Amarra best version

0 Members and 1 Guest are viewing this topic. Read 54304 times.

audioengr

Amarra best version
« on: 7 Apr 2012, 06:02 pm »
I used version 1.2 at the RMAF 2009 show with good results.  The versions 2.0-2.2 that were released after that were not satisfying IMO.

However, just before the last 2011 RMAF, Jon issued a version 2.3.2 4319.  Now this was really outstanding.  He told me that he discovered a problem that had been alluding him and it was finally a release that he liked the sound of.  I believe he introduced some kind of SQ issue into the latest release, so I am waiting for a fix to that.

For those that have other versions of full Amarra installed, here is a link that you can download version 2.3.2 4319:

www.sonicstudio.com/releases/Amarra_233_Release_4319.zip

I highly recommend that you install this on a new disk partition with a fresh Mac OS install.  I would avoid using Amarra playlist.  Just use it with iTunes.  Best thing to do is use Cache mode.  I also use some EQ for my speakers.  Make sure that you close-out of Amarra each day in order to avoid memory leaks.

Steve N.
« Last Edit: 20 Nov 2017, 09:15 pm by audioengr »

fridays

Re: Amarra best version
« Reply #1 on: 8 Apr 2012, 11:44 pm »
Steve
Do you suggest never using Amarra's Playlist no matter the release?
Their lit says playlist is suppose to sound the best
I'm a bit lazy guess I could experiment myself
Thanks
Bruce

audioengr

Re: Amarra best version
« Reply #2 on: 9 Apr 2012, 01:38 am »
Steve
Do you suggest never using Amarra's Playlist no matter the release?
Their lit says playlist is suppose to sound the best
I'm a bit lazy guess I could experiment myself
Thanks
Bruce

I could not hear any difference, and playlist is less stable than iTunes.

fridays

Re: Amarra best version
« Reply #3 on: 9 Apr 2012, 02:13 am »
I think I hear a difference using playlist albeit subtle
Thanks for the tip on closing out of Amarra daily
Bruce

REShaman

Re: Amarra best version
« Reply #4 on: 9 Apr 2012, 02:48 am »
I could not hear any difference, and playlist is less stable than iTunes.

Steve, respectfully, may I offer another perspective: As a beta tester for Amarra, while the present state of Amarra's beta is a work in progress (I'll be artfully vague as it is not my place to comment as a beta tester. But as an Amarra Full user, build 4344 has been published for several months now and while not without flaws (often system dependent), consistent with Friday's finding, the SQ Playlist mode is better sounding than Amarra mode and Playlist mode is stable. There have been problems with the "Load" function. But the "Add" function in Playlist mode works consistently. Cache mode as you indicate is preferable along with Playlist mode. I won't pretend to know more about Amarra than Jon. For me, 4317 is a long time ago, and the iterations that have succeeded are much better sounding in my opinion and many others as well. As you rightfully state based on your own interactions with Jon, some SQ issues are present. But 4344 is clearly ahead of 4317. Nothing I have just stated is intended to correct your perception but merely to add mine to the mix. Those who need to find out from their own experience are better off deciding for themselves and not from me.  :)
Best,
Richard

audioengr

Re: Amarra best version
« Reply #5 on: 9 Apr 2012, 04:44 am »
Steve, respectfully, may I offer another perspective: As a beta tester for Amarra, while the present state of Amarra's beta is a work in progress (I'll be artfully vague as it is not my place to comment as a beta tester. But as an Amarra Full user, build 4344 has been published for several months now and while not without flaws (often system dependent), consistent with Friday's finding, the SQ Playlist mode is better sounding than Amarra mode and Playlist mode is stable. There have been problems with the "Load" function. But the "Add" function in Playlist mode works consistently. Cache mode as you indicate is preferable along with Playlist mode. I won't pretend to know more about Amarra than Jon. For me, 4317 is a long time ago, and the iterations that have succeeded are much better sounding in my opinion and many others as well. As you rightfully state based on your own interactions with Jon, some SQ issues are present. But 4344 is clearly ahead of 4317. Nothing I have just stated is intended to correct your perception but merely to add mine to the mix. Those who need to find out from their own experience are better off deciding for themselves and not from me.  :)
Best,
Richard

Rich - I'll give 4344 a try.  This is not the current release is it?

Steve N.

bluesea

  • Jr. Member
  • Posts: 9
Re: Amarra best version
« Reply #6 on: 10 Apr 2012, 07:08 pm »
I have 2.3.3 (4344), which I'm assuming is the current release.

REShaman

Re: Amarra best version
« Reply #7 on: 11 Apr 2012, 02:23 am »
Rich - I'll give 4344 a try.  This is not the current release is it?

Steve N.

Sorry for the delay in replying. At present 4344 is the published build. As you know, Sonic Studio is never idle and they are working on a beta which will succeed 4344. It's best for me to keep my comments circumspect which is appropriate for a beta tester like myself. Hope all is well.
Best,
Richard

askat1988

  • Newbie
  • Posts: 1
Re: Amarra best version
« Reply #8 on: 13 Apr 2012, 01:25 am »
Link seems broken?

audioengr

Re: Amarra best version
« Reply #9 on: 19 Apr 2012, 10:45 pm »
Okay, I installed version 4344 and compared it to 4317 and 4319.

Sorry, not even close.  4344 saps the life out of the music IMO.  4319 is vivid and dynamic, with much better clarity.

I have emailed Jon to see where the archival versions are now.  He evidently moved them.

Steve N.

audioengr

Re: Amarra best version
« Reply #10 on: 20 Apr 2012, 09:03 pm »
4319 is on the Amarra server again, so you can download it:

www.sonicstudio.com/releases/Amarra_233_Release_4319.zip

Make sure you run the uninstall script first if you have another version loaded.

Enjoy :D

Steve N.

claytontstanley

Re: Amarra best version
« Reply #11 on: 21 Apr 2012, 01:14 am »
Steve,

Do you have any theories on why the different Amarra versions have varying SQ? If the USB interface is Async, how can the playback software be introducing artifacts? Floating-point error is the only thing I can think of, and that (to me) seems unlikely to have changed between different builds of the software. Is there a way to test and record the digital USB output that reaches the DAC, or maybe at least record it on the computer (before it's sent to the DAC). I think VMWare or Parallels had some tool to do this (maybe open sourced now), as I remember reading that they needed it when developing their virtualized USB interface.

-Clayton

audioengr

Re: Amarra best version
« Reply #12 on: 21 Apr 2012, 01:41 am »
Steve,

Do you have any theories on why the different Amarra versions have varying SQ? If the USB interface is Async, how can the playback software be introducing artifacts? Floating-point error is the only thing I can think of, and that (to me) seems unlikely to have changed between different builds of the software. Is there a way to test and record the digital USB output that reaches the DAC, or maybe at least record it on the computer (before it's sent to the DAC). I think VMWare or Parallels had some tool to do this (maybe open sourced now), as I remember reading that they needed it when developing their virtualized USB interface.

-Clayton

It's not only Amarra.  Pure Music and iTunes also exhibit differences from one release to the next. iTunes has improved dramatically from 2 years ago.

Jon, the Amarra designer admitted that there was some kind of rounding error or the like that alluded him until version 4317. I can only speculate.  He told me at the RMAF in 2011 that he discovered an error in the code and he was finally happy with the sound quality of 4317 and 4319.

I wish there were some solid metrics that could be used for these comparisons to tell whether they have backtracked or not with a new release.  IME, most of these S/W designers rely on people with really resolving systems to report these SQ anomalies.  Even those feedbacks can be hit-and-miss however.

Maybe you can recommend a measurement?  I dont know what to measure.  These differences are mostly in imaging and dynamics.  Very elusive.

Steve N.

claytontstanley

Re: Amarra best version
« Reply #13 on: 21 Apr 2012, 02:39 am »
Steve,

I'm definitely not an expert on audio drivers, or the USB protocol. But from what I briefly read on the protocol, it uses a technique similar to TCP; data is separated into chunks; each chunk carries its portion in the payload, and there are some header packets attached to each payload that help make sure the data isn't mangled on transport.

So the 'what' to measure would probably be the payload data. Attach 'some' software (if it were a network interface, that sw would be Wireshark or similar) to the USB interface on the computer to watch the packets. Record each payload, aggregate, and store in a file. Then offline, take that file and reproduce the original PCM file. Then do a bit comparison to the original PCM. Do that with the different versions of the build, etc.

Thing is, that sort of thing should be done by Amarra, and automated, and incorporated into the build script; so that the build fails if anything funny is going on with the data during playback, and then that version of the sw isn't yet released.

Also, since you mentioned that the effects are a bit elusive, I'd actually put my money on some sort of timing (not data) interaction between Amarra and the async USB driver. As far as what that interaction is, I'm not sure; as I don't have the low-level driver knowledge required to figure that one out. But I hope the folks at Amarra have that expertise.

It is reassuring to hear that you and the lead Amarra developer are in direct contact.

EDIT: It turns out that Wireshark can also monitor USB interfaces: http://wiki.wireshark.org/CaptureSetup/USB. Might not be OS X compatible though. But it's a start; shows that there's probably some decent software out there to monitor this interface on OS X.
« Last Edit: 21 Apr 2012, 04:08 am by claytontstanley »

audioengr

Re: Amarra best version
« Reply #14 on: 21 Apr 2012, 04:39 am »
Quote
Thing is, that sort of thing should be done by Amarra, and automated, and incorporated into the build script; so that the build fails if anything funny is going on with the data during playback, and then that version of the sw isn't yet released.

I agree, but we are in a small niche market.  Only big companies like Microsoft and Apple can afford to do this kind of testing IME, and even Apple does a poor job of this for iTunes IMO, probably because the data is not critical data like disk data etc..

Steve N.

REShaman

Re: Amarra best version
« Reply #15 on: 6 May 2012, 02:43 pm »
Okay, I installed version 4344 and compared it to 4317 and 4319.

Sorry, not even close.  4344 saps the life out of the music IMO.  4319 is vivid and dynamic, with much better clarity.

I have emailed Jon to see where the archival versions are now.  He evidently moved them.

Steve N.

Given your preference for 4319 which I would never argue with (my policy is not to argue with perception). And, as you know, I have been a beta tester (volunteer) for Sonic Studio. The new build 4374 is exceptional. Give it a test and determine for yourself.
Best,
Richard

fridays

Re: Amarra best version
« Reply #16 on: 6 May 2012, 06:19 pm »
Just installed the new release will listen later today and hear what improvements if any I notice

fridays

Re: Amarra best version
« Reply #17 on: 6 May 2012, 10:33 pm »
Well Splashtop doesn't work with this new version, when i launch it from my iPad Amara stops playing :(

fridays

Re: Amarra best version
« Reply #18 on: 6 May 2012, 10:35 pm »
Also all kinds of dropout that I dint have before mmmm

REShaman

Re: Amarra best version
« Reply #19 on: 7 May 2012, 10:45 am »
Also all kinds of dropout that I dint have before mmmm
Short version: You're experiencing the same as I which was not present in the beta version of the same build. I experienced this last Friday night with guests and I thought something was amiss with my settings as the build had never, never done that. Now with your confirmation of the same experience as I, I apologize, not for my recommendation as the SQ is the best yet despite this new bug, but for not appreciating what I had experienced. I am taken aback by this as it is so obvious!
Please excuse me.
Best,
Richard