 32 channels of nady eq's comps, and a few berhinger verbs, i would think about what really would benefit from their use, if anything at all. same for plugins. which is where i believe people alot of people miss the boat. the first thing my boss told me was to take the plug-insoff my mix i submitted for my application, i did, and it changed my mixes ever since. they were fuller raw, than hyper processed. i started using buses instead of verbs/delays on individual tracks, and leaving well recorded things alone. but thats just me everybody has their own way.

i'm a big fan of [="http://www.pensadosplace.tv/category/into-the-lair/"]Into The Lair - Pensado's Place[/]="http://www.pensados…"]Into The Lair - Pensado's Place[/] and he uses real top dollar commercial releases for his examples. notice how few plug-inshe uses, and how he employs mostly four buses, and does the bulk of the processing itb, w/ a couple pet OB pieces.

here's another one [[url=http://="http://www.youtube…"]Engineer Makes Rihanna's "Diamonds" Shine Bright - YouTube[/]="http://www.youtube…"]Engineer Makes Rihanna's "Diamonds" Shine Bright - YouTube[/] Rhianna's smash hit "diamonds" well over 80 tracks, notice how few plugins, and the mention that he also uses a few pieces of OB. i'm not making this stuff up.

as far as a null test goes, that still doesnt address anything about the subjective quality to the sound that a pluggin is adding.

imagine a ruler flat group of analog eqs for instance, with a 1k test tone running through it will show flat response at 1k, even if you stacked them 10 times in a row, and it still showed flat at 1k, it doesn't take into account the fact that the noise has stacked, and you can hear it. so even if it's still flat at 1k, when you listen to it you still hear an increase in noise floor.

so back to the bob katz thing, any time you change your digtal audio you change the code. and to me w/ plug-insthat change adds up into the things i've described. if i wanted to be a computer scientist i would've been. but technological tests, and numbers don't always relate to the subjective art. as far as i'm concerned i don't belive their is no way to eliminate the problem at this point, but only to minimize it.

ps- by the time i could afford the trinity/antelope, they'd be on trinity/antelope mk5. i'm interestid this to try and make my investments last longer before upgrade, and possibly maintain value like my amps, and mics, and guitars.

Every issue you hear and describe in this post can easily, and more likely, be attributed to "pilot error". Not compensating in your daw manually if it needs it will cause phase issues. Diligently attending to this will take care of your issue of hearing a plug in engage.

Not using a lot of plugs as you've witnessed on pensados place etc. again this is mixing skills. Over processing is a mistake many engineers make. God knows it took me years to scale back how much processing I used ITB AND OTB. This is not the plug ins fault nor a calculation error. This is the engineers fault. Less is more as we've all learned.

DonnyThompson, post: 409448 wrote: At the very least, could we all agree that no two plugs are the same when comparing apples to apples? Is the difference in the code? The processing chain? The conversion? The phase? Somewhere else?

There's no doubt in my mind (or in my ears either) that different plug manufacturers sound different from one another. I didn't say necessarily "better", I said "different", although I've heard certain plug ins sound better than others. And, this being the case, could we not also assume that it's also possible that in being different characteristically, that it will certainly effect the outcome sonically?

I agree with Paul that plugs are more "transparent" in terms of coloration in the "classic" sense. There's no way a plug will sound like a real LA2 or an 1176. One is a digital emulation and the other has electronic wires and and maybe even tubes - gear that gets hot - and when you compare the real thing to the emulation of an effect, or, the mathematical processing of to 1's and 0's.... there will be a sonic difference.
So, in that regard, it's correct to say that one is more "transparent" than the other.

I can only say that in my own experience (and ears) that I've found that plug ins treat the sonics differently than analog gear does, and accordingly, the sonics sound different, and often, unpleasant.

Hell, for that matter, I can also personally attest to and say that no two DAW programs are alike. I've used Sonar for years, I know it blindfolded.... yet when I tried Samplitude several months back, I heard a difference. Comparing two exact mix projects (apples to apples, no processing, just raw wav files played back at the exact same levels and pan settings) between Samplitude and Sonar, Samplitude sounded better to me...more defined, less "smeary". Is this a phase thing? Perhaps. It could be several different things, as mentioned above, it could be in the code itself.

Very few have the budget to buy the top notch gear, be it OB boutique or multi-thousand dollar external clocks, so they make the best of the gear that they have with their situation at hand, and, let's face it, I don't think that plug ins are going anywhere.

As Chris mentioned, the best thing to do is to be choosy about what you process.... and, how you process it. To simply throw compressor plugs on every single track, without regard to whether it's actually needed or not, is, I think, a bit foolish, if for no other reason than that you'd be taxing your processor unnecessarily.

New users in particular are more apt to go crazy with the amount of per track processing they use, either because they think that they need to, or, because they are trying to justify their purchase of a library of processors, or, simply because they are ignorant.... how many times have we all answered questions here from new users who are inserting gain reduction on every track, yet their question is always something similar to "so what does a compressor/limiter actually do?"

IMHO, I think it's naive to think that these ITB processors aren't effecting the overall sonics... after all, on one hand, they are supposed to. The problem is that, on the other hand, they are not always effecting the sonics for the better.

IMHO of course.

-d.

Well put. I agree plug ins effect the sound differently the analog gear. Not better or worse IMO. it is about how you use them. I find analog generally quicker and lean towards its sound. I do not view plugs as inferior based on results I've heard other people get with it though.

I haven't heard a difference between cubase, logic, reaper and protools Mpowered. Samplitude maybe different. I haven't tried it.

Paul999, post: 409451 wrote: Yes to to each their own of course. I thought this exact thing for years and espoused the benefits of analog summing on various forums. The more I tested and tried to prove my opinion the more I realized I couldn't. Where I've ended up is believing that digital summing doesn't have a sound and that great analog gear sounds great. I'm not asking you to believe that. I've never been able to prove digital summing even comes close to tampering with audio. Every time I think it is it turns out to be something else. How have you proven to yourself that it is digital summing has a sound?

Would you like fries with that?(I mean this in all humor and not condicention)

Here is a bunch of stuff that has been floating around my head for years. I'm not saying I have any proof other than I've lived through 4 decades of professional music and 3 of those where using a sampler and DAW of some kind.

Look at all the people having issues. Look at the list of top level engineers, all doing the round trip that are admitting something sounds better when they use an external clock. If that isn't the writing on the wall. I bet, very few ITB engineers would rant about an external clock like these testimonials.

The Round Trip and the over use of plug-ins is destined to fail with certain styles of music. As technology advances, so will the electronic level. I've been sampling since 1979 and every year electronic music has surpassed the advancement of acoustic music in sound quality and volume. I am part of the demise of real musicians. I know what I'm talking about here. But, its not all that bad if we backup and think where it is going wrong.

Electronic music will always be able to push harder because it has less limitations. The transients and real world space has been optimized and removed. Therefore, the information is better quantized and easier to keep in the time line. ( clock) following... ?

If you put acoustic information beside optimized samples, you will never compete until you clean up the noise and chop off the transients. But then, what do you have left? Too much of something and not enough of the other. Isn't that a bitch to mix lol!

This debate is all about changing the way engineers think about the DAW. The DAW is a sequencer / sampler and people are still connecting hardware to it while its spinning in real time. They are sampling in real time and expecting acoustic music to sound natural and in phase. Its that not totally stupid?

That's why electronic music is so popular and growing, and why electronic music sounds so huge compared. All the acoustic information has long been removed or synthesized. Thats what we do as electronic geeks. . When you do this, know this, you are able to push the levels up far beyond real sounding music will ever go, UNLESS! You go about it a different way and even out the playing field.
I'm convinced, if we are trying to contain acoustic information ITB and compete with samplers, you got to go about it a different way. You sure the hell aren't going to do it simultaneously! Round Trip.

DAW's are samplers - samplers work great for electronic music.
DAW's are Sequencers. They sync bytes in a time line. This is an ideal world for quantized formats. Real musicians and the real world is not quantized.

Its so obvious.

When we stop trying to make a DAW think like a musician, and control our ravish behavior with these ridiculous plug-ins better suited for electronic music, I think we will start making better music. I'm not apposed to all the special effects or electronic music, but when it comes to real live sound, OMG, this whole ITB play-station sounds like ass to me.
You cannot use the DAW like an analog console and expect it to produce the same results with acoustic music in one step.
Its our process that needs to change.

[quote=Kurt Foster, post: 409455]EXACTLY!!! add to that the more you ask of the DAW the worse it sounds.

I've seen no proof of this. How did you prove that? Where if the threshold were you hear it? Do you hear your audio degrade everytime you add an overdub from the additional processing? Do things sound worse while you record and then better when you play back because of the decreased strain on your computer? If you open a word document while playing back audio does your stereo field narrow? If you surf the internet while listening to itunes or downloading software do you hear this same effect? You should be if it comes down to computer errors from asking your computer to do more work.

take a multi track file and run it out of the box to a mixer (mackie / beheringer anything) so each track has it's own channel on the mixer (this is how the big boys do it). use a pre fader send on the DAW. now put up a simple no efx or eq mix of the file on both the DAW and on then mixer trying to keep the levels the same as best you can (remember louder always sounds better).

now tell me which one sounds better? i already know the answer.

The only thing this proves is you like the sound of the hardware you are running through. To me I much prefer Daw summing to any Mackie or Behringer I hear. Years ago I used a Mackie VLS 24 channel board as a summing solution for a few months and thought I preferred it for a while. It turned out I was fooling myself. The smear it added was not cool and I was getting no mojo from it. You may like how it sounds. Cool!

audiokid, post: 409459 wrote: Here is a bunch of stuff that has been floating around my head for years. I'm not saying I have any proof other than I've lived through 4 decades of professional music and 3 of those where using a sampler and DAW of some kind.

Look at all the people having issues. Look at the list of top level engineers, all doing the round trip that are admitting something sounds better when they use an external clock. If that isn't the writing on the wall. I bet, very few ITB engineers would rant about an external clock like these testimonials.

The Round Trip and the over use of plug-ins is destined to fail with certain styles of music. As technology advances, so will the electronic level. I've been sampling since 1979 and every year electronic music has surpassed the advancement of acoustic music in sound quality and volume. I am part of the demise of real musicians. I know what I'm talking about here. But, its not all that bad if we backup and think where it is going wrong.

Electronic music will always be able to push harder because it has less limitations. The transients and real world space has been optimized and removed. Therefore, the information is better quantized and easier to keep in the time line. ( clock) following... ?

If you put acoustic information beside optimized samples, you will never compete until you clean up the noise and chop off the transients. But then, what do you have left? Too much of something and not enough of the other. Isn't that a bitch to mix lol!

This debate is all about changing the way engineers think about the DAW. The DAW is a sequencer / sampler and people are still connecting hardware to it while its spinning in real time. They are sampling in real time and expecting acoustic music to sound natural and in phase. Its that not totally stupid?

That's why electronic music is so popular and growing, and why electronic music sounds so huge compared. All the acoustic information has long been removed or synthesized. Thats what we do as electronic geeks. . When you do this, know this, you are able to push the levels up far beyond real sounding music will ever go, UNLESS! You go about it a different way and even out the playing field.
I'm convinced, if we are trying to contain acoustic information ITB and compete with samplers, you got to go about it a different way. You sure the hell aren't going to do it simultaneously! Round Trip.

DAW's are samplers - samplers work great for electronic music.
DAW's are Sequencers. They sync bytes in a time line. This is an ideal world for quantized formats. Real musicians and the real world is not quantized.

Its so obvious.

When we stop trying to make a DAW think like a musician, and control our ravish behavior with these ridiculous plug-ins better suited for electronic music, I think we will start making better music. I'm not apposed to all the special effects or electronic music, but when it comes to real live sound, OMG, this whole ITB play-station sounds like ass to me.
You cannot use the DAW like an analog console and expect it to produce the same results with acoustic music in one step.
Its our process that needs to change.

On all this we agree. We may use the information for different purposes but we do agree. This is about engineers shaping music. When our thoughts and beliefs align with what is in front of us technologically we can create high level art.

To take a previous comment one step further I have a serious question for all those convinced digital summing has errors, round tripping causes errors and/or the more you draw on a CPU the more it interferes with your audio.

Why is it that these computers and daws all select only the audio engine to make errors with? Why does it not effect the video card. For examples I've had video glitches and freezing just as I've had audio clicks and pops but never in the history of any of these arguments has an engineer ever said "I notice the colors on my daw getting lighter as I process more audio" or "the pixels on my text are getting blurry around the edges as I sum my analog mix" or "my vu meters fade ever so slightly when I turn on a plug in". These kind of mistakes tend to be catastrophic on computers not cumulative. Against all odds though our audio seems to be different.

Why are computers so set on taking the edge off of our audio but not video? Do photos fade the more you process them? I've never heard of that.

Is is there a plausible explanation?

FWIW,

The last post and all the other ones that have been entered by me are 100% focused on improving the way we process or combine acoustic music with electronic using both analog and digital tools, including, my own personal suggestions on how I open up and capture acoustic information through uncoupled DAW's in the hybrid environment.
My personal workflow does not exclude digital summing, in fact, its where I end up. If I am working with vsti I will go directly to the mastering matrix and final sum/capture on DAW 2 which is the optimized DAW for summing and mastering.

I simply create more choices and ways to keep acoustic or digital information in or out of an analog state up until it reaches the finish line on DAW2, the gateway to the www. This is where I monitor everything.

If you are really interested in the attributes analog offers in a hybrid system, I personally think your are destined to fail at this. I feel you are spinning your wheels here.

You are exactly right which has nothing to do with analog and everything to do with math.

Paul,

I feel like I'm in a cooking class
You are claiming that you can take all the ingredients and toss it into one shinny pot, cook that baby up and slap it on a plate for $10.95
You are treating audio like its all made of beans but also want some spice and potatoes in this. To me, you are throwing potatoes in at the start when the beans need 6 hours to cook. In the end, you have a meal but its all one big mess that was cooked in one big pot.

hehe.

I'm of course having fun here. I'm sure Bos or someone with good technical knowledge will eventually answer your valid questions. In the mean time, I'm having fun poking at all this.

Round trip reminds me of - planting carrots in the spring, mid summer we pull them out and transplant them to another location in hope they are bigger at harvest time. Why not just plant them better in the first place.

And that goes for most of this nonsense.

Ha! Damn Chris those analogies are priceless. Love it!

I've transplanted carrots before. Half of them won't survive.

Round Trip

Before Quantization

After Quantization

Attached files

Paul999, post: 409460 wrote:
I've seen no proof of this. How did you prove that?

i hear it.

Paul999, post: 409460 wrote:
Where if the threshold were you hear it? Do you hear your audio degrade everytime you add an overdub from the additional processing?

yes to a degree. best case scenario is no processing in the DAW. no plugs, no eq, no dynamics. straight record / playback and edits that all. all faders set to unity gain all fades and effects dynamics etc. in analogland. a stand alone recorder or second DAW to capture the 2 mix.

Paul999, post: 409460 wrote:
Do things sound worse while you record and then better when you play back because of the decreased strain on your computer?

no because when i'm doing (an) overdub(s) i am monitoring the input signal. on playback it sounds worse. the more overdubs, the worse it sounds.

Paul999, post: 409460 wrote:
If you open a word document while playing back audio does your stereo field narrow?

well i don't know because i don't use my music computer for anything other than a DAW.

Paul999, post: 409460 wrote:
If you surf the internet while listening to itunes or downloading software do you hear this same effect?

same answer. my DAW's never see the internet. period!

Paul999, post: 409460 wrote:
The only thing this proves is you like the sound of the hardware you are running through. To me I much prefer Daw summing to any Mackie or Behringer I hear. Years ago I used a Mackie VLS 24 channel board as a summing solution for a few months and thought I preferred it for a while. It turned out I was fooling myself. The smear it added was not cool and I was getting no mojo from it. You may like how it sounds. Cool!

you mention a narrowing of the stereo field. it's not that simple. it's not just left / right it's front to rear.

well done audio can exhibit an effect of depth. even in mono, there are ways to make an instrument appear in front or to the rear. with itb it is difficult at best to pull this off ... when you mix in analog the sound just comes alive.

some people have never heard this because they have only worked on DAWs itb. . if you have never heard this phenomenon it's easy to understand why you would say the things you say.

there is a reason why all the major producers and mixers still work with a large format console for the most part. if what you are advocating was really true, API Neve and all the other console builders would be out of business.

I agree on all points. (from my perspective)

Personally, I don't think many of us have been working with analog (hybrid) correctly for years. I also think this is why many engineers with great hearing have dropped analog. But not because ITB sounds better, because they didn't take the time to do it correctly with the DAW.
(But, ITB does sound stellar for electronic music. This I cannot deny.)

I am more certain, which took having the 10M , the piece of the puzzle for me to confirm something.
I'm ready to buy a console again. My Neos of course would be hard to beat for headroom and a summing system, the ability to have a console with the basic EQ and faders in front of it would be a stellar add-on vibe. More ways to mojo

I'm going to try and make a deal with a few console makers this year. I've approached one now, I'm hopping it goes.
I seriously think we are on the cutting edge of exposing a whole bunch of clues. I know this isn't new news, but I do think the two daw system now confirms a very important aspect to better analog.

Its going to get interesting, that's for certain.

And just to clarify, I am by no means disconnecting myself from the DAW's ability to do unbelievable effects, I get that, and embrace this. The part that is most interesting is the uncoupled second DAW and the avoidance of round trip.

Analog console, here I come.

Hey Kurt. I appreciate your candour on this. Have noticed any of the these effects on any other aspect of your system. Like the video card? Any lightening or intensifying of color, dimming of monitors, text becoming less clear?

Paul999, post: 409475 wrote: Hey Kurt. I appreciate your candour on this. Have noticed any of the these effects on any other aspect of your system. Like the video card? Any lightening or intensifying of color, dimming of monitors, text becoming less clear?

humorous , shade of disrespectful or are you serious?

audiokid, post: 409476 wrote: humorous , shade of disrespectful or are you serious?

I am absolutely serious. I can't see that a computer would only make unknowing errors of omission only to audio. It would only be logical that it is system wide.

again i dedicate the bulk of resources on my DAW box to audio ... video card runs at bare minimum ... and i don't do video work so all i am looking at is the graphics that run with the DAW.

look i don't have a horse in this race. all i know is what i hear and i hear degradation in DAWs the more they are tasked with. more tracks, more degradation, more plugs , more degradation. more eq, compression effects, latency compensation ...

surely you are not suggesting a computer doesn't have finite resources? common sense suggests the more of a load that is put on a device that has a limit to how it performs the less efficient it will perform those tasks.

record 16 to 32 tracks, put different eq on eack one, different effects (with different settings on those that may be shared) different dynamics (comps gates) settings, fader settings, automation etc, then stack 2-bus processing on top of that (this is the worst) it just doesn't sound as good as it did when you first started with three or four tracks. i have noticed this all the time.

again i can't believe you won't trust your ears.

Kurt Foster, post: 409484 wrote: again i dedicate the bulk of resources on my DAW box to audio ... video card runs at bare minimum ... and i don't do video work so all i am looking at is the graphics that run with the DAW.

look i don't have a horse in this race. all i know is what i hear and i hear degradation in DAWs the more they are tasked with. more tracks, more degradation, more plugs , more degradation. more eq, compression effects, latency compensation ...

surely you are not suggesting a computer doesn't have finite resources? common sense suggests the more of a load that is put on a device that has a limit to how it performs the less efficient it will perform those tasks.

again i can't believe you won't trust your ears.

Fair enough. I am not trying to dog you or piss you off. This just doesn't seem logical. I don't trust my ears on this because it is well documented in all fields that human bias is strong. This is the reason for double blind studies etc. This is the reason we level match when we listen to things and should set up double blind listening studies when auditioning gear.

Anyway I do believe computers have finite resources. I find it odd that all these computers are targeting only the audio side of the computer to make these small cumulative errors. In your situation you are dedicated to audio and I assume you don't have a specialized video card. This would make your Daw video more susceptible to video errors. I would think as you are drawing on the finite resources of your computer it would be effecting the whole system not just the audio.

To to me this sounds we are anthropomorphizing our computers thinking that if we are working on audio that that's where it will fail.

What I've witnessed is that a computer does as it's asked and the more it's tasked the slower it gets until catastrophic bugs occur and then failure. These bugs show up in video, audio, keystrokes, everything. It is not contained to audio.

Paul is essentially correct in what he is asking. I'm a former CompSci professor, myself. His question is regarding the distribution of errors across the entire computer system. If the computer itself is generating errors, then there would be errors in other parts: like graphics, etc. It would not be confined to only the audio.

However, there are other opportunities for errors to creep in: The software, for example. Software could be making the errors. Then, the errors would be confined to "audio"; not in the strict sense though, since the errors are actually in the software. But it would APPEAR as an audio problem, since the audio software was what goofed up.

But audio hardware OUTSIDE the computer could generate errors, too. Clocks, AD/DA converters, cables, analog hardware, etc.

Can we narrow down the soure of the errors in a scientific way? Maybe you guys already have done that in this thread and I just missed it. You'll have to excuse me, it's 5 degrees here, the coldest I've been since standing watch on a Navy ship in the north Atlantic. :)

Paul999, post: 409485 wrote: I don't trust my ears on this because it is well documented in all fields that human bias is strong.

believe me, i wanted the promise of good mix's sans a huge money gobbling console to hold true but it didn't happen. if the differences i hear were miniscule i would agree with you but what i hear are drastic differences in the sound stage. audio is audio and we listen with our ears. many of the greatest have retired at a point because their ears are degrading and they can't trust them any longer. if you can't trust your ears and you rely on what meters and displays tell you to mix audio then you are pissing up a tree. it would be the same as listening to audio program content to switch and mix video. and if you are not confident enough in you opinions and decisions then you should step aside and allow someone else who is to do it.

Paul999, post: 409485 wrote: Anyway I do believe computers have finite resources.

Thank you. however it is a fact, not hypothesis. a bucket can only hold so much water and if you punch a hole in the bucket it will only leak out so fast. these system absolutely have limits. analog is infinite in its speed ( oh well the speed of light). computers operate at a much slower speed than that.

Paul999, post: 409485 wrote: I find it odd that all these computers are targeting only the audio side of the computer to make these small cumulative errors. In your situation you are dedicated to audio and I assume you don't have a specialized video card. This would make your Daw video more susceptible to video errors. I would think as you are drawing on the finite resources of your computer it would be effecting the whole system not just the audio.

all my DAWS have NVIDIA cards. again as i don't perform any other tasks with them i can't say if i have that problem with my DAWS. however i experience these issues everyday when using my internet and office processing computers. the more windows tabs open the slower it gets. the more content included on a page the slower it opens and sometimes they just clog up and the screen dims while the computer takes a breath and works through the load.

i would also point to the massive increase of error on broadcast and cable Television as an example. in the 50's /60's / 70's and early 80's analog broadcast TV was perfect. it was very rare that any errors got through and if they did you knew it was human error and someone was going to be fired. these days pixelization, switching errors, drop outs are abundant.

Paul999, post: 409485 wrote: To to me this sounds we are anthropomorphizing our computers thinking that if we are working on audio that that's where it will fail.

What I've witnessed is that a computer does as it's asked and the more it's tasked the slower it gets until catastrophic bugs occur and then failure. These bugs show up in video, audio, keystrokes, everything. It is not contained to audio.

audio is far more complex than a video signal, especially when you are talking about multi track with multiple plugs and instances of eq all performing at different settings and levels. there was already very complex and powerful systems to do video way back when you could only get at best 2 channels or DAW processing (remember Sound Tools?)

i would agree that these issues are not only a problem of the DAW but also in part a problem of interfacing and design of peripheral equipment, clocks , converters, control surfaces etc. bottom line however is analog serves us best to sum multi channel audio. DAW is king for editing and convenience. i advocate the use of both for the best sound.

It seems to me that the point at issue is not that a computer could be making arithmetic errors - nulling tests can show that, provided you keep the demands on processing power within the range that the particular computer can offer, all processes can be completed without digital error - but that the process of anti-alias filtering and the conversion between analog and digital domains does change the signal being processed. This cannot be tested within the computer.

We've had threads in these forums in the past about the difficulty of repeatability of analog sources, and how it's not easy to perform analog nulling tests (as opposed to digital) between analog signals and their converted representations. However, some years ago I did one set of tests to demonstrate a particular point. These involved recording audio from a microphone + pre-amp to a hard disk recorder (Alesis HD24XR) at 24-bit 96KHz and then replaying the recorded signal in analog and re-capturing it on a DAW at 24-bit 44.1KHz. With appropriate synchronisation, I could re-play the original 96K recording simultaneously with the 44.1 KHz recording and attempt an analog null. There were significant differences, not only the expected ones in the range 20 - 40 KHz, but also other more subtle ones in the range 10 - 20 KHz.

This test did not in any way show that the 96KHz recording was indistinguishable from the analog original, but it did demostrate that digital recordings do not necessarily hold an accurate representation of signals that have significant energy in the octave immediately below the Nyquist frequency. In this case, I have to assume that the band 20-40KHz in the original 96KHz recording was equally suspect as the 10-20KHz band of the 44.1KHz recording, but because there was significant acoustic energy at 10-20KHz in the original signal, the 96KHz recording made a better job of capture and replay of this band.

I am well aware that the sampling rates I used were not a factor of 2 different, but the original tests were concerned with CD vs DVD-Audio sonic quality, and I did use 24-bit sampling in both these cases. Subsequent tests included reduction to 16-bit 96KHz and (naturally) to 16-bit 44.1KHz in an attempt to tease out the different contributions of sampling rate and wordlength. I ought to update these tests with various MP3 resolutions, since I was sufficiently alarmed by what standard CD quality failed to represent. A further extension involving comparisons with DSD recording would also be interesting.

What this does show is that, once you have digitised an analog signal, you can only work with the representation you have, and there may be more difference between this representation and the original than all the small changes you subsequently make in the digital domain.

Nice one Bos, always the best when you chime in.

I just posted this in our news section, it is relevant in an analog sort of way.

I've been doing it like this for years (30 at least). I mean, I dragged around my studio and used it live. The only difference for me today is I'm in my studio now ( not performing) and using the DAW instead of sequencers and samplers.

I'm inclined to move parts of this thread into its own new thread as its pretty far off of the virtual instruments category (or is it?).

If or before that ever happened (yes/no and/or title suggestions welcome?) this is something I'm considering investing in which is relevant to the OP's Vocal Rider questions, but on the analog side of the track.

See the Dangerous Compressor .
[[url=http://[/URL]="http://dangerousmus…"]The Dangerous Compressor - Dangerous Products - Dangerous Music[/]="http://dangerousmus…"]The Dangerous Compressor - Dangerous Products - Dangerous Music[/]

independent L/R SC:

http://dangerousmusic.com/media/compressor-product-photo-rear.jpg

This is an interesting read as well. And don't forget, not all of us who use the word analog are talking about tape either.
[[url=http://[/URL]="http://en.wikipedia…"]Comparison of analog and digital recording - Wikipedia, the free encyclopedia[/]="http://en.wikipedia…"]Comparison of analog and digital recording - Wikipedia, the free encyclopedia[/]

If I understand you correctly Boz the oversimplified Coles notes version of what your saying is: A null test will show if calculation errors are happening. Also that calculation errors are likely to happen when your computer is taxed beyond a certain point. And that it is resonable that the 200+ plugs plus 20 tracks was a resonable load in my experiment.

Your further point is that issues inherent in ADDA conversion will be causing much more havoc then any of this computer error business we've been torturing for 3 pages.

If I've got it right it makes perfect sense to me and I can't find an issue in your methodology.

audiokid, post: 409503 wrote: This is an interesting read as well. And don't forget, not all of us who use the word analog are talking about tape either.
[="http://en.wikipedia.org/wiki/Comparison_of_analog_and_digital_recording"]Comparison of analog and digital recording - Wikipedia, the free encyclopedia[/]="http://en.wikipedia…"]Comparison of analog and digital recording - Wikipedia, the free encyclopedia[/]

Thanks for posting this! ...I've been looking for references to this part of Dr. Diamond's work for decades.
"Dr [[url=http://="http://en.wikipedia…"]John Diamond[/]="http://en.wikipedia…"]John Diamond[/] 1980 article Human Stress Provoked by Digitalized Recordings"
[[url=http://[/URL]="http://web.archive…"]The Diamond Center: Digital stress[/]="http://web.archive…"]The Diamond Center: Digital stress[/]

Paul999, post: 409528 wrote: If I understand you correctly Boz the oversimplified Coles notes version of what your saying is: A null test will show if calculation errors are happening. Also that calculation errors are likely to happen when your computer is taxed beyond a certain point. And that it is resonable that the 200+ plugs plus 20 tracks was a resonable load in my experiment.

Your further point is that issues inherent in ADDA conversion will be causing much more havoc then any of this computer error business we've been torturing for 3 pages.

If I've got it right it makes perfect sense to me and I can't find an issue in your methodology.

Pretty much, yes, Paul, especially as I think the computer error discussion is a completely separate issue.

Any A-D and D-A conversion processes are going to have some of each of errors due to imperfections in the amplitude and timing of the sampling but also "errors" due to signal changes needed to constrain the signal to fit into the finite information content of a digital representation. This means that a signal fed into a perfect A-D and then out of a perfect D-A, but clocked at a real-world sampling rate with appropriate anti-aliaising filters, is going to show differences from the original signal on a nulling test. I'm not saying that these differences are audible as distortion, but they are what they are: differences. Since these differences show mainly in the top octave of the digitised bandwidth, they can be reduced as far as the human ear is concerned by raising the sampling clock rate.

It's still my belief that the actual digitisation errors (in amplitude and timing) cause more of the audible problems than the artifacts attributable to the process of digitisation. As a crude demonstration of this, you can record a highly-harmonic instrument such as an oboe and then randomise the phase of the harmonics relative to the fundamental. On a nulling test, this would result in gross differences between the reconstructed signal and the original, yet the human ear is remarkably insensitive to what has been done to the waveform.

I've said several times in these forums that digital mixing of tracks all recorded at 44.1KHz compounds the top-octave problems. I have also (rather light-heartedly) drawn a comparison with the bedroom-studio technique of recording each track separately using the same one microphone, resulting in the microphone's imperfections being reinforced in the mix.

One experiment I have never done but could do relatively easily when I have time is to use one of my 96KHz HD24XR multitrack recordings replayed via analog and then perform both an analog mixdown with 44.1KHz conversion of the mix and also a digital mixdown where all tracks are digitised at 44.1KHz at the input. It would have to be done with careful matching of the track amplitudes and using no EQ, no dynamics and no effects. It would then be interesting to compare the 44.1KHz results on a listening test, since a null test would not be useful in this case.

I can see errors during playback. I get them when my VST metere pings in the red: I get skipping and stuttering.

But what about during mixdown. On Cubase, that is done off-line, and it takes considerably longer. Surely no errors then, since the computer won't overtax itself? It may max out, but it doesn't have to "drop" anything to keep the speed up.

Maybe I've lost what we are talking about? I'm primarily an artist with a good ear. Some of the highly technical stuff is beyond me at this point. I'm learning, but not there yet.