Thanks DJ Ricochet, you're correct that work is being done to find the root cause of this and measures to fix it are already being discussed. There will be a more thorough explanation and proposed schedule of fixing this next week, so keep an eye on this space!

is this an issue only on the PC platform? i haven't noticed this on my Mac. i actually have been able to pitchlock the same song for the entire length of the song without any noticeable drift. plus i have used SL on multiple 5 hr gigs without any problems with beatmatching. anyone else care to add any comments on this?

feniks,

The problem will not show itself if you are matching the same song at the same point. I have tested the vinyl quite a bit and the info is on the last few pages of the unofficial forum. Im not going to get into my hypothesis here because I want Serato to give an official response, but there is something up

mac or pc? also, one would think that the problem would show itself if you played the same song "pitchlocked" on both A and B decks and listened to see if they drift out of sync...no?

PC and MAC regardless - the root of the problem is thought to be possibly the pressing of the timecode on the vinyl which would make it a universal problem, but we'll see.

and yes if you read my original post you'll see what i suggested trying with 2 copies of the same track.

I can confirm that the effect is real, and is due to pitch variations introduced by the vinyl mastering process. It is possible to compensate for the pitch variations in software, which should be available as a free update in the next couple of weeks.

Our preliminary tests show the first 8 minutes of the 15 minute side have the most consistent pitch, so sticking to that section of the records will minimize the drift.

But are the CD present same problem ?
If the update compsate timecode of vinyl , it will report problem on CD ?

The CD has no problem. The software compensation will be switchable.

Shouldn't be much trouble to differentiate CDs from line level TT's... either that or just include a preference?

And please guys, there's no "timecode" anywhere in SSL! ^_^

Ok.

When we'll select Phono , compensation will be enable and when we select Line, compensation will be disable ?

At first I thought that would be the solution, but there's the problem that some TT's these days have switchable phono or line level output.

Ah yes !
A button near Line selctor will be enable compensation too then ?

Couin

Ah very good! The problem will be solved! Great news! Without that dirft the system will be nearly perfect! Is the drift at any timecode-vinyl or will be pressed new ones in the future that will not have that fault?


Pete

Josh~ Maybe one more time will do the trick ;)

For those who are interested, I am putting together an explanation of how Scratch LIVE determines absolute position on the vinyl record / CD, and how our system is superior to "timecode".

My german product brochure uses several time the term "timecode" :P

Shaun, how do I quote on this board? Did you get mail, btw?

Sounds like a mistranslation or misunderstanding by the German distributor.

Quotes are wrapped in [ quote ] and [ /quote ] without the spaces.

Steve, ok thx, the quote function seemed to be disabled before, was just wondered that it works now.

amen to that, looking forward to this and I agree that the system will be pretty much flawless with this fixed!

I have the impression of fine pitch decrasing with time (I must pitch + my SL1200 when playing to keep beatmatching with same mp3 in internal mode on the other deck), but more on the 10 minutes side than 15 minutes side.
Will the compensation be switchable between 3 postitions :
None - 10 Min side - 15 Min side
?

That's the idea. The plan is also for it to switch automatically between 10 and 15.

Huhuuu :-)
Then , I'm wainting for this update :-P
Couin

it can differentiate between different sides of the record?! interesting encoding of the vinyls then...

how so? I think steve means that it will automatically switch _when_ it knows it's definitely on the other side i.e it goes past 10 minutes or it hits the scroll section.

But before the 10 minutes, how does it reconize side of vinyl ?
Couin

When we were laying out the record we didn't think there would be any need to distinguish the two sides of the record, so the first 10 minutes of the two sides are the same. The 10 minute side has a scroll section which is unique to that side, and the last 5 minutes of the 15 minute side is unique. So for the software to know that you've flipped the record, you would need to briefly put the needle in one of those two areas.

Our location system is flexible enough that for the next record pressing we can easily make the two sides unique.

To reconize side ?
Okayy :-)

Next record will be locatable ? Like another label ? If I buy new records, I want to be sure to have new records, not old records.
But in first time, I would like to get update for correcting actual records ;-)))

Couin

Dont you need to disinguis between the two sides even in the first ten minutes as I guess that both are pressed in a diefferent way so the way and flutter (indocued by tha lathe+master) which needs to be compnsated through the software is most certainly different?

That's why...

Any news on this? an eta? just curious.. im pretty anxious to get this fixed lol :}

goodluck guys

i think u guys should fix this problem with an optional download... and also release the corrected vinyl that does not require a software download... if u downloaded the optional software fix and then got the corrected vinyl the software should know that the corrected vinyl is playing at anypoint on that vinyl and "switch off" the whatever the software is doing... am i making any sense?

software as in the downloaded fix.

i donno.. i dont think fixing this in the software is going to be easy.

This problem is still being worked on, although we've made some good progress IMO.

I think he means: what happens when the actual vinyl gets fixed and there is also a software fix what will happen with the software fix and fixed vinyl?

The software will know that the vinyl record is a different pressing.

nice

whats new? anything discloseable?

Basically we're proceeding as planned...

oww kayyy

What exactly do you want to know??? The rest of this thread since then seems to cover everything which we know and and are planning to fix this.

well like it says.. "proposed schedule of fixing"

any eta on new vinyls?
any eta on the fix? Steve mentioned a couple weeks and that time is coming up soon, is that still reasonable?

i for one cant really use SSL well till this is fixed so the timeframe is very important to some of us

Right my bad, I should have said 'proposed method of fixing' since that schedule by it's very nature has a changeable timescale...

as per the help faq:



The fixes are coming, but I can't say when exactly at this stage, if I could, I would have already.

If Steve's general estimate changes due to something cropping up, I'll make a post to let you know.

Rest assured, we appreciate you're anxious to get this fix and as such are working very very hard to get it out.

Will a software fix also increase the systems latency?

Will it require more processing?

Do you guys know exactly what the drift is on the faulty vinyls?

Also, will re deciphering the signal from the faulty to the correct pitch open up the possibility of more audio artifacts introduced?

I was discussing this today with some other serato owners and we were starting to feel a bit umcomfortable with the idea of a software fix. Let me know if you guys can answer any of these questions. Thanks

hey yuri,



nope



a verrrrry neglible amount, wouldn't be noticed.



yes



more? you get artifacts?

I am very interested

Me too ;-)

indeed

Well I am an FS user and have done been finding alot of the problems with FS that Stanton and NI will not admit to, but somehow fix. The thing called TIMECODE on the FS encoder records were pressed by a company called TRUTONE. I havent seen an SSL record but im pretty sure that if there is an issue with drifting, you all might be able to look for yourself..

Now if you hold your SSL vinyl up to the light and angle it to see the "THING THATS NOT TIMECODE" you might see lines on the record from the outter edge to the inner..Now if you are looking at the 33 side. You will notice these lines are spread wider apart but as you get more into the song.. or futher down the record.. you will see these lines become closer.This intern will make the current song playing ..speed up.. and the new song on the other deck is spinning the code that is further apart..and wont be able to catch up.. even thou you rub the platter to slow it down.. or twist the rod to speed it up.. it still drifts off.

if you look at the record in the light you will see lines like this \\\\\\\\ .. when in fact they should look like this ||||| all the way from the outter edge .. to the center.. that is evenly spaced "THING NOT CALLED TIMECODE".

now on the 45 side.. the signal is pressed closer together from start to finish.. even thou its still incorrect.. the signal is close enough not to pose a huge offset in the pitch.. so it will take longer for the songs to run away from eachother.. so thats why alot of you are probably spinning the 45 side .. am I right.

well I hope i havent wasted your time with this post...like I said I havent seen or touched the SSL records or system.. but this was a shot in the dark to try to help you SSL users out..


thanks for your time..

Thanks for the input. I dont know if you know, but its on the way getting fixed, Serato/Rane is aware of the problem, check

www.scratchlive.net

and

forums.seratoscratchlive.com

It has been confirmed that its the press of the vinyls. Your conclusion about the 45 rpm is not right, as there is no dedicated 45rmp control-signal side, its all done with _one_ control signal. But you are right that the relative pitch fluctuation playing at 45 rpm is of course less than playing at 33 rpm.

oh thats a cool feature..I dont know why FS even bothered making both 33 and 45 sides seeing that we only really need one lol..

thanks for your response. and not bashing me lol.. i was really expecting that.

DJ ESP, if you're talking about the pattern on a FS record that repeats every 1/4 inch or so, then take a look at some Scratch LIVE vinyl - it isn't there. There is no "timecode", just a 1kHz carrier tone with bandwidth limited random noise modulated onto it. The software comes with a copy of that noise, so when you drop the needle somewhere it can pattern-match the incoming noise against its copy in memory and very quickly determine the location of the needle.

The obvious advantage over timecode is that as the signal degrades, the pattern matching system is much less likely to deduce a false position than a digital timecode system, where even a 1 bit error is disastrous.

The problem with the record pressing is that the rotational speed of the lathe was not within spec, so there is a drift pressed into every record. We have measured the drift and are successfully compensating for it in software in the lab here at Serato.

also, do you know if you are able to use the SSL vinyl with FS interface?

You guys guys are awesome...like DJ Esp said, "I am an FS user and have done been finding alot of the problems with FS that Stanton and NI will not admit to" That's why I'm now a current SSL user!

thats cool.. so if there is say maybe a small scratch in the record.. the software compensates and for that and adds it in from memory?

yeah and nice to see someone actually responding with some actual answers and not..

email us with your platform. phone number. and description of problem..

and then get back .. check your wires.. clean your needles and records.. make sure you have the latest updates.

Wow that is very interesting! Thanks for enlightening us Steve!

So I guess you're saying it's like a time "pattern" rather than a time "code"... it's about the noise on the carrier signal at the current time as well as the noise around that point in time which determines the exact position within the song rather than every single point in time having a pre-determined "code". so there's built-in ability to make a "best guess" at where the needle is rather than instantly assuming the needle was moved if a single bit of bad-data is received?

thx for the info - and still checking anxiously for the fix!

No, I wouldn't call it a time "pattern" either. What is on the record has nothing to do with time. It is much more like a landscape, and the memory of it inside the software is like a map. Internally we sometimes refer to it as a NoiseMap.

It works exactly like what a person does when they are in unfamiliar territory and have a map handy. It looks around for local features and tries to find them on its map. As soon as it spots enough features to be sure of its location, it moves the play position to that place. From there it follows the map in the direction of travel and ignores slight variations, only changing again if it sees enough features that match somewhere else on the record.

Yeah random noise is about as far from "pattern" as you can get.

My sources say no.

um....i'd say so. The first day that I got my SLL set up, something fell on one of my records still unopened and put a crack in it that went from one side to the other. I freaked out because of my experiences w/ the crap-tacular timecode on final crash. I tested the record after i got everything set up and running and to my surprise, there was NO DISTORTION @ ALL when the needle went over that spot. No audio blurp, no slowdown, no nothing....just solid music. now that's impressive...
eder

Steve is the magic 8-ball.

LoL @

well yeah i guess it is if it were the case that it was a different random noise pattern every time you played it? but since it's always the same series of noise that gets played from the SSL vinyl then the "finding the place on the map" operation that you talked about is essentially matching the known "noise pattern" against the random landscape to find where it is...

sounds very cool in any case, and it sure works well!

hi is the the software fix going to be a standalone fix or part of an upgrade.. hence 1108?

It will be part of a free upgrade

yes, the version with the fix for drifting (currently in beta) is 1108.

will the version 1108 correct the problem of list sorting in import screen with Fat32 file system?

No, but it is fixed in 1.2

and do you know an approximative date for the release of 1.2?

yea, it's at the bottom of this thread:

www.scratchlive.net

lol

don't see it....thanks!

read it again :-p

Hi,

Because launching the software fix is taking so long, I have to ask:

Could it be possible for you to somehow document the "drifting" pattern? With that it could maybe be possible to apply that same (or actually - negative) pitch variation pattern to audio files e.g. with sound forge - and thereby eliminate the drifting.

Just a desperate thought. But what do you think?

Uh, ville e, do you really want to timestretch your mp3s wiht that pattern? That would fuck up all your mp3s! Plus its a lot of work+processing... I dont recommend it. The better fix is what they are trying internally, do eliminate the drifting by using the pitch variation pattern on the control signal.

Of course it would be better to do it internally - the only problem is that nobody knows when the fix comes out. I'm doing at least 2 house gigs a week with SSL, and I don't have cd players - that's why I need solutions fast. I've even thought of making own _driftless_ dubplates of SSL control signal. Time stretching the audio files would be less expensive...

i was also going to enquire to see how much it would cost to get the audio from the cd pressed onto vinyl. Trouble is, a dubplate wouldn't last that long, especially as it would get HOURS of constant use rather then the normal 5-10 minutes per set.

I'm waiting for the news on the fixed vinyl vinyls

s
;-)

don't you buy vinyl? *sniffs* smells like an mp3 pirate to me, matey.

Actually, he could have switched to buying CD's and converting them to MP3 to play in SSL. I've only bought a total of 5 vinyl records that I transposed to MP3, since I first was using FS. And if he doesn't have those songs on vinyl, only on CD, then that may be his troubles.

Hi SpinThis,

I've been actively djing now for four years and I've never pirated a single song. I have a major collection of vinyl - I think bigger than most of the average djs have. That collection has been built with thousands and thousands of euros.

The reason why serato is currently vital for me is that major part of the stuff I currently spin (club's musical policy: upfront vocal and tech house) I have in digital only format. Every single track has been bought from legal wav/mp3 shops with real money. And of course I also have many many producer friends giving me wav/cd promos.

I think calling me "mp3 pirate" was not very nice.

br,

ville

i apologize if that's the case. it's just that a lot of folks I know (or have talked to) are switching to digital because they don't have to buy music anymore where they can just download it. i think that's sad. plenty of labels (including, for example,hooj) have gotten jacked because people weren't buying vinyl (and very few actually bought the cd releases because they downloaded them). let's hope some of that changes.

os9? :-P

I know this issue is being actively worked on and I'm not trying to hassle you for a release date for the patch, but I would like an update on how it is going?

In your testing which is going on are you successfully finding that you can eliminate this drift and have tracks stay in synch for 1min+ without tweaking the platter by using the software patch?

thanks.

oops sorry - IE stuffed up and posted twice whilst showing me a blank screen. Glad that serato software works better than M$ software ;)

fixed... have you tried mozilla? it rocks.

Josh, I still get a blank screen after posting, using IE.

lol :D Proof that double posting bug exists ;)

i'm using apple safari and i'm getting that bug too...

the bugs gone... dope...

hey nik39, sorry I meant that I'd deleted ricochet's extra post, not solved the blank screen thing.

no word on any success so far with a software patch?

Is it possible for you guys to release betas/alphas? I would love to be your guinea pig and help you out with bugs! Specially the pitch drift that is driving me nuts after so many weeks :-D

At first I thought it wasn't a big deal, but switching back to regular vinyl after using Serato is almost like a dream without drift hehehe

ahh sorry - i found that you're updating this in the "Known Issues" section. Thanks for the updates

Drift correction is planned for release this Friday.

wooohoooo very good news!!

can't wait to try it out. :)

me too :DDD

After friday there will truly be no reason to own Final Scratch then...good stuff

cant wait to get my hands on ssl

For testing correction do you :

1) Record a vinyl with a turntable and make an mp3 of the song
2) Mount the mp3 on the two deck
3) Launch one deck in internal mode
4) Launch the over deck in abs mode
5) Sync the two deck onne time at start
6) Look if the two deck stay synchronized (normaly, yes ?)

?

Thanks !
Couin

Step 4 : With the same turntable of step 1
Couin

I dont wanna make you regret letting us know in advance when the upgrade is coming, but is there any way you can tell us what all is fixed on it? is it just the pitch drift? Im holding off on ssl til the bpm field issue gets fixed.... Also, any new toys to play with in this next batch?????

Yuuuuuuuupie!!!! Yuuuuuuhuiuuuu! Great! Great! Great! Yeeeeeeees! I´m soooo happy!
Friday I make holiday! :-)

LOL Pete is HAPPY FACE !!!!

kakfdiunfaingfnriarnjgrnojqowerf

yes...i take day off

@ couin
every tt has a drifting so ur test would not fold cause the internal mode would have no drifting and the tt itself would have a little bit.

Yes, but it is the same for a normal vinyl .

Check how much wow/drift you can get with a tech 1200 with quartz lock on, its less than 0.0025% !
But basically as Couin has pointed out, this is the way you should test it.

For exemple, if we have an little oscillation of pitch on a tt in one rotation, the global pitch is constant.
Witch an internal mode, @0% pitch , we have normaly the same global pitch ... and then , we can compare the two modes (with a little tolerance)

Thanks nik39 !
Globaly, I'll be satisfied if the same mp3 (made from the same mk2 and in internal mode and abs (or rel) mode can stay synchronised.

I *think* the BPM field issue is addressed in version 1.2. Hopefully, it will be released mid-September--which is approaching faster than I thought it would.

dynamight.... so is this update just the pitch drift issue then, you think?

Someone from the SSL team wrote in a different thread here that the next update will include the bpm field adress issue.

I just used the search feature of this forum for the first time, and I found a post by Sam --> www.scratchlive.net <-- from a few weeks ago. I'm not sure if that's the post you're referering to, nik, but it's not clear if "next release" means 1.1110 or 1.2.

I would bet my 5 cent in my pocket that it is in the next release... and I am not talking about 1.2 ... you wanna bet? :-)

hehe, you sound pretty confident about it, maybe I'll just have to take your word. If you're correct, I'll be a happy mo-fo because that is the only thing I've been patiently waiting for to be fixed!! :-D

me too.... I think im in the minority, but Im more excited about the bpm fix than the pitch drifting

No one wanna bet against me? :)

I think they'll add more features than only pitch drifting correction because a moderator wrote pitch drifting will be corrected in 1.1.08 ... and announced version is 1.1.10 ...

Couin !

Not me. Call it sixth sense :)

anyone have any speculation on autogain? I might have missed it, but I dont remember anyone at ssl saying that it was coming anytime soon....did they?

No exact date, IIRC. But in the meantime, check this thread

www.scratchlive.net

i actually had a question about that.... did i do something wrong, or does it use the comments field on the id3 tag for its autogain? I tried it on 5 tracks and they all had some code in the comments field, so I didnt use it because I didnt want it to erase my comments or cloud them at all..... did you notice that? or am i making it up?

It uses some mp3 tag field, but I have listed the tags on those files which I used mp3gain on, but they were not listed. Its unlikely that it would store it in the comment tag... But I cant tell you for sure. Winamp reads the adjusted volume, and I cant imagine it gets it from the comment field, but again, I am just assuming. Maybe someone else here knows details about it?

overall, my question is; has anyone used this app with mp3s that are extensively tagged and if so, have they noticed anything weird in itunes or ssl??? or does it just do its autogain thing and I wont know the difference?

I don't believe mp3gain uses tags for the actual adjustment - that's done by modifying gain settings within the audio data itself. Whether it writes tags for some other reason I don't know.

huhu !

What relation between pitch drifting and autogain ?

;-)

Couin

Radish, would you please either read my post or the links I provided before commenting on this issue? Its not that I put this links there just for nothing. They contain information, and yes, I also dont want my mp3s to be altered, or (god save) reencoded.

Check mp3gain.sourceforge.net
I dont know about the practical impacts, havent used mp3gain much yet.

heh, totally gotta be in the minority. I can beat match without knowing the proper BPM...but I can't really beat match when the song's flyin all over the place! :)

Does anyone know if the CD's that shipped with SSL will have a problem once they initiate this pitch flux patch? I'm thinkin of usin the control CD's at an outdoor party so I don't F-up my records.

I'm just worried cause theoretically the noisemap is the same on the control CD's and the record right?

Radish, check

www.hydrogenaudio.org

www.hydrogenaudio.org

djpetey, check

www.hydrogenaudio.org

www.hydrogenaudio.org

www.hydrogenaudio.org

mp3gain FAQ www.hydrogenaudio.org

BTW mp3gains help file is great.

It would make sense to have an option whether you want the pitch drift correction to be applied or not, for the case using CDs.



It is, but only the vinyls noisemap "wobbles" and drifts, due to the bad pressing.

don't mean to nit pick---not a bad pressing, a bad master
and the cd's are fine

You're right. I was referring to the whole process (recording master... etc.).

the issue isn't about needing the BPM to beatmix, but rather to have the BPM field of the id3 tag remain intact after SSL plays a song that contains BPM data.

SSL versions up to build 1107 truncate a BPM value of 11025 to 110 once the file is loaded. Some of us use other mp3 mixing applications other than SSL, and full functionality is lost b/c the BPM data has been altered. Other people simply don't like SSL changing any info in their tags...right, nik? :)

the pitch issue (from what im hearing) is mainly after the 7 minute mark or something like that? therefore, it doesnt affect me ever..... plus, im very used to spinning with pitch changes.... about half the records I spin dont have constant tempo..... this is why the bpm field is my first priority.... but, obviously, if both get fixed correctly, everyone can win....

I guess I just don't know how what kind of programing they're doing around it...my first guess is that their doing a reverse pitch whereby at a certain point in the noisemap, the program knows the drift happens at XX:XX in the timecode and thereby applies a negative drift to even out the problem...But this seems a little too easy, and further thought would suggest that the program is "listening" for the drift instead of prempting it.. Can a moderator elaborate on how the drift fix works?

Ahh, yea, makes sense if you're not spining a genre with constant pitch. But I think there are drift issues throughout the record to, and there's an overall drift on the record.

Yep.

Thankyou, these exactly match my memory (and description) of how mp3gain works. Nik39, please try not to be so dismissive. You asked how it worked, you said "Maybe someone else here knows details about it?" and I provided some more (correct) information. Maybe I should have gone into more detail, but I considered it off topic - sorry. I actually know a lot about this subject (I have written mp3 encoders/decoders in the past). As I commented, it works by adjusting the master gain in the audio portion of the file. It doesn't, as I again stated, use tags to control gain. One thing I did discover by reading the links you provided is that it writes a tag with the old gain value, so it can undo. This doesn't however contradict what I said.

Youre absolutely right, and I apologize, for being such a dick and sounding that harsh. What I thought you did, that did I, I didnt read properly. I thought your were saying the gain is adjusted by altering the audio data, which is not correct.

I just wanted to point out that the audio data itself (all the coeffs etc. you know about these better than I do) is not touched at all, which to me means that the audio data itself (I dont consider the gain setting to be a part of the audio data and the compression, which might be wrong from the mpeg audio layers definition point of view) is not touched, not in a way that can not be easily reverted.

Bottomline is, the sound quality wont be degraded, all the changes can be reverted. Thats not the case with reencoding.

radish, can you explain how mp3gain stores the old gain settings? As I understand it, it stores it somewhere in the tag fields. Is it in the comment field, and if so can this easily be undone?

Thanks for your help and pushing me back to the friendly track again ;-)

The drift, since it's pressed into the vinyl, is the same on every record, so the application doesn't "listen" and compensate, what we did was take numerous precise measurements of the drift (i.e. the reported position vs real-time) averaged them to cancel turntable wow and implemented compensation in software.

Ok, that's what I thought. Now that being said, I'll ask 2 more questions:

1) isn't the noisemap the same on both the CD's and Vinyl? (I know that the new vinyl's have a different noise map)

2) Assuming the above is true, would the CD's have drift due to the software compensation? or does the program know internally when you're playin control vinyl VS. control CD's?

1) correct

2) The drift compensation is switchable, using CDs you would want to turn it off...

But as per nik39s feature suggestion, we could release CDs with a new noisemap, thus making it recognizable...

Sweet.. you guys thought of everything. Now, my final question is this:
begin rant--

WHY THE HELL COULDN'T NI AND STANTON PUT THIS KIND OF THOUGHT INTO THEIR PRODUCT!!!!!!!!

--end rant :)

Of course I'm ever so happy to be an SSL customer, I just needed to get that last bit of anger outta my system ;)

haha we're like therapy for some of you guys I swear...

LOL...good one Josh.

Please proceed to the download area.