Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Curve Dominant]

Bob Olhsson wrote on Sat, 05 November 2005 22:18

where we're at is that best case, digital gear needs to be aligned to the limitations of the analog gear it is being used with. Worst case the analog circuit of some digital gear isn't up to working optimally with some analog gear.

That sums this whole issue up very succinctly.

My concern is that this test is perhaps an excercise in looking for "sonics" in an obscure example, and then drawing overly-broad conclusions from that very narrow application.

Then there is the whole issue of: Who will decide what the results are? How will the results be measured? Can you graph the baseline standard of "lacking balls" along with corresponding standard deviation points?

And as Bob has pointed out, there are matices of standard deviations on both the analog side and the digital side. Adds up to lots of variables.

The big question mark in this whole excercise is: How is the outcome measured and documented? That whole side of this equation doesn't seem to have been addressed at all so far. The technical side has this great team of engineers in place to sync the machines, and that's the essential first step.

But how in the hell are we going to process the outcome?

I propose we work that out before the test is conducted.

[Fletcher]

The Resonater wrote on Sat, 05 November 2005 20:20

RE: the clocking of the various devices...as you probably know, Dan Lavry swears up and down that nearly all digital devices work optimally when using the internal clocks.  FWIW...

Yes, he does... and he may very well be 1000% correct... does that mean we shouldn't experiment?  Take one learned man's statement as "gospel truth" and move on?  If that's the way you want to work, go for it... there are a plethora of "red states"... me, I live in a "blue state" and question constantly but that could just be a personal problem.

RKrizman wrote on Sat, 05 November 2005 21:10

Fletcher wrote on Sat, 05 November 2005 14:26 I am working from the presupposition that there is going to be a pretty large difference when the audio is played back from tape and when the audio is played back from PT... and less of a difference when the audio is played back from RADAR... so, my question [at least one of the questions I have] is whether these differences will stem from the analog electronics in both units [converters], or if this difference will stem from the clocking source. First, I wouldn't presuppose that at all.  Secondly, if there is a problem it might stem from the interface between the devices, as Bob O has suggested.

Hence the concept of doing "straight wire transfers", hence the concept of employing different recording / repro levels.

If issues of "loading" are removed from the picture, and if issues of "interface bewteen devices" is removed from the picture [as a variable]... then you're left with "machine vs. machine".

RKrizman wrote on Sat, 05 November 2005 21:14

Fletcher wrote on Sat, 05 November 2005 14:26 We could also go for 18 and 24 as far as I'm concerned... but I would like to walk in with a solid and agreed upon methodology for the levels at which these tests are recorded... I DO NOT want this particular arguement/discussion/whatever you want to call it while we're all standing in the middle of a control room. This needs to be agreed upon well in advance of the test(s) performed. That's fine, but realize that in the act of doing this if you are getting digital overs then you have to back it down, no matter what levels you decided on in advance, otherwise you're just misusing the equipment.  So I'd add that contingency to your methodology. Should this be in the methodology thread?

A method of dealing with this should be defined here [then agreed upon in the "methodology" thread... it's a close call].  

While there is a distinct possibility of running into "overs" at 14 or 16, I would hope that possibility will be reduced at 18, 20 or 24.  If SMPTE recommends 20, and Digi ships their units at 18, and RADAR can roll with any of 4 choices... then all we need to do is pick two from within the capabilities of the machines.

So, what are the proposals of the day? [Those should be discussed in the "methodology" thread... the idea being to keep that thread defining the actual methodology, while this thread discusses "merits"... yeah, I know its a confusing pain in the ass... but I would like to be able to used the "methodology thread" as a reference/procedure manual during the tests so we don't do something out of turn or that wasn't previously defined and agreed upon].

I'm good with 16 and 22... and see who runs out of "breathing room" first?  The upper area of the headroom of tape is in the +15 area [though you can hit it harder and get into the "compression" thing... but as Mr. Albini has pointed out previously... he likes to stay out of that level area when he's recording].

[Fletcher]

Curve Dominant wrote on Sun, 06 November 2005 03:51

My concern is that this test is perhaps an excercise in looking for "sonics" in an obscure example, and then drawing overly-broad conclusions from that very narrow application.

Could you explain this statement in a familiar language... like English?

Quote:

Then there is the whole issue of: Who will decide what the results are?

The people "in the room"... and if anyone has their shit together enough to post the results on a website or burn CD's of the event... y'all can play along at home.

Quote:

How will the results be measured?

I dunno... it's music... why don't we say that the results will measured "subjectively", much like we measure what it is a "good song" and what is a "bad song".  Just a thought for ya.

Quote:

Can you graph the baseline standard of "lacking balls" along with corresponding standard deviation points?

I don't think it can be graphed... but I'm sure a picture could be drawn

Quote:

And as Bob has pointed out, there are matices of standard deviations on both the analog side and the digital side. Adds up to lots of variables. The big question mark in this whole excercise is: How is the outcome measured and documented? That whole side of this equation doesn't seem to have been addressed at all so far. The technical side has this great team of engineers in place to sync the machines, and that's the essential first step. But how in the hell are we going to process the outcome? I propose we work that out before the test is conducted.

Hence the purpose of this thread and the "Methodology" thread.  

The outcome will be measured and documented in the following manner... we're going to listen to the shit side by side... there will be some people who've been around for a while and might be considered to have a fairly good set of "ears" attending the event.

To recap... we're going through wires, from analog deck to digital machine.  A rough mix will be set on the desk, the analog playback will occur, while that playback is recorded through some pretty high resolution converters to a storage medium.  Remove the wires between that machine and the console, hook up a new machine [PT / RADAR / whatever ya got] to the console directly with wires, verify that the level returning from the new machine is identical to the levels that were returned from the previous machine [I think a 1kHz test tone will work splendidly for level verification], listen to and print the same "mix"... lather, rinse, repeat until all formats have been recorded and played back at various input levels, sampling rates, possibly with different clocks... whatever is the agreed up methodology that appears in the "methodology" thread.


The "outcome" will be recorded, and possibly that recording will be distributed in some manner... maybe over the internet, maybe by mail... that has yet to be determined.  From there, each listener can draw their own conclusions.

Any problems with any of this Eric?  Any positive/potentially productive comments or thoughts?  The questions you have raised have all been covered several times over the course of this discussion... if you're going to play along, please cover all the previous information before providing further redundant questions.

Peace.

[Mixerman]

RKrizman wrote on Sat, 05 November 2005 18:10

Fletcher wrote on Sat, 05 November 2005 14:26 I am working from the presupposition that there is going to be a pretty large difference when the audio is played back from tape and when the audio is played back from PT... and less of a difference when the audio is played back from RADAR... First, I wouldn't presuppose that at all. -R

That is precisely the presupposition that should be made.

The whole impetus of this little shindig, is my claim that there is a severe loss of low end in this sort of transfer when made to HD, a loss that does not occur when the same transfer is made to Radar.

The whole purpose of this "test" is to prove or disprove my statement, once and for all. I find it humorous that there is an enormous amount of hedging going on around here, and I might point out, none of it is coming from me.

There have been a number of nquiries here as to my operating level on the HD unit, when I originally discovered the problem. I'm usually operating on the HD at an input of -18, as anything hotter than that usually hits the console too hard. However, I do not operate under any hard and fast rules, particularly since input level is wholly dependant on how hot I've printed the source material to tape.

I have transferred into HD and Radar at varying levels, and the loss of low-end in HD is apparent regardless of level, while the Radar always performs well in comparison.

Mixerman

[rnicklaus]

Mixerman,

What kind of hedging is being done here?

Also to presuppose you are right before the test is done is as "dangerous" as to presuppose you are wrong.  For anyone to go into the test with a presupposed outcome adds a flaw to the test IMHO.

This should not be a test of RADAR.  This is about the low end of a transfer into Pro Tools from what you said was 8 or 9 tracks of bass and drums from 2". The test has grown from that target already.

RADAR is a great machine.  I used the Otari RADAR 1 on a long project in '98 and bought 2 IZ RADARs for the followup project in 2000.  I am no stranger to the format.

Like it or not Pro Tools and the rest are here to stay.

[Fletcher]

The purpose of this test is to prove or disprove MM's original assertion that PT loses bottom, hence the presuppostion that this will occur is natural.  This is not to say that this test should not be performed in the most neutral of testing proceedure... hence why there is a seemingly very nice fellow named Gannon who works for Digi-Design involved in the process of methodology definition.

If you have 3 or 4 factions that disagree on a premise, and try to set up a test where each of these 3 or 4 parties has their own agenda... then, when you form the final methodology with all of these various factions in agreement, you can pretty well rest assured that the test procedure will not be skewed in a manner that will support any one agenda.

So, while my presuppostion is that PT is gonna suck dogs... that ain't necesarily so.  To say I'm walking into this without a presuppostion would be bullshit... and I'm just not too long on that at the moment.

[malice]

R.Nicklaus wrote on Sun, 06 November 2005 19:44

Mixerman, What kind of hedging is being done here? Also to presuppose you are right before the test is done is as "dangerous" as to presuppose you are wrong.  For anyone to go into the test with a presupposed outcome adds a flaw to the test IMHO.

Dear Randy,

Imho, this is exactly the contrary. Presupposition is the base of scientific experiments. Without presupposition, you wouldn't make test to corroborate your theories.

If you don't want to prove anything, why would you spend time for this ?

malice

[rnicklaus]

malice wrote on Sun, 06 November 2005 12:01

R.Nicklaus wrote on Sun, 06 November 2005 19:44 Mixerman, What kind of hedging is being done here? Also to presuppose you are right before the test is done is as "dangerous" as to presuppose you are wrong.  For anyone to go into the test with a presupposed outcome adds a flaw to the test IMHO. Dear Randy, Imho, this is exactly the contrary. Presupposition is the base of scientific experiments. Without presupposition, you wouldn't make test to corroborate your theories. If you don't want to prove anything, why would you spend time for this ? malice

I have no presupposition that the test will go either way.  It's a test.

Mixerman is not doing this test. This is being done by others.

We can agree to disagree but I will still stick to the presupposition that people doing the testing should be going in with an open mind.  To presuppose that just because Mixerman said it's so therefore makes it likely is interesting in that he has made a side career in bashing the format.

Pretty simple really.  

[Mixerman]

R.Nicklaus wrote on Sun, 06 November 2005 10:44

Mixerman, What kind of hedging is being done here?

You are not familiar with some of the history of this debate, therefore, the hedging would not be so obvious to you. I would imagine that the large majority of posters here have a genuine and unbiased interest in the results.

Quote:

Also to presuppose you are right before the test is done is as "dangerous" as to presuppose you are wrong.  For anyone to go into the test with a presupposed outcome adds a flaw to the test IMHO.

Presupposing I am right gives a purpose for the test in the first place. There have been claims that this is merely a converter shootout. This is NOT a converter shootout. This is an attempt to disprove my claims, that have been supported by two other independant and respectable sources, Slipperman and Bob Ohlsson. The claim being: There is a significant and unacceptable loss of low-end when a transfer is made from 2" to HD, a loss that does not occur in all digital mutitrack machines. The assumption is that there IS as loss, and the goal is for an independent group to attempt to DISPROVE the assumption.

An assumption is not an agenda. It's merely a place to work from. I expect everyone to go in with an open mind. Frankly, the large majority of the people that will be at this particular test, are going in with the presupposition that I will be disproved.

This concept falls under scientific law and theory. Theories are always assumed true with the goal of disproving the theory. Repeatability offers the comunity a way of making the theory more concise and accurate. When I first discovered the problem, it was a hypothesis. Since Slipperman and Bob Olhsson have also discovered the problem independently of me, my hypothesis becomes a theory. If you read the link above about theory, you will see that a theory is accepted as true until it's disproven.

Quote:

This should not be a test of RADAR.  This is about the low end of a transfer into Pro Tools from what you said was 8 or 9 tracks of bass and drums from 2". The test has grown from that target already.

I used bass and drums. Adding some instruments isn't going to change the results. But this is a good example of why the assumption is so important. IF the group transfers an entire production and notices no loss of low-end, then the next test should be totransfer bass and drums only. The assumption allows for that next logical step. IF the group discovers precisely the low-end lop-off that I have described using more than bass and drums, THEN the theory remains valid, and the rsults have had repeatability. This doesn't change anything.

Quote:

RADAR is a great machine.  I used the Otari RADAR 1 on a long project in '98 and bought 2 IZ RADARs for the followup project in 2000.  I am no stranger to the format.

I am not trying to prove or disprove that Radar is a great machine. That is a subjective analysis that must be made by the user, which includes paramaters that go far beyond sound quality.  The Radar is there to prove this isn't a "digital issue," an argument that is consistently brought up when I discuss this problem. The Radar does not play back, with absolute accuracy, the 2" transfer. I don't know a digital multitrack in the world that could. What the Radar does, is illustrate a digital transfer from an analog source, can be "reasonably close." For some it might not be close enough, but there will always be some level of subjectivity to all this. After all, those who have a particular disdain for low-end, could argue that the HD sounds better after the transfer.

Depsite some level of subjecctivity, I am confident that the amount of destruction I have discovered, will be unacceptable to the community at large.

Quote:

Like it or not Pro Tools and the rest are here to stay.

I don't like or dislike Pro Tools. I have stated on many, many occasions, I have and do use the platform. I have been firm that a GREAT record can be made on the platform. Neither one of these statements has anything to do with the issue at hand. A loss of low-end when making a simple transfer from analog to HD. My goal and purpose here, is to force Digidesign to fix a glaring and inexcusable problem. We as a community have an obligation to force manufacturers to keep their gear up to spec.

Mixerman

[djui5]

-18 and -24 sound good to me


192's have an A and B calibration level. So you can easily switch between the 2.

Set A for -18 and B for -24

[Mixerman]

R.Nicklaus wrote on Sun, 06 November 2005 12:10

malice wrote on Sun, 06 November 2005 12:01 R.Nicklaus wrote on Sun, 06 November 2005 19:44 Mixerman, What kind of hedging is being done here? Also to presuppose you are right before the test is done is as "dangerous" as to presuppose you are wrong.  For anyone to go into the test with a presupposed outcome adds a flaw to the test IMHO. Dear Randy, Imho, this is exactly the contrary. Presupposition is the base of scientific experiments. Without presupposition, you wouldn't make test to corroborate your theories. If you don't want to prove anything, why would you spend time for this ? malice I have no presupposition that the test will go either way.  It's a test. Mixerman is not doing this test. This is being done by others. We can agree to disagree but I will still stick to the presupposition that people doing the testing should be going in with an open mind.  To presuppose that just because Mixerman said it's so therefore makes it likely is interesting in that he has made a side career in bashing the format. Pretty simple really.

You cannot disprove something that you do not assume to be true in the first place.

This is not a test. It is an effort at either disproving a previous test or to show repeatability of it.

You are accusing me of "bashing" a product, but I am defying an independant group to "disprove" my findings. How is that bashing?

I have ceratinly been critical of the product and, more importantly, the manufacturer. But I have always laid out clear and concise arguments for my statements of both fact and opinion, and I have always encouraged others to investigate my claims on their own. Expressing a well-informed and well grounded opinion based on years of experience, can hardly be represented as "bashing."

Mixerman

[rnicklaus]

Mixerman wrote on Sun, 06 November 2005 13:36

R.Nicklaus wrote on Sun, 06 November 2005 12:10 malice wrote on Sun, 06 November 2005 12:01 R.Nicklaus wrote on Sun, 06 November 2005 19:44 Mixerman, What kind of hedging is being done here? Also to presuppose you are right before the test is done is as "dangerous" as to presuppose you are wrong.  For anyone to go into the test with a presupposed outcome adds a flaw to the test IMHO. Dear Randy, Imho, this is exactly the contrary. Presupposition is the base of scientific experiments. Without presupposition, you wouldn't make test to corroborate your theories. If you don't want to prove anything, why would you spend time for this ? malice I have no presupposition that the test will go either way.  It's a test. Mixerman is not doing this test. This is being done by others. We can agree to disagree but I will still stick to the presupposition that people doing the testing should be going in with an open mind.  To presuppose that just because Mixerman said it's so therefore makes it likely is interesting in that he has made a side career in bashing the format. Pretty simple really.   You cannot disprove something that you do not assume to be true in the first place. This is not a test. It is an effort at either disproving a previous test or showing repeatability of it. You are accusing me of "bashing" a product, but I am defying an independant group to "disprove" my findings. How is that bashing? I have ceratinly been critical of the product and, more importantly, the manufacturer. But I have always laid out clear and concise arguments for my statements of both fact and opinion, and I have always encouraged others to investigate my claims on their own. Expressing a well-informed and well grounded opinion based on years of experience can hardly be represented as "bashing." Mixerman

Of course this is a test.

Let's however call this what it is.  When I read the mixerman diaries that were posted online and read your forum, there is plenty of "bashing" going on IMHO.  This is my word, not yours.

I can change bash to comically critical if the word bash came off harsh to you.

My point was more along the lines of someone isn't going to ask Ralph Nadar to judge the seat quality of the Corvair or how long the headlights last on a Pinto.

[Mixerman]

R.Nicklaus wrote on Sun, 06 November 2005 13:55

My point was more along the lines of someone isn't going to ask Ralph Nadar to judge the seat quality of the Corvair or how long the headlights last on a Pinto.

But I'm not the one judging this. I will be 2200 miles away when this is going down. So, you're point is lost on me.

Quote:

I can change bash to comically critical if the word bash came off harsh to you.

Not harsh. Innaccurate.

And comically critical? I have been comically critical about far more than just Alsihad. My use of the terms Alsihah (one who loves Alsihad) and Luddite (one who loves 2") is meant to poke fun at our tendencies as humans to draw hard and firm lines on our positions based purely on our emotional and monetary investments. The fact that so many people take umbrage to these terms, only makes it all the funnier.

Personally, I think I've been comically critical about life in general. The fact that I prefer to use satire to make my point, only illustrates my disdain for dry and uninteresting discussion. It hardly proves me an unworthy source of information, and it hardly dismantles my credibility to make an evaluation of a transfer.

BTW: I do believe Ralph Nader was right about many of the issues that he brought to light. Like those pesky seatbelts, for instance.

Ahem.

Mixerman

[rnicklaus]

Mixerman wrote on Sun, 06 November 2005 14:12

R.Nicklaus wrote on Sun, 06 November 2005 13:55 My point was more along the lines of someone isn't going to ask Ralph Nadar to judge the seat quality of the Corvair or how long the headlights last on a Pinto. But I'm not the one judging this. I will be 2200 miles away when this is going down. So, you're point is lost on me. Quote: I can change bash to comically critical if the word bash came off harsh to you. Not harsh. Innaccurate. And comically critical? I have been comically critical about far more than just Alsihad. My use of the terms Alsihah (one who loves Alsihad) and Luddite (one who loves 2") is meant to poke fun at our tendencies as humans to draw hard and firm lines on our positions based purely on our emotional and monetary investments. The fact that so many people take umbrage to these terms, only makes it all the funnier. Personally, I think I've been comically critical about life in general. The fact that I prefer to use satire to make my point, only illustrates my disdain for dry and uninteresting discussion. It hardly proves me an unworthy source of information, and it hardly dismantles my credibility to make an evaluation of a transfer. BTW: I do believe Ralph Nader was right about many of the issues that he brought to light. Like those pesky seatbelts, for instance. Ahem. Mixerman

I knew when I brought the Ralph Nadar example into this what the response would be.

Mixerman, I have never posted that that you have no credibility.  This is about presupposition, not Mixerman.

People may not realize that we know each other and I like your work.  This is just discussion.

Going back to Ralph Nadar, even if he was right about the Corvair and the Pinto, the suppliers of the upholstery and headlights would not want him testing them while on those cars.

[The Resonater]

Fletcher wrote on Sun, 06 November 2005 12:14

The Resonater wrote on Sat, 05 November 2005 20:20 RE: the clocking of the various devices...as you probably know, Dan Lavry swears up and down that nearly all digital devices work optimally when using the internal clocks.  FWIW... Yes, he does... and he may very well be 1000% correct... does that mean we shouldn't experiment?  Take one learned man's statement as "gospel truth" and move on?].

No, I totally support you looking at it from both directions.  I was more thinking about the industry wide "presumption" that external clocks are likely to make an improvement over lesser grade internal clocks.  I mean, in general, don't we typically presume this?  I know I have.  Yet, Dan's argument is likely coming from a true feeling that this is typically NOT the case...that clocks of most any quality will almost always run *better* driven internally.  I was just throwing it out there to say that not everyone expects to hear improvements when convertors are clocked externally.  But I agree that both approaches should be exercised if possible.

Carry on...