John T
Max Output Level: -7.5 dBFS
- Total Posts : 6783
- Joined: 2006/06/12 10:24:39
- Status: offline
Re:Sample rate conversion question.
2012/04/02 11:18:16
(permalink)
Well, a 64th note in 4/4 at 120bpm is 0.031 seconds long. So let's call that about 3/100ths of a second. A single cycle at 44.1khz is, by definition, 1/44100th of a second, with 48khz being 1/48000th of a second. You're simply not getting anywhere near the realm of audible or musical differences at these rates.
http://johntatlockaudio.com/Self-build PC // 16GB RAM // i7 3770k @ 3.5 Ghz // Nofan 0dB cooler // ASUS P8-Z77 V Pro motherboard // Intel x-25m SSD System Drive // Seagate RAID Array Audio Drive // Windows 10 64 bit // Sonar Platinum (64 bit) // Sonar VS-700 // M-Audio Keystation Pro 88 // KRK RP-6 Monitors // and a bunch of other stuff
|
bapu
Max Output Level: 0 dBFS
- Total Posts : 86000
- Joined: 2006/11/25 21:23:28
- Location: Thousand Oaks, CA
- Status: offline
Re:Sample rate conversion question.
2012/04/02 11:19:28
(permalink)
|
drewfx1
Max Output Level: -9.5 dBFS
- Total Posts : 6585
- Joined: 2008/08/04 16:19:11
- Status: offline
Re:Sample rate conversion question.
2012/04/02 11:38:28
(permalink)
mike_mccue
Here is a website that claims to have collected SRC data from more apps than I can count:
http://src.infinitewave.ca/
I selected the "sweep" analysis and compared R8brain Pro, Minimum Phase to SONAR 8.5.
The results seem to close, but there seems to be enough difference to encourage curiosity.
best regards,
mike
Mike, if you'd been paying attention, you'd have seen that I posted that link a long time ago (post #22).  You can learn a lot from that site, but people should keep in mind that when the differences are far from being audible (note the color scale on the right for the sweep test), "better" is a questionable term. Having said that, note that: 1. The iZotope 64 filters are almost scary-perfect. You can use those as reference standards. 2. If you compare Sonar's SRC to competing DAW's, you'll get an idea of just how good Sonar's SRC is (though I'm not sure how up to date the versions tested there are).
 In order, then, to discover the limit of deepest tones, it is necessary not only to produce very violent agitations in the air but to give these the form of simple pendular vibrations. - Hermann von Helmholtz, predicting the role of the electric bassist in 1877.
|
Jonbouy
Max Output Level: 0 dBFS
- Total Posts : 22562
- Joined: 2008/04/14 13:47:39
- Location: England's Sunshine South Coast
- Status: offline
Re:Sample rate conversion question.
2012/04/02 12:03:09
(permalink)
Bapu, no it won't make any difference.
HTH
"We can't do anything to change the world until capitalism crumbles. In the meantime we should all go shopping to console ourselves" - Banksy
|
drewfx1
Max Output Level: -9.5 dBFS
- Total Posts : 6585
- Joined: 2008/08/04 16:19:11
- Status: offline
Re:Sample rate conversion question.
2012/04/02 12:04:28
(permalink)
Danny Danzi
There would be a few slight timing issues once they brought my stuff into their projects to where I'd have to show them where they needed to align something.
From a theory standpoint, we generally assume the programmers didn't screw things up. But it's always possible that something like, say, the delay compensation of the SRC routine wasn't done correctly. But it's hard to say what happened without looking at the specific stuff in detail. I can't come up with any reason why a properly implemented SRC would have timing differences even remotely close to being audible.
In order, then, to discover the limit of deepest tones, it is necessary not only to produce very violent agitations in the air but to give these the form of simple pendular vibrations. - Hermann von Helmholtz, predicting the role of the electric bassist in 1877.
|
js516
Max Output Level: -84 dBFS
- Total Posts : 347
- Joined: 2006/05/17 15:14:53
- Status: offline
Re:Sample rate conversion question.
2012/04/02 12:12:31
(permalink)
☄ Helpful
|
bapu
Max Output Level: 0 dBFS
- Total Posts : 86000
- Joined: 2006/11/25 21:23:28
- Location: Thousand Oaks, CA
- Status: offline
Re:Sample rate conversion question.
2012/04/02 12:26:02
(permalink)
Jonbouy
Bapu, no it won't make any difference.
HTH
Got it.
|
bitflipper
01100010 01101001 01110100 01100110 01101100 01101
- Total Posts : 26036
- Joined: 2006/09/17 11:23:23
- Location: Everett, WA USA
- Status: offline
Re:Sample rate conversion question.
2012/04/02 12:50:13
(permalink)
I just noticed I have some lint in my belly button.
 All else is in doubt, so this is the truth I cling to. My Stuff
|
bapu
Max Output Level: 0 dBFS
- Total Posts : 86000
- Joined: 2006/11/25 21:23:28
- Location: Thousand Oaks, CA
- Status: offline
Re:Sample rate conversion question.
2012/04/02 12:51:04
(permalink)
bitflipper
I just noticed I have some lint in my belly button.
24/48K or 24/44.1K?
|
Beagle
Max Output Level: 0 dBFS
- Total Posts : 50621
- Joined: 2006/03/29 11:03:12
- Location: Fort Worth, TX
- Status: offline
Re:Sample rate conversion question.
2012/04/02 12:55:08
(permalink)
bapu
bitflipper
I just noticed I have some lint in my belly button.
24/48K or 24/44.1K?
24 carat. gold (plated)
|
Jonbouy
Max Output Level: 0 dBFS
- Total Posts : 22562
- Joined: 2008/04/14 13:47:39
- Location: England's Sunshine South Coast
- Status: offline
Re:Sample rate conversion question.
2012/04/02 13:22:54
(permalink)
bitflipper
I just noticed I have some lint in my belly button.
Are you sure it's not just the debris from all the farts that have been skinned during the making of this thread?
"We can't do anything to change the world until capitalism crumbles. In the meantime we should all go shopping to console ourselves" - Banksy
|
SCorey
Max Output Level: -80 dBFS
- Total Posts : 538
- Joined: 2011/04/26 15:13:14
- Location: Salt Lake City, UT
- Status: offline
Re:Sample rate conversion question.
2012/04/02 14:23:12
(permalink)
Did it occur to anyone to just test it? I did a quickie test with a 1 minute long series of pulses every second, with L and R offset by .25 sec. The SRC pulse train lined up exactly with the non-SRC train. No drift whatsoever.
|
Jonbouy
Max Output Level: 0 dBFS
- Total Posts : 22562
- Joined: 2008/04/14 13:47:39
- Location: England's Sunshine South Coast
- Status: offline
Re:Sample rate conversion question.
2012/04/02 14:37:48
(permalink)
☄ Helpful
SCorey
Did it occur to anyone to just test it? I did a quickie test with a 1 minute long series of pulses every second, with L and R offset by .25 sec. The SRC pulse train lined up exactly with the non-SRC train. No drift whatsoever.
I've been real world testing it for the last few years... 
"We can't do anything to change the world until capitalism crumbles. In the meantime we should all go shopping to console ourselves" - Banksy
|
Danny Danzi
Moderator
- Total Posts : 5810
- Joined: 2006/10/05 13:42:39
- Location: DanziLand, NJ
- Status: offline
Re:Sample rate conversion question.
2012/04/03 02:22:19
(permalink)
drewfx1
Danny Danzi
There would be a few slight timing issues once they brought my stuff into their projects to where I'd have to show them where they needed to align something.
From a theory standpoint, we generally assume the programmers didn't screw things up. But it's always possible that something like, say, the delay compensation of the SRC routine wasn't done correctly. But it's hard to say what happened without looking at the specific stuff in detail. I can't come up with any reason why a properly implemented SRC would have timing differences even remotely close to being audible.
Couldn't it be possible it was just the two soundcards not quite syncing up, Drew? I never thought for a second I was hearing a sample rate issue, but something was weird. Like the parts didn't hit on the beat like they did when they were in MY version. Not terribly off to where a normal person listening to music could hear them, but definitely enough for the people that frequent this site to hear. It just sounded loose...not horribly out of time. You know when you play a drum kit and sometimes we may not hit a kick drum and crash at the same time and we SHOULD be? It's kinda like off by that much. Like certain vocals I did hit right on the beat on my end yet when the client got them, they didn't hit on the beat on their end and we both used the same tracking files. Whatever it was, it only happened when I used a bigger sample rate than they did and it's happened like 3 times so far with 3 different people. The next time it happens, I'll see if I can grab the parts and post them up. Maybe we can see what the deal is. -Danny
My Site Fractal Audio Endorsed Artist & Beta Tester
|
The Maillard Reaction
Max Output Level: 0 dBFS
- Total Posts : 31918
- Joined: 2004/07/09 20:02:20
- Status: offline
Re:Sample rate conversion question.
2012/04/03 07:36:10
(permalink)
drewfx1
mike_mccue
Here is a website that claims to have collected SRC data from more apps than I can count:
http://src.infinitewave.ca/
I selected the "sweep" analysis and compared R8brain Pro, Minimum Phase to SONAR 8.5.
The results seem to close, but there seems to be enough difference to encourage curiosity.
best regards,
mike
Mike, if you'd been paying attention, you'd have seen that I posted that link a long time ago (post #22).
You can learn a lot from that site, but people should keep in mind that when the differences are far from being audible (note the color scale on the right for the sweep test), "better" is a questionable term.
Having said that, note that:
1. The iZotope 64 filters are almost scary-perfect. You can use those as reference standards.
2. If you compare Sonar's SRC to competing DAW's, you'll get an idea of just how good Sonar's SRC is (though I'm not sure how up to date the versions tested there are).
:-) Drew, I think we have the proof... I wasn't paying attention. :-)
|
js516
Max Output Level: -84 dBFS
- Total Posts : 347
- Joined: 2006/05/17 15:14:53
- Status: offline
Re:Sample rate conversion question.
2012/04/03 15:11:52
(permalink)
Danny Danzi
Couldn't it be possible it was just the two soundcards not quite syncing up, Drew? -Danny
That sounds more likely. I don't think this is an SRC issue. If there was a problem with the SRC the tracks would still be in time with each other. However if you compare the imported tracks to the original files, all the tracks would be off by the same amount since its the same code performing the same calculations with a different data set. If you are importing the project, you may run into problems with any offsets applied to clips that were compensating for the original sound card for the original sample rate. I donot know what Sonar does at the project/clip level, but if there is any compensations put in, then they will be off when you change both the project sample rate and the sound card. If you are importing tracks individually without a broadcast timestamp (where you rely on the first sample of the clips to line things up), you may end up with timing issues because you loose any compensation that was part of the original project. Again, I'm assuming Sonar saves this sort of information.
|