Forum Replies Created
-
Posts
-
a screenshot would be way easier, because as far as I know (other than enabling the "depreciated" presets) there’s nothing you can do to the sound in the INI that isn’t already shown in the GUI. please correct me if i’m wrong.
It is 96 kbps yes (data rate). But it is also 88.2 kHz too (sampling rate). They are not the same thing. 😉
You probably should be using 44.1kHz if you’re playing music that’s 44/48, since not everyone has soundcards that will play above 48kHz sampling rate, and also the codec isn’t optimized for your source content at all that way. Literally the whole SBR is completely unused.
So you want 96 kbps, 44.1 kHz. 😉
I see your (Modulator) post as being a compliment, not a slam. 8) I applaud the use of ABBA in your post too. 🙂 Well crafted, if intentional.
@ the S sound… There is a crash that I’ve discovered that’s being fixed (of course) by Leif (top priority) which is preventing me from setting two things in particular which would not only partially prevent that (one half) but enhance other transients more consistently (the other half). So I’ve had to compromise on the settings a little bit.
Also to remember… this isn’t setup ideally for how it would be if it were a stand-alone end user Hi-Fi product. It is absolutely intended and optimized for use before broadcast audio processing. Not necessarily anything heavy, in fact I like using it the best with the most open preset for Breakaway/Omnia.9 dynamics core there is so far… Passive Aggressor. Just the same, that does still allow you to "flub" certain things like that, in the name of greater consistency and improved accomplishment of goals.
I would rather upload these samples without any additional processing although, so people really get an idea of the output of the Undo processing, and can also try the before & after with their own existing broadcast audio processing. You *will* hear it improve over the course to this summer, and I’m happy to start sharing where it’s already at. Remember, this has never been done before, and despite over a decade of on-air experience of Leif, and half a decade of R&D & use of this kind of expansion (although not automated) by myself in my own mastering…
(drum roll please) Undo is not even 3 months old, as of the time I’m writing this post. It’s totally incredible what Leif has put together here, in such a short time, and is a massive testament to how awesome his new processor creation framework is. Even the "development GUI" itself looks like a finished product, which is a joy to use even without hearing what it’s doing.
If you do hear anything else "odd" like that, please talk about it here. There could be stuff that we’re either not noticing, or not "annoyed" by, like some other people would be. This is a valuable part of the process of improving this component.
—
As far as if Undo will be included in the software products… that is the wish of Leif the last time he mentioned it. I can’t say for sure of course, but he wanted to. There are some limitations to that although. Part of it is licensed, so at least part of it will never make it to the very bottom rung products. But I would expect to see it fully (no promises) in products like Breakaway Live and Breakaway Broadcast.
You might need a scope for it. That’s one thing the calibration tool can’t show you… is if your audio INPUT used in this process actually has proper frequency response. If the soundcard you’re using for input isn’t flat (only tilted is ok since that can be fixed) then you obviously can’t tune the frequency response of the output either. Rose colored glasses, as the saying goes.
It’s not a bad idea to have a scope around anyways, since you can get them on eBay for under $50 USD.
Things like multi-phase power, higher end capacitors that handle more heat and last way longer, "over" engineered power sections, BIOS settings that are made more for the long haul as far as stability and how fast certain timings are pushed. Generally the chipsets that are chosen and the things enabled with them are geared for general servers too, not multimedia daily use, and that alone can stabilize not just the hardware, but the OS too.
Some of that stuff you’re starting to see in even some more budget-friendly "gaming" boards, from companies like ASRock (ASUS). There’s some gaming boards now with 16 phase power sections, higher end caps, several heat pipe coolers on chipsets and power regulation amps, etc… so at least OEM built home computers are catching up, if you’re willing to spend a few bucks more (and end up with a product 2-3 times as good as brand name pre-built).
I’ll respectfully disagree about loudness. 8) Backing off of the final limiter/clipper generates much more ATSL. When I was handling radioio (#1 online for 2.5 years straight, Arbitron/Ando) we were 4 times as "sticky" as anything else in the top 20, with what Leif described as "a really good sounding EQ". Very little dynamics processing, because my processing was designed to do spectral processing without changing the dynamics (maybe 1-2dB here and there). We had a slow transparent lookahead AGC for dynamics processing, and that’s it.
It does depend on the format, but I still don’t agree that the final loudness should be pushed very hard. The average can certainly be brought up a bunch, and that’s where it counts, but to smash that last 4-5dB of dynamics up into limiting/clipping when you only gain a higher perception of loudness for 7 seconds (the amount of time the average person can remember how loud something before it was)… So basically after 7 seconds, the only thing you actually gain is increased listening fatigue. 😐
It depends on several factors that, as usual, are not technical in nature, but business. I know that Leif wanted to put at least part of Undo even in the bottom level version. No promises, but it is possible that at least some of it could make it all the way down to the ground floor.
It’s too early to bet just yet though.
The MPX clipper however probably will not be introduced in the software, even in the possible software (again, no promises) above the current top shelf BBP+ASIO.
The declipper itself is actually before the multiband dynamics restoration stuff. The dynamics restoration works to a varying degree (completely based on program material) even without clipping or hard limiting present in the original. Even though it doesn’t do much for stuff with already great dynamics (like Donald Fagen from the early 80s), it’s still going to bring out detail in the recordings if it’s "lacking" anywhere.
The way it works is absolutely uncanny. Of course it’s more complicated than it looks.
[quote author=”radio oude stijl”][quote author=”Boki”]What’s Omnia 9 ? 😯[/quote]The new Breakawaybox = Omnia.9 It will cost about $10.000 and offers built-in encoders for streaming (Leifcast!) and a de-clipping section which is called "undo" Release will be this summer. So far for an affordable hardware version…[/quote]
As far as "Undo" being a declipper, it’s far more than that… I’m not sure what I can let out of the bag yet though.
But now that the Omnia.9 cat is out of the bag… I can be free to say that Frank Foti is completely blown away by Undo. His first two responses upon hearing it were limited to one word each!! (you have no idea how rare that is, lol) It literally makes me GIDDY when I’m hearing the output, raw or through further processing, when I’m toggling the bypass button on it. It’s that huge of a leap in broadcast processing, if you play modern smashed stuff or not. Actually it’s even more of a leap forward if you play a variety of different eras and "loudness" of recordings.
Images speak more than words though sometimes… and since we don’t all yet have one of these to play with…
Most people really won’t believe it can even be done, much less done properly, until they hear it. I realize this. 😆
[quote author=”Guillou”]Would you tell us that version posted above is better sounding that your version ?[/quote]
michi95 already said they output the exact same mp3 file. Reread his second post.The file I’ve uploaded is not my own build. It’s from rarewares years ago, and I *think* it was built by John33. Please upload any builds you have of it too, and people can see if it’s faster for their particular CPU.
My new preset called "Passive Aggressor", combined with a new thing that will be introduced at a processing "drag race" from Netherlands called Freakdag (Freakday)… is nothing short of astonishing. So completely wide open dynamics, audio that is always very detailed, and always at the perfect loudness and spectral balance.
If I had to put a number on how dynamic it is, somewhere between actual signal and subjective dynamics, I would guess that it’s about 5dB more dynamic than the Reference preset (almost twice as much at times). But with consistency that is unmatched by any other preset I’ve tried. It’s hard to explain but the consistency you do get with a preset/sound that’s very hard loud processing is only tricking the ears in some ways into thinking that it is consistent. The difference in listenability is not even worth comparing, a wide open yet consistent sound is unchallenged. The problem is that it’s very difficult to do, and until Breakaway a processor hasn’t had a few key features to make it possible without having big problems with certain audio events.
The Omnia.11 AGC i have not heard yet, but I’ve heard it’s very good. Perhaps equally as "uncanny" as Breakaway’s or even better, I dunno. I would love to get a unit for a few months and do some presets with it.
