|
Title: Crack Detect trips in 0.9y-4 Post by: arvind on November 01, 2009, 01:22:39 pm Hi Peter,
Detected a very strange problem today; while playing the album Dire Straits - Money for Nothing I got crack detect problem on only one track "Money For Nothing". All other tracks do not give this problem & the entire album has been ripped at the same time. My native files are 16/44.1 Wav. The windows read as follows: 24 bit Second Stage Crack Detect -->22 88200/529200/40/5000000. I tried to play the same track thrice & all three times I got the crack detect at the same time on the OSD track time showing 2.21. Arvind Title: Re: Crack Detect trips in 0.9y-4 Post by: PeterSt on November 01, 2009, 01:40:45 pm Hi Arvind,
This is a known problem for (and since) Arc Prediction and is caused by files which have steep transients in the first place which just get a tad steeper because of (say) that was reality. The number 22 tells that the treshold was exceeded by 2 (this is set to 20) and indicates that a total of 22 samples were found in the file (part) to be too steep to be normal. At 44100 samples per second this is "nothing" of course, and not caused by wrong calcultations etc., what the Crack Detect actually is for. This all means that the treshold must be uplifted somewhat, which will be the case in 0.9y-5. When you digitally attenuate a bit (6dB should be sufficient) it will go for 0.9y-4. Peter Title: Re: Crack Detect trips in 0.9y-4 Post by: AUDIODIDAKT on November 01, 2009, 05:13:47 pm I have this problem too sometimes, (last night actually)(but its rare)
Thnx Peter for explaination, will keep track of this. :) Title: Re: Crack Detect trips in 0.9y-4 Post by: AUDIODIDAKT on December 11, 2009, 10:59:18 am This is a very high crack,
You know what, I turn it off, happens too often now a days. Roy Title: Re: Crack Detect trips in 0.9y-4 Post by: PeterSt on December 11, 2009, 08:39:25 pm That amount is still normal (but already in the range you won't know in advance).
0.9y-5 will be better on this ! |