More > JRiver Media Center 21 for Windows
NEW: Volume Leveling uses the additional Headroom provided by Internal Volume
mwillems:
--- Quote from: lisbethfox on August 14, 2015, 12:32:53 pm ---This is a really weird issue that had persisted throughout many updates, eventually I just figured Id put clip protection on but honestly I'd rather not settle for that.
--- End quote ---
Well you could probably "solve it" by using internal volume and setting your max volume at 95% (or just using volume leveling, which will solve the problem as well), but I agree it would be better to run this to ground somehow.
I just reproduced what you're seeing by recreating your DSP chain exactly and playing an album with a 0dB peak value. Of interest, even with 2dB of attenuation from internal volume, I was still getting occasional overflows, which instantly went away when disabling the high pass. I'm not sure what's causing it to happen, but adding the high pass is definitely pushing things into clipping which is bizarro. It's also odd that it will report overflows even when analyzer is reporting that the RMS volume is still much lower.
This is definitely one for the devs I think.
lisbethfox:
I do no internal volume (at least on main speakers) all my amps (Its 2 for each front and then one for each center and rear) are pro amps so they have gain controls and would rather not. In the mean time I trust the T6's internal crossovers. This could be DEVASTATING to someone believing they could use JRiver to derive actively crossed over signals and bypass the analog ones.
Im really glad I found this.
Even more curious to find out why it happened.
lisbethfox:
Having said that when I go up to 14+ channels I will be using internal for global volume control.
mattkhan:
Fwiw there was a bug report a while ago that showed jriver was not handling a convolution filter with magnitude > 1 even if the end result was ok. This kind of sounds like the same sort of magnitude issue.
mwillems:
--- Quote from: lisbethfox on August 14, 2015, 12:56:39 pm ---This could be DEVASTATING to someone believing they could use JRiver to derive actively crossed over signals and bypass the analog ones.
--- End quote ---
Let's not get carried away. If clip protection is engaged (which is the default and recommended setting) it's not really an issue (JRiver just quietly turns things down a dB or two), and even then it's only an issue when you have the internal volume cranked all the way up above 95%. I've been running JRiver as my active crossover into block amplifiers using internal volume exclusively for 2 years and have never had a blowthrough.
I agree that it's anomalous, but you have to bypass a few safety features (like clip protection) in order to create a real problem. Additionally, I didn't actually hear any actual clipping with my test rig while testing (despite seeing "overflows" reported as you described), so it may also be falsely reporting clipping. Clipping isn't super subtle, so it should be obvious when it's happening.
--- Quote ---Im really glad I found this.
Even more curious to find out why it happened.
--- End quote ---
Me too.
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version