Forum Replies Created
-
Posts
-
Bad choice. 1010LT does not support 192k at all, so it’s unusable for MPX.
The best choice is this card:
http://marian.de/en/products/trace8
Low latency, DC straight output (no tilt), 8 channels in and out.
It’s €399, though.
If you want a cheaper card, buy an ESI Juli@. It’s also a good card, but it’s not DC straight output so it does need tilt correction, and it’s only 2 channels.
Best,
///Leif[quote author=”yorkie98″]I’ve listened to them on headphones, at various different levels.I notice very quickly exaggerated background noise and analogue tape hiss, apart from this, was there something else?[/quote]
No, that’s it.
Some people love over-the-top fast AGCs like this one 🙂. Other people do not.
Breakaway has proven it can do transparent levelling since the very beginning. Intentionally audible leveling, on the other hand, has never available before. For some people, this appears to have been what was missing 🙂.
[quote author=”yorkie98″]Same here, something I mentioned some time ago and also asked if it’s possible to have seperate AGC and MBL speed controls.. Any chance of this Leif? This is still a major disadvantage of BBP against other products in my opinion.[/quote]
This may possibly come. No promises though.
Those other products have inferior back-ends which end up distorting the audio when doing peak control. I guess one has to weigh the pros and cons.
///Leif
Howdy!
This behaviour is by design. BBP always does pre-emphasis. You choose how much — 15 to 75us. On the main output, you can choose whether to de-emphasize or not (as it could be feeding an external stereo generator at which point you’d want de-emphasis OFF).
The internal encoder output (plug-in), on the other hand, ALWAYS receives de-emphasized output, as there are no codecs that need pre-emphasis.
Crashing, on the other hand, should not happen. EdCast must be set to Record from DSP! The Live recording should be crossed out, as follows:
Does this solve the problem?
Best,
///LeifSure, choose Breakaway Pipeline 2 as the output device in Breakaway, and then choose Breakaway Pipeline 2 as the sound card to record from in WM Encoder.
///Leif
[quote author=”TDCat”]Hehe!! I’m afraid it’s a bit too rich for me 😀
My personal preference is that the AGC should always be ‘a slow hand on the pot’ for processing with the multiband offering fast release times.
[/quote]A lot of people agree with you. For those people, I recommend almost any preset except eruption. 😀
///Leif
[quote author=”maczrool”]My new favorite preset with Plutonium now a distant second 😀 .
[/quote]Wow, thanks! 🙂
And yes, ceaudio.com is a little easier than claessonedwards.com — who can even spell that, anyway 😉. We’ve got breakawayaudio.com too.
///Leif
Yes, the latency times are much longer than the delay times, but we don’t care how long the latency is as long as it’s the same at all transmitters (as long as it’s reasonable).
What we care about is consistency.
For example, if a box sometimes has 2ms of delay, and sometimes 4ms of delay (randomly selected on boot-up probably), then that would be really bad.
On the other hand, if a box has 17.000 ms of delay every time it boots up, all the time, then it would be possible to use this for SFN.
///Leif
Interesting story!
Regular BBP will not work. Due to the buffering and the asynchronous SRCs used, latency will never be predictable enough.
BBP ASIO in low latency mode could work, though. In this mode it has no extraneous buffering, and all processing is done in the ASIO thread, so it’s possible that the audio from two PCs with identical ASIO sound cards, fed from the exact same audio source, would come out at exactly the same time. I haven’t tried this though! That’d be the first thing to test.
Sub-sample adjustable delay is easy! If that’s the only thing standing in the way, I could definitely implement it. Let’s make sure it’s possible to even get predictable delay through a PC first, though.
///Leif
The answer, for perfectly flat frequency response and no output tilt, is:
Marian Trace 8
http://marian.de/en/products/trace8
I don’t know of a sound card without input tilt, though. I also don’t know of a tuner without tilt, except for the Belar Wizard.
///Leif
Whoa, that’s heavy stuff man 🙂.
I’ve heard many audiophile grade explanations.. This one isn’t the worst by far, but that doesn’t make it true, and it just muddies the waters.
The temporal resolution you mention, is phase, and can be accurately reproduced as long as you’re far enough away for nyquist for the reconstruction filter to deal with it. Far enough usually means ~10% (20 kHz out of 22.050 kHz).
Any "detail" with higher frequencies further than that, would actually be *out of band* material. If that "detail" was in at the original signal before sampling, and somehow got through the input filtering in the ADC, it would be folded down into the audible band as aliasing (yikes). If that "detail" or out-of-band crap is present at the output but not at the input, then the DAC is faulty.
If by poor temporal resolution you meen poor phase response (think phase tornado) then yes, this is possible for a DAC, although I have never seen it is a 192k capable dac. Also, this problem would be dead easy to spot, as any tiny little modification to a peak-controlled signal (such as the one coming out of Breakaway Broadcast) would cause overshoots from the harmonics (some of which resulting from clipping) to no longer line up with the fundamentals, and thus no longer do their job of keeping the peak frequencies down.
In conclusion, to check if your sound card is good enough, check the peak control on the output with a scope, after having done Tilt and EQ correction. Turn off the Pilot and look at MPX, so that you see the edge of the signal more clearly. Is the peak control perfect, revealing a sharp signal edge? Congratulations! Your audio card is good enough for these purposes and will yield excellent fidelity.
Let’s try to stick to science, and keep the smoke and mirrors to a minimum.
Best,
///LeifHowdy!
The Nyquist theorem is not the problem. This theorem states any frequency below 96 kHz is perfectly reproducable with 192k sampling, and it’s right. 57k is well below 96k 🙂. Problems arise only from d/a converters with erroneously designed reconstruction filters.
Every Realtek HD audio dac i’ve tried (the ones that support 192k) has given me perfect sinewave output at any frequency up to 60 kHz. (I haven’t tested higher).
VIA Vinyl VT1708B on the other hand, not the case. In fact, i don’t know what they did to make it so bad!
Juli@ is a really nice and linear card, especially by soldering in bigger capacitors to improve the low frequency response. For software, take a look at my program MpxTool – http://mpxtool.com .
Best,
///LeifIt may be, but I don’t know — I’ve never tried it. It may be worth just waiting for the Juli@, or ordering somewhere else?
///Leif
