Is there a remote possibility of a live DSD/PCM comparison?

[Inky Goddess]

i love you all! i love the sanity of this forum!

i never knew, until i wandered about a bit in GS yesterday, that there are MADMEN in here with us!!! praise you, my friends, for being here to bolster me...

i refer to the chaos going on over at shipley's thread Pyramix etc. lord have mercy! unless you have and are used to eight perpetually bickering children, don't go there...you'll never get out alive.

let me just post here what i posted there. it'll be pretty self-explanatory, and it'll save you a horrific trip:

Originally posted by Curve Dominant
I asked a simple question on page 2 of this thread: Why do you prefer to record through 1-bit convertors, and find that "superior"?

Originally posted by Brad Blackwood
Nika, let's not assume anything. I like DSD, but have not claimed it superior in any way to PCM. I just don't like people arguing against something based on theory, not on how it sounds. Some people prefer DSD and choose to record that way, and I say that's fine - it's not a black and white issue…And I don't see how it's hard to see which one sounds better - why not record some live music to PCM and DSD at the same time listen to playback?

Originally posted by Nika Aldrich
You make that sound easy to do
Not only is it difficult to get together the equipment to do such a comparison test, but if the goal is to objectively determine which one inherently sounds better then the circuit design of the analog portion of the converters should be nearly identical and the filters in the PCM version should be amongst the best available. If it were only possible to do as you've suggested that would be widely done and we'd be discussing those tests.

curve, as nika points out, there hasn't been any effective third-party testing along the taste-test lines that brad has suggested; therefore, the best answer you're going to get here is an opinion or opinions. there's no point in debating this any longer, as the only reason i've seen here is "it just sounds better." you won't accept that, and that's fine, but until you hear it and judge with your own ears, you can't tell these other professionals, as you have been, that their opinion is not good enough. personally, i trust their judgement! this is what they do for a living, and people pay them for using not only their gear, but their ears and experience as well.

audio recording isn't a clinical procedure, it's an artform. all the pictures of signals you want to post are fine, and your points may very well be valid, but there is so much aesthetic creation going on when you record audio, especially music, which is its own artform, that opinion must count, as it is indigineous to the entire creative process.

sooooooo...

i'm taking the whole issue over to the remote guys in Remote Possibilities. those cowboys might be up to putting together such a comparison, since they can drive to one another with whatever equipment they do have. then anybody who wanted to participate, a.k.a. those that really wanna know how the formats compare against each other, could join them for a listen, and judge for themselves.

i expect that if they take me up on this, all those that have put so many hours into this thread will attend, by hook or by crook.

...and they call me a bitch like it's a bad thing, sheesh!

and that was all she wrote over there

now what i wanna know from you guys is is there a way to make this happen? i think it'd be a great way to get some slutz together for a productive activity that's business-related read "write-off"...and perhaps i could help solicit some media coverage read "the girl gets a write-off too."

what say you, my sane and beautiful friends? am i out of my mind? is this a ridiculous idea? i think it sounds like a blast.

[Curve Dominant]

Yeah, I agree. A good scientific test would be cool.

Is it OK if the mad children come along??

I'll bring beer...

We don't get out much, ya know.

[djui5]

Originally posted by Inky Goddess
now what i wanna know from you guys is is there a way to make this happen? i think it'd be a great way to get some slutz together for a productive activity that's business-related read "write-off"...and perhaps i could help solicit some media coverage read "the girl gets a write-off too."

what say you, my sane and beautiful friends? am i out of my mind? is this a ridiculous idea? i think it sounds like a blast.

I'm down.....I'd love to see this one!

I haven't done a side by side comparison ovbiously...but have listened to SACD and do enjoy it....

So..who's gonna call Sony and arrange a rental of required DSD gear?

[djui5]

Originally posted by Curve Dominant
I'll bring beer...

Only if you bring Tucher......

[0VU]

About four or five years ago I did some comparison recordings using DSD and various PCM formats. We were about to start recording a bunch of string quartet discs and had the opportunity to choose any format we wanted to use so we checked out a few options.

I hasten to add that there was nothing particularly scientific about the test methods, nor were they particularly exhaustive; the object was to give the artists and producer (and us) the chance to hear the possibilities. The recording machines and media were different between the DSD and PCM formats and the D-A converters were different (Meitner / dCS954) though for comparison we subsequently listened back to the DSD via the dCS D-A converters. The idea wasn't really a shootout between DSD and PCM, more between the specific complete systems we had available for the project.

We simply made parallel recordings from the outputs of the desk to DSD (via the supplied Sony (Meitner) converters), 24/176.4 PCM (via dCS 904 A-D), 24/88.2 PCM (dCS 904 A-D) and 24/44.1 (dCS 904 again). We did a second set of recordings to DSD and to 24/48, 96, 192 for comparison. We also tried several different filter setups on the PCM converters, the DSD stuff at the time was less flexible in that respect so we were stuck with what we had.

I might still have the test recordings somewhere (if I can find them ). I also still have access to the hardware (Sonoma/Meitner DSD system and dCS 904/954 converters (also DSD capable)) so I guess that I could repeat the tests in a more scientific fashion. If someone would like to suggest an appropriate/acceptable test methodology I could maybe have a look at setting something up.

[Inky Goddess]

thanks 0VU, for that generous offer. i'm going to start making some inquiries about this. if i get some more positive reaction and a little more committment from a few more slutz, this could be happening! i'm jazzed.

you other critters...of course this will be open to all gearslutz, but you will NOT embarrass me or any other gearslut, especially if there are corporate reps there. if you start bickering like old biddies or brawling like drunken bikers, you're eighty-sixed. additionally, the alcohol and other party matériel you brought stays with the party, not with you. do i make myself clear?

[djui5]

Originally posted by Inky Goddess
you other critters...of course this will be open to all gearslutz, but you will NOT embarrass me or any other gearslut, especially if there are corporate reps there. if you start bickering like old biddies or brawling like drunken bikers, you're eighty-sixed. additionally, the alcohol and other party matériel you brought stays with the party, not with you. do i make myself clear?

Yes Ma'am.....

[Dave Martin]

Originally posted by Inky Goddess
do i make myself clear?

Yes Ma'am... But what if Randy starts it?

[Inky Goddess]

so mr. martin, i assume you're considering attending, since you're worried about somebody starting something?

to answer your question, i really don't care who starts it...if you're involved, you'll have to go stand in the corner!

and randy, i'm willing to organize everything, but i'll need to find a location first. i have a place in mind...i'm still looking into it.

any other slutz interested? even mildly?

[0VU]

..er.. just in case anyone (Miss Goddess?) is starting to plan something based on my offer of setting up a comparison, it might be worth noting that I'm in the UK. Whilst I do travel to the US and Canada quite often, and I've been known to bring my dCS converters with me I'm not trying to get a Sonoma through US Customs; apart from the physical size and weight, it's hard enough as it is, convincing them that my mics won't explode or turn into sharp objects when I want to get on a plane with them
That said, I'm happy to have a go at some kind of testing with anyone who can get to the UK to do it.

[Inky Goddess]

fear not, 0VU...i'm thinkin' y'all'd like to see the bleeding-edge stuff, so i'm gonna try to get the company reps to give demos. the only gear we'll need for this will be locally available, so plan on packing light.

[hollywood_steve]

Um, maybe I'm missing something here, but I don't get Nika's position that setting up this kind of A/B test would be that difficult. My Genex GX9000 offers PCM and DSD recording and the analog section is the same so the only difference is the digital recording technology (PCM vs DSD) . Doesn't this take care of his major concerns, or am I missing something?

FWIW, I've done the test a few times and I find DSD recording to be a very worthwhile undertaking. And the Genex makes it a risk free proposition: if DSD is outlawed tomorrow, I just use the PCM converters for all future work. More folks should invest a few hours and actually listen to DSD for themselves; its not that difficult.

[djui5]

Originally posted by Dave Martin
Yes Ma'am... But what if Randy starts it?

If I start it no-one will ever know as you'll be tied up on a closet.....muuuaahhhhhh ahhhh ahhhhhhhhhhhhhhhhhhhhhhhhhhh











kidding...love ya DM

[Inky Goddess]

hollywood, the discussion was specifically about the pyramix system. also, nika was getting attacked by hostile GS's that were actually just mad that (s?)he was taking up for another GS that was stirring up the shite like nobody's business. soooo, the a/b comparison nika's on about refers specifically to PCM side-by-side with pyramix.

plus, it seems nika's written a book on the topic that's pending publication, so (s)he was ardently defending the thesis of that work.

hope that clears it up for you.

Originally posted by hollywood steve
More folks should invest a few hours and actually listen to DSD for themselves; its not that difficult.

and this is what i'd like to accomplish with this get-together i'm proposing. i believe with enough of a show of committment from us gearslutz, we can get one or more DSD manufacturers to come on down and demonstrate their stuff, and also have some of the trade mags down there to cover it. everybody wins.

and i make the most bitchin' margaritas you ever tasted. no shite!

[djui5]

Inky....

Where are you thinking about holding it?

[Inky Goddess]

the venue is still very undecided, dj. i'm thinking about several different sites, all in new york city.

[Ted Nightshade]

Nika has stated repeatedly that he believes (although he states it as objective fact) that DSD, categorically, is a marketing ripoff.

So I think the Genex test would be more than adequate. Only, Nika would say that Genex intentionally designed it so DSD sounds better, and so higher sample rates sound better, to sell people on DSD and high sample rates. Really, it's pointless to try to prove anything to the guy, so whatever test is set-up needs to be really worth it to those involved for non-Nika reasons.

[Ted Nightshade]

Like margaritas.

[Inky Goddess]

Originally posted by Ted Nightshade
Really, it's pointless to try to prove anything to the guy, so whatever test is set-up needs to be really worth it to those involved for non-Nika reasons.

hey ted, if you wanna know my real reason for wanting to host this a/b soundoff, it's this: i haven't thrown a party in i-don't-know-how-long. i bartended for about thirteen years, and hence worked in a party, but those days are gone. i miss it a bit.

hey, can you imagine the amazing single-bit karaoke we could have going on? in 7.1 surround? what a friggin' blast!

seriously though...i believe there are enough AE's and other interested parties that would like to compare the formats side-by-side to make this a fair-sized gathering; i think that this event will rate somewhere along the lines of a mini trade convention or a seminar for tax purposes, so you can write it off; i'm sure it will be educational; i believe i can get the manufacturers and media involved, which means fifteen minutes for many, and nobody having to haul their gear to the venue...and i think a lot of other stuff that i can't commit to at the moment, so i'll just keep at it til i can tell you more.

the pyramix etc. thread was the catalyst. if you wanna get technical, it was really Brad Blackwood's idea to do an a/b. dsd is clearly a popular topic, there's a lot of curiosity about it, and we'd be professionally negligent not to examine it further. i'm not trying to help anybody prove a point; i'm trying to afford anybody who's professionally interested the opportunity to gather with colleagues to hear the new technology, to compare it with what they're already familiar with, and be able to discuss their observations in a relaxed, friendly environment.

now is that enough to get you to say maybe?

[djui5]

Originally posted by Ted Nightshade
Nika has stated repeatedly that he believes (although he states it as objective fact) that DSD, categorically, is a marketing ripoff.

So I think the Genex test would be more than adequate. Only, Nika would say that Genex intentionally designed it so DSD sounds better, and so higher sample rates sound better, to sell people on DSD and high sample rates. Really, it's pointless to try to prove anything to the guy, so whatever test is set-up needs to be really worth it to those involved for non-Nika reasons.

Ok...I understand Nika is highly intelligent and knows more about digital audio than just about anyone I know but...........................................................
................................................................
................................................................
................................................................





If we all do tests to "please" Nika's beliefs and "guidelines"...then Nika might as well do all of our work for us.

We should do the test for the sake of comparing DSD to PCM.....and that's it. Nothing more...nothing less. And maybe a little drinking

Nika...I meant nothing personal........

[Ted Nightshade]

Inky and Randy- a worthy cause, for sure, and if I were several thousand miles closer to NY I would be there! It's just so important with Science not to go in there with a dissertation to defend... I say add the Genex test to the menu- a couple different implementations of PCM/DSD to compare, and make some moquitos (sp.?) too.

[Inky Goddess]

ted, i understandably thought you were in virginia, but after your "several thousand mile" remark, i searched "state of jefferson" and came up with this informative article. beautiful country, dude. i've only seen pictures, though. one day i'll definitely cruise through there, probably when i get my next harley.

to all those who can't attend: we'll definitely post pics afterward, and maybe there'll be an informal cd or dvd for gearslutz, who knows? if we do this, we know there'll be a lot of slutz not present in body, but definitely in spirit. can i get an amen?

ted, i completely agree with your assertion regarding maintaining neutrality in empirical research. in researching venues, those i've approached have been receptive to the concept. all the locations i've inquired of are such that the equipment can be set up in discrete control rooms, allowing solo monitoring of each format, with the systems interfacing in a large "A/B" control room. maybe "A/B/C," if we can get more than one DSD manufacturer involved.

of course, if anybody wants to submit a paper on the outcome of the testing to some journal, fine with me, but that's not my intention here. the environment isn't going to be perfectly sterile. i do however want all your remarks on how you slutz think the gear should be set up to make the systems as parallel in performance as possible. just post all your suggestions here; thanks.

the time frame we're looking at is late autumn or early winter, if that helps you guys.
goshdang, this is exciting.

[Nika Aldrich]

To all those involved,

Please, don't do any of this for my benefit, as you've itterated.

The reason I said it will be difficult to conduct this test is that you will have a difficult time drawing any helpful conclusions unless the test is so rigorous that you negate the test itself. Let's say, for example, that you use the Genex machine for both the PCM and the DSD versions. That's fine, but if the Genex uses the same DSM for both systems then you are inherently putting the PCM at a disadvantage, because PCM can benefit from having more than a single bit DSM. So if you do this test then you are likely to only be able to draw a conclusion that is so specific as to not be terribly helpful, such as "On the Genex we could/n't tell a difference and liked/did not like this one better." But that only tests one machine - one protocol, and not the format unto itself.

So lets say you use a multi-level DSM device for the PCM and for the DSD. That's fine, but what are the filters on the PCM side, and how much do we know about those filters?

And then there's always the results of the test that indicate you like one better than the other. That's fine, but which is more accurate?

What you really need to do in order to test the formats is find a PCM converter that is audibly transparent enough that you can sense no difference between input and output. Paul Frindle has some of these that have passed his ABX testing lab at Sony. There are other ones on the market, I'm led to believe. So find one of these converters that is absolutely totally transparent on input vs. output under any test signal you can run through it - full bandwidth, single sines, localization tests, etc. Then test THAT converter against a DSD converter that meets the same requirements. Now you have a real format test.

Therein lies the problem. If the PCM signal is totally flat such that you can't tell a difference at all between the converted signal and the pass-through signal then.... what are you testing for? If PCM meets all requirements of the human auditory system then why are we looking for a "better" system, and what could possibly make it "better?"

Also, in order to do a good, well prepared listening test it would be necessary to do some ear-training first. Any auditory tester will tell you that you have to teach someone specifically what to listen for if you want to do a good ABX test, unless you're looking for an unbiased first reaction at a popular level. So in doing listening tests you will need to find out from those that object to DSD exactly what they hear as a problem. My belief (that I often state as fact) is that when you really talk to these people and figure out what the mathematical problems are so that you can ascertain how, exactly, to train your ears to listen to the defects that are spoken of you are inherently going to find out that developing an audibly transparent 1 bit converter is impossible.

In the end, a little bit of mental masturbation about the testing itself goes a long way to determining the results. The results are either going to be "the test was useful for some things but essentially inconclusive with respect to the format issue," or "we found a perfectly transparent PCM converter and failed to find a perfectly transparent DSD converter. Ergo..."

Anyway, that's my perspective and I hope it is helpful. I certainly am a fan of listening and studying and drawing your own conclusions, but in owning those conclusions make sure you state them for what they are - which is often only conclusive about very specific things, like "on this equipment with this signal..."

Cheers!
Nika.

[Jose Mrochek]

Why don't we just claim piramix is better, have Inky bring some cutie girlfriends, open up some bottles and party.


P.S. Inky, can you give me the receipe for liquid cocaine shots ???

thanks!

[djui5]

Originally posted by Nika Aldrich
What you really need to do in order to test the formats is find a PCM converter that is audibly transparent enough that you can sense no difference between input and output. Paul Frindle has some of these that have passed his ABX testing lab at Sony. There are other ones on the market, I'm led to believe. So find one of these converters that is absolutely totally transparent on input vs. output under any test signal you can run through it - full bandwidth, single sines, localization tests, etc. Then test THAT converter against a DSD converter that meets the same requirements. Now you have a real format test.

Therein lies the problem. If the PCM signal is totally flat such that you can't tell a difference at all between the converted signal and the pass-through signal then.... what are you testing for? If PCM meets all requirements of the human auditory system then why are we looking for a "better" system, and what could possibly make it "better?"

Cheers!
Nika.

What if we are testing systems currently in the market.....say...HD w/192's and a Pyramix DSD recorder like loud said.....
Personally I think your thinking about it too much......the converters are part of what make up the system......you can't have DSD without a converter.....and in todays state how that converter reacts is part of what makes one format different from the next......this might change down the road but today this is how it is and how we should be comparing the systems....
And to make it fair I would recommend tracking a band/group whatever....multing the multi-track sends/outboard chain outputs into both systems so that both systems receive the same signal....and monitoring the output of both systems through the same console.....being able to switch between them for a comparison (sending a stereo out of both systems).
How do you feel about this?

[Nika Aldrich]

So long as you state your conclusions accurately I don't mind at all. E.g.: The Pyramix sounded ... in comparison to Protools HD using the following converters on subsequent systems, etc.

I don't see anything wrong with that type of answer.

Nika.

[Inky Goddess]

jose, liquid cocaine tastes like shit, and i won't give you the recipe for a shitty tasting drink. make an oatmeal cookie instead: any glass 3/4 full of bailey's, fill almost to the top with goldschlager, and then top it off with a float of jaegermeister & stir. ice is very optional. it tastes yummy, just like an oatmeal cookie.

nika, your diligence and attention to detail is amazing. this get-together is definitely more of an informational than a scientific event. mr. shipley and others have stated that DSD sounds better to them. this is just a chance for other people to experience it side-by-side with PCM so they can form their own opinions and conclusions. nobody should be expecting lab results.

i've requested everyone's input, but ultimately i'm going to give the suggestions to the company reps that provide the equipment and let them use whatever they want. i believe these reps, who are showing their stuff to twenty or fifty potential customers, are going to set their stuff up to its maximum advantage, so i really don't think we need to worry about those details too much. i'll be the one to work out with them how much gear we need to provide, et cetera, blah blah blah, with some AE's i know in new york to provide technical guidance. and of course, all the gear involved will be listed on a program of some sort so everyone can discuss and evaluate it as much or as little as pleases them.

so, let's try to stay on topic here and leave the debate over at shipley's, where it is doing marvelously and has a life of its own. this is more about professional camaraderie and fun...kinda like gearslutz.com, imagine that! thanks everyone.

[kevinc]

Originally posted by Nika Aldrich

What you really need to do in order to test the formats is find a PCM converter that is audibly transparent enough that you can sense no difference between input and output. Paul Frindle has some of these that have passed his ABX testing lab at Sony. There are other ones on the market, I'm led to believe. So find one of these converters that is absolutely totally transparent on input vs. output under any test signal you can run through it - full bandwidth, single sines, localization tests, etc. Then test THAT converter against a DSD converter that meets the same requirements. Now you have a real format test.

O.K. So what your saying right here is that the system you`ve been talking about (DSD) which is according to you flawed in some way can`t be tested against something like a Lavry or Prism converter. Converters which are thought of around here as being pretty damn good and extremey expensive.

You need something better that`s still being tested that we would have a hard time even ataining ?

How is that a fair test ?

You should be testing this system against very good PCM converters. End of story.

If it beats out the Lavrys and DCS and whatever other converter on the market currently you`d have to be an idiot to sit there and say they aren`t on to something.


It`s like your trying to woo us with vocabulary to hide the fact that you arent using any common sense.

[kevinc]

Originally posted by Nika Aldrich


Also, in order to do a good, well prepared listening test it would be necessary to do some ear-training first. Any auditory tester will tell you that you have to teach someone specifically what to listen for if you want to do a good ABX test, unless you're looking for an unbiased first reaction at a popular level. So in doing listening tests you will need to find out from those that object to DSD exactly what they hear as a problem. My belief (that I often state as fact) is that when you really talk to these people and figure out what the mathematical problems are so that you can ascertain how, exactly, to train your ears to listen to the defects that are spoken of you are inherently going to find out that developing an audibly transparent 1 bit converter is impossible.

Another load of crap.

Your saying that an experienced AE or ME wouldn`t have the ear to tell whether one format is capable of giving them a more accurate listening experiance than another.

You need training for that ?

You mean if I listen to the 2 and decide I like DSD better it`s because I haven`t trained my ears to hear the flaws in the 2 systems. That`s absolutely rediculous.

If it sounds better to me IT SOUNDS BETTER TO ME. Period.

[Ted Nightshade]

well there is something to being hip to what to listen for- remember how clear and clean the first CD's sounded? Some things are great at first listen and then get old- getting some pointers of what to listen for would be very useful, IF it goes both ways- things to listen for in PCM and DSD.

It could get pretty petty- like looking at a gorgeous photograph and having someone there ignoring the overall and continually pointing out some minor ignorable and inevitable flaw- you can play that way with anybody or anything. Sure Ray Charles is great (collosal understatement), but any classical singer can point out myriad flaws in his technique- some of us wish those classical singers would ever develop those particular flaws...

I think repeated and careful listening pretty much will cover that- maybe not possible under the influence of much conviviality and many margaritas, but you got to start somewhere.

I do agree that as long as the results, such as they are, don't claim to be anything they aren't, they will be completely valid, as far as they go. You just can't prove any sweeping thing conclusively in one go-round.

I would say if possible, throw in a few PCM contenders- PTHD, something ultra-pristine like Lavry or Prism, and something vibey like SLAM! or HEDD. And get whatever DSD contenders you can together. That way, if it turns out that some folks prefer some DSD to some PCM but not all PCM, or whatever, you can tell, and that may well be the case- implementation is everything, and taste is taste. I really doubt there's one "correct" answer.

It's almost as if we were comparing oil paints to watercolors, or it might be, I don't know, I haven't tried it yet!