DJM 900 isolator behaviour

Tarek, my djm-900 does exactly the same thing. Really don't know why it does this. When I record on audacity, and switch from EQ to ISOLATOR the wave changes too, as you say, much more erratic and there is a jump in volume. Would be interested to know if this happens on all the 900s or just some.

Hrm, this is the first I've heard of it ... I'll look into it and ask the engineers about it.

Thanks Pulse, also, when in EQ mode, when you take one of the freq bands out, for example bass, the db level goes up too. The LED's increase by about 2-3 db. Confusing, as you are taking out 26db of bass, why would the song increase in db?

@Stereojack > Thats not a bug. Remember, if you are tweaking one end of the frequency spectrum thats going to have an inverse effect on the other end. (Therefore, decreasing the low frequencies, increases the high frequencies).

Try it with the Isolator and you will see different results.

Yeah, I'm not sure how the level monitor reports the track levels, but I'm pretty sure there's no "averaging" or compensation going on.  What Gavin suggests doesn't make sense either (sorry Gav!) as a removal of all low frequencies would also lower the overall volume levels of the signal, not make the high frequencies louder.  ;)

Hmmm, puzzling. Pulse, did the engineers say anything?

@Stereojack > Currently awaiting a response.

I believe I read somewhere that Isolator's have control over the volume of the frequencies as well as the curve of them. Not as in "Duh, I turn it clockwise and it gets louder",  but the EQ uses the curves with High pass, Low Pass, and Band pass and only changes the level of the curve to alter the sound, but with an Isolator, I believe it controls the curve as well as the volume of that frequency in order the alter the sound.

I don't believe any technical numbers will be generated as it has to do with specifications that could be copied by competitors.

If you see a +4 - +7 db increase in volume I'm assuming you have the levels on your channels turned up a decent amount?

When I replicated it, I only saw +2 - +4 db of an increase, which I believe is because I keep my levels really low and it doesn't add the extra volume from the isolator (if this is true). I keep my levels in the green and mostly max them at +2 db. lol I don't need the extra volume at all, all my gear tuned on to mid range is more than enough power for my tiny room! lol I don't want to start an argument of what your channel levels and Master levels should be, but lets just say you can "Christmas Tree" the DJM-900 pretty good before clipping. I just choose not to. :)

But let me know your thoughts about it since I'm not sure and it's just something I read...

Thanks,

Foster 

Not totally sure I understand what you mean Foster : / sorry, maybe i'm being an idiot. I keep my channel levels around 0db too, meaning my master output is usually around -15db, but I don't see how the level on your channels is going to make any difference.

@Stereojack > I think it's actually a good thing that you didn't understand it because now I don't really know what I was going on about myself! lol It kinda made sense to me, but it goes both ways to be honest, so I'm not even going to bother trying to explain myself. lol It was something I read and I thought it made sense at the time, until I tried explaining it myself... (fail) lol

I guess it probably just has to do with the way the curve is laid out between the EQ and the ISO. In order to get that complete kill, the ISO has to have a steep curve at the end, but as to why it's louder near 12 O'Clock... beats me for the time being, or the ISO curve simply levels out  slightly higher in the db range than where the EQ levels out.

I'll test out that audacity wave change too, it sounds cool. lol I've never bothered to switch the EQ and ISO during a recording.

Sorry for the confusion,

Foster 

Any progress on this guys? Foster, did you check it out in Audacity?

I did some measurements -- will show in a bit, need to put it to visuals for you.

Okay, colours are a bit funny on the graphs because of some funky Photoshop overlaying I had to do in order to get this to work, but here it is...

--note-- The images are wider than this forum; right-click on them to view them in another tab/window.

This is an overlay showing a spectrum analysis of white noise recorded over 30 seconds, each dot represents the peak of that particular frequency for that 30s period.

The colours are:

Black - NO EQ

Green - Low, ISO mode

Orange - Low, EQ mode

Purple - Mid, ISO mode

Yellow - Mid, EQ mode

Red - Hi, ISO mode

Blue - Hi, EQ mode

Here's a breakdown of that graph comparing each EQ side-by-side:

The lows:

Mids:

 I think there may have been a bump or other anomaly in the mid-EQ (yellow) recording (because it is lower than expected), but I wasn't about to redo them all after I noticed it.

And the Highs:

 You can see through these four graphs that at no point, regardless of the ISO/EQ mode selection, does the level peak above the mean un-EQ'd black line.  As such, it's my conclusion that the metering on the DJM does not necessarily represent the actual output and that there may be some advanced method of measurement being used (although I'm not sure what or why).

Wow, thanks for taking the time to do this Pulse. Yeah I can't actually 'hear' any difference in sound levels with my ears when I play in EQ and ISO mode, and these graphs prove it. Still strange however, that when in audacity the db level does record differently depending on which mode you're in, and also, as you said Pulse, the metering on the DJM does not represent the actual output levels, although strange why it wouldn't.

Yeah, I'm not sure just how that works, but don't let your ears deceive you!  ;)

I asked the engineers and finally got a very technical answer -- and it has to do with IIR (Infinite impulse response) filters.  The short answer is:

"The levels of some bands are raised because of the unstable phase in some frequency bands, which is displayed in the level meter."

Bottom line is back to my last reply; use your ears.  ;) 

Thanks for the explanation Pulse. Not that I understand it but at least the engineers know what's up lol. 

I echo Tarek's thoughts. Cheers Pulse.