Repository navigation
Player reports for baken expressport (out of beta since 4.3.0; CDJ-3000 and CDJ-2000NXS2 confirmed) #116
Description
Activity
- pinned this issue
on Sep 18, 2026 - added a commit that references this issue
on Sep 18, 2026 - changed the title
[-]Testers wanted: CDJ and XDJ owners, for an experimental USB export[/-][+]Testers wanted: CDJ and XDJ owners, for baken expressport (beta in v3.5.0)[/+]on Sep 18, 2026 First volunteer, from #114: @Alex2Code, who has CDJ-3000, CDJ-2000NXS2, CDJ-3000X and XDJ-AZ and offered to test when time allows.
Which of those answer what:
- CDJ-3000 and CDJ-2000NXS2 are the beta. These two are the go/no-go this thread exists for, and CDJ-3000 is the strictest reader of
export.pdbwe know of, so a pass there is worth more than a pass anywhere else. The checklist in the first post applies as written. - CDJ-3000X and XDJ-AZ are OneLibrary players, which the beta does not target. A result from them is still worth having, because nobody has measured what a OneLibrary player does when it finds a legacy library and no
exportLibrary.db: read it, read it partially, or refuse the stick. "Database not found" there is the expected outcome, not a bug report. That question now has its own issue, expressport: OneLibrary players (CDJ-3000X, XDJ-AZ, OPUS-QUAD, OMNIS-DUO) #139.
Same rule for all four: a spare stick, never the one you play from.
- CDJ-3000 and CDJ-2000NXS2 are the beta. These two are the go/no-go this thread exists for, and CDJ-3000 is the strictest reader of
A film for this call is built and its sources are committed (M-Igashi/baken-mac
378aff5,promo/testers/). Nine lines in English and Japanese, narration done with the cloned voice, every scene drawn so there is no screen recording in its pipeline.It is deliberately not a product film. The last scene is a custom call to action pointing at this issue rather than the shared end card, which sells the Mac App Store app, and the script leads with the thing that actually needs a human: the bytes already match a real rekordbox export, and nobody has put one in a player. "Use a spare stick, never the one you play from" is its own scene rather than a footnote.
It also states the offer plainly, which the written post does not: baken stays MIT and free, and a Mac tester gets a Bake'n Deck promo code.
Not published yet. Distribution is tracked in M-Igashi/baken-mac#29 along with the rest of the series.
Unrelated to the film, for anyone reading this issue from the app side: the app is not getting expressport in 1.0.2, which shipped today as a small release. M-Igashi/baken-mac#31 is the app surface and it deliberately does not start before this issue's second spike clears on real hardware, for the reason in its body: on a feature like this, being early with the claim is worse than being late with the feature.
- changed the title
[-]Testers wanted: CDJ and XDJ owners, for baken expressport (beta in v3.5.0)[/-][+]Testers wanted: CDJ and XDJ owners, for baken expressport (beta)[/+]on Sep 21, 2026 - changed the title
[-]Testers wanted: CDJ and XDJ owners, for baken expressport (beta)[/-][+]Testers wanted: CDJ and XDJ owners, for baken expressport (beta). This is the one thing left before release[/+]on Sep 22, 2026 I prepared a stick with this method to test if the CDJ3000 reads it correctly. I will test it this friday and come back to you with the results.
@Alex2Code Thanks, that is exactly step 1 of the list above, and the CDJ-3000 is the strictest reader we know of, so it is the most valuable single result there is. A stick prepared with 3.7.0 is fine to test as it is: 3.8.0 (out today) changes nothing for a copied-analysis stick, it only adds
--generate-analysisfor tracks rekordbox never analysed. No need to rebuild it.Two things worth doing while the stick is in the player, beyond the checklist: note whether the colour waveform appears immediately or after a moment (the moment would mean the player re-analysed), and check afterwards whether the player wrote any
ANLZ0001.*file next to ours underPIONEER/USBANLZ. If the stick is refused,PIONEER/rekordbox/export.pdb.bakandRBFLTR.DATfrom it are the two files that let me bisect.For the record, the volunteer table above now has both Friday tests: yours (step 1, CDJ-3000) and @sairutra's from #145 (step 2, a generated-analysis stick from a Mixxx library on Ubuntu, player to be confirmed). If both come back, expressport's beta question is answered from two directions in one day.
Reacted by Alex2CodeMost of the internal testing has been done. I restructured issues to clarify what is remaining for external testers. I expect the expressport bata to production would be in couple of weeks after final integration. Huge thanks to volunteers.
@sairutra , if you'd like to use Mac App (GUI, cache & preferences are saved in Mac), I'd give you a promo code for the complement.I tried today to get the .DAT files from a usb export of a friend that runs rekordbox 7 on MacOS. However, there was no DEVSETTINGS.DAT file on the USB.
I tried to use this .DAT file I found online. However it gave this message
Tomorrow, I have somebody coming by will bring a USB that has a Microsoft export (i believe). Otherwise, there might be somebody else that will help me with laptop and rekordbox on Microsoft. I am saying this as it might be something do with MacOS or maybe newer versions of rekordbox don't export DEVSETTINGS.DAT? (just a suspicion, not based on any knowledge).
@M-Igashi Would you maybe know why the DEVSETTINGS.DAT did not get exported to the USB? Also the CDJs that we have are CDJ-2000NXS2.
P.S. we also tried to look at the filesystem in MacOS if the file would be there with the rekordbox application but it was not.
I prepared 2 USB sticks for my tests tomorrow, one with the tracks already analysed in Rekordbox and one with tracks being analysed by baken with --generate-analysis, so I will test both tomorrow.
I got no errors in CLI when doing this command on my end.
I will test the 2 sticks on the following players: CDJ-2000NXS2, CDJ-3000 and CDJ-3000X.
Will come back with a detailed report for each stick on each player and will also upload the content of the USB sticks in its before and after state so it can be compared if the CDJs modified or wrote something to the stick.
@sairutra Thanks for the detail, the screenshot says exactly what happened:
DEVSETTING.DAT: too short (39 bytes). A real one is 140 bytes (a 104-byte header, a 32-byte payload, a CRC16 and two zero bytes), so the file from that GitHub repo is not a settings file at all, probably a broken download or a placeholder. The check is there on purpose: these four files are what the player loads as its preferences, and a stick that resets a CDJ to factory settings, or makes it refuse the stick, is worse than no stick. The message now says what size it expects (#157).Two things to look at on your friend's export:
- The name is
DEVSETTING.DAT, without the S, next toMYSETTING.DAT,MYSETTING2.DATandDJMMYSETTING.DATinPIONEER/at the root of the stick. rekordbox 7 on macOS writes all four when it exports; a player may rewriteDEVSETTING.DATon first use, but it does not remove it. If none of the four is there, see 2. - rekordbox only creates the files after Preferences > DJ System > My Settings has been opened once. On macOS they live in
~/Library/Application Support/Pioneer/rekordbox6/(rekordbox 7 still uses therekordbox6folder), not inside the app bundle. Your friend can open that preference pane once, export again, and either the stick'sPIONEER/folder or that local folder is a valid--settings-dir.
A Windows export works the same way, the format is identical, so tomorrow's stick is fine too. And a CDJ-2000NXS2 is one of the two players the beta is waiting on, so your result counts as step 1 or step 2 in the list above, depending on the stick.
@Alex2Code Perfect: those two sticks are steps 1 and 2 exactly, and the before-and-after copies of the stick are the most useful thing you can bring back, since they show whether the player wrote anything next to our files. Nothing in 4.0.0 or 4.0.1 touched
expressport, so the sticks you prepared are current.For anyone on a Mac: Bake'n Deck 1.1.0 is uploaded to TestFlight as build 122 and is waiting for Beta App Review; the External test group gets it automatically once that clears. It carries the same baken-core 4.0.1, an analysis cache that no longer needs a re-analysis when the ceiling changes, and the Delete Backup fix for ExFAT libraries. What to test and the join link are in #109. It does not include the USB export; that page waits for this issue.
- The name is
34 remaining items
- added a commit that references this issue
on Sep 28, 2026 @Alex2Code Thanks for checking. Since the CDJ prints whatever key name
export.pdbholds, I compared baken against a stick rekordbox itself exported with the key display set to Classic. rekordbox writes the names it displays (Cm,Ab, ...) and lists only the keys the exported tracks use.#190 now does the same: the key names on the stick are taken from the XML exactly as rekordbox spelled them, so a Classic export shows
Cmon the player and an Alphanumeric one shows5A. On a 35-track playlist, baken's Classic stick gave the same key names as rekordbox's own export of that playlist (the only two exceptions were tracks rekordbox had rewritten in a later export). #189 and #190 will both be in the next release.The E-8307 question from my earlier comment is still open, whenever you get a chance to try a stick exported by rekordbox itself.
baken 4.2.0 is out (Homebrew, crates.io, binaries; winget pending review): https://github.com/M-Igashi/baken/releases/tag/v4.2.0
For the tests here:
- Keys (Read Classic key notation in rbsort and expressport (#116) #189, expressport: write key names as the XML spells them (#116) #190): a Classic XML (
Cm,Ab) now gives the tracks their keys on the stick, and the player shows them the way the XML spells them, as on a stick rekordbox exports:Cmfrom a Classic XML,5Afrom an Alphanumeric one. - Stick format (expressport: warn before writing when the stick's filesystem or partition table rules out older players #184): before copying anything,
expressportwarns when the stick is exFAT, NTFS or APFS, or on a GPT disk. A CDJ-2000NXS2 or older does not mount exFAT or GPT. --generate-analysisnow runs alongside the copy (expressport: prepare tracks ahead of the stick writes #170). What is written to the stick is unchanged.
@Alex2Code this is the last check before the beta label comes off. Please write USB1 again with 4.2.0 from the Classic XML (the same command; audio already on the stick is not copied again) and check whether the key shows on the CDJ-2000NXS2 and the CDJ-3000, and in which form (
Cmor5A). If it does, the next patch release drops the beta (4.2.1 is already out with an unrelated--prunefix, #192, so 4.2.1 or later is just as good for this check). The E-8307 question is still open whenever you get to it.@sairutra 4.2.0 is the release to use for your checklist run.
- Keys (Read Classic key notation in rbsort and expressport (#116) #189, expressport: write key names as the XML spells them (#116) #190): a Classic XML (
TEST - baken 4.2.0 - CDJ-2000NXS2 on 1.85 - Ubuntu 24.04 - 1121 tracks
Context:
.xml generated from mixxx2rekordbox to be converted to CDJ-ready USB without rekordbox.
Steps taken:
1. Download 4.2.0 Linux-x86_64 version of baken 2. mount usb to desired location in filesystem (FAT32 MBR USB) 3. Ran following command “./baken expressport mixxx2rekordbox/rekordbox.xml --device /mnt/rekordbox-usb --playlist "playlist-name" --generate-analysis --no-settings –dry-run”Note: I put all tracks in one playlist to test entire library export, do not know if there is just an option to export entire .xml without defining a playlist.
Output of command:4. Ran same command without –dry-runOutput of command:
time command output:
( ./baken expressport ../mixxx2rkxbox-pr/rekordbox.xml --device --playlist ) 481.46s user 26.02s system 8% cpu 1:40:37.91 total
The checklist
Result Stick recognised (no "rekordbox database not found" or other message) ✅ Playlists appear with the right names, folders and order ✅ A track loads and plays; title, artist, BPM, key, colour, rating shown ✅ Beat grid on the beat ✅ Waveform shown, and it is the colour one, without a delay (a plain waveform first and the colour one later means the player is re-analysing: say so) waveform is plain blue one, but stays that way (has always been for me) Hot cues and memory cues where they belong ✅ Player comes up with your My Settings (play mode, quantize, hot cue colour), not its own; with --no-settingsit keeps its own and does not askused --no-settings and it was good Search by artist and by key works ✅ What the player changed or added under PIONEER/USBANLZ, compared with a copy from before (ANLZ0001.*files thatbakenwrote for tracks sharing a folder are normal)Forgot to copy before plugging in. Added file below that could be found with a diff of before copy --generate-analysisstick only: the same rows for a generated track, plus whether the phrase display behaves with no phrase data✅ --cdjsafestick only: a FLAC source, now an MP3, loads with grid and cues alignedDo, I also need to test this with my setup since all analysis comes from --generate-analysis and not from rekordbox, if so I can update later Based on later comment:
"What the player changed: without the before copy, this tells most of it: ls /mnt/rekordbox-usb/PIONEER (a CDJ folder, RBFLTR.DAT and LOG are the player's) and find /mnt/rekordbox-usb/PIONEER/USBANLZ -newer /mnt/rekordbox-usb/PIONEER/rekordbox/export.pdb | head (anything listed was rewritten by the player after the export)."Only RBLTR could be found, CDJ folder was empty and LOG could not be found on stick.
find /mnt/rekordbox-usb/PIONEER/USBANLZ -newer /mnt/rekordbox-usb/PIONEER/rekordbox/export.pdb | headdid not return any outputupdate 30-09-2026
Now used
find /mnt/rekordbox-usb/PIONEER -type f ! -newermt 2020-01-01instead and the result is the following:/mnt/rekordbox-usb/PIONEER/rekordbox/RBFLTR.DAT
Thank you, this is the first full-library stick (1121 tracks, every analysis generated) reported from a player. Could you add the player model and firmware to the report (the checklist asks for it; from your earlier runs I take it to be the CDJ-2000NXS2 on 1.85)?
On the numbers: 1:40:37 for 1121 tracks is 5.4 s per track, of which 0.43 s is baken's own work (481 s of CPU for the analysis; at 8% CPU the process was waiting for the stick the rest of the time). So about 5 s per track went to the stick, against 2.7 s per track for the 10 tracks in #173: sticks slow down as they fill and warm up, and
flushsends every closed file to the device at once (4 files and up to 4 folders per track). A re-run after a playlist change writes only what changed, as your 9 kept tracks show. The one thing that could still cut the first run on your side is the mount withoutflushfrom #173; #196 and #197 will not move it much, since your source is a local disk and you do not use--cdjsafe.Whole library: leaving out
--playlistexports every playlist in the XML, but there is no switch for every track in the collection; the one playlist with everything is the way to do it. If that is a common need with mixxx2rekordbox I can add one.Blue waveform: the colour is a per-stick setting the player stores itself, not part of the analysis (which does carry the colour waveform). With
--no-settingsthe stick has none, so the player uses its default, blue. On the CDJ-2000NXS2, with the stick in: [MENU/UTILITY], then [WAVE FORM COLOR] under [MY SETTINGS], choose RGB (manual DRI1290-A p.34, "Changing the color set in a storage device"). The player writes the choice to the stick, so it stays.--cdjsafe: not needed on an NXS2, which plays FLAC, AIFF and WAV; the row is for pre-NXS2 players, and it is about cue alignment after the transcode, not about where the analysis came from. Skip it unless you plan to use the stick on an older player.What the player changed: without the before copy, this tells most of it:
ls /mnt/rekordbox-usb/PIONEER(aCDJfolder,RBFLTR.DATandLOGare the player's) andfind /mnt/rekordbox-usb/PIONEER/USBANLZ -newer /mnt/rekordbox-usb/PIONEER/rekordbox/export.pdb | head(anything listed was rewritten by the player after the export).- added a commit that references this issue
on Sep 29, 2026 - changed the title
[-]Testers wanted: CDJ and XDJ owners, for baken expressport (beta). This is the one thing left before release[/-][+]Player reports for baken expressport (out of beta since 4.3.0; CDJ-3000 and CDJ-2000NXS2 confirmed)[/+]on Sep 29, 2026 Closing this with 4.3.0, which takes
expressportout of beta: https://github.com/M-Igashi/baken/releases/tag/v4.3.0What the beta waited for is in this thread: @Alex2Code's runs on a CDJ-3000 (3.22) and a CDJ-2000NXS2 (1.87), which found the NXS2 freeze (#168) and the blank Classic keys (#189, #190), and @sairutra's 4.2.0 stick with 1121 generated-analysis tracks on a CDJ-2000NXS2 (1.85), which showed keys and key search working. Thank you both; the README's expressport section now names those players.
Still open elsewhere: OneLibrary players (CDJ-3000X, XDJ-AZ, OPUS-QUAD, OMNIS-DUO) in #139, and the two rows nobody could fill yet (what a player changes under
PIONEER/USBANLZ, and--cdjsafegrid and cue alignment on a real player). Reports for other players or firmware are still welcome, here or through the issue form.Reacting on:
"Whole library: leaving out --playlist exports every playlist in the XML, but there is no switch for every track in the collection; the one playlist with everything is the way to do it. If that is a common need with mixxx2rekordbox I can add one."
- I indeed put everything in one playlist for these tests because had not tried to test out without using --playlist. - I now see it can be done. - I now also see that in mixxx2rekordbox it is a bit clunky to export all playlists so I will change [that](https://github.com/dimashenme/mixxx2rekordbox/issues/8) - What do you mean with this sentence though "but there is no switch for every track in the collection; the one playlist with everything is the way to do it" - I don't quite understand what you mean with a switch for every track in a one playlist stick
What the player changed: without the before copy, this tells most of it: ls /mnt/rekordbox-usb/PIONEER (a CDJ folder, RBFLTR.DAT and LOG are the player's) and find /mnt/rekordbox-usb/PIONEER/USBANLZ -newer /mnt/rekordbox-usb/PIONEER/rekordbox/export.pdb | head (anything listed was rewritten by the player after the export).
- updated the report with this. Also added the OS, and baken version etc.Thanks for adding the OS and versions to the report.
The "switch" sentence: sorry, that was unclear. "Switch" meant a command-line option.
baken expressportexports either the playlists you name (--playlist, repeat it for several) or, without--playlist, every playlist in the XML, each as its own playlist on the stick with all its tracks. What it cannot do is export tracks that are in<COLLECTION>but in no playlist, and an option for that is what I offered to add. mixxx2rekordbox only puts the tracks of the crates and playlists it exports into<COLLECTION>, so in its XML every playlist together already covers every track. Once dimashenme/mixxx2rekordbox#8 exports everything by default, this puts the whole library on the stick, with no extra option needed:./baken expressport rekordbox.xml --device /mnt/rekordbox-usb --generate-analysis --no-settingsWhat the player changed: the
RBFLTR.DATyou attached explains the emptyfind, and the mistake is mine. It saysCDJ-2000NXS21.85, but it is dated 2015-01-01 01:00: the player's clock is not set, so everything it writes on the stick is dated 2015, and-newer export.pdbcan never list those files. If the stick has not been exported to since, this lists the files that carry the player's date instead:find /mnt/rekordbox-usb/PIONEER -type f ! -newermt 2020-01-01If nothing under
USBANLZshows up, the player left the analysis alone, as on your 4.0.2 stick, where the before/after diff showed changes only toexport.pdb, the newCDJandMPJfolders andRBFLTR.DAT. It is optional: it only sees files whose date the player sets, so that diff remains the better evidence. The missingLOGfolder is worth noting too: Alex2Code's NXS2 on 1.87 wrotePIONEER/LOG/*-alog_1.jsonand yours on 1.85 wrote none on either stick, so that may be a firmware difference.Hot cues D to H: 4.3.1 writes them where rekordbox does (#205). If you get to try a few on the NXS2, a line here or in dimashenme/mixxx2rekordbox#7 settles it on hardware.
Switch
For the "switch", personally I tried to organize everything via crates and playlists in mixxx and if it is only in the library then for me it is too be sorted. So for me this limitation is fine as it works for my workflow. I will not change it in mixxx2rekordbox for now since there is no demand for it. However, if there are users in the future that would like such a feature then we can come back to this.
update 30-09-2026
mixxx2rekordbox now exports both crates and playlist in .xml and baken 4.3.1 correctly exports everything in .xml without --playlist flag.
Player changed files
I used the new
findcommand you pointed up above and added an update to the report.
Hotcues
I will test out now and give an update to corresponding issue in both baken and mixxx2rekordbox
@sairutra Thanks for both updates.
- Whole library: good to hear that the XML from mixxx2rekordbox, with crates and playlists, goes onto the stick with 4.3.1 without
--playlist. Agreed on the option for tracks in no playlist: it stays out until someone asks for it. - What the player changed:
RBFLTR.DATbeing the only file with the player's date means nothing underUSBANLZwas rewritten by your CDJ-2000NXS2 on 1.85, as on your 4.0.2 stick. That fills the row that was still open when this issue was closed, for a generated-analysis stick on that player. - Hot cues: confirmed in expressport: hot cues D to H are written where only A to C belong #205, thanks. Active loops are answered in expressport investigate if it is possible to add active loops to a rekordbox.xml file only export #210.
- Whole library: good to hear that the XML from mixxx2rekordbox, with crates and playlists, goes onto the stick with 4.3.1 without
Use a spare USB stick. Never test this with the stick you play from. The point is to find out what a player rejects, and a rejected stick at a gig is not a finding.
Where
expressportstands (updated 2026-09-29)baken expressportwrites a rekordbox USB export straight fromcollection.xml:export.pdb, the analysis files, the audio and your My Settings, without launching rekordbox and without reading its database. It has shipped as a beta since v3.5.0. The implementation is complete (design record: #115, closed). Every byte it writes has been checked against real rekordbox 7 exports: the database pages match, and the analysis files it copies are identical to what rekordbox put on a stick for hundreds of tracks.Where the tests stand. @Alex2Code ran a copied-analysis stick written by 4.1.1 on a CDJ-3000 (fw 3.22) and a CDJ-2000NXS2 (fw 1.87) on 2026-09-28: recognised, playlists in order, search, beat grid, waveform, tempo and hot cues fine on both, and no freeze on the NXS2 now that #168 aligns the strings. One row failed: the key was blank on both players, because the XML was exported with rekordbox's key display set to Classic and 4.1.1 read only Alphanumeric. 4.2.0 reads both and writes the key names the way rekordbox does (#189, #190). What the beta waits for now is a player showing the keys from a stick written by 4.2.0. This issue is the single place for reports; those from #114, #139 and #145 all land here.
Test with 4.2.0 or later. 4.1.0 writes
export.pdbstrings and the track count the way rekordbox does (#168) and numbers tracks that share an analysis folder (#176); 4.1.1 stops a track from taking another same-named file's analysis (#180); 4.2.0 reads Classic key names (Cm) as well as Alphanumeric ones (5A) and shows them on the player as rekordbox would, and warns when the stick's format rules out a player (#184). A stick written with an earlier version only needs the same command run again with 4.2.0; audio already on the stick is not copied again.What to test, in priority order
bakencopies its analysis. One pass on each of these two models takesexpressportout of beta.--generate-analysis, expressport: generate the analysis files, for libraries rekordbox has never analysed #147). For tracks rekordbox never analysed,bakencomputes the waveforms itself and leaves out the phrase analysis. Only a player can tell whether it minds. Test this only after step 1 passed on the same player, so a failure has one cause.bakenlooks for the analysis cache in%APPDATA%\Pioneer\rekordbox\share\PIONEER\USBANLZand the settings in%APPDATA%\Pioneer\rekordbox6. Both are inferred from the macOS layout and unconfirmed. Ifbaken expressport --dry-runfinds them on your Windows machine, say so; if not,--anlz-dirand--settings-dirfix it and the real paths fix it for everyone.--cdjsafestick on a pre-NXS2 player (CDJ-2000NXS, CDJ-900NXS, XDJ-1000): every track transcoded to 320 kbps CBR MP3 with the source track's analysis. Grid and cues must still line up.Not on the path to release, still worth a report: OneLibrary players (CDJ-3000X, XDJ-AZ, OPUS-QUAD, OMNIS-DUO). The beta writes the legacy library only. A CDJ-3000X shows
OneLibrary not found, as expected; the others are unmeasured, and the answers go to #139.Install
brew install M-Igashi/tap/baken(orbrew upgrade baken)winget install M-Igashi.bakencargo install bakenTake the newest release. ffmpeg is only needed for
--cdjsafe.Bake'n Deck for Mac 1.2.0 (Mac App Store, 2026-10-01) runs
baken-core4.4.0 and has a USB Export page on the same export code asexpressport. A player reading a stick the app wrote, rather than the CLI, is the check scheduled for 2026-10-07 (#109).The test
DEVSETTING.DATis optional, and--no-settingswrites none so the player keeps its own. Export the collection: File > Export Collection in xml format.bakennever deletes what players leave behind. If rekordbox is running, eject the stick inside rekordbox first, or quit rekordbox: it keepsexport.pdbopen while attached.--dry-runshows what would be written and which directories were found.--anlz-dirpoints at rekordbox'sPIONEER/USBANLZdirectory if auto-detection fails (macOS:~/Library/Pioneer/rekordbox/share/PIONEER/USBANLZ, orPIONEER/Master/share/PIONEER/USBANLZon a library drive).The checklist
Please state the player model and firmware version, your OS, the
bakenversion and roughly how many tracks. Then:--no-settingsit keeps its own and does not askPIONEER/USBANLZ, compared with a copy from before (ANLZ0001.*files thatbakenwrote for tracks sharing a folder are normal)--generate-analysisstick only: the same rows for a generated track, plus whether the phrase display behaves with no phrase data--cdjsafestick only: a FLAC source, now an MP3, loads with grid and cues alignedA photo of the player screen is worth more than a description of it.
If anything fails, please also attach what the player wrote to the stick:
PIONEER/rekordbox/export.pdb.bak, any analysis file underPIONEER/USBANLZthat the player changed or added, andPIONEER/rekordbox/RBFLTR.DAT. Those tell more than the error message does.What ends the beta
Known limits of the beta
PSSI) is copied when rekordbox made it and left out whenbakengenerates the analysis.E-8307 DEVICE NO RESPONSEon the first insertion of abakenstick (with 3.8.0 and 4.1.1), not on re-insertion. Whether a stick rekordbox exported does the same on that player is not known yet.Volunteers so far
OneLibrary not found. Retest with the #168-patchedexport.pdb: no freeze on the NXS2, ARTIST list drawn, My Settings loaded, grid, BPM, key and hot cues on USB1. Step 1 with 4.1.1 on the CDJ-3000 and the NXS2: everything but the key (Classic XML, fixed in 4.2.0); E-8307 on the CDJ-3000's first insertion. A key check with 4.2.0 nextTEMPO. A full-library stick with 4.2.0 nextTwo more players, or the same players with the other stick type, would still help.
What you get
bakenis MIT-licensed and free, and stays that way. If you run a real test on real hardware and you are on a Mac, say so and I will send you a promo code for Bake'n Deck for Mac, the App Store app built on this same core, as thanks for the trouble. Codes go out privately, never in this thread.