- This topic has 83 replies, 18 voices, and was last updated 7 years, 10 months ago by
.
-
Posts
-
Scotty, add an attenuator as the first link in the DSP chain. That will take care of it.
http://www.claessonedwards.com/plugins/
///Leif
Yeah, to fix that, I control the input to BB, by turning down the wave output within the windows mixer, adjusting down the breakaway 1 pipeline, which is how I port the audio from winamp to bb. I like that preset on 70’s/80’s stuff.
If I’m processing today’s stuff, and competing against an Omnia…I just turn the bass drive down a little bit and crank up the final drive…using the same preset. I might switch to Plutonium instead of Plutonium NR.
Dang, I cannot wait to have a full control version. When it’s for sale, I’m buying.
That’s even better, Sgierk. That way you also ensure that the input isn’t clipping (pre-attenuation).
///Leif
Leif,
[quote author=”Leif”]Scotty, add an attenuator as the first link in the DSP chain. That will take care of it.[/quote]I already have an attenuator in the DSP chain. I don’t have any input overload causing the meters to flash until I use that setting with those plugins. Because it uses the bass efx and the impact/clunk plugins do I need to attenuate further?
Scott
Yes. Increase attenuation until the flashing stops.
Make sure you’re not overloading the actual sound card input, too.
///Leif
Hi Scott,
[quote author=”ScottyJames”]I already have an attenuator in the DSP chain. I don’t have any input overload causing the meters to flash until I use that setting with those plugins. Because it uses the bass efx and the impact/clunk plugins do I need to attenuate further?[/quote]
My audio files in the database are .wav files normalised to 95%.
I use the following plugins in this order:1 – 5dB Attenuator
2 – Impact/Clunk (Clunk 1 Slam 2)
3 – Bass-EFX (Level 2)This seems to be enough attenuation for Impact/Clunk and Bass-EFX following behind it.
Now if I were to stick in a Stereo Enhancer Plugin (Spartacus for example) to the chain:1 – 6dB Attenuator
2 – Spartacus (Stereo Enhancer)
3 – Impact/Clunk (Clunk 1 Slam 2)
4 – Bass-EFX (Level 2)Notice how I attenuate further (6dB Attenuator) so as not to overload the input.
As Sgeirk and Leif said above, an even better option is to enable the volume control of Breakaway Pipeline 1 (in the control panel) and turn it down a bit so as not to overload the input until the meters stop flashing red.
The AGC will make up for your slight loss in input volume. 8)
You can then turn up range slightly from say 50 to 60 if you like… 🙂Hope this helps,
Best regards,
Conor HylandActually turning the pipeline volume control could wouldn’t help — you could still be clipping the input of the pipeline, and then simply hiding the problem by attenuating afterwards. The meters would stop flashing, but the distortion would still be there!
The very best solution is to lower the gain before any clipping.
Using a sound card with an analog input, this translates to lowering the sensitivity (i.e. raising the clip level).
Using a player on the computer, feeding a pipeline, this translates to decreasing the MP3 encoders gain! As an example, in Winamp, this can be done by enabling Fast Layer 3 EQ in the in_mp3 configuration and then turning on the EQ and adjusting the PRE-AMP slider down. Unfortunately, most players do not have this functionality, which is why clipped CDs get clipped a second time when decoding MP3s recorded from them.
As long as you make sure that there’s no clipping further in the chain, though, using the attenuator is perfectly acceptable.
Perhaps I should modify the clip detector to make it detect clipping both before and after the plug-in chain.. This way one could really be certain. Input clipping is the worst possible thing to do to an audio processor (whether done intentionally while mastering a CD, or by accident in the program chain). It adds no loudness whatsoever (any loudness increase gets undone by the processing itself), but adds plenty of grunge and distortion. The clipper at the end of an FM processor is *designed* to clip audio with a minimum of artifacts, but clipping the input (squaring off the samples) is not at all the same thing, and sounds a hundred times worse.
///Leif
Hi Leif,
[quote author=”Leif”]Actually turning the pipeline volume control wouldn’t help — you could still be clipping the input of the pipeline, and then simply hiding the problem by attenuating afterwards. The meters would stop flashing, but the distortion would still be there!
The very best solution is to lower the gain before any clipping.
Using a player on the computer, feeding a pipeline, this translates to decreasing the MP3 encoders gain![/quote]
That is a valid point above.
I always wondered about that.I use .wav files normalised to 95%.
The Radio Automation software I use is Jazler Simplepack.
Jazler Simplepack’s main output to Breakaway Pipeline 1.When adding files to the database, it analyses the queue point at which a song should start.
I have it set to queue the start of a song when the song level rises to or above -26dB.It also analyses the fade out of a track/queue point at which the next track should start.
I have this set to -17dB.
It does this by starting at the end of the track and searching backwards until the level rises to or above -17dB.So lets say the first song is playing and it fades out to -17dB, it starts the next track.
If the second track starts with a beat, the sum of the 2 songs could clip the input of Breakaway Pipeline 1.I assumed that if I use the clip detector on Breakaway’s input without an attenuator or any of the plugins to check for input overload and if the meters didn’t flash red on song changes like the one above, then no clipping was occurring at Breakaway Pipeline 1 input.
Is that a safe asumption? ❓Jazler Radiostar One has an option to reduce the gain on it’s main output, but unfortunately Jazler Simplepack doesn’t… 🙁
Thanks Leif, 🙂
Best regards,
Conor HylandHi Conor!
Correct — if you run without any attenuators or plug-ins, then the clip detector will be looking at exactly what comes out of the pipeline (which is exactly what went into the pipeline). Thus, no red flashing = no clipping at the input.
By the way, regarding normalizing to 95%:
95% translates to 0.45dB.
0.45dB is not a lot of headroom, Conor 😉.
Although, for normalizing a wave file I suppose it’s fine, but you might as well have normalized to 99 or 100% really — 0.45dB is barely enough to be audible.
Best,
///LeifLeif,
[quote author=”Leif”]Yes. Increase attenuation until the flashing stops.
Make sure you’re not overloading the actual sound card input, too.
///Leif[/quote]I run direct out of NexGen Digital to a VAC then BBP (or Live) so I have no control over the levels. I did use the 6db attenuator and that solved the problem when I was trying sgeirk’s suggested settings. From that result I couldn’t figure out where the DSPs are in relation to what I am seeing on the Input meter. The BBP input meter showed my levels to be accordingly lower and my guess is the AGC makes up for that attenuation, so where does the affect of the plugins bass efx and mpact/clunk (that caused the level increase that caused the overload) come into play? Prior to the Multiband?
I’m just not sure where the overload was being introduced.
Conor,
Thanks for your reply.
[quote author=”Conor Hyland”]My audio files in the database are .wav files normalised to 95%.[/quote]All the files I use are .WAV files as well.
I see how you use more attenuation based on the added DSPs but I wasn’t sure where the overloading was being introduced.
[quote author=”Conor Hyland”]As Sgeirk and Leif said above, an even better option is to enable the volume control of Breakaway Pipeline 1 (in the control panel) and turn it down a bit so as not to overload the input until the meters stop flashing red.[/quote]I thought I had read in one post that Leif suggested that the Volume Control option not be checked and if it was that it be left at 100%. Actually I think I checked it once and it did nothing to change the volume.
The 6db attenuator did the trick anyway. 🙂
Scott
Hi Scott,
[quote author=”ScottyJames”]I see how you use more attenuation based on the added DSPs but I wasn’t sure where the overloading was being introduced.[/quote]
The overloading was being introduced before the AGC. 🙂
I misread what Sgeirk was saying above – silly me!
I see now what he was saying: 🙂[attachment=0:gejxfdey]Master Volume.JPG[/attachment:gejxfdey]
Best regards,
Conor HylandHi Leif,
[quote author=”Leif”]Perhaps I should modify the clip detector to make it detect clipping both before and after the plug-in chain.[/quote]
What a great idea!
Even better would be to introduce a 2 or 3 color clip detector! 💡
One color (say blue) to detect clipping before the plug-in chain.
Red to detect clipping after the plug-in chain.
Maybe even a third color (say purple) if clipping is occurring both before and after…This will make it easier to determine where the clipping is occurring in the chain! 🙂
Best regards,
Conor Hyland[quote author=”Conor Hyland”]The overloading was being introduced before the AGC. 🙂
I misread what Sgeirk was saying above – silly me!
I see now what he was saying: 🙂[/quote]That will do the same thing that an attenuator plugin would be doing, possibly in lower quality, I’m not sure how VAC handles that.
Like Leif said, if there’s clipping coming from the mp3 decoding before it even has a chance to get attenuated, then you need to fix it in that codec pair first… by reducing encoder input, adjusting the file, or using a decoder (like MAD) which supports input attenuation via mp3 gain…
If you’re playing back files as opposed to streams, check out mp3Gain
http://mp3gain.sourceforge.net/
it will losslessly apply ReplayGain to all of your mp3 files, so that they don’t clip on playback, it works with all mp3 players of any kind, and is even reversible after the fact.Try a bunch of your mp3s in there, preferably the ones with the most dynamic range. Those will be a good representation of the lowest average levels in your collection and about where you should be setting your reference decibel level to avoid clipping. Then mp3gain your entire collection at the same reference decibel level.
If you want to go by industry standards, then set the reference decibel level to 83dB… "shockingly low" to many "modern" producers and even some engineers, but 83dB is industry standard for a LOOOOONG time, and it’s quite sad to see the total & reckless abandon of calibration (one might even say total lack of knowledge of) in the industry over the last 20 years.
Film and mastering engineers who know what they are doing seem to be the last parts of the entire industry that still enjoy the benefits of a calibrated monitoring system. And in the case of film, actually keeping those calibrations onto the destination medium. It’s too bad CDs were not done this way still. (see also: Volume War)
</rant>
Hi Scotty!
quote :I run direct out of NexGen Digital to a VAC then BBP (or Live) so I have no control over the levels. I did use the 6db attenuator and that solved the problem when I was trying sgeirk’s suggested settings. From that result I couldn’t figure out where the DSPs are in relation to what I am seeing on the Input meter. The BBP input meter showed my levels to be accordingly lower and my guess is the AGC makes up for that attenuation, so where does the affect of the plugins bass efx and mpact/clunk (that caused the level increase that caused the overload) come into play? Prior to the Multiband?The DSPs are before the Input meter, and thus before the clip detector. The plug-ins are called as if you were building a chain of compressors in hardware, i.e. the only place to plug something in is before the input.
The Bass-EFX and Impact/Clunk presets do not increase average levels! They do not increase the energy of the audio.
However, in the case of audio that was peak limited or clipped on the CD, the energy level will be artificially high while the peak level is low, and when it goes through either of these plug-ins, some of the peak (as it was before clipping / limiting) will be restored. Those peaks will clip again if there’s no headroom, and that’s extra bad. 😉
Please note that even though some peaks are restored, any distortion which was in the original will still remain. It’s not a clip-restoration plug-in — perfect clip restoration is conceptually impossible (once information is lost it cannot be retrieved).
///Leif
Leif,
in the newar future …many imposible things will become…possible.Maiby you create an algorythm who will be able to recognize clipping on the SOURCE cimpletely and be able to manage the re-creation of the most "ear-tasty" frequencies.
like…let’s say…..highs and bass 🙂)))I have many recordings with…clipped bass ,
imagine , a clipped and bad recorded bass …which goes through a ..high density overdrive…
the worse becomes….more and more worse ..if it’s bad mastered..will YOU be the first to create that re-generator …or BOB ORBAN ?
who knows…….
….take care of yourself ! we need you 🙂
- The forum ‘Breakaway Professional Products – [discontinued]’ is closed to new topics and replies.