Forum Replies Created
-
Posts
-
Hi Dave!
I believe understand the problem.
Breakaway was never designed to be an audio switcher. What you’d want to change is the Input Device in Breakaway Broadcast, and leave everything the same, but doing so would mean you’d have to go into the I/O configuration every time (and take your station off the air for a few seconds).How about a nice old fashioned analog solution?
Use a mixer (or just a switch box), feed it the Winamp output and the lectures. Feed the output of the mixer to the sound card Breakaway is listening to. Feed Breakaway’s output to a Pipeline, and then have the encoders listen to the Pipeline.
For MP3, I recommend switching from Shoutcast to Edcast. Breakaway can call the encoder directly — please follow the instructions in the "recommended mp3 encoder settings" at the very top of the forum index.
Best,
///LeifDepends on how you’ve connected Breakaway into OTSAV.
If you’re using the LiveLink DSP plug-in, then the problem is OTSAV — you’ll have to ask the support how to record DSP processed audio.
If you’re using Pipelines, then you can set Breakaway to output to a pipeline, and record from that pipeline with any audio recorder.
///Leif
The other controls are just as complicated, yes. Each control adjusts multiple parameters inside the core. Each preset sets 150 (!) parameters, and the gui sliders then make adjustments on multiple controls from there, based on rules and extensive listening tests.
For example, when adjusting Speed downwards, not only does the attack and release for every band (as well as the input agc) slow down, it also raises the multiband limiter thresholds, and progressive release increases.
As an example, here’s the Plutonium preset:
code :presetname=Plutonium
gating_threshold=-30
freeze_threshold=-50
mb_gating_threshold=-30
mb_freeze_threshold=-50
inagc_ratio[0]=100
inagc_range[0]=50
inagc_atk[0]=22
inagc_rel[0]=24
inagc_threshold[0]=27
inagc_filter[0]=1
peq_level[0]=100
peq_freq[0]=4
peq_q[0]=30
peq_level[1]=50
peq_freq[1]=71
peq_q[1]=20
peq_level[2]=50
peq_freq[2]=2
peq_q[2]=30
ng_range=0
num_bands=6
mbagc_ratio=75
mbagc_range=58
mbagc_progressive=40
ng_threshold[0]=-50
ng_atk[0]=50
ng_rel[0]=50
inf_ratio_above_thresh[0]=0
mbagc_atk[0]=39
mbagc_rel[0]=55
mbagc_threshold[0]=50
mblim_threshold[0]=58
outmix[0]=70
mbclip_threshold[0]=66
ng_threshold[1]=-50
ng_atk[1]=50
ng_rel[1]=50
inf_ratio_above_thresh[1]=1
mbagc_atk[1]=44
mbagc_rel[1]=65
mbagc_threshold[1]=50
mblim_threshold[1]=47
outmix[1]=53
mbclip_threshold[1]=66
ng_threshold[2]=-50
ng_atk[2]=50
ng_rel[2]=50
inf_ratio_above_thresh[2]=1
mbagc_atk[2]=48
mbagc_rel[2]=65
mbagc_threshold[2]=50
mblim_threshold[2]=58
outmix[2]=50
ng_threshold[3]=-50
ng_atk[3]=50
ng_rel[3]=50
inf_ratio_above_thresh[3]=1
mbagc_atk[3]=48
mbagc_rel[3]=65
mbagc_threshold[3]=50
mblim_threshold[3]=60
outmix[3]=52
ng_threshold[4]=-50
ng_atk[4]=50
ng_rel[4]=50
inf_ratio_above_thresh[4]=0
mbagc_atk[4]=48
mbagc_rel[4]=70
mbagc_threshold[4]=60
mblim_threshold[4]=60
outmix[4]=52
ng_threshold[5]=-45
ng_atk[5]=50
ng_rel[5]=50
inf_ratio_above_thresh[5]=0
mbagc_atk[5]=48
mbagc_rel[5]=75
mbagc_threshold[5]=58
mblim_threshold[5]=33
outmix[5]=63
ng_threshold[6]=-50
ng_atk[6]=50
ng_rel[6]=50
inf_ratio_above_thresh[6]=0
mbagc_atk[6]=4
mbagc_rel[6]=5
mbagc_threshold[6]=50
mblim_threshold[6]=0
outmix[6]=50
final_lim_drive=55
output_level=106
bass_clipper_enable=1
bass_clipper_threshold=20
preset_version=3
user[0]=50
user[1]=50
user[2]=50
user[3]=50
user[4]=50
user[5]=50
user[6]=50
user[7]=50
user[8]=50
user[9]=50
user[10]=50
user[11]=50
user[12]=50
user[13]=50
user[14]=50
user[15]=50
wb1.delay=0
wb1.enable=1
wb1.atk=50
wb1.rel=50
wb1.progressive=25
wb1.range=0
wb1.maxgr=100
wb1.ratio=67
wb1.threshold=75
wb1.gating_threshold=-30
wb1.freeze_threshold=-40
wb1.peq_level[0]=29
wb1.peq_freq[0]=15
wb1.peq_q[0]=30
wb1.peq_level[1]=50
wb1.peq_freq[1]=20
wb1.peq_q[1]=20
wb1.peq_level[2]=50
wb1.peq_freq[2]=20
wb1.peq_q[2]=20
wb2.delay=0
wb2.enable=0
wb2.atk=50
wb2.rel=50
wb2.progressive=50
wb2.range=25
wb2.maxgr=25
wb2.ratio=50
wb2.threshold=50
wb2.gating_threshold=-30
wb2.freeze_threshold=-40
wb2.peq_level[0]=50
wb2.peq_freq[0]=20
wb2.peq_q[0]=20
wb2.peq_level[1]=50
wb2.peq_freq[1]=20
wb2.peq_q[1]=20
wb2.peq_level[2]=50
wb2.peq_freq[2]=20
wb2.peq_q[2]=20///Leif
I concur with Jesse’s recommendation.
Also, if you’re running Breakaway Broadcast on the same machine, you can use LiveLink to get audio from Breakaway Live to Breakaway Broadcast — this way you would only be using the Output of the Xonar Essence, the Input wouldn’t need to be connected anywhere.
///Leif
With ASIO, the lowest buffer size the Breakaway Core will support, is 48 samples, and in this mode the core will run in Extra Low Latency mode. 32 samples will not work under any circumstances.
With Kernel Streaming, 96/3 is the lowest I can recommend, and even that is not supported by most configurations.
///Leif
It did!
5.1 surround sound, breakaway processed.. It sounded incredible 😉.
Unfortunately I dismantled the whole system before I left the United States. (This was in Lancaster, Pennsylvania.)
I gave away the screen, and sold all the other parts.
When I have a house with a yard in Thailand I might just do it all over again though!
///Leif
Hi Dave,
The CGSmooth preset uses very low ratio settings, to retain some dynamics. It’s an awesome sounding preset, but by default (as designed by Cornelius for Breakaway Live) it’s not very loud — so I make up for this with the clipper at the end in Breakaway Broadcast. However, even the advanced distortion-masking clipper in Breakaway can only do so much. How about turning down the Final Drive? It should cure the problem.
///Leif
You got it 🙂.
When cutting bass, EQ isn’t affected. Instead, the B1 and B2 multiband agc thresholds are lowered (compared to where they are by default in the selected preset).
Bass Cut -50, Shape +50: B1 -12, B2 -12
Bass Cut -50, Shape 0: B1 -12, B2 -6
Bass Cut -50, Shape -50: B1 -12, B2 0Bass Cut -25 gives you half the cut (i.e. -6, -3 for shape 0).
Bass Cut is a great way to clean up the sound, because it still retains the bass boost inherent to many presets, while lowering the thresholds for essentially "more bass compression".
If you BOOST, on the other hand, several things happen — more and more depending on how far you turn it up. Be careful, there’s a lot of available power here 😉.
Assuming Bass Boost +50 (bassclippingorama):
Shape 0: B1 out mix +12, B2 out mix +3, B1/B2 agc thresh: +7.25, and PEQ +14.5dB, 23 Hz, 3.0 Octaves
Yes, that’s insane. I never intended for people to actually use Bass Boost +50, I just had to make sure the maximum setting was sufficiently over the top, to be able to tell anyone deaf enough to want more to go away, or buy a subwoofer.Looking at a more sane bass boost setting of +15:
Shape 0: B1 out mix +4.75, B2 out mix +1.25, B1/B2 agc thresh 0 (the thresholds don’t start rising until bass boost is ABOVE +20), and PEQ +4.75dB, 23 Hz, 3.0 octaves
Let’s stick to bass boost +15 and explore other Bass Shape options (agc threshold omitted since they’re not affected at this setting):
Shape -50: B1 out mix +4.75, B2 out mix +0.00, PEQ +4.75dB, 40 Hz, 1.0 octaves
Shape -33: B1 out mix +4.75, B2 out mix +0.00, PEQ +4.75dB, 35 Hz, 1.6 octaves
Shape -16: B1 out mix +4.75, B2 out mix +0.25, PEQ +4.75dB, 28 Hz, 2.3 octaves
Shape 0: B1 out mix +4.75, B2 out mix +1.25, PEQ +4.75dB, 23 Hz, 3.0 octaves
Shape 16: B1 out mix +4.75, B2 out mix +2.0, PEQ +4.75dB, 33 Hz, 3.0 octaves
Shape 33: B1 out mix +3.0, B2 out mix +3.0, PEQ +4.75dB, 50 Hz, 3.0 octaves
Shape 50: B1 out mix +0.0, B2 out mix +4.75, PEQ +4.75dB, 71 Hz, 3.0 octavesBy the way, this exact algorithm is also in Breakaway Personal.
People have no idea how much effort they’re getting for that measly $29.95 😉.
///Leif
Latency will not be fast enough for realtime monitoring when chaining multiple apps together.
Realtime monitoring latency = less than 20ms. That includes everything — adc/dac delay, buffering, processing. Adding two more buffering steps will certainly push it over the edge. 🙁
So basically what I’m saying is, it can’t be done. Realtime monitoring and pipelines simply do not mix.
///Leif
If I had to make a really really long cable run, I’d probably spring for Belden 1694A. (That’s Belden = The real thing, not Belkin = the overpriced consumer garbage). Belden 1694A is technically RG6, but a particularly good, low-loss kind. I used it to make a 60 meter (3×60) component run, 720p, for my outdoor HDTV home theater. That was a fun project 😉.
This is an actual photo of the screen! Image is composed from two exposures — one without flash (for the screen photo) and one with flash (for everything else).
For short cable runs (a couple of meters), I’ve successfully used whatever cheap RCA cable was laying around. The longer your cable, the more important the quality of the cable, but for a couple of meters I personally wouldn’t worry.
///Leif
Scotty, YES. The latency is a product of Sound Card input buffering, output buffering, SRC thread buffering, and the core latency itself. They all add up!
With the core in Low Latency core, the core only adds 4ms of latency (in phase linear, the core adds 100ms).
By using very small Kernel Streaming buffer settings, (for example 96×3 for input and output, if your sound card drivers can handle it) you can get reasonable latency, but it will still be 20ms or so at best with kernel streaming.
For true low latency, you need an ASIO capable sound card. Most semi-pro and pro cards are ASIO capable, and using ASIO you can have a total throughput of 10ms or less.
The cheapest ASIO capable card I know is ASUS Xonar D2X. That’s pro-sumer, and you can have one for under $100.
For semi-pro, look at ESI Juli@ or M Audio’s Delta / Audiophile cards ($150 or so). For pro, the sky is the limit (both functionality wise and price-wise).
Best,
///LeifNo no no no no no NO.
It sounds EXACTLY as bad. Come on, man! Are you making a joke? Are you pulling my leg here?
That’s not healthy audio I’m seeing (or hearing), Cam.
I think you need to start over with a factory preset, it seems these settings are beyond repair.
Try New York, Final Drive -0.5, Range 40, Power 40, Speed 50, Bass Normal, Bass Shape 0.
Then, leave it like that so that I can listen and verify that it sounds okay. If it doesn’t, we’ll know it’s a problem in the chain either before or after BBP.///Leif
Camclone, yes it does. It sounds just as distorted on FM broadcast, it’s just that you’re not hearing it through the radio you’re testing with, or when comparing with other stations (which are even more distorted). One thing is for sure though — FM transmitters, airwaves and radios do not magically improve the sound. What you get out of the speaker of a radio CAN NOT POSSIBLY be better than what the audio processor put out! It can only get worse.
This is why other audio processing manufacturers marketing strategies aggravate me to such a serious degree. This whole "take it home, hook it up to your transmitter, you’ve gotta hear how it sounds in YOUR environment" is pure unadulterated BS. It just makes it impossible to listen properly. An audio processor needs to be tested on the bench, and the real-world installation needs to be made to emulate the bench sound — not the other way around.
It’s really pretty simple — if there is ANY modification to the processed output signal (further clipping, phase errors, non-flat frequency response, tilt etc) it will show up as *loss of peak control* on an oscilloscope, and/or *spectral contamination* on a spectrum analyzer. Thus, if you’re seeing a tight signal off the airwaves (using a modulation monitor), the only conclusion is that the signal *is not being modified* (if it was, you wouldn’t be have tight peak control), and thus what you get on the air is exactly what the audio processor puts out… which brings us back to the bench test — it’s valid.
Using a high quality encoder to stream the output of an FM processor, while not representative of peak control, IS a valid representation of the sound for listening purposes. The streaming encoder can not possibly create this kind of distortion, thus we must conclude that the distortion was already there at the output of the audio processor, and that it is in fact on your airwaves too.
Any audio processor can be set up to sound bad — including BBP. If a processor cannot, that means the control range is too narrow. It’s possible to crank it — but that doesn’t mean that you *should*.
Camclone, what settings are you running now?
INI file settings do not matter (they are only a representation of the on-screen sliders, there are no hidden settings accessible from the ini file), so you can just read the on-screen slider values as usual. If you don’t want to reveal the settings, PM me instead, I’ll read it and make a recommendation. (My gut feeling is that I’ll be recommending the final drive to be turned down a lot). 😉
Best,
///LeifThe encoding quality is better than before, but the audio still sounds AWFUL.
Seriously, you should be comparing the cleanliness of the sound to Reference Heavy or at least Plutonium, not distorted off-the-air stations. Just because they’re deaf doesn’t mean your station has to sound bad.
To my ears, Plutonium is as loud as things can go without making (to me) unacceptable sacrifices. It’s not a coincidence that Plutonium is clean and relatively loud at the default settings — I deliberately tuned it this way! It’s also not a coincidence that it’s the default preset in BBP.
Look at the oscilloscope, for crying out loud. Even if you ears aren’t hearing that it’s distorted, how can you not see it?? 😯
///Leif
