The No.1 Website for Pro Audio
Best converter that you have heard
Old 18th May 2006
  #1
Lives for gear
 

Best converter that you have heard

Sorry, this topic is really getting on the nerves of some of you guys, so go to the next thread. It is a while since I've done MIDI sequencing, but this evening I setup my MIDI interface and did some MIDI recording with my Yamaha Tyros. WHAT A SOUND! I had almost forgot how good it sounds when no converter is present in the signal chain...! It was musical sounding, I did some rhodes solos and stuff, it was really fun to record...! It was like playing in a real band... The depth in the drums was awesome, the kick drum was like the best I've ever heard in my studio...! The softness was incredible! I converted the signal with the Fireface and the best was gone... I am currently saving money for a new converter, I need to do something about this, it's just too frustrating to lose that much beautiful things about the signal...!

But I'm interested in what converter is the best YOU have HEARD in real life and especially what converter have you noticed can translate keyboard sounds really well? My biggest dream would be to be able to capture my MIDI sequences with such a quality that I wouldn't notice any difference... Actually that was my intention when I changed from a cheap portable tape recorder into an expensive digital audio interface, I thought I would be able to capture THAT sound. But far from it, too far from it... I must say my creative process suffers from this, because I kind of lose the vibe when I'm tracking to the DAW. When I'm doing MIDI sequencing it feels awesome. I'm able to be really creative, record things that feel perfect, enjoying a beautiful sound etc. For this reason I think I will start making complete MIDI mixes and convert the whole song at once. Too bad it needs to be converted to be captured on a CD...!! *arghhh*

It's something about that latency that kills the sound. I think it is due to the fact that I'm recording an instrument at a time during the playback of other tracks. Those latency samples seem to be causing phase shifting that kills the vibe. What is the best sounding buffer size?! How many samples should be used as space between tracks? How can I better control this in order to make the harmonic content better? I think I have phase shifting problems but I don't really know how to target them...

Let's say I have something like 128 samples between each track. If I record 4 instruments the difference in samples between the first and the last track will be much bigger than 128... I'm interested in what kind of damage a thing like this can bring to a mix... I guess it's some sort of complex phase shifting that kills the vibe.
Old 18th May 2006
  #2
Lives for gear
 

Sigh ... i'm not sure if you're trolling or what ... but in the spirit of what the heck ...

CrazyRainbow - it is 100% impossible to listen to your Tyros with "no converter present in the signal chain".

Quite apart from the fact that all the samples on board were converted from analog to digital via a converter in the first place ...

If you were actually hearing any sound at all, you were hearing it through a D/A converter that you own. Maybe not the Tyros converter (if you were using s/pdif), but through something. Even if you burnt an audio CD from your s/pdif input - you would be hearing it through the converter of your CD player.

So IF, as you say, it was such a joy to hear the damn thing, then you must be happy with the D/A converter you were using. Whatever that was.

RainbowMan, it is 100% possible to make CD's of virtual synth music with No converters at all. But these will still be heard via the CD players converter - which in most cases is very sub standard. But it usually doesn't matter to the listener.


OK - back to your original question. I personally I don't own a Lavry Gold or Cranesong, but i'm sure those that do will tell you they are the best. There are plenty of good converters out there.

Yesterday, I ordered a Benchmark DAC-1 - and i'm sure that would be adequate for most gearslutz. Recently i've been using my Kurzweil Rumour, and that's not bad either - but the specs for the Benchmark are much better.
Old 18th May 2006
  #3
Lives for gear
 

TonyCrazyStorm - are you sure you aren't Walters?

"Those latency samples seem to be causing phase shifting that kills the vibe. What is the best sounding buffer size?! How many samples should be used as space between tracks? How can I better control this in order to make the harmonic content better? I think I have phase shifting problems but I don't really know how to target them..."

That's pure Walters-speak.

It should be fairly obvious that - after tracking - buffer size/latency shouldn't make any difference in audio quality. Unless your CPU is busting a gut, and you can improve peformance with a greater latency.

Are you having us on?
Old 18th May 2006
  #4
You don't say what DAW software you're using (I'm assuming it's a computer based DAW)...

This is NOT an endorsement of Tracktion -- I'm merely using it here to illustrate the issue a bit:

If you're using Mackie Tracktion there's a track alignment calibration utility (in version one you access it in the track view with an analog input/track selected). With it, you can test your overdubbed materials time-alignment with playback tracks. If they don't align, it will fill in an offset time that will automatically realign your track. (I believe Cubase's external FX delay compensation ping utility works about the same. Just as you can test a whole chain, you can test the direct output-to-input loop.)

[Whatever you do -- do NOT monitor live source while you're doing a loopback test or you will end up with the feeback loop from HELL.]

Here's a screen shot of the help balloon for the Tracktion alignment utility:



When you click Auto-Detect it pops another dialog that tells you to connect your output to your input (one side, left or right, seems sufficient) and hit the test button. When you do, it sends some tones through, measures the offset between and puts a default value in the Time Adjust field.

So, you can use it to set proper alignment for just your interface, or for a whole analog audio chain (for, say, a real-time analog FX send).

I'm pretty sure the Cubase ping utility is probably pretty similar. If you don't have any outboard FX in your loop, you'll simply be testing the alignment of new tracks against old ones. In a perfect recording system, they should line up precisely.

But on many real worls systems, the results may not be so good.

In my own recent tests I found these consistent misalignments in two different DAWs with my MOTU 828mkII Firewire interface:

Sonar
WDM (actually WDM/KS for Sonar): 8.1 ms
ASIO drivers 2.3 ms

Tracktion
(no WDM available)
ASIO 4.08 ms
ASIO Direct 1.17 ms

Now, again, you have the ability to have Tracktion auto-correct the alignment. But in Sonar, I have to nudge (move) all new tracks 'left' by 356 samples (8.1 ms) when using WDM drivers. I'm experimenting with switching to the ASIO -- but I still plan on nudging, even though it's a relatively ignorable 2.3 ms. I'm in the habit. Why shouldn't it be dead on?

Anyhow, it seems like there is an increasing awareness of this issue and I'm wondering if it's hitting critical mass now because so many people now have USB and Firewire interfaces which CAN have a number of latency-related issues. (Mind you, THIS track alignment issue is related to playback/monitor latency -- but it does NOT appear to be directly correlative. Some people report almost no track misalignment, even though their interface buffers are just as big and their playback latency is about the same.)

Just one more thing to worry about... but SINCE you mentioned it...


Anyhow, even if you don't have Tracktion (or Cubase?) you can set up a loopback test on any DAW software to find out if such an alignment problem is affecting you.

Again, while making SURE you do not accidentally monitor your live source (feedback loop from HELL): connect your DAW output to your DAW input.

Take a previously recorded track with a good, identifiable transient. (You can even chop a sound to get a hard transient.) Play it back and record in on a new track.

Zoom WAY IN... so you can see individual sample steps and count the sample addresses in your ruler/timeline.

If everything is absolutely hunky dory, the original and the copy should line up perfectly on the time axis. (The waves may not be identical unless you carefully matched level. You don't care about that, though. You just want them to start at precisely the same time.)

But for a LOT of folks, they won't. It might be anywhere from a ms or two on up.

Hopefully, the offset will be consistent. For most folks it seems to be.

(I have a USB mic that comes up with different times -- but that's more complex since it's 'one-way' and that means it's using the built-in interface for playback. Still they SHOULD match up, so I DO recalibrate each time I use it. For that reason I restrict it to working with Tracktion.)


At any rate, once you determine what your offset is, you can decide whether or not you want to worry about it. I suspect most folks who only have a couple ms, maybe less than 5 ms, probably ignore the issue. If, OTOH, it's getting up closer to 10 ms you really are talking about an amount of time that can contribute to rhythmic slur, a rhythmic 'fog' if you will, perhaps, as successive tracks may make time reference to previous tracks... if everything is dominated by a central drum track, your rhythmic center may not drift all that much. But if you listen to the bass (and ignore the drums) when you track guitar and then the guitar when you track keyboards, your worst case scenario with a track misalignment of 10ms might be something like:

Drums '0' (this is our reference)
Bass 10 ms behind Drums
Guitar 20 ms behind drums (if he cues off the bass)
Keyboard 30 ms behind drums (if he cues off the guitar)

This is an extreme example, of course, since most folks will probably refer to the drums as a primary reference... but you can see how it COULD contribute to rhythmic imprecision...


Once you determine the amount of any misalignment you MAY have, if you decide you want to deal with it, it's relatively easy if you have a "nudge" command similar to that in Pro Tools and Sonar (and undoubtedly others)... you set a nudge value for the precise amount of samples of offset and get into the habit of doing it to new tracks.

(In Sonar, you can't nudge a track 'left' if it's already up against the beginning of the project -- 0:00. So I just slip trim the edge away from 0. And, THAT way I know any audio tracks that go all the way to 0:00 probably haven't been adjusted.)

If you don't have a nudge that can move precise sample amounts, then see if you can nudge/move by a ms value. Samples divided by 44100 (at 44.1) will return the ms value accurately enough for these purposes. If the only move you have is in MIDI ticks, you may what to create a little cheat sheet for some typical times, say 90 bpm, 120, maybe 140 if you're a modern dance producer, whatever typical project tempos you might encounter, with an equivalency for the amount you need to move your new audio tracks.


Anyhow... I must surely be burning off nervous energy from putting off REAL work, so I'm gonna leave this here before it drifts off into quantum mechanics or something.


Hope that's food for thought, anyhow.
Old 19th May 2006
  #5


The best converters are the ones you don't hear at all.

By the way, if you have a decent recording set-up, the latencey will be compensated for when multitracking - just don't try to monitor the source though an ouput on the recording program.





-tINY

Old 19th May 2006
  #6
Lives for gear
 
synthoid's Avatar
 

RainbowStorm,

Dude, honestly, you are confusing converter quality with latency problems. I don't know which one you have, but they are really different. Like the difference between athlete's foot and the flu.

You would be really happy I think with an MPC4000 and a dedicated hard disk recorder like the HD24XR or a Radar. The timing of the MPC is rock-solid and if you track into a simple dedicated recorder you will minimize latency issues. When it's mix time, copy .wav files into your computer and mix away.

-synthoid
Old 19th May 2006
  #7
Lives for gear
 

If you use zero latency monitoring, it's just not an issue.

I really think this must be some operator error or DAW setup problem that's bugging Rainbow.

For example - I wonder if he's got a midi feedback problem? When I was learning midi stuff, that was one of the first issues I had. If connect your keyboard in to your sequencer out, and your sequencer in to your keyboard out ...

and if you don't disable Local ... then you get a really nasty midi note doubling, which is the worst imaginable phase issue you can have.

What is the DAW again? I use Cubase SX3, and it compensates for latency. Although when I first set it up, I had a wierd timing issue that took me a while to resolve.

If Rainbow is monitoring both the input and the output, the latency will cause similar flanging/phase problems.

There's got to be some simple basic reason - unless Rainbow is really Walters.
Old 19th May 2006
  #8
Lives for gear
 
dreamsongs's Avatar
 

Kiwi, just curious...

I have SX3 as well. Do you experience phasing problems when sending/returning with outboard gear on SX3 while live monitoring ?

Or, do you use the "external fx" bus and insert outboard gear that way ?

With the latter, I find that once I use any fx outboard processor as an insert, I can't use it anymore during that mixing session as it is "tied up" on that particular channel/track ...

I've only had SX3 for 6 months and it has been fine, but I still haven't found the best way to incorporate my OB gear...
Old 19th May 2006
  #9
Lives for gear
 
synthoid's Avatar
 

>> If you use zero latency monitoring, it's just not an issue.

what's not an issue? I think he is complaining about the mis-alignment in time of two recorded tracks. It sounds like he has variable latency going from inputs to tracks. Then again, I'm not altogether sure what problem he is describing...

-synthoid
Old 19th May 2006
  #10
Lives for gear
 

At the moment I stay in the box. My outboard is only used while tracking, so I don't use the external FX bus.

That's going to change very soon - i've ordered a Benchmark DAC-1 to make the trip outside the box worthwhile.

My plan is simply to re-record onto another track and then realign the new track to the old. I like the trick of inverting the phase, and then adjusting for the maximum cancellation, to find the perfect sync, before flipping the phase back.

I expect I'll learn the exact number of samples to pull it forward, and should be able to do it without thinking about it.
Old 19th May 2006
  #11
Lives for gear
 

Quote:
Originally Posted by synthoid
>> If you use zero latency monitoring, it's just not an issue.

what's not an issue? I think he is complaining about the mis-alignment in time of two recorded tracks. It sounds like he has variable latency going from inputs to tracks. Then again, I'm not altogether sure what problem he is describing...

-synthoid
Well i'm still not sure if Rainbow is serious or not ... Walters leaves, Rainbow arrives ... seem to have similar rhetoric. He isn't easy to understand, and maybe that's on purpose.

All we know is that he loves the sound of his Tyros, but it all turns sour when he records it. It is obviously not the sound of his converters that is his problem - because he loves the sound first up. So what is he doing wrong? Stuffed if I know - i'm just trying to give his some clues, in case he's genuine.
Old 19th May 2006
  #12
Gear Nut
 

Quote:
Kiwiburger: All we know is that he loves the sound of his Tyros, but it all turns sour when he records it. It is obviously not the sound of his converters that is his problem - because he loves the sound first up.
Actually I think his problem is he's unhappy with his analog-digital conversion when he's recording his Tyros. The Tyros is a digital keyboard/synth that has analog outputs which must be digitally converted before it's digitally recorded. That latter step is where he's finding the reduction in sound quality.

Frankly I have the same problem myself (though I realize latency ain't the issue). Sometimes when playing with my hardware synths I'm really happy with their sound, but find that once the synths get recorded there's a noticeable difference in the quality of sound (which may not be entirely explained by poor a/d conversion). Not quite enough for me to really care all that much, but the difference is there.

tINY's advice is helpful: "the best converters are the ones you don't hear at all."

So I think what RainbowStorm is asking for are recommendations for a/d convertors that "you don't hear at all" (i.e. so what he records will sound just like what is outputted from his Tyros).
Old 19th May 2006
  #13
Lives for gear
 

Now you're just teasing the poor lad.

Analog tape has so much hiss, i'm sure Rainbow would not like to hear his Tyros coming back from tape.


OK - I don't own a Fireface, but the specs look pretty good:

Quote:
Eight balanced line inputs and outputs with software controlled switching of the reference levels (-10 dBV, +4 dBu, HiGain), of course realized discretely in the analog domain, guarantee highest dynamic range and highest fidelity. Apart from the levels of the microphone pre-amps and the headphone outputs, all device settings are software controlled. Equipped with the latest A/D and D/A converter chips, all I/Os operate up to 192 kHz and reach 119 dBA dynamic range on playback - even the headphone output!
The Tyros only has unbalanced outs. Maybe he is mismatching unbalanced with balanced?
Would a Radial JDI Duplex help match up to his Fireface?
I see the gain is electronically controlled - is he mismatching -10 with +4, or "high gain"?
Is he tracking too hot?
Does he have a midi loop, causing flanging?
Is he monitoring both inputs and outputs and hearing flanging due to latency? Just monitor the input, and the latency won't worry you at all.

The more I think about this, the more I suspect operator error. Plenty of room to screw up.
I don't expect the Tryos converters to be better than the Fireface, so it should be possible to record this with no discernible loss in quality.
Old 19th May 2006
  #14
Lives for gear
 

Thanks all for trying to help out here, especially you theblue1 for that very long post, but of course I appreciate all of your help!

I have the following DAW software:

Cakewalk Sonar
Cubase SX v2.something
Steinberg Nuendo v2.2

Currently I have serious sync problems with my USB MIDI interface in both Cubase SX and Nuendo, so I'm working on that. But that's a completely different thing, I'm talking about direct audio recording right now (with no MIDI involved)... (I play with the local disabled, to not get double signal.)
I'm using Cubase SX right now because it is the best sounding (for some odd reason).

But, is very small track timing misalignment even a problem? I don't have any problems with the rhythm, I'm only concirned about what effect this unnatural sample offset difference between tracks will have, especially related to the digital summing process and the overall vibe. I've just thought that maybe the vibe gets killed due to the misalignment, because I think I don't have the tracks lined up on a sample based accuracy, I'm mostly recording at 44,1KHz... Maybe I should try to record at 176,4 KHz and set the buffer size as low as possible... I'll try that for now and then I'll try to align the tracks manually to see what kind of effect it has...
Old 19th May 2006
  #15
Gear Nut
 
johnnyvince's Avatar
 

Quote:
Originally Posted by rufus13
are found inside AMPEX AG-series recorders: Ferric-to-voltage converters. Pretty nice, when used with the right talent. Some use those big hot glass circuits, with heavy wire-wound inductive audio couplers. Later models use hot little metal toobs and couple right to the outputs.

I don't know how they make money on these things; they're hand-built.

rufus13
Reminds me of a clever photo caption in the latest issue of Make Magazine that reads: "This wireless local-area audio network server broadcasts music via a protocal known as "FM".
Old 19th May 2006
  #16
Lives for gear
 
Harley-OIART's Avatar
 

Quote:
Originally Posted by tINY


The best converters are the ones you don't hear at all.
-tINY

2 x thumbsup
Old 19th May 2006
  #17
Lives for gear
 
lawrence_o's Avatar
Best converters: by far, Prismsound. Most natural, don't colour.
You AD a sound and DA it and you don't hear the difference.

Cheers.
Old 19th May 2006
  #18
Lives for gear
 

Quote:
Originally Posted by johnnie red
Actually I think his problem is he's unhappy with his analog-digital conversion when he's recording his Tyros. The Tyros is a digital keyboard/synth that has analog outputs which must be digitally converted before it's digitally recorded. That latter step is where he's finding the reduction in sound quality.

Frankly I have the same problem myself (though I realize latency ain't the issue). Sometimes when playing with my hardware synths I'm really happy with their sound, but find that once the synths get recorded there's a noticeable difference in the quality of sound (which may not be entirely explained by poor a/d conversion). Not quite enough for me to really care all that much, but the difference is there.

tINY's advice is helpful: "the best converters are the ones you don't hear at all."

So I think what RainbowStorm is asking for are recommendations for a/d convertors that "you don't hear at all" (i.e. so what he records will sound just like what is outputted from his Tyros).
That's right... The difference is not very huge, but you can hear it and especially feel it... The midi version is clean, the audio version is dirty. Both are equally loud. I simply want a cleaner signal. When the signal is clean I can feel the harmonic content of each instrument much better. This is why I have suspected the latency, because that could make the signal dirty due to bad track alignment and mess up beautiful frequencies... I have the Yamaha Tyros connected directly to the interface and I'm using no additional audio interface mixer... Currently I've been using 176,[email protected], but it's the same with 44,[email protected] I guess the converter can't convert the signal clean enough for my taste, maybe due to the crystal reference clock in the converter. The signal might be finally killed during the digital summing process...

When audio is clean it is pleasant to listen to. Each instrument has a smooth edge and it sounds natural. The high end is thin due to no noise artifacts, the low mids are clear and the overall sound is relaxing to listen to. It also feels a little lighter overall, but still with a lot of color. The reverb tail/decay on one instrument is not messed up by another instrument's reverb tail/decay in the mix. Each instrument simply shines in a special unique way... Awesome!

I know this is hard for you to be able to sort out, but I think trying a new converter would give me some useful perspective...

So what's the best converter you've heard?
Old 19th May 2006
  #19
Lives for gear
 
funka's Avatar
 

Quote:
Best converters: by far, Prismsound. Most natural, don't colour.
You AD a sound and DA it and you don't hear the difference.
I have quite the same thing with aurora16. I have extracted a track from a CD, played it thru a pair of aurora D/A then back to a pair of aurora A/D. Then compared the original to the recorded track(aurora 16 D/A->A/D) thru a benchmark DAC1, and I heard no difference. Doing the same thing with RME ADI 8 DS, I heard a difference(mainly stereo imaging).

So, for me, aurora16 are the best I know, clear and neutral.
Old 19th May 2006
  #20
Lives for gear
 

Quote:
Originally Posted by lawrence_o
Best converters: by far, Prismsound. Most natural, don't colour.
You AD a sound and DA it and you don't hear the difference.

Cheers.
That's exactly what I want. I don't want to hear any difference at all, I want the input to tell me what it will sound like by 100%... In that way I can focus all my energy on tracking well...! lawrence_o, have you compared the RME Fireface 800 with the Prismsound? Is it converting the signal cleaner?
Old 19th May 2006
  #21
Lives for gear
 

Quote:
Originally Posted by funka
I have quite the same thing with aurora16. I have extracted a track from a CD, played it thru a pair of aurora D/A then back to a pair of aurora A/D. Then compared the original to the recorded track(aurora 16 D/A->A/D) thru a benchmark DAC1, and I heard not difference. Doing the same thing with RME ADI 8 DS, I heard the difference.

So, for me, aurora16 are the best I know, clear and neutral.
Interesting...! Have you compared it to any Prismsound converter?
Old 19th May 2006
  #22
Lives for gear
 
Harley-OIART's Avatar
 

If u get a chance... Listen to RADAR (http://www.izcorp.com/mainframe.asp)

Sweet conversion to my ears.
Old 19th May 2006
  #23
Lives for gear
 

Quote:
Originally Posted by Harley-OIART
If u get a chance... Listen to RADAR (http://www.izcorp.com/mainframe.asp)

Sweet conversion to my ears.
Hmm... So iZ makes converters as well, or what kind of converter can be found in a RADAR DAW?
Old 19th May 2006
  #24
Lives for gear
 
funka's Avatar
 

Quote:
Interesting...! Have you compared it to any Prismsound converter?
No, I have not.
Just had compared it with the other gear I have, masterlink converters(pretty good A/D but D/A is crap IMO), converters in my digital console, soundcraft 328, dac1 and RME ADI 8 DS.
From now, the aurora 16 is a true reference for me.
But I will keep the DAC1 for its monitoring facilities and conversion quality(have not compared it with the aurora yet...).
Old 19th May 2006
  #25
Lives for gear
 
synthoid's Avatar
 

Well, the way I read Rainbow's post, he is worried about latency and track-to-track offset in his DAW. This is a very understandable concern and I hate this problem too, I like hardware recorders -- digital or analog -- for tracking because you get an absolutely consistent and simple latency model. I know you can get this with a DAW, but I also know enough about device drivers and OSes and timing of interrupts and whatnot to know that no matter how much you fiddle, you won't match the simple reliability of a dedicated hardware recorder for tracking. So I gotta say again -- you should look at something like the Radar, not only for it's A/D conversion (I don't believe that's your problem, honestly, I have no idea why the RME fireface gets a bad rap around here for its A/D conversion) but especially because you can track into the Radar without going nuts about latency. For my money, I'd rather have a couple or three or four HD24XRs than a Radar, but many folks feel that the Radar is a higher-quality piece of gear with better converters.

-synthoid
Old 19th May 2006
  #26
Lives for gear
 
pingu's Avatar
 

Quote:
Originally Posted by theblue1
You don't say what DAW software you're using (I'm assuming it's a computer based DAW)...

This is NOT an endorsement of Tracktion -- I'm merely using it here to illustrate the issue a bit:

If you're using Mackie Tracktion there's a track alignment calibration utility (in version one you access it in the track view with an analog input/track selected). With it, you can test your overdubbed materials time-alignment with playback tracks. If they don't align, it will fill in an offset time that will automatically realign your track. (I believe Cubase's external FX delay compensation ping utility works about the same. Just as you can test a whole chain, you can test the direct output-to-input loop.)

[Whatever you do -- do NOT monitor live source while you're doing a loopback test or you will end up with the feeback loop from HELL.]

Here's a screen shot of the help balloon for the Tracktion alignment utility:



When you click Auto-Detect it pops another dialog that tells you to connect your output to your input (one side, left or right, seems sufficient) and hit the test button. When you do, it sends some tones through, measures the offset between and puts a default value in the Time Adjust field.

So, you can use it to set proper alignment for just your interface, or for a whole analog audio chain (for, say, a real-time analog FX send).

I'm pretty sure the Cubase ping utility is probably pretty similar. If you don't have any outboard FX in your loop, you'll simply be testing the alignment of new tracks against old ones. In a perfect recording system, they should line up precisely.

But on many real worls systems, the results may not be so good.

In my own recent tests I found these consistent misalignments in two different DAWs with my MOTU 828mkII Firewire interface:

Sonar
WDM (actually WDM/KS for Sonar): 8.1 ms
ASIO drivers 2.3 ms

Tracktion
(no WDM available)
ASIO 4.08 ms
ASIO Direct 1.17 ms

Now, again, you have the ability to have Tracktion auto-correct the alignment. But in Sonar, I have to nudge (move) all new tracks 'left' by 356 samples (8.1 ms) when using WDM drivers. I'm experimenting with switching to the ASIO -- but I still plan on nudging, even though it's a relatively ignorable 2.3 ms. I'm in the habit. Why shouldn't it be dead on?

Anyhow, it seems like there is an increasing awareness of this issue and I'm wondering if it's hitting critical mass now because so many people now have USB and Firewire interfaces which CAN have a number of latency-related issues. (Mind you, THIS track alignment issue is related to playback/monitor latency -- but it does NOT appear to be directly correlative. Some people report almost no track misalignment, even though their interface buffers are just as big and their playback latency is about the same.)

Just one more thing to worry about... but SINCE you mentioned it...


Anyhow, even if you don't have Tracktion (or Cubase?) you can set up a loopback test on any DAW software to find out if such an alignment problem is affecting you.

Again, while making SURE you do not accidentally monitor your live source (feedback loop from HELL): connect your DAW output to your DAW input.

Take a previously recorded track with a good, identifiable transient. (You can even chop a sound to get a hard transient.) Play it back and record in on a new track.

Zoom WAY IN... so you can see individual sample steps and count the sample addresses in your ruler/timeline.

If everything is absolutely hunky dory, the original and the copy should line up perfectly on the time axis. (The waves may not be identical unless you carefully matched level. You don't care about that, though. You just want them to start at precisely the same time.)

But for a LOT of folks, they won't. It might be anywhere from a ms or two on up.

Hopefully, the offset will be consistent. For most folks it seems to be.

(I have a USB mic that comes up with different times -- but that's more complex since it's 'one-way' and that means it's using the built-in interface for playback. Still they SHOULD match up, so I DO recalibrate each time I use it. For that reason I restrict it to working with Tracktion.)


At any rate, once you determine what your offset is, you can decide whether or not you want to worry about it. I suspect most folks who only have a couple ms, maybe less than 5 ms, probably ignore the issue. If, OTOH, it's getting up closer to 10 ms you really are talking about an amount of time that can contribute to rhythmic slur, a rhythmic 'fog' if you will, perhaps, as successive tracks may make time reference to previous tracks... if everything is dominated by a central drum track, your rhythmic center may not drift all that much. But if you listen to the bass (and ignore the drums) when you track guitar and then the guitar when you track keyboards, your worst case scenario with a track misalignment of 10ms might be something like:

Drums '0' (this is our reference)
Bass 10 ms behind Drums
Guitar 20 ms behind drums (if he cues off the bass)
Keyboard 30 ms behind drums (if he cues off the guitar)

This is an extreme example, of course, since most folks will probably refer to the drums as a primary reference... but you can see how it COULD contribute to rhythmic imprecision...


Once you determine the amount of any misalignment you MAY have, if you decide you want to deal with it, it's relatively easy if you have a "nudge" command similar to that in Pro Tools and Sonar (and undoubtedly others)... you set a nudge value for the precise amount of samples of offset and get into the habit of doing it to new tracks.

(In Sonar, you can't nudge a track 'left' if it's already up against the beginning of the project -- 0:00. So I just slip trim the edge away from 0. And, THAT way I know any audio tracks that go all the way to 0:00 probably haven't been adjusted.)

If you don't have a nudge that can move precise sample amounts, then see if you can nudge/move by a ms value. Samples divided by 44100 (at 44.1) will return the ms value accurately enough for these purposes. If the only move you have is in MIDI ticks, you may what to create a little cheat sheet for some typical times, say 90 bpm, 120, maybe 140 if you're a modern dance producer, whatever typical project tempos you might encounter, with an equivalency for the amount you need to move your new audio tracks.


Anyhow... I must surely be burning off nervous energy from putting off REAL work, so I'm gonna leave this here before it drifts off into quantum mechanics or something.


Hope that's food for thought, anyhow.



If everyone was as helpful as you mate i wouldnt have dreams and aspirations of hacking every single forum to death.


Good one man.


its too bad your the minority.
Old 19th May 2006
  #27
Lives for gear
 
Albert's Avatar
 

I think the original question was pretty good.

Quote:

"But I'm interested in what converter is the best YOU have HEARD in real life and especially what converter have you noticed can translate keyboard sounds really well?"

That's very specific, and he doesn't want any regurgitation of marketing hype or third person "I heard that brand X is good" kind of stuff.

So to directly answer the question, the best converters I've heard are the Benchmark DAC1, and the Universal Audio 2192. The 2192 is excellent on synths, I use it all the time record synths and samplers to disk.
Old 19th May 2006
  #28
Lives for gear
 

Quote:
Originally Posted by Albert
I think the original question was pretty good.

Quote:

"But I'm interested in what converter is the best YOU have HEARD in real life and especially what converter have you noticed can translate keyboard sounds really well?"

That's very specific, and he doesn't want any regurgitation of marketing hype or third person "I heard that brand X is good" kind of stuff.

So to directly answer the question, the best converters I've heard are the Benchmark DAC1, and the Universal Audio 2192. The 2192 is excellent on synths, I use it all the time record synths and samplers to disk.
That's really interesting! Maybe I should focus on the Universal Audio 2192 since my mixes are heavily keyboard oriented... Thanks for the advice!
Old 19th May 2006
  #29
Best A/D I've heard is the Crystal CS5381 ADC. Load it with a very fast transconductance opamp front end and it's very pleasing, not euphonic at all.

Best D/A is the BurrBrown PCM/DSD1792, by far. -129 db THD+noise is very hard to beat. With direct coupled, class A transconductance I/V stages and output filters, it's like digging down to the next level. You really start to hear new things on familiar recordings with this level of conversion.

Everything else is second banana to me. Tracking the A/D into a Alesis HD24XR is very hard to beat, especially with the PCM1792 on the outputs.

Jim Williams
Audio Upgrades
Old 19th May 2006
  #30
Lives for gear
 

Quote:
Originally Posted by Jim Williams
Best A/D I've heard is the Crystal CS5381 ADC. Load it with a very fast transconductance opamp front end and it's very pleasing, not euphonic at all.

Best D/A is the BurrBrown PCM/DSD1792, by far. -129 db THD+noise is very hard to beat. With direct coupled, class A transconductance I/V stages and output filters, it's like digging down to the next level. You really start to hear new things on familiar recordings with this level of conversion.

Everything else is second banana to me. Tracking the A/D into a Alesis HD24XR is very hard to beat, especially with the PCM1792 on the outputs.

Jim Williams
Audio Upgrades
Jim, you seem to have some interesting insight into converter chips. Do you know what converter chip RME Fireface 800 is utilizing and how it compares to the chips in the Apogee/Lynx/Universal Audio? You mentioned the Crystal CS5381 ADC chip, what commercial units utilize that chip?

BTW, "harmonic bloom" sounds like something for me...!!

Check it out! (UA 2192)

(2:29 - 3:26)
http://websrvr25ca.audiovideoweb.com...o/WMV/2192.wmv
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