Qwerty69
Max Output Level: -62 dBFS
- Total Posts : 1435
- Joined: 2004/02/19 17:44:10
- Status: offline
Audiosnap Bug: Manually moved transients don't bounce correctly
Following on from my other thread, I finally worked out how to preserve the phase relationships between two bass tracks I've been playing with while changing the tempo. Given that success, I thought I'd tidy up the tracks a bit before I bounced and settled on the final tracks to mix. So I've fiddled and faddled, pushed and pulled the odd beat here and there and got everything super tight and sexy. If I look at the track, everything lines up neatly. If I turn Audiosnap off, I can see the track goes back to its former state. So I choose BOUNCE TO CLIP or BOUNCE TO TRACK. ...and I get the original audio - no edits. WTF???!!! Any ideas what's going on? Happening in both the 32 and 64 bit versions of X1c and 8.5.3. Help!! Q.
post edited by Qwerty69 - 2011/10/20 16:46:03
|
Rick O Shay
Max Output Level: -88 dBFS
- Total Posts : 115
- Joined: 2005/01/26 21:35:59
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/08/22 18:05:58
(permalink)
When I finish marking and editing transients, I don't even breathe on it, let alone turn AudioSnap off and then on again.
My workflow is this:
Mark Transients
Tweak transient markers
Audio Snap
Listen - if OK, freeze or mix down to track
During this process I don't dare touch ANY of the settings in the AudioSnap panel. I also make sure not to click on any other tracks. I've found that doing so has either changed some of my snap settings or reset them altogether.
Yeah, yeah, I know. It shouldn't be that way...but it is.
|
Bub 
Max Output Level: -3.5 dBFS
- Total Posts : 7196
- Joined: 2010/10/25 10:22:13
- Location: Sneaking up behind you!
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/08/22 18:54:58
(permalink)
Now this is my kind of thread! LOL!
"I pulled the head off Elvis, filled Fred up to his pelvis, yaba daba do, the King is gone, and so are you."
|
Qwerty69
Max Output Level: -62 dBFS
- Total Posts : 1435
- Joined: 2004/02/19 17:44:10
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/08/23 04:54:20
(permalink)
Rick O Shay
When I finish marking and editing transients, I don't even breathe on it, let alone turn AudioSnap off and then on again.
My workflow is this:
Mark Transients
Tweak transient markers
Audio Snap
Listen - if OK, freeze or mix down to track
During this process I don't dare touch ANY of the settings in the AudioSnap panel. I also make sure not to click on any other tracks. I've found that doing so has either changed some of my snap settings or reset them altogether.
Yeah, yeah, I know. It shouldn't be that way...but it is.
Thanks for the reply Rick. I only did the turn on/turn off thing after I'd been sent to "WTF?!?!" mode to check I wasn't totally insane. I think that an inadvertent click on another track is the cause. No way back that I can see. Bugger this for a game of jelly-beans. I'm going back to old school - manual cuts and cross-fades. Wasted too much time on this bogus "feature" as it is. Q.
|
Qwerty69
Max Output Level: -62 dBFS
- Total Posts : 1435
- Joined: 2004/02/19 17:44:10
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/08/23 05:07:51
(permalink)
Anyone know if this is working properly in Cubase or Studio One or something? Phase accurate with no impact to tone of recorded signal is what I'm after.
Sick of wasting time with this toy.
Q.
|
michaelalala
Max Output Level: -88 dBFS
- Total Posts : 126
- Joined: 2007/07/26 12:11:42
- Location: Unionville, CT USA
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/08/23 12:43:37
(permalink)
Audiosnap is a very frustrating tool. I've spent weeks on it over the past few years and still can't do anything repeatably. I'm sure they're working on 3.0!! It's going to be so cool.
|
Qwerty69
Max Output Level: -62 dBFS
- Total Posts : 1435
- Joined: 2004/02/19 17:44:10
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/08/24 10:01:27
(permalink)
Ha!
I'm now not sure what the hell is going on... I cannot get a quantized audio track of ANY sort to bounce and keep changes.
Luckily I'm paid by Cakewalk to beta test their software - oh, wait...
Q.
|
brundlefly
Max Output Level: 0 dBFS
- Total Posts : 14250
- Joined: 2007/09/14 14:57:59
- Location: Manitou Spgs, Colorado
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/08/24 10:33:24
(permalink)
I'm not sure what going on, either. I can't reproduce a problem either way. AS modifications are preserved after disable/enable on a stretched clip, and stretches/quantizing bounce down as expected
Keep in mind, as mentioned before, that after bouncing, the transient markers will be recalculated - complete with the same detection errors made the first time. So the transient markers might be out of position, but the audio transients themselves should be good, notwithstanding the slight alteration of transient timing in favor of stretching quality when using algorithms other than Percussion, and maybe Groove-Clip - also discussed earlier.
I think the algorithm that can stretch audio completely transparently "with no impact to tone" does not - and likely never will - exist, though there' certainly room for improvement.
SONAR Platinum x64, 2x MOTU 2408/PCIe-424 (24-bit, 48kHz) Win10, I7-6700K @ 4.0GHz, 24GB DDR4, 2TB HDD, 32GB SSD Cache, GeForce GTX 750Ti, 2x 24" 16:10 IPS Monitors
|
Qwerty69
Max Output Level: -62 dBFS
- Total Posts : 1435
- Joined: 2004/02/19 17:44:10
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/08/25 18:37:45
(permalink)
Something definitely weird going on with this...
I exported the AudioSnap disabled version of one of the tracks out of Sonar and then created a new blank file and re-imported the track into it.
Here's where it gets weird -
If I just hit QUANTIZE and don't fiddle with any transient marker, then it will bounce down perfectly and preserve changes. If I move any transient market, then the bounce down changes are skewed all over the place - in some case before the beat, in other cases way after the beat.
About to do a clean install of X1c on another PC and see what results I get there.
Ciao,
Q.
|
Qwerty69
Max Output Level: -62 dBFS
- Total Posts : 1435
- Joined: 2004/02/19 17:44:10
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/08/26 06:18:28
(permalink)
OK -
Replicated the same results on another fresh install of X1c on a different PC.
- Quantize and bounce = OK, (if using Radius Mix algorithm)
- Quantize and bounce = Occassionally weird, (if using Solo Bass algorithm)
- Quantize + manually move a few beats around by dragging transient markers and then bounce = ROOTED
All tests being done at 175bpm with a stereo 32-bit .WAV, (any .WAV - have tried multiple).
Anyone replicate or advise what to look at?
Q.
|
Qwerty69
Max Output Level: -62 dBFS
- Total Posts : 1435
- Joined: 2004/02/19 17:44:10
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/08/26 06:41:11
(permalink)
Qwerty69
OK -
Replicated the same results on another fresh install of X1c on a different PC.
- Quantize and bounce = OK, (if using Radius Mix algorithm)
- Quantize and bounce = Occassionally weird, (if using Solo Bass algorithm)
- Quantize + manually move a few beats around by dragging transient markers and then bounce = ROOTED
All tests being done at 175bpm with a stereo 32-bit .WAV, (any .WAV - have tried multiple).
Anyone replicate or advise what to look at?
Q.
More detail - Here's what I get if I quantize, manually move some beats and then bounce using the Solo Bass algorithm (original marked up track with AudioSnap transients + bounced track) -  Here's what I get using the Radius Mix algorithm -  Now that looks ok doesn't it, but if you zoom into the beat on 267:01:000 you can see the transient is getting squished and it is still out of time -  Now replicated on three PC's - main DAW plus two laptops with fresh installs of X1 + X1c patch, (although 8.5.3 is doing the same thing on my main DAW). I've tried high latency, low latency settings for the ASIO driver, small and large playback and record buffer sizes, Trigger and Freewheel vs. Full Chase Lock settings and mucked around with the BounceBufSizeMsec= setting in the .INI file. No dice. Same situation all round. Clues, theories, idle speculation, jokes, class actions all welcomed. Thanks, Q.
post edited by Qwerty69 - 2011/08/29 17:26:09
|
Qwerty69
Max Output Level: -62 dBFS
- Total Posts : 1435
- Joined: 2004/02/19 17:44:10
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/08/26 06:53:04
(permalink)
 And for the sake of completeness - Radius Mix Advanced - same story as the normal Mix algorithm.
post edited by Qwerty69 - 2011/08/26 06:54:16
|
Dave Modisette
Max Output Level: 0 dBFS
- Total Posts : 11050
- Joined: 2003/11/13 22:12:55
- Location: Brandon, Florida
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/08/26 07:54:51
(permalink)
I think you should try using the SPLIT at TRANSIENTS approach. I think that asking an alogorithm to stretch and crossfade bits of audio will never work to a satisfactory level of someone with a critical ear. First of all there isn't enough fine tuning controls for the transient detection. Reaper's Dynamic Split tool has twice the control that Audio Snap has and you still have to really work a track to get it to detect a transient on audio information that doesn't have sharp spikes on the attack.
I find it's better to split the tracks and then manually use slip editing to crossfade gaps and overlapping. In this regard, Reaper's tools are ahead of SONAR's once again and I get a more satisfactory result quicker because of the ability to define how it approaches gapping and overlapping. IOW, you can set crossfade lengths as well as how much to automatically slip edit a clip when a gap occurs.
|
Qwerty69
Max Output Level: -62 dBFS
- Total Posts : 1435
- Joined: 2004/02/19 17:44:10
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/08/26 09:41:41
(permalink)
I hear ya Dave. Normally when I do this, I do the manual split, cross-fade thing, check for clicks and pops and move on to the next wee edit point. Only reason I started down this path was due to the tempo change I was looking to accomplish.
I did try and use AudioSnap to automate the "split at transients approach" but wasn't too succesful at that either - cross-fades being ignored as per my other thread on the subject...
I do agree that I'm probably asking too much of this functionality, but a) it is an advertised feature presented as though it works, (for the last 5 years), and b) the un-rendered track state is correct!! All I want is that to be correctly bounced out to a track or clip!
To be honest, I did go down the Reaper path and while it handles this with aplomb -- far simpler and easier than X1, the resulting tonal change to my ears at least is too much of a compromise. It's another case of so close, but so damn far that is frustrating me to distraction. That and the fact that I've tweaked these damn transient markers so much, I can't bear to turn around now and re-do it all manually!
C'est la vie!
:) Q.
|
Qwerty69
Max Output Level: -62 dBFS
- Total Posts : 1435
- Joined: 2004/02/19 17:44:10
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/08/27 03:40:07
(permalink)
Just fiddled about with Ableton Live 8. Seems to work properly and is reasonably tone neutral.
Q.
|
Qwerty69
Max Output Level: -62 dBFS
- Total Posts : 1435
- Joined: 2004/02/19 17:44:10
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/08/29 05:23:42
(permalink)
Qwerty69
And for the sake of completeness - Radius Mix Advanced - same story as the normal Mix algorithm.
...actually, not exactly the same story. Setting the PHASE COHERENCE to 100% retains the transients as per the original however their placement is still random - some correct others incorrect. Who knows??????????? Q.
|
brundlefly
Max Output Level: 0 dBFS
- Total Posts : 14250
- Joined: 2007/09/14 14:57:59
- Location: Manitou Spgs, Colorado
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/08/29 13:06:42
(permalink)
Referring to post #10, the captions for both of the first two screesnhots say "using Radius Mix". Was the second one actually Groove Clip (using Same as Online) or Percussion or...? Incidentally, I've noticed that when you set Offline to Same as Online, you have to re-select the online algotihm to get the Offline filed to update. I'm not sure whether that's just a display bug, or whether it will actually affect which algorithm gets used; I'll have to check.
The transients in the bounced clip in the first screenshot all look to be late by about the same amount, and by a lot more than could normally be attributed to algorithmic wonkiness (to use the technical term). How does that compare to the original un-snapped clip? It almost looks like the changes just fail to bounce for some reason.
In the zoomed screenshot, the bounced signal seems to have a really pronounced high-freq. tremolo/warble riding on the original signal. Does it sound as bad as it looks? I've heard stretching results that sound like that waveform looks, so I'm not too surprised - just curious.
I'd be interested to see what the Ableton-corrected track looks like next to the original when imported into SONAR and zoomed to that same level.
SONAR Platinum x64, 2x MOTU 2408/PCIe-424 (24-bit, 48kHz) Win10, I7-6700K @ 4.0GHz, 24GB DDR4, 2TB HDD, 32GB SSD Cache, GeForce GTX 750Ti, 2x 24" 16:10 IPS Monitors
|
brundlefly
Max Output Level: 0 dBFS
- Total Posts : 14250
- Joined: 2007/09/14 14:57:59
- Location: Manitou Spgs, Colorado
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/08/29 13:59:07
(permalink)
I just did a couple quick tests with a 125BPM click track quantized to a 120BPM project (so some transients get moved one way, and some the other, and some can't both be quantized because they would collide and exceed the 25% "compression" limit).
The bounce with Groove Clip algorithm nulls to silence when phase-inverted against with the original clip with online Groove Clip stretching.
The bounce with Radius Mix at 100% phase coherence bounced with most transients located about 11 samples late with a variability of maybe ±2 samples, though a couple transients were right on the money. Basically the same as you described in post #16. I don't find it too surprising that the non-percussion-oriented algorithms would sacrifice sample-accurate phase coherence in favor of preserving overall audio quality in more complex material.
I'll try again with some different rendered synth tracks using more sustained sounds and see what happens, but if you can share the audio file you're working with, I'd be interested to see if my environment produces the same errors.
SONAR Platinum x64, 2x MOTU 2408/PCIe-424 (24-bit, 48kHz) Win10, I7-6700K @ 4.0GHz, 24GB DDR4, 2TB HDD, 32GB SSD Cache, GeForce GTX 750Ti, 2x 24" 16:10 IPS Monitors
|
Qwerty69
Max Output Level: -62 dBFS
- Total Posts : 1435
- Joined: 2004/02/19 17:44:10
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/08/29 17:36:33
(permalink)
Hi - Firstly, thanks for the response. To address some of your points - - The first image was incorrectly labelled. It was the Solo Bass algorithm. I have updated the post.
- In each image, the track at the top of screen is the marked up file being snapped.
- The track at the bottom of the image is the result of the BOUNCE TO TRACK command.
- Unfortunately I was using the demo version of Ableton Live so can't save anything. I was making the changes in that program and then looping the audio through RME's TotalMix to record the result. As such, I don't have a before vs. after.
- I have not done any testing on SAME AS ONLINE or the PERCUSSION algorithms as I am using a bass guitar's track.
With the RADIUS MIX ADVANCED algorithm set to 100% phase coherence, the transients do not get smashed but there is some relatively noticeable modulation or audio warbling. What is strange is that I see the beginning of the file render out correctly - alignment is perfect - and then a few bars into the track the delay issue pops up suddenly. I think this may have something to do with disabled transient markers or a combination of that plus changes in the direction of moved transients, ie. when I've been pulling to the left and then pull one to the right. I can't reproduce a recipe for that 100% of the time though, so am a little unsure. Regards, Q.
|
brundlefly
Max Output Level: 0 dBFS
- Total Posts : 14250
- Joined: 2007/09/14 14:57:59
- Location: Manitou Spgs, Colorado
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/08/29 17:47:44
(permalink)
In each image, the track at the top of screen is the marked up file being snapped. Just to clarify: When I asked how the bounced clip compares to the "un-snapped" clip, I meant the original raw clip before it was "marked up" (i.e. transients quantized and/or dragged). You mentioned things getting wonky later in the clip. I must confess I have a bad habit of testing this stuff using short clips to save time. I'll be sure to use song-length clips in the future.
SONAR Platinum x64, 2x MOTU 2408/PCIe-424 (24-bit, 48kHz) Win10, I7-6700K @ 4.0GHz, 24GB DDR4, 2TB HDD, 32GB SSD Cache, GeForce GTX 750Ti, 2x 24" 16:10 IPS Monitors
|
Anderton
Max Output Level: 0 dBFS
- Total Posts : 14070
- Joined: 2003/11/06 14:02:03
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/08/30 01:21:18
(permalink)
[So I keep trying to edit this to put in paragraphs, yet the text keeps running into one long line. What am I doing wrong?!?] I needed to snap electronic kick drums to a real drummer, so of course, that meant either AudioSnap or cutting each electronic kick into its own event and lining it up visually... I dreaded using AS because I hadn't had much success in the past. But this time, things worked well. However, I did find out a few things in the process of doing this. First, AudioSnap is an "automated" process only for the simplest material. When I converted a human-played drum clip to create a Pool, zoomed out it looked okay, but zoomed in revealed that quite a few transients were off, and there were sort of "ghost" transients. I spent as much time getting the transient markers correct as I would have if I'd been acidizing it. Once that was squared away, getting the kicks to match up was another issue. Most of the time, it worked great. But, quite a few of the kicks locked to the "ghost" transient I thought I had deleted. That's when I discovered the right-click/snap to nearest transient command - a real life-saver. Doing that once or twice would usually lock it to the right transient. When the kicks lined up, I bounced and they lined up as they were supposed to. I'll have to take more careful note of what I did, maybe that would help. The bottom line for me was this. First, quantizing the electronic kick to the human part sounded FAR better than quantizing the human kick to the electronic, metronomic kicks (not surprising). Second, there was a lot of manual labor involved in making a grid based on the humanized playing. However, once that was out of the way, things worked smoothly. We'll see what happens with more complex material...I'm not a big fan of audio snapping anyway, so I don't know how much more I'll be pushing it.
post edited by Anderton - 2011/08/30 01:24:48
|
audiyo
Max Output Level: -87 dBFS
- Total Posts : 159
- Joined: 2011/08/05 19:51:34
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/08/30 01:31:04
(permalink)
Anderton
[So I keep trying to edit this to put in paragraphs, yet the text keeps running into one long line. What am I doing wrong?!?] I needed to snap electronic kick drums to a real drummer, so of course, that meant either AudioSnap or cutting each electronic kick into its own event and lining it up visually... I dreaded using AS because I hadn't had much success in the past. But this time, things worked well. However, I did find out a few things in the process of doing this. First, AudioSnap is an "automated" process only for the simplest material. When I converted a human-played drum clip to create a Pool, zoomed out it looked okay, but zoomed in revealed that quite a few transients were off, and there were sort of "ghost" transients. I spent as much time getting the transient markers correct as I would have if I'd been acidizing it. Once that was squared away, getting the kicks to match up was another issue. Most of the time, it worked great. But, quite a few of the kicks locked to the "ghost" transient I thought I had deleted. That's when I discovered the right-click/snap to nearest transient command - a real life-saver. Doing that once or twice would usually lock it to the right transient. When the kicks lined up, I bounced and they lined up as they were supposed to. I'll have to take more careful note of what I did, maybe that would help. The bottom line for me was this. First, quantizing the electronic kick to the human part sounded FAR better than quantizing the human kick to the electronic, metronomic kicks (not surprising). Second, there was a lot of manual labor involved in making a grid based on the humanized playing. However, once that was out of the way, things worked smoothly. We'll see what happens with more complex material...I'm not a big fan of audio snapping anyway, so I don't know how much more I'll be pushing it.
You're a victim of the dreaded Firefox 6 issue that many are having with this forum. Chrome seems to be ok.
Sonar X2a Producer (64 bit), Win 8 Pro (64 bit), Intel Core i7 920@2.67Ghz, Asus P6T, 6GB DDR3 Ram, Komplete Audio 6, Fender BT Jazzmaster, AKAI MPK Mini
|
Qwerty69
Max Output Level: -62 dBFS
- Total Posts : 1435
- Joined: 2004/02/19 17:44:10
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/08/30 05:51:59
(permalink)
brundlefly
In each image, the track at the top of screen is the marked up file being snapped.
Just to clarify: When I asked how the bounced clip compares to the "un-snapped" clip, I meant the original raw clip before it was "marked up" (i.e. transients quantized and/or dragged).
Gotcha, (now...). It looks as though there isn't a direct correlation. Sometimes the bounced file's transients line up with the original, un-snapped recording and other times it is just something else entirely - not the original track, not the snapped version. Q.
|
Qwerty69
Max Output Level: -62 dBFS
- Total Posts : 1435
- Joined: 2004/02/19 17:44:10
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/08/30 05:56:52
(permalink)
Anderton
[So I keep trying to edit this to put in paragraphs, yet the text keeps running into one long line. What am I doing wrong?!?] I needed to snap electronic kick drums to a real drummer, so of course, that meant either AudioSnap or cutting each electronic kick into its own event and lining it up visually... I dreaded using AS because I hadn't had much success in the past. But this time, things worked well. However, I did find out a few things in the process of doing this. First, AudioSnap is an "automated" process only for the simplest material. When I converted a human-played drum clip to create a Pool, zoomed out it looked okay, but zoomed in revealed that quite a few transients were off, and there were sort of "ghost" transients. I spent as much time getting the transient markers correct as I would have if I'd been acidizing it. Once that was squared away, getting the kicks to match up was another issue. Most of the time, it worked great. But, quite a few of the kicks locked to the "ghost" transient I thought I had deleted. That's when I discovered the right-click/snap to nearest transient command - a real life-saver. Doing that once or twice would usually lock it to the right transient. When the kicks lined up, I bounced and they lined up as they were supposed to. I'll have to take more careful note of what I did, maybe that would help. The bottom line for me was this. First, quantizing the electronic kick to the human part sounded FAR better than quantizing the human kick to the electronic, metronomic kicks (not surprising). Second, there was a lot of manual labor involved in making a grid based on the humanized playing. However, once that was out of the way, things worked smoothly. We'll see what happens with more complex material...I'm not a big fan of audio snapping anyway, so I don't know how much more I'll be pushing it.
Interesting experiences Craig... I did start down this path trying to extract a groove from a human kick/snare dump track and use that to match to this bass track. As I failed miserably at that process, I simplified and have been trying the process documented in this thread using just the metronome. Either way, there's too many eighth notes at 200bpm in this little beastie for me to play 100% so it was either this process or the old cut-drag-fade routine... Thanks for the snap to transient tip though! Q.
|
Qwerty69
Max Output Level: -62 dBFS
- Total Posts : 1435
- Joined: 2004/02/19 17:44:10
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/10/20 16:44:24
(permalink)
OK.
Just had this confirmed as a bug by Cakewalk which has been submitted to development - CWBRN-5855
Hate to think how many recordings have been smeared since AudioSnap was first released... Glad to know I wasn't going insane or doing something wrong...
Q.
|
brundlefly
Max Output Level: 0 dBFS
- Total Posts : 14250
- Joined: 2007/09/14 14:57:59
- Location: Manitou Spgs, Colorado
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/10/20 18:00:33
(permalink)
You wrote in an earlier post "I can't reproduce a recipe for that 100% of the time though, so am a little unsure." What recipe did you submit to Cakewalk? Was it based on a submitted project file? Just curious, since I wasn't able to reproduce this problem, though there are plenty of other AS issues to go around. 
SONAR Platinum x64, 2x MOTU 2408/PCIe-424 (24-bit, 48kHz) Win10, I7-6700K @ 4.0GHz, 24GB DDR4, 2TB HDD, 32GB SSD Cache, GeForce GTX 750Ti, 2x 24" 16:10 IPS Monitors
|
Qwerty69
Max Output Level: -62 dBFS
- Total Posts : 1435
- Joined: 2004/02/19 17:44:10
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/10/20 21:50:40
(permalink)
The recipe above gives the problem 100% of the time. The bit where I said "I can't reproduce..." meant that I couldn't, (and still can't), predict exactly where or when the transient misalignment will come into the track, however the problem still comes.
The test conditions above were with 4 minute audio tracks.
Q.
|
pianodano
Max Output Level: -67 dBFS
- Total Posts : 1160
- Joined: 2004/01/11 18:54:38
- Location: Va Beach Virginia
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/10/20 22:34:12
(permalink)
Fwiw, I know what the OP states to be a fact.
I spent 4 days last week working on snapping drum tracks. What a crock. I could not under any conditions get the last track to maintain it's snapped settings.
Also, I have never ever been able to freeze tracks and get them to reliably play what the transients show. I have found that the only acceptable (for me) method is to ouput the tracks to the console and buss them back into Sonar to record new tracks.
Even then, I sometimes have to cut and slide things to line up hits here and there.
Audio snap = audio crap
Best, Danny Core I7, win XP pro, 3 gig ram, 3 drives- Lynx Aurora firewire- Roll around 27 inch monitor, 42 inch console monitor- Motif xs controller - Networked P4's and FX Teleport for samples- Muse Receptor VIA Uniwire for samples and plugs- UAD QUAD Neve - UAD 1- Sonar X1 but favor 8.5 GUI - Toft ATB 32 - Vintage hardware - Tascam MS-16 synched via Timeline Microlynx -Toft ATB32 console
|
brundlefly
Max Output Level: 0 dBFS
- Total Posts : 14250
- Joined: 2007/09/14 14:57:59
- Location: Manitou Spgs, Colorado
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/10/21 02:06:46
(permalink)
Also, I have never ever been able to freeze tracks and get them to reliably play what the transients show. I have found that the only acceptable (for me) method is to ouput the tracks to the console and buss them back into Sonar to record new tracks. There are only two online algorithms, Groove Clip and Percussion. If you set the offline algorithm to be Same as Online, you'll get a bounce that is identical to the live playback (see my post #18). Using any other algorithm will necessarily result in some changes. Transient location accuracy generally takes a back seat to overall audio quality preservation with the offline algorithms, but the timing changes shouldn't be that significant. Keep in mind that the transients get re-detected after bouncing, so some if the discrepancies you see are likely just mis-located markers. I'm certainly not saying Audiosnap doesn't exhibit some anomalies. I've reported my fair share, and they've all been confirmed. But to say it's all "crap", and "never" bounces down correctly is not accurate.
SONAR Platinum x64, 2x MOTU 2408/PCIe-424 (24-bit, 48kHz) Win10, I7-6700K @ 4.0GHz, 24GB DDR4, 2TB HDD, 32GB SSD Cache, GeForce GTX 750Ti, 2x 24" 16:10 IPS Monitors
|
Qwerty69
Max Output Level: -62 dBFS
- Total Posts : 1435
- Joined: 2004/02/19 17:44:10
- Status: offline
Re:Audiosnap: Urge to Kill... RISING!!
2011/10/21 09:07:31
(permalink)
brundlefly
But to say it's all "crap", and "never" bounces down correctly is not accurate.
Umm... No. That's exactly what the confirmed bug report says. Smeared transients, regardless of how good the tone of the algorithm, are still inaccurate - AND reproducible according to the bakers... Marketing a feature for six or so years that doesn't actually perform as advertised? I'd say that also qualifies as pretty "crap" - or at least deceptive advertising. Q.
|