We would like to ask for evaluation of our tool for sound quality assessment and approach for automated sound clip compression optimization. Relevant documents are attached.
Are you doing this with an eye to solving the ACX/AudioBook presenter sound problem? We get many posters that have been rejected from ACX standards and compliance and want to know how to fix their submission.
Thank you for your question. In fact we are trying to solve a general problem of those who have to choose audio compression parameters to achieve the best quality and smallest files size. So, from ACX AudioBook perspective this seems not directly applied as f.e. they demand 192kbps MP3 audio or higher, but our tool will let compress and audio book with bitrate much lower preserving high quality, because the process is fully automated relying on AQuA technology of sound quality analysis.
However, if you could please give some more information about the problem users face with ACX we would do our best to examine the problem in more details and offer a solution.
Our goals may not be converging after all then. The reason ACX requires such high quality MP3 submission is so they can resubmit later at a lower bit rate. If a speaker/presenter delivers at the lowest possible rate, that’s the end of the story. The show can’t be adapted to any other application or product and the work stream drops dead.
Our 10,000 foot goal is a product that users can apply to their show in order to meet ACX compliance.
You will find that there’s significant problem with finding out what ACX compliance is and further, how to measure it, and then, way at the end, how to correct the work.
As if that wasn’t enough, we have the problem of people producing a show that will never make compliance whether through annoying echoes, high noise level or excessive distortion. What do you do then?
Here’s a mere 30 chapter forum discussion of the problems Ian had in getting Submission Ready. It’s not the only one.
Thanks! The problem there is not in the number of requirements ACX presents, but to follow them if one is not recording in a professional studio. Measuring noise level, checking or inserting room tone at the head and tail, checking RMS etc can be more or less easily automated in a stand alone application or a tool that checks recording and points out that it does not fit requirements. However, if we check just this requirement
Each uploaded file must be free of extraneous sounds such as plosives, mic pops, mouse clicks, excessive mouth noise, and outtakes.
… this looks like a disaster to both ACX and the recording party. I wonder how ACX searches audio for these impairments. Considering the amount of sound it is almost impossible to make somebody listen through it. Well, one can search for obvious impairments like clicks, clipping, but they also may be of different nature, with higher or lower level. We have a tool that finds impairments in call recordings such as f.e. noise, clicking, clipping, fuzzy voice, but at the moment it’s working for 8kHz, 16 bit, Mono.
My point is that creating a full scale application is quite a project and providing a simple tool to check basics may not be enough for majority. I will talk to our development and see if we can offer at least some basic tool to the community to ease the life and post here what we can do.
I spoke to our development and management and there is interest to contribute to the community. However, we don’t be able to create full scale utility that will 100% qualify audio to fit ACX requirements as it’s quite an investment project. Would be good to collect information what are the main checks users like to have automated. Our plan is to make a utility that will analyze noise and signal energy according to ACX requirements. But please understand that it’s just a good will and I can’t promise when the utility will be ready etc as commercial projects influence roadmap for noncommercial.
Ok Yes, it’s cross platform, but currently available for Windows, CentOS, Ubuntu. This is due to main customer base. However, we can assemble it for MAC too. See, the technology has proved to work well in telecom, but it can do more, f.e. optimize audio clips compression. This is the reason for my post about evaluation.
OK, thanks, but in that case I don’t see why it is appropriate to ask on this (Audacity support) forum, for our (open source) community members to contribute to your closed source commercial product. On the other hand, we do welcome contributions to Audacity, whether that be in terms of development, testing, documentation, translation, donation, helping out on this forum, or other contributions, as Audacity is a community product, created and supported by, and for the benefit of Audacity users. “Open Source” is about sharing knowledge and not just what you can get for nothing.
Ok, as you could see from the upper posts we wanted to contribute upon feedback from Koz. However, if that’s not the right place to ask for evaluation of our technology I am sorry and welcome to delete this discussion branch. There are things that don’t let everything to be open source… these are IPRs, patents, decades of investments in technology without any contribution or donation, and if looking deeper open source is also just a yet another way of making money.
Although this topic is not strictly within the scope of this forum I don’t think that we need to delete it. it may be of interest to some of our users, and I believe that this conversation is the first contact between your company and ourselves, so I am happy to leave it up as a mark of good will. If you release free or open source software that you feel may be of interest to Audacity users, you are welcome to start a new topic about that.