The No.1 Website for Pro Audio
What is it about RADAR?
Old 22nd March 2007
  #211
Gear Nut
 
RedEar's Avatar
 

Quote:
Originally Posted by RKrizman View Post
It's not a conjecture? Then what's with all the disclaimers? Do you know this to be the case or not?

As I posted before, in a digital mixer numbers are expanded and then dithered back to 24 bits all the time. I've demonstrated to myself and others many times that this can happen without creating any AUDIBLE difference whatsoever. So the silence test, regardless of its outcome, doesn't shed any light on this.

If somebody else thinks it's meaningful in any way, then I also invite them to do the test and make a case for what the results signify.

-R
***** I have changed my tune. Please see below. I did say this but you know how things are... **********

Dude, if you were to spend the time spent on any one of your many posts either proving to the readers your "conjecture" (that is to say there is no difference) or testing the simple test provided here perhaps you would not have to get all tessty on this site? Either try it or dont worry abou it. If you are not worried about it then let it go. Maby someone else (me) wants to see the results without haveing to wade through your onset of bad blood.
If I was able to right now I would test this but I am not. I would love to see the results though. I would love to hear your test too.

***** I have changed my tune. Please see below. I did say this but you know how things are... **********
Old 23rd March 2007
  #212
Gear Nut
 

Quote:
Originally Posted by RKrizman View Post
It's not a conjecture? Then what's with all the disclaimers? Do you know this to be the case or not?

As I posted before, in a digital mixer numbers are expanded and then dithered back to 24 bits all the time. I've demonstrated to myself and others many times that this can happen without creating any AUDIBLE difference whatsoever. So the silence test, regardless of its outcome, doesn't shed any light on this.

If somebody else thinks it's meaningful in any way, then I also invite them to do the test and make a case for what the results signify.

-R
Good point. Somehow I missed that in your posts. If digital mixers don't sound different between in and out, then neither should DAWs.

Do you think it is possible that, for lack of any other explanation, digital mixers have tighter control on the processing due to their dedicated hardware?

I've heard this RADAR/DAW disk playback difference claim from reputable users so many times over the past 14 years that I've concluded there must be something to it - hence my conjectures. Do you really believe that these users are imagining it?
Old 23rd March 2007
  #213
Lives for gear
 
Tony Shepperd's Avatar
Uh oh... That sound that you just heard was a can of worms opening up. LOL
Old 23rd March 2007
  #214
Lives for gear
 
RKrizman's Avatar
 

Quote:
Originally Posted by bhenderson View Post
Good point. Somehow I missed that in your posts. If digital mixers don't sound different between in and out, then neither should DAWs.

Do you think it is possible that, for lack of any other explanation, digital mixers have tighter control on the processing due to their dedicated hardware?

I've heard this RADAR/DAW disk playback difference claim from reputable users so many times over the past 14 years that I've concluded there must be something to it - hence my conjectures. Do you really believe that these users are imagining it?
LOL, by digital mixer I meant DAW. What's the difference where the chip resides? PT has its own dedicated cards for this, using the host computer for housekeeping. It's just a packaging difference, IMO. One could do a good job or a bad job of designing either kind of system.

As to your last paragraph, who can say whether people have been imagining this or not, or ascribing the correct reasons? The impression I've always gotten was that people thought it improved PT by using Radar as a front end, and I even toyed around with checking this out for myself. But now you're suggesting there would be no benefit in that because the DAW itself degrades the audio and the external clocking from Radar would be inferior. This is all news to me.

If so many people have heard the difference you're talking about, where files played off of PT through a Radar differ significantly from the same files played right of the Radar, then it shouldn't be hard to come up with some comparison files. Then we can see if it's real or imagined. If it's real, then it makes sense to try to figure out why, and I'll be the first one into the trench.

Looking at the bits and detecting an academic difference doesn't tell us anything. Like I said, there will be technically a difference between a signal before it goes into Radar and after it comes out. We can examine those bits too and verify that they are different, but it doesn't really get at the issue.

If I don't answer any of this it's because I'm off to China for a couple of weeks, I'm not just being rude. I find it all very interesting, and would love to come back and see that someone has posted some (audible) comparison files.

-R
Old 23rd March 2007
  #215
Lives for gear
 
RKrizman's Avatar
 

Quote:
Originally Posted by RedEar View Post
Dude, if you were to spend the time spent on any one of your many posts either proving to the readers your "conjecture" (that is to say there is no difference) or testing the simple test provided here perhaps you would not have to get all tessty on this site? Either try it or dont worry abou it. If you are not worried about it then let it go. Maby someone else (me) wants to see the results without haveing to wade through your onset of bad blood.
If I was able to right now I would test this but I am not. I would love to see the results though. I would love to hear your test too.
There's no bad blood here. I have the highest respect for Barry, and I also have the highest respect for the quality of discourse on this board. If you don't want to wade through all the details, fine. BTW, I assume you own a DAW, otherwise you wouldn't be following this, so here's something you can try. Record something into your DAW. Now create a bunch of aux inputs and bus your recorded track in series through all of them (into one, out of that into the next one, etc.) Now alternately raise and lower the faders on this series of aux channnels, but do so so that the net sum is zero. Now record that to another track. So now you've performed a bunch of math operations on your original signal, and have a recording of the results, at the same volume as the original. Now listen and compare those two files. Blind test. Flip the phase (make sure they're lined up sample-wise, there will be latency from the auxes) What do you hear? This is the sound of doing math to a 24 bit signal.


For years I've posted innumerable files here and elsewhere that deal with a lof of this stuff, and was part of the inception of of Lynn Fuston's Awesum-DAWsum CD and one of the contributors. You can check that one out at his site. That was the one where people were hearing all sorts of astounding differences between files that later proved to be bit identical.

Carry on.
R
Old 23rd March 2007
  #216
Gear Nut
 

Quote:
Originally Posted by RKrizman View Post
LOL, by digital mixer I meant DAW.-R
Oops. Sorry. So back to the original problem....
Old 24th March 2007
  #217
Gear Nut
 
1954U1's Avatar
 

Quote:
Originally Posted by RKrizman View Post
Looking at the bits and detecting an academic difference doesn't tell us anything. Like I said, there will be technically a difference between a signal before it goes into Radar and after it comes out. We can examine those bits too and verify that they are different, but it doesn't really get at the issue.
I dont understand, here.
Do you think that the test proposed by bhenderson is nonsense?
The test is very simple, IMO.
I now dont have a DAW, but you have it.
So, why not doing the test and posting the results,
and _then_ we can debate if the test is meaning or not..?

Last edited by 1954U1; 24th March 2007 at 02:01 AM.. Reason: typo
Old 24th March 2007
  #218
Gear Nut
 
RedEar's Avatar
 

Quote:
Originally Posted by RKrizman View Post
There's no bad blood here. I have the highest respect for Barry, and I also have the highest respect for the quality of discourse on this board. If you don't want to wade through all the details, fine. BTW, I assume you own a DAW, otherwise you wouldn't be following this, so here's something you can try. Record something into your DAW. Now create a bunch of aux inputs and bus your recorded track in series through all of them (into one, out of that into the next one, etc.) Now alternately raise and lower the faders on this series of aux channnels, but do so so that the net sum is zero. Now record that to another track. So now you've performed a bunch of math operations on your original signal, and have a recording of the results, at the same volume as the original. Now listen and compare those two files. Blind test. Flip the phase (make sure they're lined up sample-wise, there will be latency from the auxes) What do you hear? This is the sound of doing math to a 24 bit signal.


For years I've posted innumerable files here and elsewhere that deal with a lof of this stuff, and was part of the inception of of Lynn Fuston's Awesum-DAWsum CD and one of the contributors. You can check that one out at his site. That was the one where people were hearing all sorts of astounding differences between files that later proved to be bit identical.

Carry on.
R
I will say this, after reading your latest post, it apears I misread you. So I appologise.
And I do have a DAW, and have been concidering getting Radar as well. That is why I am following so closely this thread. I will do that test, thanks for posting it. In all honesty, while I do think gear makes a difference I am also focusing on myself and how I react to and adjust what I hear. In the end that is how truely good work gets done.
So anyway, sorry for misunderstanding.
Old 24th March 2007
  #219
Lives for gear
 
adam_f's Avatar
FWIW:

I mixed a project not too long ago that was tracked to RADAR 24.

The mix was done in an SSL room with PT+Apogee converters.

We had the RADAR, Pro Tools and 1/2" Tape to mix down to.

We did an A/B monitoring of the mix while going down.

We monitored the mix coming off the console = SOUNDED GREAT

We monitored the 1/2" Tape = Did what you'd expect = FAT

We monitored the mix sent to 2 tracks on RADAR 24 = VERY, VERY Close in sound to the actual mix comming off the SSL = FANTASTIC

We monitored the same mix sent 2 tracks on PT = TOTAL COLLAPSE

The PT mix was so bad that we were left stunned, slack jawed actually.

The only analogy I can make is that everyting else sounded HUGE compared to Pro Tools, that Pro Tools in comparison sounded like a cassette.

I HOPE THIS HELPS.
Old 24th March 2007
  #220
Gear Nut
 

Quote:
Originally Posted by tubehead View Post
FWIW:

I mixed a project not too long ago that was tracked to RADAR 24. The mix was done in an SSL room with PT+Apogee converters. We had the RADAR, Pro Tools and 1/2" Tape to mix down to. We did an A/B monitoring of the mix while going down. We monitored the mix coming off the console = SOUNDED GREAT. We monitored the 1/2" Tape = Did what you'd expect = FAT

We monitored the mix sent to 2 tracks on RADAR 24 = VERY, VERY Close in sound to the actual mix comming off the SSL = FANTASTIC

We monitored the same mix sent 2 tracks on PT = TOTAL COLLAPSE

The PT mix was so bad that we were left stunned, slack jawed actually. The only analogy I can make is that everyting else sounded HUGE compared to Pro Tools, that Pro Tools in comparison sounded like a cassette. I HOPE THIS HELPS.
Was it PT HD?
Old 24th March 2007
  #221
Lives for gear
I just did the test last night, but not quite in the way described as it had to be a rush job and the results were pretty bad for all DAWs we tried.

I do not have thew time to post the results here, but we tested Nuendo, PT-Lite, Audition and CuBase. I could not do any others as they were being used.

If you create a 0 file, most gave us a zero file back, but the moment we processed something that was not zero, but for example DC or a digitally set sine wave or sq.wave (Rhode & Schwartz sig-gen digital out) the results were very strange indeed.

For example linear normalising a sq.wave that went to 0dBFS should have given us exactly the same sq.wave. It did not!

Some time next week, I shall post the results and do tests on Soundscape and PT-Heavy - but in the mean time, the few results we got last night confirmed what I have always suspected, that nearly all pug-ins and other bundled processes are poorly designed and poorly written.
Old 24th March 2007
  #222
Gear Nut
 

Quote:
Originally Posted by The Byre View Post
I just did the test last night, but not quite in the way described as it had to be a rush job and the results were pretty bad for all DAWs we tried.

I do not have thew time to post the results here, but we tested Nuendo, PT-Lite, Audition and CuBase. I could not do any others as they were being used.

If you create a 0 file, most gave us a zero file back, but the moment we processed something that was not zero, but for example DC or a digitally set sine wave or sq.wave (Rhode & Schwartz sig-gen digital out) the results were very strange indeed.

For example linear normalising a sq.wave that went to 0dBFS should have given us exactly the same sq.wave. It did not!

Some time next week, I shall post the results and do tests on Soundscape and PT-Heavy - but in the mean time, the few results we got last night confirmed what I have always suspected, that nearly all pug-ins and other bundled processes are poorly designed and poorly written.
Byre. Thanks for doing this. I suspected that the "Silence Test" of all zeros might cause the DAW to automatically mute since it's below the noise floor. Did you loop it out and back through AES?

What happens when you make a file of half 1s and half 0s, for example, every 24 bit word set to "000000000000111111111111". The perfect DAW should not change this when looped out and back in and recorded.

Look forward to seeing your results next week.
Old 24th March 2007
  #223
Lives for gear
Don't hold your breath Barry! We are very busy right now and so finding the time and getting access to the machines is not that simple, even for a simple thing like this. I can't say to a customer, "Ooo-er! Could you stop for a while, whilst we play silly games testing a theory from some mad Canadian bloke!"

To be fair to the whole thing I would like to test Soundscape and PT as they use almost the identicle chip set (or crisps if you live in the UK - not to be confused with French fries if you are an American). So any difference between the two would be interesting.

All we did last night was to punch up some theoretical signals from the sig-gen and also a 0-wave and see what the kit in the video room did with them. For example, if you mix zero with zero, you should get zero, but in some of the DAW software, you do not. (And before anybody asks, dither did not play a role in any of this.) Normalising zero gave mostly (but not always) zero - but normalising a 0dBFS signal that should have remained exactly as it was, gave some strange results.

The behavior of some of the effects plugins was so strange that definately some further investigation is required.

I must point out that this was a very rough stab at things and involved a certain amount of beer, so no screen dumps or anything like that.
Old 24th March 2007
  #224
Gear Nut
 

Ok. I've created a -17.5dB DC audio file that contains all 1s HEX, so viewed on a Hex (binary file) editor it looks like thumbnail attached. I've also uploaded the file for anyone to test their DAW. Do this:

1) Loop digital out to digital in.
2) Play back the file and simultaneously record the digital on another channel.
3) Determine if the recorded file is different than the playback file by opening the two files with a Hex editor such as, http://www.hhdsoftware.com/Products/...itor-free.html and by scrolling through the data looking for differences. The playback file will appear as all 11 11 11 etc, so it is really easy to see a change in data.
4) Post the resulting recorded file so the rest of us can see the results

NB: If anyone can't download the Hex editor or does not wish to scroll through a list of 1s, let me know and I'll go through the uploaded recorded file for you and post jpegs of the results.
Attached Thumbnails
What is it about RADAR?-dc-11-hex-binary-view.jpg  
Attached Files

DC Audio Signal - 11 HEX -17.5 dB.wav (5.15 MB, 116 views)

Old 24th March 2007
  #225
Gear Nut
 

Here is the first test result done on Sonar by StormKloud. He recorded from SONAR through digital out into RADAR. I've posted both a JPEG of the binary vile view as well as the audio file. The two files are completely different.

I've asked him to do the same test looping out/in to Sonar to take RADAR out of the equation, and send me the file.
Attached Thumbnails
What is it about RADAR?-dc-11-hex-binary-view-original-vs-sonar.jpg  
Attached Files

DC Audio 11 HEX -17.5 - Sonor.WAV (5.16 MB, 187 views)

Old 25th March 2007
  #226
Lives for gear
 
Tony Shepperd's Avatar
Quote:
Originally Posted by tubehead View Post
FWIW:

I mixed a project not too long ago that was tracked to RADAR 24.

The mix was done in an SSL room with PT+Apogee converters.

We had the RADAR, Pro Tools and 1/2" Tape to mix down to.

We did an A/B monitoring of the mix while going down.

We monitored the mix coming off the console = SOUNDED GREAT

We monitored the 1/2" Tape = Did what you'd expect = FAT

We monitored the mix sent to 2 tracks on RADAR 24 = VERY, VERY Close in sound to the actual mix coming off the SSL = FANTASTIC

We monitored the same mix sent 2 tracks on PT = TOTAL COLLAPSE

The PT mix was so bad that we were left stunned, slack jawed actually.

The only analogy I can make is that everything else sounded HUGE compared to Pro Tools, that Pro Tools in comparison sounded like a cassette.

I HOPE THIS HELPS.

I must live on the opposite end of your world.
A few months ago I got a a call to remix a new artist.
The mixes originally were done on a SSL 9k (label wasn't happy), then to a different studio with another engineer and another 9k.
The label was disappointed TWICE with the different mixes of the singles of the new artist.

I mixed the same song in the box (Pro Tools HD Accel) and they flipped out over it.
So much so that they asked me to remix the entire record.

I HOPE THIS HELPS.

Last edited by Tony Shepperd; 25th March 2007 at 11:09 AM.. Reason: clarification of question
Old 25th March 2007
  #227
Lives for gear
 
Fleaman's Avatar
 

Quote:
Originally Posted by Tony Shepperd View Post
I must live on the opposite end of your world.
A few months ago I got a a call to remix a new artist.
The mixes originally were done on a SSL 9k (label wasn't happy), then to a different studio with another engineer and another 9k.
The label was disappointed TWICE with the different mixes of the singles of the new artist.

I mixed the same song in the box (Pro Tools HD Accel) and they flipped out over it.
So much so that they asked me to remix the entire record.

I HOPE THIS HELPS.
But that was 2 different mix engineers

Obviously the label like Your mixes.

Tubehead's comparison was the same mix in the same room off different systems.

I think you were being a bit sarcastic here though
Old 25th March 2007
  #228
Lives for gear
 
Tony Shepperd's Avatar
Yes, I am being sarcastic, that's why I ended it by saying... I HOPE THIS HELPS!
Did I need to put on a LOL.

It struck me as funny that somehow his comments were helping.
Seriously, we all know RADAR sounds great, it's a great sounding box!
Sonically outstanding, but his Pro Tools comments were not necessary.
Old 25th March 2007
  #229
Lives for gear
 
indie's Avatar
 

Just curious...what was not necessary about his comment? He was telling us his experience in the studio, how printing the same 2mix to PT didn't sound good in comparison. It was the point of his story -- maybe I'm missing something.
Old 25th March 2007
  #230
Gear Nut
 

RADAR DC Audio Loop Back Test

Tonight I went to the office and did the RADAR DC File loop back test in an attempt to understand why some RADAR users are saying that when they play audio files from the hard disk of a DAW, through AES (or some other digital I/O) into RADAR and then out RADAR's converters, it sounds different than when the exact same file is played from the RADAR's hard disk out through it's converters. Is it imagined, or is it real?

The two paths are:

Path 1: Audio file played from DAW disk-> DAW Software-> DAW AES Out-> RADAR AES In-> RADAR Software-> RADAR Converters/Analog Out - RADAR is the clock.

Path 2: Audio file played from RADAR disk-> RADAR Software-> RADAR Converters/Analog Out. RADAR is the clock.

Theoretically, Path 1 and Path 2 should yield the same sound since, in both cases, the same clock and converters are used. However, users are insisting that there is a difference. The only way this is possible is if the audio data is changing in some way between the two paths. To try to understand where the difference is coming from I set up the following test:

First, I created a BWAV audio file containing all 1s (Hex) so that when you scroll through the contents of the file with a Hex editor, all you see is 1s, making it easy to spot differences when comparing to another file that is supposed to be identical. I've attached the created file, "DC Audio 11 Hex - Original.WAV", to this post. I opened the file with a audio editor that I downloaded called GoldWave. The waveforms below show what the file looks like in the time domain - just a block at -17.5 dB:

Zoomed in to 132 ms across the screen it looks like this:


Zoomed in a bit more to 18 ms across the screen it looks like this:


I played the audio file off the RADAR hard disk on track 1, sent it out the 2 channel AES/EBU port and then back into the AES In. During play back, I recorded the incoming signal back onto RADAR's audio hard disk on track 3 using Autopunch to make sure the two files were the same length. I've also attached the resulting recorded file, "DC Audio 11 Hex - Recorded through AES.WAV", to this post. Then I scrolled through both files and up loaded a screen shot of the two files being compared side by side. The files are identical and are shown below:


This test can be performed on any recorder. Earlier I posted the results of a Sonar test sent to me by a user, Stormkloud, and the comparison of the files showed that they were significantly different. I've re-posted it case any of you missed it:


These files look significantly different, so I'm wondering if the test was done right. If the test was performed correctly, then it would appear that the DAW is modifying audio data, just by playing back the file.
Attached Files

DC Audio 11 Hex - Original.WAV (4.88 MB, 87 views)

DC Audio 11 Hex - Recorded Through AES.WAV (4.88 MB, 80 views)

Old 25th March 2007
  #231
Gear Addict
 
lefthando's Avatar
 

Quote:
Originally Posted by bhenderson View Post
The two paths are:

Path 1: Audio file played from DAW disk-> DAW Software-> DAW AES Out-> RADAR AES In-> RADAR Software-> RADAR Converters/Analog Out - RADAR is the clock.

Path 2: Audio file played from RADAR disk-> RADAR Software-> RADAR Converters/Analog Out. RADAR is the clock.
What do you mean by "Analog Out"? How is that part of the path? Did you re-record the RADAR analog signal to another machine?
Old 25th March 2007
  #232
Gear Nut
 

Quote:
Originally Posted by lefthando View Post
What do you mean by "Analog Out"? How is that part of the path? Did you re-record the RADAR analog signal to another machine?
The claim that some users have made is that when they play audio from a DAW's hard drive, through digital outputs into RADAR and monitor through RADAR's converters and using RADAR's clock, i.e. analog output, it doesn't sound as good as when they play the same files from RADAR's hard drive and monitor through the same analog outputs, obviously still using RADAR's clock.

Others on this thread have suggested that those claims are illusory, or psychological in nature, since all things being equal, there is no plausible explanation.

I have suggested that one possible explanation is that the DAW is somehow modifying the data through mathematical operations that are necessary for it to work the way it was designed. If this is true, I can't imagine the data modification would be significant, so then it remains to be determined if such modification of the data would actually affect sound quality.

I therefore proposed an objective test that can be performed on any DAW including RADAR to see if it modifies the data. The test is to create a DC audio file (a file where all the data bytes are identical so it's easy to compare), play it from the DAW through AES and record it to RADAR, then perform a bit-for-bit comparison by scrolling through the binary file looking for differences. The file I created is all 1s (Hex) so if there is the slightest modification to the data of even 1 bit, the 1 Hex would look significantly different and be easy to detect.

This takes converters and all subjectivity (user perception) out of the equation.
Old 25th March 2007
  #233
Gear Addict
 
lefthando's Avatar
 

Yeah, I've been following the thread, but I just don't understand how the RADAR's analog output would affect the files you uploading for analysis.

Or are you merely pointing this out with regard to how a listener would percieve it?

Paul
Old 25th March 2007
  #234
Gear Nut
 

Quote:
Originally Posted by lefthando View Post
Yeah, I've been following the thread, but I just don't understand how the RADAR's analog output would affect the files you uploading for analysis.

Or are you merely pointing this out with regard to how a listener would percieve it?

Paul
Correct. Users have to use the converters cause that's how they detect the difference, but that is rather subjective and I was trying to come up with an objective test, hence the DC test. If the DC test proves there is no possibility of data modification, then I can't see how there could be a difference in the sound when they do go through the converters cause everything else is in the setup is the same.
Old 25th March 2007
  #235
Gear Nut
 

Sonar Update

I think I figured out the Sonar effect. In the previous Sonar test, the panner was set hard left on the channel 1 output so the level was 3 dB high, hence the data looked really strange. We did the test again with the right level setting. However, Sonar IS changing the data by changing the overall level.

The original DC test file has a level of exactly -17.50 dB. Below is a screenshot:



This file was played back off Sonar, out AES, back in through AES, and recorded as a new file. The Sonar recorded file however, is at a slightly lower level than the original playback file. The Sonar recorded file is at almost -17.53 (-0.03 db lower) as shown:



No matter what we did, we could not get Sonar to loop back record a file without changing the level, but as you can see from the zoom in of the waveform, it is perfectly flat so every byte is identical horizontally.

DISCLAIMER: I am not a Sonar or DAW expert. There may be some Sonarites out there that can set me straight and show us how to do the DC loopback test without level degradation. Please feel free to provide direction in this regard.

This test clearly proves that Sonar is consistent, but it is definately changing the overall level down a bit, which may explain why some users think it sounds better when played back from RADAR - it could be simply because when the audio is played back off of RADAR the level is a little hotter - which is sometimes perceived as bigger or fuller. This can happen because a slightly higher level can cause people to perceive sounds, harmonics etc. that they might not percieve at slightly lower levels.

Therefore, there is a real change to the data, AND, there may be a perceived difference in sound. Below is the binary file view showing that the data pattern is consistent, every 24 bits (3 pairs of Hex data) is the same. I scrolled through the entire file and did not see any inconsistencies, so in Sonar, at least there are no artifacts - just a slightly lower level. In the original file, Hex 11 11 11 = binary 000100010001000100010001 = decimal 1,118,481. In the Sonar recorded file, Hex 97 03 11 is displayed in what is called "little endian", which is a Microsoft WAV file specification, so these bytes must be reversed to the "Big endian" format in order to convert the value to binary = Hex 11 03 97 = binary 000100010000001110010111 = decimal 1,115,031 which is 99.69% of the original file level or 0.31% low. 0.9969 = -0.0269681 dB down, which when subtracted from -17.5 = -17.5269, which is what we are seeing in the above waveform file, and in the binary file shown below:



Below is the waveform of the RADAR loopback test, showing that the level is exactly -17.5 dB as with the original playback file:



...and here is the binary file view, showing that the data is identical:



I've also attached the Sonar recorded file, "dc test again !-Sonar DC test aes 3rd try ! lol-2.wav", (see bottom of this post) in case anyone wants to take a look at it.

It would be great if some other DAW users would do the DC test. It would be good to know if other DAWs change the level, or modify the data in any way.

The test is really simple to do.

1) Go to post #230 of this thread: https://www.gearslutz.com/board/showp...&postcount=230 and download the attached file, "DC Audio 11 Hex - Original.WAV" (at the bottom of the post) into your DAW.

2) Play the file out AES(or SPDIF, Lightpipe, or TDIF), loop it back in and record it on another track. Make sure there are no plugs running, all the faders are at zero, and pan is centered, and ensure that the playback and record levels match on the meters.

3) Upload the recorded file to this thread (or email it to me) and I'll post the results using Goldwave for the waveform display and HHD Free Hex Editor for the binary file view.
Attached Files

dc test again !-Sonar DC test aes 3rd try ! lol-2.wav (4.94 MB, 135 views)

Old 26th March 2007
  #236
Gear Nut
 
1954U1's Avatar
 

Quote:
Originally Posted by bhenderson View Post
This test clearly proves that Sonar is consistent, but it is definately changing the overall level down a bit
Not so correct!

This test proves that:
- Sonar is consistent, dealing with "1111111111111...." DC file.
- Sonar is changing [only] the overall level, dealing with "1111111111111...." DC file.

But, what about a real audio file?
I say this, because I think that is different, for an algo, to work with "11111111111..." [-> constant value], instead of "real" values.
Maybe we can try to use a "diff" utility on 2 real audio files instead of watching/scrolling..

That said, if the above test has been correctly done, at least it means that,
with a "11111111..." file, Radar leaves it intact, and Sonar lowers it a bit.

And big respect to Barry, copying/pasting in the w/ends.
He loves his gear.
Old 26th March 2007
  #237
Gear Nut
 

Quote:
Originally Posted by 1954U1 View Post
Not so correct!

This test proves that:
- Sonar is consistent, dealing with "1111111111111...." DC file.
- Sonar is changing [only] the overall level, dealing with "1111111111111...." DC file.

But, what about a real audio file?
I say this, because I think that is different, for an algo, to work with "11111111111..." [-> constant value], instead of "real" values.
Maybe we can try to use a "diff" utility on 2 real audio files instead of watching/scrolling..

That said, if the above test has been correctly done, at least it means that,
with a "11111111..." file, Radar leaves it intact, and Sonar lowers it a bit.

And big respect to Barry, copying/pasting in the w/ends.
He loves his gear.
Yeah. I wondered if DAWs and plugs anticipate DC signals - if not, you could be right. However, when it comes to performing the function of a recorder/player, a DAW should not discriminate - it should faithful play back whatever it is given.

A couple of examples come to mind. Three years ago we installed a RADAR in the Vancouver planetarium. They needed to play normal bandwidth audio but they also needed to record and play stepper motor control voltages so that the lasers were in sync with the music. The lasers were controlled by small mirrors on the ends of the stepper motor shafts which deflected the lasers to shine about 100 feet across the planetarium onto the walls and ceiling. If the control voltages coming off the RADAR, which often go right down to DC, had been out by a small fraction of a millivolt, the lasers would not intersect properly and the patterns would look fuzzy.

Another example is IMAX. They are now installing RADARs in all of their theatres world-wide as players. IMAX demands extremely wide bandwidth audio to drive their $300,000 super high bandwidth speaker systems. The system has specs that go almost right down to DC.

As for "Sonar". As I said before, I like the name cause it sounds like sorta like "RADAR" and I would like to have them be more compatible. Perhaps if some users contact Cakewalk they would listen!

Cheers.
Old 26th March 2007
  #238
Lives for gear
 
Sofa King's Avatar
 

Quote:
Originally Posted by RKrizman View Post
Hey Sean, you'd be the best guy to answer the question we've been kicking around. When you play your tracks off PT through the Radar converters do they sound worse than when you just play them straight off the Radar?

-R
Hey Rick,
I wish I could help, but I never go in that direction.

I either play out from PT through the console, OR
Radar through the console, OR
Radar through PT through the console.

But never PT back through Radar through the console.



Though I will fuel or water to fire by adding this....
Tracks recorded into Radar, and then imported into PT, sound a slight bit different [perhaps better], than when I just track through Radars converters and print directly to PT.

The smart guys here may have an explanation for this, or I could just be entirely insane.

BTW, everything is being clocked by Big Ben, so I dont think its a clocking issue.


All the best,
Sean
Old 26th March 2007
  #239
Gear Nut
 
1954U1's Avatar
 

Quote:
Originally Posted by Sofa King View Post
Though I will fuel or water to fire by adding this....
Tracks recorded into Radar, and then imported into PT, sound a slight bit different [perhaps better], than when I just track through Radars converters and print directly to PT.
Do the D/As or console you use for listening, in the 2 cases, are the sames?
If so, then your ears are saying that the Radar's internal path sound different & perhaps better [more true?] than PT's internal path.

But the point here, as I can understand, is to establish if _objectively_ there are diffs in:
Radar internal audio processing pre&post [it seems there arent]
vs
Daw internal audio processing pre&post [it seems in Sonar (and maybe also in PT?) there are]

So, we need more tests!
Old 26th March 2007
  #240
Lives for gear
 
RKrizman's Avatar
 

[QUOTE=bhenderson;1199232] Make sure there are no plugs running, all the faders are at zero, and pan is centered, and ensure that the playback and record levels match on the meters.

[/QUOTE

Actually, there should be no panning at all. It needs to be a mono track sent to its own output, otherwise you are applying processing and are subject to the particular DAW's panning law.

So my understanding is that doing this test with Sonar results in the file being slightly softer but still exactly the same in every other respect? So Sonar can apply some sort of math function(lowering the amplitude) and still retain signal integrity?

Nice to know I can follow this from my hotel in Beijing, but the connection's a little slow, so I'm mostly just enjoying the read. I think you guys are onto something here. Thanks, Barry, for carrying the torch.

Sofaking, I know you're a busy professional and this is strictly extracurricular screwing around, but if you took a few minutes to patch up your gear like we're suggesting and ran a couple of tracks it could really add to the discussion.

Look at the numbers, AND listen to the sound. Seems fair.

-R
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump