SC5000 High End Roll-off


#44

Hi you guys. The readout from the analyzer suggests, that there is a low pass filter in the dsp signal chain of the sc5000. There would be many reasons to do this. But the drop off does look a bit early and bit steep. Clearly the filter is audible.


#45

It’s funny because a guys last years find the same thing with Traktor Pro 2 @Gee_DenonDJ it’s possible to improve this in the future update ?


#46

Ultimately, this is the player spotting a track, or sections in the track, about to exceed digital zero ( 0 dbfs - which is a fatal sin in terms of digital audio ) and applying safety, but only as required.

A lot of meters, even the correctly and professionally calibrated hardware meters, often have smoothing, weighting, bias or averaging applied, and may not show brief moments where the overall sound, or particular frequencies, exceed digital zero, but the player sees/hears them and responds accordingly on the rare occurences where certain tunes/re-rips/tweaked copies contain such excesses.


#47

So if I understand you correctly, the drop off of high frequencies is related to potential clipping and it should not occur with enough headroom in the sound material?


#48

That means the Player work like a limiter ?


#49

As mentioned above, it’s only there on tracks which would otherwise exceed digital zero.

I wouldnt consider, or call it a limiter. That definition, to me, would be a piece of kit that effectively turns down the gain across all frequencies in the track, regardless of where in the spectrum the sound file exceeded a threshold, rather than a gentle, graduated roll off.


#50

I would rather have no round-off at all, to be honest, because I “master” and check all my files for correct volume, silence before-after and faults of a track.

Is this something a firmware could resolve or is it hardware design? I hope firmware, because then the units would really go “off the scale” (no pun intended) in sound purity terms!

Pleasssse?! :stuck_out_tongue_winking_eye:


#51

Most short, cheap RCA line-level phono cables are just coaxial shielded type cables and not a twisted pair, though their termination impedance may not be exactly 75 ohm. Many “digital” coax cables aren’t actually exactly 75 ohm, either, but this usually isn’t an issue until long runs of cable.

FWI, with lossless files the Pioneer CDJs are bit-perfect out the SPDIF when the pitch fader is at zero. InMusic should shoot for that. With regards to the SC5000 processing’s intolerance of hot masters, the easiest method would be to achieve bit-perfect decoding & processing at zero pitch first as proof-of-capability and then add an optional blanket -6dBFS digital domain pad in the settings (obviously no longer quite bit perfect, but close enough) so as to prevent not only traditional clipping but intersample distortion, which requires over -3dB magnitude attenuation to always prevent. The -6dBFS pad, which you should probably then make the default setting in options/utility, would provide a little more headroom for your key correction algorithm than even that and is mathematically flawless and computationally easy. This would obviously be more transparent and less corrupting of the musical content than your current solution.


#52

Sounds a little like Torq 2 with a few days of listening to the SC5000. I need to get around to doing some measurements and spectrum analysis, as well as look at how the SC5000 is processing the standard test waveforms. Torq 2 would mutilate waveforms and overall tended to sound, dark, compressed, and sort of rolled off on top. Worked for very wide dynamics old funk recordings a little and basically made everything else sound worse. Besides working on the general transparency of the processing, another option to a blanket -6dB pad might be doing the DSP stuff in the players at 32bit float and then bump the volumes down by half just prior to the sample rate conversion & dither to the SPDIF and prior to the DA converters.


#54

http://forum.djtechtools.com/showthread.php?t=93197&page=10&p=765578&viewfull=1#post765578


#55

Wow even on that forum you keep replying to yourself. :joy:


#56

You’re posts are getting boring mate ! we understand now et maybe it’s time to enjoy no ?


#57

As a potential buyer, I find them quite informative.


#58

Not sure what you mean by “replying to yourself”.


#59

Thanks for trying to figure this stuff out, Reticuli, and explaining it to us. Learning a lot from this. Thank you!


#60

This is a serious technical thread. Please do not distract with unrelated discussions.


#61

Yeah. Never mind me, I’ll go and have a technical discussion with myself on every forum I visit… :crazy_face:


#62

All of this information matters. Even for someone maybe not as technical like me I was able to hear something that just sonically wasn’t what I was used to when I did a side by side comparison. So if any of this information in the thread helps bring something better via firmware/software updates, then I gladly welcome it.


#63

Of course it’s matters ! Reticuli are right ! I just hope that @Gee_DenonDJ will take into account this kind of problem with a possible correction because for high end media it’s not normal


#64

I have finally come around to doing the following test: white noise at -40dB 24bit 96kHz. At this kind of audio level most dj mixers don’t even register a signal as present - so any kind of dynamic protection should not kick in at all. All audio connections I used were digital. The result of this test is the very same as it was with all tests before: there is definitely a low pass filter present. At 21kHz the signal is down by a whopping -24dB, the -3dB point is at around 16kHz. I would like to share my analyzer read outs with you, but as a “new” user I am unable to upload any images.

To summarize, there is a static low pass filter, it’s audible and many users would like to deactivate it. For studio users it is essential to turn this filter off - there must not be unwanted and uncontrolled equalization.