I’ve seen a few “similar” posts but had come to the conclusion that bugs had been fixed…
Today I opened a film post-production mix that I’d finished on 23rd Dec.
The director has sent a list of final tweaks so I re-opened said session.
I’m using SpectralLayers via ARA in Pro Tools and even though my session loads all tracks that were previously manipulated (into Layers), any processing appears to have vanished.
I know that ARA data gets stored within the PT Session folder.
But under preferences I noted a “Cache” path /Users/blakeyman/Library/Caches/Steinberg/SpectraLayers 11.0
And also a “Max: 5 days”
Is this the reason that “data” is getting lost when I re-open?
Hi, I don’t know why you had this issue, but the Cache path (and its time limit) is unrelated, the Cache path is only used to store data while you’re working on a session, and all data stored here are deleted once you close the session.
@Robin_Lobel Thanks for clarifying how the cache operates.
It would appear that Pro Tools (latest version) and the ARA integration with SpectralLayers Pro (latest version) is still buggy…
Render on the go!
I appreciate the replies. I’ll definitely be rendering on the go from now on. For what it’s worth, Melodyne ARA saves/recalls in Pro Tools 100% of the time for me. Still not sure whether that means it’s explicitly Spectralayers fault, but just more info to add.
There is a known issue with ARA and macOS, without system details it’s hard to suggest possible work arounds/etc. Look over on Logic Pro Help site…HTH…/s~
As written above I am on Pro tools 24.10.1.204, Spectralayers 11.0.50.402, macOS Ventura 13.7. Also a Macbook Pro M1Pro. Happy to provide any other info that might be helpful. Not too sure the Logic Pro site will be of much use, but I’ll take a look. thanks!
Got it…I suggested the LPH site because there are a lot of discussions about ARA/DAW issues, and the underlying OS, which is also going to affect your situation. I use a number of different DAWs depending the client, and ARA is “no bueno” for all of them. There have been a lot of “upgrades” to the OS that seemed affect ARA on Mac…/s~
Well they are the ones that created ARA, so to me it’s no surprise their software works well and the rest no so much.
But of course I can be wrong and the issue is more on SL side. And AAP2 side too. And Acon Acoustica side as well.
We as customers should never accept this as “is what it is.” The total price you pay for the product includes the development time spent on the features advertised. If those features do not function, you are NOT getting what you pay for. Therefore we as as customers should define this as “it is what SHOULD WORK.”
I would love to use SL more often, but because the VST-interface is very bad implemented, SL has lost its value for me. In fact, as it stands today, VST with SL is practically unusable for me. This is not acceptable.
Another interesting quirk of SL:
When creating a repeating sound pattern with copy and paste, the program becomes awefully slow after some inserts.
For instance: I make a sweep, that is 5 seconds long and need to repeat this for lets say an hour.
Copy and paste exponentially. 1,2,4,8,…etc to get the full hour .
After a while, the insert becomes at least quadratically slow. (O(n^2))
This suggests suboptimal datahandling/datastructures.
This is simply not acceptable.
Another thing is normalization, which is terrible slow when compared to Wavelab or Audacity.
The selection tools “Lasso” and “Magic Wand” causing crashes after a while. Memory leaks, maybe? (The kind of magic wand selection has always been tricky. I remember this at Adobe Photoshop or other paint programs, who offered this selection method for their first time - they crashed so often. Also, in analogy, the recursive filling algorithm used up a lot of memory decades ago. Maybe something similar happens in SL today…)
Allow me to digress.
I have been a SL customer since first versions and when every new one comes out, brings surprising (sometimes amazing) features.
Considering that these functions have, many of them, been pioneering if not radically new in the field… and this for a field that has been created along, then one would expect some inconveniences once and then.
Of course, once a new version is released it is expected to work on all kinds of differing PC systems and varied configurations, so bugs are expected. More so at a so complex tool as SL has become.
Personally, I vote for continuing the impressive cadence of improvement and innovative development. In all fairness it is accompanied with the main developer involvement and consequent post release responsiveness, accommodating many interesting requests and suggestions from this thriving users’ community. Let alone the vital in between versions frequent upgrades, workflow improvements and coming new features!
In this perspective, and in view of all the amazing tasks many at this forum and I, are able to daily achieve using SL at very real artistic and professional fields, then calling SL a beta release is a gross misnomer. Something that you are entitled to think, respectfully say and accordingly act upon.
Sure, SL is an ingenious concept made in software. No question about that.
SL is one of those rare pieces of software, that are kind of groundbreaking. Decades ago, I had similar ideas of spectral editing, talkking with friends in a pub while being students of math, physics and computer science. But we never got into it. And than came SL -just wow!
The point is, that SL lacks solid quality on the codebases. Even though we do not know anything about the underlying code in an exact way, one with some knowledge of programming and some expertise in FFT can see, where some coding problems might be. This is not meant to be negative or blaming, it’s just a fact. SL has some obvious bugs and quirks, which may be caused by old and not optimal code, that has not been updated or quality checked by peer review, for instance. The internal datahandling of SL might not be state of the art. And when it comes to the mentioned selection tasks,… year, this should be no issue at all, but it is.
The VST-interface is lousy, looks rather like a quick and dirty implementaion to proof, that it is possible. For me, the actual state of VST as today, is practically unusable. Which is sad, because SL has so much potential on this side.
