• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

DDJ-400 Measurments

kipman725

Active Member
Joined
Jul 6, 2020
Messages
255
Likes
224

Intro:​

The DDJ-400 is an entry level DJ controller with master out (consumer line level single ended) and headphone out. Most of the functionality is through the REKORDBOX software with almost all controls and audio processing within that software with the exception of the Master level control. My controller is modified from stock in that the I2S lines to the DAC are also connected to an SPDIF/AES output board I have mounted internally to bypass the DAC. I don't think this will affect the DAC output as the twisted pairs to the SPDIF board are connected via 100ohm resistors but performance degradation is possible from this modification and I would like to check my results with a stock unit. The Firmware is the latest 1.03.

1640880370744.png

Tests have been conducted using my TASCAM US-2x2HR interface which you can see loopback of here. Mostly test tones (pink, 1kHz Sine) have been generated and are played back from Rekordbox 6.5.3 unless otherwise stated.

Output levels and THD+N:​

Clipping test:​

Mixer headroom set to NONE, 0dBFs sine wave on Deck 1, Master output set to FULL channel trim at +9dB, 2.239Vrms left channel, 2.234Vrms right channel (limiter has engaged), -13.4dBfs on input meters (9.22dBu).

1640880407543.png

Setting Mixer headroom to 9dB results in almost the same level of 8.45dBu suggesting we are running into Rekordbox limiter.

THD+1 1kHz normal output:​

Setting trim to mid point (0dB) and keeping output level at full and 9dB mixer headroom reduces level to -0.26dBu (752mV) the top red level light on the channel level meter on the controller is no longer lit. This seems an appropriate setting to use the controller in as the level has reduced by ~9dB as the trim setting indicates and so we are confident we have the specified mixer headroom before limiting.

1640880433928.png


Noise dominated (spurs of fundamental). With no signal the noise goes down the -101dBu. Perhaps the spurs are a hardware processing artefact or due to jitter? I notice spectral peaks centered around 48kHz so perhaps the internal sample rate of the DDJ-400 is related to 48KHz and we are seeing clocking errors or resampling?

1640880445649.png

The spurs are also present when FOOBAR200 is used for playback and so are not a Rekordbox playback issue.

Effect of output level control:​

Master Level PositionOuptut [mV]Output [dBu]
0dB (max output)753-0.25
End of white section751-0.27
Halfway white section504-3.73
Start of white section332-7.36
Dot 6210-11.3
Dot 5 (midpoint)101-17.7
Dot 446.1-24.5
Dot 316.2-33.6
Dot 24.02-45.7
Dot 11.09-57.0

For 9dB of headroom with 1Vrms max input devices like 3E audio DSP output level control should be set to start of white section (-7.36dBu)

For 9dB headroom with 2Vrms max input devices (consumer line level) output level should be set to max output (~8.5dB headroom) or mid white section (11dB headroom).

For devices that can accept high amplitude signals (mixing desk, audio interface etc.) the Master output should be set to maximum to optimise SNR.

Channel level red lights will engage the Rekordbox limiter even if the Master level control is below maximum as the Master level control just scales the DAC input data in this controller and the Rekordbox software has no knowledge of its setting. Therefore even if there is DAC headroom available too high channel level will cause the Rekordbox limiter to engage.

Frequency response:​

Frequency response measured using full range pink noise. Trim set midpoint, level set to maximum, 9dB mixer headroom. No differences noted at lower master out level than maximum.
1640880678366.png


Occasionally anomalous frequency response appears from DECK 1 on left and right channel. Not observed on DECK 2 and intermittent on DECK 1. Appears to be processing bug in Rekordbox? Using FOOBAR2000 to output the pink noise shows higher level but no anomalous behaviour.

Filters NXS2 ISOLATOR mode:​

1640880706231.png


Conclusions:​

The hardware potentially looks to have good enough performance (around CD quality) but there is some very odd behaviour such as the DECK1 frequency response issue and the spurs on the 1kHz test tone that suggest Firmware/Software issues. As my unit has SPDIF capability I will conduct some of the tests using digital output when I have SPDIF input hardware. I am also keen to test another unit that isn't modified.
 

AnalogSteph

Major Contributor
Joined
Nov 6, 2018
Messages
3,334
Likes
3,278
Location
.de
Perhaps the spurs are a hardware processing artefact or due to jitter?
Looks like jitter... ~90 Hz is an odd frequency, USB packet noise perhaps? I'm also fairly convinced that sidebands -70 dB down at 1 kHz would equate to a crapton of jitter. In a device like this I would be looking at speed / pitch related features as a potential culprit.

I notice spectral peaks centered around 48kHz so perhaps the internal sample rate of the DDJ-400 is related to 48KHz and we are seeing clocking errors or resampling?
Much simpler, these are clearly the images of your 1 kHz test tone after going through reconstruction + analog filtering. (Not sure what is being used as a DAC, but around -70 dB would match the filter performance of a CS4344 or similar.) Yes, clearly 48 kHz at work there.
 
OP
K

kipman725

Active Member
Joined
Jul 6, 2020
Messages
255
Likes
224
The side bands I don't think where caused by speed change features as they where also present when I used foobar2000 for playback. The speed changing is performed in software by Rekordbox and not on the controller.

I'm not sure if this is a useful test in this circumstance but I played back the REW J-TEST signal at 24bit (Rekordbox):
1641083574214.png


and 16bit (Rekordbox):
1641083488864.png


I also tried loading ASIO4ALL and using foobar2000 for playback of the 24bit signal:
1641084795255.png


Limited to two averages as REW has a bug only saving 60s long files.
 
OP
K

kipman725

Active Member
Joined
Jul 6, 2020
Messages
255
Likes
224
As a sanity check I re-ran the 1kHz test tone with ASIO4ALL and Foobar, the sidebands disapeared! (24bit WAV):
1641085185265.png


and without ASIO4 all:
1641085349129.png
 

Jakkookkaj

New Member
Joined
Feb 14, 2023
Messages
2
Likes
0
I am a little late to the party, but the problem you describe is a general problem using ALL pioneer hardware as a USB interface.
Even the new DJM V-10 seems to clip before the signal shows up on the channel mixer: https://www.reddit.com/r/PioneerDJ/comments/ux9g4o
I thought it was only rekordbox, but it seems all programs using ASIO drivers show clipping at the programs master led indicators.
Started a topic on Pioneer forums about it here: https://forums.pioneerdj.com/hc/en-...-USB-audio-interface-on-ALL-Pioneer-hardware-
 
OP
K

kipman725

Active Member
Joined
Jul 6, 2020
Messages
255
Likes
224
I am a little late to the party, but the problem you describe is a general problem using ALL pioneer hardware as a USB interface.
Even the new DJM V-10 seems to clip before the signal shows up on the channel mixer: https://www.reddit.com/r/PioneerDJ/comments/ux9g4o

I performed these measurements because I thought the DDJ-400 sounded worse than the stand alone pioneer mixer+decks. Some of it is very obvious such as the crippling of the low/highpass filters and reverb effect, but in the course of the measurements I have found very odd behavior:
1) Early limiter engagement unrelated to hardware headroom
2) sometimes unrequested filters are been applied (this one is especially suspect)
3) nasty side-bands that can be made to disappear using ASIO4ALL
I'm wondering is some of this deliberate so these controllers can't be used for 'serious' gigs?

in terms of mitigation, I think turning the channel trims down a bit should be fine? do these problems occour if using an external 4ch interface? (I haven't got one to test).
 

ofrappier

Active Member
Joined
Jul 20, 2018
Messages
122
Likes
74
Location
France

Intro:​

The DDJ-400 is an entry level DJ controller with master out (consumer line level single ended) and headphone out. Most of the functionality is through the REKORDBOX software with almost all controls and audio processing within that software with the exception of the Master level control. My controller is modified from stock in that the I2S lines to the DAC are also connected to an SPDIF/AES output board I have mounted internally to bypass the DAC. I don't think this will affect the DAC output as the twisted pairs to the SPDIF board are connected via 100ohm resistors but performance degradation is possible from this modification and I would like to check my results with a stock unit. The Firmware is the latest 1.03.

View attachment 175570
Tests have been conducted using my TASCAM US-2x2HR interface which you can see loopback of here. Mostly test tones (pink, 1kHz Sine) have been generated and are played back from Rekordbox 6.5.3 unless otherwise stated.

Output levels and THD+N:​

Clipping test:​

Mixer headroom set to NONE, 0dBFs sine wave on Deck 1, Master output set to FULL channel trim at +9dB, 2.239Vrms left channel, 2.234Vrms right channel (limiter has engaged), -13.4dBfs on input meters (9.22dBu).

View attachment 175571
Setting Mixer headroom to 9dB results in almost the same level of 8.45dBu suggesting we are running into Rekordbox limiter.

THD+1 1kHz normal output:​

Setting trim to mid point (0dB) and keeping output level at full and 9dB mixer headroom reduces level to -0.26dBu (752mV) the top red level light on the channel level meter on the controller is no longer lit. This seems an appropriate setting to use the controller in as the level has reduced by ~9dB as the trim setting indicates and so we are confident we have the specified mixer headroom before limiting.

View attachment 175573

Noise dominated (spurs of fundamental). With no signal the noise goes down the -101dBu. Perhaps the spurs are a hardware processing artefact or due to jitter? I notice spectral peaks centered around 48kHz so perhaps the internal sample rate of the DDJ-400 is related to 48KHz and we are seeing clocking errors or resampling?

View attachment 175574
The spurs are also present when FOOBAR200 is used for playback and so are not a Rekordbox playback issue.

Effect of output level control:​

Master Level PositionOuptut [mV]Output [dBu]
0dB (max output)753-0.25
End of white section751-0.27
Halfway white section504-3.73
Start of white section332-7.36
Dot 6210-11.3
Dot 5 (midpoint)101-17.7
Dot 446.1-24.5
Dot 316.2-33.6
Dot 24.02-45.7
Dot 11.09-57.0

For 9dB of headroom with 1Vrms max input devices like 3E audio DSP output level control should be set to start of white section (-7.36dBu)

For 9dB headroom with 2Vrms max input devices (consumer line level) output level should be set to max output (~8.5dB headroom) or mid white section (11dB headroom).

For devices that can accept high amplitude signals (mixing desk, audio interface etc.) the Master output should be set to maximum to optimise SNR.

Channel level red lights will engage the Rekordbox limiter even if the Master level control is below maximum as the Master level control just scales the DAC input data in this controller and the Rekordbox software has no knowledge of its setting. Therefore even if there is DAC headroom available too high channel level will cause the Rekordbox limiter to engage.

Frequency response:​

Frequency response measured using full range pink noise. Trim set midpoint, level set to maximum, 9dB mixer headroom. No differences noted at lower master out level than maximum.
View attachment 175575

Occasionally anomalous frequency response appears from DECK 1 on left and right channel. Not observed on DECK 2 and intermittent on DECK 1. Appears to be processing bug in Rekordbox? Using FOOBAR2000 to output the pink noise shows higher level but no anomalous behaviour.

Filters NXS2 ISOLATOR mode:​

View attachment 175576

Conclusions:​

The hardware potentially looks to have good enough performance (around CD quality) but there is some very odd behaviour such as the DECK1 frequency response issue and the spurs on the 1kHz test tone that suggest Firmware/Software issues. As my unit has SPDIF capability I will conduct some of the tests using digital output when I have SPDIF input hardware. I am also keen to test another unit that isn't modified.
Warning

This device is 48000 sample Rate only with Bad src when playing file at 44100.

Rew and all measurements was done with re war 48000 ? And Test Tone at 48000 ?
 
OP
K

kipman725

Active Member
Joined
Jul 6, 2020
Messages
255
Likes
224
The test tones where all 44.1kHz sample rate. REW was run asyncronously on another computer at 192kHz sample rate. Could be a resampling issue its posible that the clean spectrum was with 48KHz resampler on FOOBAR. Can investigate in a few weeks.
 
OP
K

kipman725

Active Member
Joined
Jul 6, 2020
Messages
255
Likes
224
Some users asked me about the SPDIF/AES mod. Here are some pictures I added this circuit board in tapped into the DAC I2S lines. Something from Aliexpress/ebay. The lines are tapped with twisted pairs cut to equal length with a resistor on one leg to reduce loading. One side of the twisted pair connects to ground and the other the signal of interest via the resistor. Can't recall the resistor size, probably about 100ohms. There must be a power connection in there somewhere as well. The plastic is cut away to fit the board in. I'm not sure I would bother though with all these processing bugs....
 

Attachments

  • IMG_20210411_010245152.jpg
    IMG_20210411_010245152.jpg
    268.2 KB · Views: 88
  • IMG_20210411_005835616.jpg
    IMG_20210411_005835616.jpg
    341.9 KB · Views: 92

ofrappier

Active Member
Joined
Jul 20, 2018
Messages
122
Likes
74
Location
France
The test tones where all 44.1kHz sample rate. REW was run asyncronously on another computer at 192kHz sample rate. Could be a resampling issue its posible that the clean spectrum was with 48KHz resampler on FOOBAR. Can investigate in a few weeks.
Hi,

Because i've made measurements with my DDJ1000 and Rekordbox 6.6 on Mini mac M2 pro (MAC OS), the SRC in Mac is perfect with Rekordbox.

Rekordbox works like a charm at 44100/48000/96000 on my OS Ventura. (On MAC, on Windows i don't know)

In the past i had the same jitter like you :

i've deleted the Rekordbox settings files, restart Rekordbox, and never add this Jitter.... (after all reset Rekordbox audio settings)

i suspect Master tempo /Key transpose shift.

and now i've good result without jitter even with -1 +1 key transpose, In Rekordbox. Strange.

Rekordbox have very good THD at any Samplerate on OS X, i will post result soon.

It is not the cas eof Serato, that have a non Working pitch in time (even when you don't use -1 + 1 key with pitch n time is activated )

See Serato Drama 'pitch' in time bug (that Rekordbox and VirtualDj 2023 don't have) :


It show that it's often a Software or OS SRC problem, before Hardware.
(And i trust more OS X SRC than Windows SRC... Check the software output with a virtual cable between the Dj software and REW)

Regards,

Olivier F.
 
Last edited:

Jakkookkaj

New Member
Joined
Feb 14, 2023
Messages
2
Likes
0
I performed these measurements because I thought the DDJ-400 sounded worse than the stand alone pioneer mixer+decks. Some of it is very obvious such as the crippling of the low/highpass filters and reverb effect, but in the course of the measurements I have found very odd behavior:
1) Early limiter engagement unrelated to hardware headroom
2) sometimes unrequested filters are been applied (this one is especially suspect)
3) nasty side-bands that can be made to disappear using ASIO4ALL
I'm wondering is some of this deliberate so these controllers can't be used for 'serious' gigs?

in terms of mitigation, I think turning the channel trims down a bit should be fine? do these problems occour if using an external 4ch interface? (I haven't got one to test).
So I have always felt like there was a big difference between DJing with a laptop using Pioneers internal audio interface with ASIO drivers, or using a usb stick in one of the pioneer CDJ decks, and your findings seem to confirm and explain that.
The early limiter engagement is especially strange. Rekordbox in external mode shows channel clipping, where all other Audio interfaces playing the same thing at the same volumes and settings show no such thing.
If we are not changing anything about a tracks volume, why is the track internally clipping in rekordbox, and why do no other USB interfaces show this behavior?

A kind of harshness and ear unfriendly effect in the higher frequencies that I noticed could also be explaiend by the nasty side-bands you are talking about.

Since I bought a Fireface UCX II, I no longer have any of these issues, I connect the fireface to the mixer with analog cables, and all seems solved.

I think the problem can be narrowed down to (ASIO) driver issues. The weird internal clipping seems to also happen in other software than rekordbox.
 
Top Bottom