Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Advanced ABX (Read 18721 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Advanced ABX

Reply #25
I just returned from AES London where most of the focus seemed to be on issues related to of hearing/testing high resolution audio (e.g. 24/96) -- some panelists (e.g. Peter Craven) expressed the idea that current scientific testing (e.g. ABX) methods are too flawed to measure the benefits of 24/96. Here we go again

See:  - How Do We Evaluate High Resolution Formats for Digital Audio?

Cheers
Sean Olive
Audio Musings



Ha, I see Milind Kunchur was on that one....too bad JJ wasn't there too.  ;>

And Wieslaw Woszczyk is John Atkinson's go-to academic when JA wants to push the high-rez audibility meme.

Must've been an interesting session.


Advanced ABX

Reply #26
This is my perspective: First, in my free time I develop stuff for my own interest. And usually I just develop as far as I need it. The only aspect, where "other people" come in at this stage, is virtual, respectively manifested in how I modularize my code.

After that I might share my work with a possibly interested audience. Then one of two things can happen. Either I get the possibility to connect with (an)other developer(s) with similar interest and we start collaborating. Or it does not happen, which is fine. And then I move on.


I developed a mIRC script back in 1998-2000. I started alone, and had other people interested and collaborated on making it better (we developed mirc windows with controls before mirc allowed them natively!). Like everything, when interest was lost (or other better scripts appeared), it faded away.

During 7 years, I've also been the main developer under an open source music composition program. I've worked with many people that liked it, improved on it and sometimes they did even more than what I could even imagine to do.  Nowadays, my interests have shifted again.

I've also made the java version of lossywav. Nothing special and always one foot behind the actual lossywav, but hey, I found it interesting at first.

So if you like to code in your free time, and then share it, at least do not expect people to show admiration for no special reason. Like you say, I may not see the benefits of it, but that's the way I see it.

Currently, my next interest is comparing high resolution tracks* with a native Mac ABX tool, basically a port of AA to Objective-C. I have gotten less and less satisfied with Java's abilities to handle anything else but 44.1kHz/16 bit on different platforms, and even the latter may suck depending on your setup.


What a shift of interests... to develop in a platform that is considered "make once, run everywhere", to a "build for Apple, run on Apple" one..  Of course, if you are an Apple guy, let be it, but be prepared to face it:  Less people knowing it, and much less people interested in doing it for free.


Advanced ABX

Reply #27
googlebot,

First I would like to thank you for your work.

You seem to pretty hastily think that no one is interested about the inventions you have made. You released this project only a month ago. Surely a developer may be interested to use your source code when the time is right. That may happen tomorrow, next month, next year... (of course "never" is also possible). I am not a developer, but I had a very busy and exhausting last month and I didn't have any time/inspiration/strength for participating the HA activities.

Personally I don't find the demo application usable for testing lossy encodings because it hasn't a possibility to specify the start and end points. I always isolate one potentional problem passage at a time and ABX it separately. It can be as short as a single cymbal crash or something like that. If the start or end position produces a pop that bothers me I can usually fix that by moving the start or end marker very slightly. The current implementation is more like a switchbox for a speaker test, i.e. you let a track play continuously and switch between the "speaker pairs".

In addition, it doesn't work if the tested audio samples don't have identical durations. Lossy decoders don't always decode the original sample amount. The feature in ABC/HR for Java that can remove silent samples from the beginning works fine for practical purposes. It doesn't matter if one of the tested sample files have a few additional samples in the end.

It would be great if someone would combine your work and the now quite abandoned ABC/HR for Java and keep the development active.

Advanced ABX

Reply #28
if this thread is going to spread in general ABX thinking, I'll throw that that there is space for mobile (java based) ABX software that can convince me that nero hev2 is different than lossless on my phone while I ride my bicycle

Advanced ABX

Reply #29
I didn't really expect this. Maybe software based ABXing in the field of audio isn't just that important anymore. Or it is just more talked about than actually conducted. Or maybe correlated switching artifacts are just not considered that serious, so that simpler approaches embedded into applications with integrated converters as Foobar are just more convenient for most users.


Thanks for the reply. I understand your reasoning for not continuing development, if there is little interest. Perhaps you are right, that the demand to ABX lossy audio CODECS is diminishing as more people are using lossless audio.

I just returned from AES London where most of the focus seemed to be on issues related to of hearing/testing high resolution audio (e.g. 24/96) -- some panelists (e.g. Peter Craven) expressed the idea that current scientific testing (e.g. ABX) methods are too flawed to measure the benefits of 24/96. Here we go again

See:  - How Do We Evaluate High Resolution Formats for Digital Audio?


Apparently there was a presentation from someone from McGill University that claims reliable detection of 24/192 versus 16/44 using her own orchestral recordings. Unfortunately, there appear to have been a number of differences in the respective signal paths.

Advanced ABX

Reply #30
I just returned from AES London where most of the focus seemed to be on issues related to of hearing/testing high resolution audio (e.g. 24/96) -- some panelists (e.g. Peter Craven) expressed the idea that current scientific testing (e.g. ABX) methods are too flawed to measure the benefits of 24/96. Here we go again

See:  - How Do We Evaluate High Resolution Formats for Digital Audio?

Cheers
Sean Olive
Audio Musings



Ha, I see Milind Kunchur was on that one....too bad JJ wasn't there too.  ;>


While he is not as well known, one of the ABX co-developers *was* present for much of these shennanigans, accoding to what he told me last Satruday.

Quote
And Wieslaw Woszczyk is John Atkinson's go-to academic when JA wants to push the high-rez audibility meme.


Check out this:

A bit of WW's CV

Quote
Must've been an interesting session.


We've been here before, and not just a few times. It is just "all audio compnents have a characteristic sound" all over again. Atkinson is still turning that misapprehension into gold.

Advanced ABX

Reply #31
Hi googlebot,

just read this thread for the first time. Was away from HA for the last few weeks due to the same reasons as Alex.

Thanks very much for your work! I work in Fraunhofer's audio coding department, and I can tell you that artifact-free switching in a blind test software makes testing much easier. However:

Quote
AA is basically just a tool with feature completeness already in sight. One may add a couple of automatic decoding filters, maybe automatic track alignment (encoding delay correction) a more complete UI and that's about it.

Sorry, but that's incorrect. Think of all the things that could go wrong. For example, what if the compared WAV files have differing
  • lengths, as mentioned? Solution: take length of shortest file
  • offsets? Solution: automatic sample-accurate delay alignment, see Synchrotron
  • sampling rates? Solution: error message, or automatic resampling of lower-sampling-rate files to highest sampling rate in the test (in combination with previous points, quite some work!)
  • Bit depths? Solution: increase word length of lower-rate files to highest word length in the test
  • Number of channels? Similar solution as above


These and Alex's feature description regarding segmenting and looping (which I strongly support) are only some of the features that would attract interest away from ABC-HR and the foobar plug-in to your software. Here is another thing which hardly anyone talks about on this forum but which I'd like you to consider.

The scientific community has been using the MUSHRA test methodology in the evaluation of audio codecs for years. It's also used for the majority of internal tests at Fraunhofer. I disagree that it's only useful for "intermediate audio quality". I successfully used it for AAC tests between 160 and 256 kbps stereo (for example, with less than 10 listeners, it revealed that a particular codec was not transparent at 256 kbps). Take a look at the screenshot in this thread on the RateIt tool written by Jean-Marc Valin. Fraunhofer uses a self-made MUSHRA tool for testing (no, sorry, can't share it), which looks very similar to RateIt but has the additional feature of looped playback, with hotkeys for setting loop start and end points for faster changing of loops during playback. Send me a PM if you'd like to know further details. Why don't you consider turning your AA tool into a MUSHRA tool? Artifact-free switching is just as important for MUSHRA as it is for ABX!

I'm not surprised that nobody contributed any code to your project so far. It's much less work asking for improvements than implementing them  Same with me: I program for a living during day time, I don't want to continue doing it at night. But if you can use some software engineering support (requirements, planning, GUI design mock-ups, ...), I'd be willing to contribute. Last note: notice how nobody replied to Jean-Marc's RateIt thread! You're not the only one with a good idea but little (initial) support

Best,

Chris
If I don't reply to your reply, it means I agree with you.

Advanced ABX

Reply #32
@googlebot

It is going to be very hard to get people to commit to a project that you aren't even committed to! You don't get an instant reaction that you like and you are ready to quit. Even worthwhile projects take time to develop and gain popularity.
Glass half full!

Advanced ABX

Reply #33
@bilbo

I'm not interested in "How to bootstrap an open source community project for dummies" kind of advice. If that had been my goal, I would be sitting here churning out feature request implementations, so that everybody would feel involved and love me. Eventually other altruists would chime in and also show off what they can do.

Sorry, I'm not that type and that hasn't been my goal from the beginning. Many much more rewarding things for my free time come to my mind instantly.

I can only repeat: The main motivation for AA was personal, technical interest (and not becoming an admired project maintainer). I shared it on Hydrogenaudio with the anticipation of professional exchange and collaboration. I don't think I owe anyone any more commitment than what I have already delivered to justify that anticipation posteriorly. And it is really no big deal, when no one finds the time to contribute anything right now. I guess I would have done it anyway.

@all

Admittedly I expected more initial interest for a topic so close to HA's Magna Carta. But I have got the impression now, that I may really just have hit the wrong time. Thank you for the recent feedback. I agree with C.R. Helmrich, that there are plenty more potential features left, that would be worth adding. So maybe AA isn't really 95% done but 50%. Any commits are still welcome! I'm willing to help. The next thing I'm going to focus on is marking/looping. But I can't promise when I will find the time.

@C.R.Helmrich

While MUSHRA is certainly interesting, I'm not planning to implement it right now. For anyone interested in the artifact-free-switching part of AA, it is cleanly encapsulated within the SwichtableArrayClip class (in another thread I have also posted a solution for streams). It might be pretty easy to integrate into your own applications and is both suited for pushing and pulling audio pipelines.

Advanced ABX

Reply #34
Currently, my next interest is comparing high resolution tracks* with a native Mac ABX tool, basically a port of AA to Objective-C. I have gotten less and less satisfied with Java's abilities to handle anything else but 44.1kHz/16 bit on different platforms, and even the latter may suck depending on your setup.

Now, that would be "something". 
All the trouble with ABC/HR Java some years back have made me shy away from Java in mix with audio. A native Mac OS X application on the other hand would score a lot more interest from me...

5 years since I first asked for one -> Seeking ABX software for Mac OS X, any good one out there?