CDJ Pitch Accuracy

DJ Gear & Hardware CDJ & Vinyl DJing
1

CDJ Pitch Accuracy

OK guys.. after a bit of help here. Have tested a few cheaper CDJs and have been finding that the pitch ONLY changes when the CDJ’s on-‘screen’ %age value changes (e.g. from 0.1%>0.2%).

To test, you need TSP & a timecode CD
I ran a CDJ (with Traktor timecode CD loaded) into an Audio 8, and changed the pitch on the CDJ. The tempo in Traktor ONLY changed when the %age readout on the CDJ in question (American Audio Radius 1000) changed. So, I can therefore get 0.5% and 0.6% tempo on the CDJ, BUT NOT 0.55%. I also tested this on my Citronic MP-X10 (USB stick player) after ripping the Traktor timecode CD to USB stick (legal? :confused:slight_smile:, and got exactly the same result (although admittedly the AA Radius and Citronic firmwares are similar).

Back to using CDJs and mixer (no Traktor), 0.1% accuracy is ok. You can mix with it, but you can’t lock tunes in like you can with Technics. I am often finding that one value is too slow and the next is too fast. I always thought it was my pitch slider control when I couldn’t get the tunes to lock, until I tested their actual resolution using Traktor.

On the AA Radius, on 4% pitch range, there are only 81 steps of accuracy/resolution (-4.0->+4.0)… people complain about 7 bit midi resolution (127 steps) using controllers, but 81 is far less than 127 (although it is still, technically, 7 bit! :confused:slight_smile:. 7 bit is not enough for MIDI controllers, and therefore definitely not enough for CDJs!

I am currently thinking of buying a pair of new CDJs (Numark NDX900 - liking the double interface), but am now worried of them having this ‘feature’. Is this normal to ALL CDJs, or just the cheaper ones? (I think I know the answer to this, but would like to know at what price point/quality that the accuracy improves)

Any chance of any of ‘you lot’ with various CDJs can test and see how common this is? I expect that it is manufacturer specific (as manufacturers will tend to use the same [pitch slider control] code in a lot of different products). Any CDJ that cannot give me more accuracy than 0.1% should be given to the toddlers to play with (should really be 0.01% resolution or less).

I know 0.1% doesn’t sound like much, but when you are mid-way between the 2 values, it does notice a lot. It’s enough to distract you so that you have to worry about tempos and nudging it every phrase or so. This takes brain CPU cycles, which results in me doing less in the mix, which results in a more boring set, which results in me getting less bookings (& therefore, money), & which results in me having to buy cheap CDJs that only have 0.1% accuracy!!! :eek:

Many thanks if you can add your CDJs accuracy to this thread, and some sort of accuracy indicator. If your CDJs are too sensitive to give an accuracy, then great, they are the ones I want!

Thanyovermuch

2

I’ll try to test my CDJs 350 tomorrow if i can remind myself (wtf is that a correct sentence ? hope you got the point).

Even though they are cheap CDJs, it’s still from the best CDJ manufacturer so we’ll see how it goes.

3

Thx

Many, many thanks dope. Hopefully we can get most brands tested (i.e. other people please help!! :slight_smile: ). The accuracy thing is something you wouldn’t know unless you do the Traktor (or Serato I suppose) test - you would just think that your mixing is a little off, which it would always be on those particular CDJs.

4

For CDJ-2000s/1000s/900s/850s/400s/350s/200s:

+/- 6% : 0.02%
+/- 10%: 0.05%
+/- 16%: 0.05%
+/- 100% (WIDE) : 0.5% (if the cdj even had wide)

CDJ-800:

+/- 10% : 0.05%
+/- 100% (WIDE) : 0.5% for cds, 0.1% for mp3

5

Thx TSB. 0.02% is what I would accept as just about ok. I am surprised by the 0.05% figs on anything other than 6% range. It simply isn’t accurate enough to get tight mixes that require no extra fiddling. Admittedly, with vinyl rips and older stuff, the bpm varies slightly anyway, but getting a mix locked in on newer digital stuff allows you to mess around with FX, EQ or other tricks to add a little something to the mix.

I am surprised that no-one has bought this up before (especially trance jocks with their never ending mixes :slight_smile: ). Searched DJTT forums, but couldn’t find any mention of it.

Hoping someone can test the Numark NDX800 or 900 :smiley:. Am thinking of getting the NDX900 (CD/USB/midi player with separate soundcard for Traktor, and a sorta slip mode :slight_smile: - god I love slip mode:stuck_out_tongue:), but if it’s 0.1% only, then I am having serious second thoughts. It’s a joke that my USB player is far more accurate in MIDI mode (10bit res=1024 steps) than native USB MP3 mode! (@ 6% range=121 steps).

6

These are numbers when using the CDJs with Traktor or as a standalone device?

7

if you’re using timecodes with traktor, yes since it still reads cds the same way. i would imagine hid would be the same, havent tried that though

8

Do you know what the increments are for when you’re just using the CDJs and a mixer without traktor?

9

Those ARE the increments when using it without traktor.

10

Having done it for a very long time, you’re completely wrong.

First, you’ll always have to fiddle.

Second, I used to worry about things like that, just like you are, but Pioneer CDJs are plenty good enough. .05 is already more accurate than any Technics turntable and perfectly fine to mix on.

And before you say “wait, the pitch faders on the technics were analog”……analog doesn’t necessarily mean higher resolution…they’re still stepped values, but they’re stepped kinda randomly based on the last time the fader was cleaned..

11

+1 Lol i’m glad someone said it.

12

Become more comfortable with beatmatching and mixing your favourite music genre, and pitch increments will become the least of your concerns.

13

You’ll never “lock in” with Technics. You have to contend with wow & flutter - so you’ll always be giving it the odd nudge or brush.

Like mostapha said - try not to worry too much about it. It’ll come with practice.

14

Practise?? :slight_smile: 18 years’ worth not enough then?? Technics can lock in, yeah, maybe the tiniest play on the spindle required to keep it locked (yeah, ok, so it’s not locked, but it’s damn close), but with 0.1% on accuracy, you are correcting every 16 or 32 beats (which is FAR more of an adjustment than required on technics). Maybe I am getting too finicky, but had a tune the other night, I needed 0.95%, but could only get 0.9 or 1.0%. Was annoying as hell to mix, and I longed for a more accurate pitch.

If I should ‘just get used to it’ then why are DJTT (& Ean) among others, raving about 14 bit midi on pitch faders?? - surely the 7 bit should do just fine then, and that people should just ‘get used to it’?

My point is that if I am going to drop some money on equipment, this sort of half arsed implementation is what really pisses me off!.. the manufacturers add just enough for it to look good on- and to shift off the sales floor, only for you to take it home and find it’s shit, when you use it ‘properly’, rather than playing in a shop.

FYI I CAN happily mix with 0.1% accuracy on the pitch, and likewise I CAN happily mix on some soundlab belt drives, but I’d PREFER technics, and i’d PREFER more accuracy on CDJs.

Yes, I KNOW I can nudge it into time… (as that IS what I do to correct anyways) but it’s just too coarse for when your tune needs that 0.05% and you’re mixing some long-ass techno intro while sucking on a balloon (now this is no ordinary balloon parents… :slight_smile: ) Just sayin that I don’t want to pay money for crap resolution… have been there, and it’s shit…

Oh, and thx for everyone who’s bashing me here, but let’s keep it light shall we folks, as that’s what this excellent forum is all about.

15

Numark’s old Axis 8s were about the same res as all the top pioneer decks, even though they only showed 0.1 increments. You could tell by the way it changed around the zero point with + and -. I also had the correct pitch range from the firmware engineer who stated I think that it was like 0.03 or something at one of the ranges between 10 or 12% and the lowest one. So it didn’t just pop up to .05 after you switched above the lowest range. Anyone find out about the pitch res on the NDX-900? By the way, 10-bit res is the minimum you want. The fader length should be about 60mm for that to get proper change at any motion. Shorter and you’re wasting res. Longer, and small motions won’t register… like on the Xponent. Not a terrible thing, but you just have to watch the screen. 10-bit will give you at least .05 at the 8, 10, 12, and 16% ranges. It will actually be better in software at lower ranges than 16% than on the Pioneers. 14-bit is obviously the best. Buying new stuff nowadays, you should accept no substitutes. Don’t let companies rip you off. Make them source quality 14-bit pots for their 100mm faders. And like 10-bit, any shorter than 100mm will waste res. A little longer might not hurt, but 100mm 14-bit pots is a sweetspot.

16

Blah. The NDX only supports MP3 over USB drives. That’s a deal breaker. Like the digital out. Need at least wav support, and flac would be even better.

17

Your math is right, but misleading.

Max drift speed: .025 % * 120 BPM = 0.06 BPM → 33 minutes, 20 seconds to drift 1 beat, so that’s correct.

But, it seems like drift becomes audible at about 1/64th to 1/32 beat if not earlier, so that’s actually more like 15 to 30 seconds to be audible in the worst case scenario.

Still…a tiny nudge on a platter or tiny tweak of a pitch fader a few times a minute is nothing.

18

:stuck_out_tongue:

Your definition of average is incorrect. And I was sure you knew all of this. Just kinda felt like being pedantic while stuck doing things I didn’t want to do to keep myself awake.

No offense intended.

19

And that’s assuming the tracks are perfect in timing to begin with. Add actual timing variations within tracks, and a 0.05 pitch resolution in internal USB mode or CD playback or on a 10-bit fader in the most used pitch ranges is still functionally inferior to 0.01 pitch resolution that we seem to get with 14-bit pitch faders in the usual ranges… assuming you are careful in your original blend. 10-bit is the minimum you want in a pitch fader for mixing. 7/8-bit is garbage, unless you plan on mixing only tracks that are nearly the same BPM and you use, say, 2-4% pitch range. And if you are constraining yourself to that, then you’re probably someone who has no problem just using Live and autosyncing after analyzing all your tracks and setting warp markers, or something. I didn’t get into DJing to have a damn computer mix for me, though, after I do all the grunt prep work. 14-bit is obviously better and with how cheap these pots have now become, there’s really no reason for manufacturers to be duping you out them if you’re buying a new model. And avoid any manufacturer who can’t even post such specs such as bit of the pitch fader or pitch resolution while in internal USB or CD playback modes (they are not always the same). Demand 100mm 14-bit pitch faders. If you’re buying used or the unit is cheap, yes, you can get away with 60mm 10-bit faders. Realize on a Versadeck or an Xponent, very small fader changes will do nothing because they are 10-bit 100mm faders.

20

Thanks guys for that, was getting slated at the start of this thread… yeah, I CAN mix with shitty pitch controls and nudge, but if I’m spending money on my own kit, I don’t wanna have to make do. Thanks for doing the 31 secs math, as I couldn’t be arsed to work it out :slight_smile: