good question!
i was considering switching to a mac, but that has put me off the idea. i dont want to be dealing with any lag if at all possible,
and i have tried macs with the buffer as low as it will go, and the difference is noticeable, and feels very different to my pc with the buffer at the lowest setting

It must be all in your head j cue, all the next buffer up does is add 1 more ms. It's virtually impossible to notice. You see DJs like Revolution etc. absolutely tearing it up with SSL on a mac, and they have no problems, at all.

I still hope they get this thing sorted out though.

Peace

Mac have less system latency by design, so they have in fact less total latency than PC's have. The USB buffer size is not the same as the total system latency.

yeah, i hear ya man, but lets face it,, im not revolution!!! and it will be a long time before im half as good as him!!
he played at our venue once and totally tore shit up!! using a mac, so i guess what your saying is true,, but for me, its noticeable!

My understanding of this was that the usb buffer was directly related to the level of latency in Scratch Live, and that the actual latency was limited by the software, not the OS. Even if there is less syetem latency on a mac, does Scracth Live in OSX actually run at this minimum level...?

Perhaps someone could tell me more about how system latency in Scratch Live OSX compares to system latency in Scratch Live XP. Can someone provide some figures (ie actual latency of a windows machine running SL at lowest Buffer of 1 compared to actual latency of a mac running SL at lowest Buffer of 2).

I agree with j cue... I can tell the difference in latency between using XP on my Macbook Pro and OSX on the same machine.

SSL uses the OS... so any limitation from the OS will also affect SSL.



AFAIK the total system latency with the best setting is ms less on a mac than on a PC.

I would love to find out that you're right Nik... I preffer OSX but when using it with SL I have to put up with occasional drop-outs and it just feels slower responding to scratching than XP.

Perhaps I'm wrong but I had always believed that latency and buffer size in SL were directly related. Clearly latency is increased by raising the buffer but perhaps this relationship is not that simple.

I'm a bit confused. Nobspangle told me in another post that 'when the usb buffer is set all the way to the left (1), SL is using 1 ms for the usb bufer and the rest of the latency comes from the speed of the program and not the Operating System'. What I glean from this is that even if Macs offer a lower overall system latency, when the buffer is set to the left in Scratch Live, OSX cannot make Scratch Live run any faster than it is configured to. In other words SL is not necessarily running at the lowest capable latency for a macintosh.

Any further input would be greatly appreciated.

Well, it is. The higher the buffer, the higher the latency. But the overall latency is also determined by various buffers on the OS and USB handling. It seems like Mac OSX has less buffers resulting in an overall lower latency than windows.



AFAIK, that was not quite right. There are various buffers involved from the OS.

So Gnosis, j cue, I honestly beleive it's you guys knowing that the buffer is set higher in the software, making you guys think it feels slower, but it really isn't. I could be mistaken, but ANYTHING under 10ms is indestinguishable from the human ear. If what nik39 is saying is true about the Macs having an all around lower latency, then it should be on par with PCs.


Peace

I'm going to try SSL on my PC and see if I notice any difference.



Peace

Wow, didn't think I would be saying this, but SSL does perform better on my PC 0_0. No worries, until they fix the Mac issue, I'll be just fine using Windows until then :).


Peace

Scratch that, haha, it really doesn't perform better on the PC, it feels exactly the same. I was just feeling that placebo affect or w/e you call it. I think using it on a Mac might have a lower latency actually.

Peace

the 1 ms performance boost lol

I simply cannot find a post discussing this but I thought it was accepted that OSX ran Scratch Live with 2 ms more latency than XP. From searching the forum this seems to be a point of contention. Is anyone in a position to confirm or deny this?

Please, I'm not concerned with what is audible to the human ear.

What I'm really looking for is a measurement of latency from an intel-mac running Scratch Live in Mac OSX and also a measurement running in Windows XP via bootcamp on the same machine.

How to measure:
www.scratchlive.net

Has anyone run the above test to measure the latency of their Mac OS compared to that of XP?

bump

Can anyone please tell me...

What is the best possible system latency running Scratch Live on a mac?

What is the best possible system latency running Scratch Live on a pc?

8ms

9ms

Thanks Nik, but with respect, how did you get these figures? Again, I am getting conflicting information from you and other posts on the forum.



www.scratchlive.net

I can't find a result to similar testing for a mac anywhere on the forum. Have you done this Nik? Anyone?

It looks like I'll be measuring the latency myself but I won't have access to a studio for a couple of weeks.

I've measured it on a PC.

I'll try it out on the mac side in a few weeks... It will be interesting to compare the results comparing XP to OSX on the same machine.

In any case, my initial question still stands, can mac users still expect a future version of Scratch Live to impliment a buffer setting of 1? Any news?

Rane/Serato says in future releases, Mac users will be able to set the buffer size to 1.

Again, scratch-scratch that, I think I DO feel a difference between a mac and a pc. I'm going to stick with Windows until Mac users can set the USB buffer to 1.


Peace

i used vinyl for 13 years before i switched to serato's vinyl emulation. anything that will allow ssl to more fully capture the feel of vinyl, i fully appreciate. i can understand how improvements on latency could be a factor in making the feel of ssl more accurate to that of real records. i'm a hip hop dj. i scratch a lot. i've been incorporating all types of sratching into my sets. but i am reading things in this thread about people feeling as though latency is a factor in their ability to scratch. i just wanted to bring up the fact that a huge factor in how scratching sounds, is not just the record movement but also the use of the cross fader. the crossfader isn't effected by latency. another factor... if you're talking about repetitive rhythmic patterns ... well... if you are trying to execute a phrase, you're going to naturally compensate for any delay with your body movement. that doesn't mean you have to scratch any faster. you don't have to constantly play catch up with the computer. if i am moving my hand back and forth on the record with a baby scratch at 100 bpms, once i hear the pattern, i am not moving my hand any faster than if i were scratching a real record.
i don't know if i'm making myself clear. i just got off work. i'm gonna pass out. time to go to sleep.

airforce, with the latency setting to 1 in SSL, there is absolutely NO latency. So if you're used to the way vinyl feels, SSL will feel no different at all.


Peace

really? no latency? cool! thanx.

how can there be no latency? please explain in detail. there is around 8ms +-1 at best depending on mac or pc.

i do think its tiny and most people wont notice, i cant. but that statement of NO latency is misleading.

That's what I meant, there ofcourse is a minute latency of 7ms on the pc, and 8-9ms on the mac. What I meant by no latency is: you can't feel it or hear it. Technically it's there, but it feels exactly like vinyl in terms of response. 7ms or 1ms latency, it will feel exactly the same. 7ms of latency or regular vinyl, it will feel exactly the same.

Basically, no latency that makes it any different from regular vinyl.

Question to all SSL users: Does your sound stutter/jitter, as well as the wave form if you backspin the control vinyl too fast? On the setup screen grid, it shows the control signal image turning into a square like doohickey, must mean the signal isn't being read properly by the needle.

Anyone else exprience this?

Peace

When I said "minute", I don't mean minutes, seconds, hours. I mean minute (pronounced "minoot") as in extremely small.

Thus stutter thing only happens when I backspin very sharply and quickly, and even then there are plenty of times when it doesn't occur. When it does occur, it happens for a single second, sometimes less.

I should probably start a new thread, but this would be just a quicky. :)

Any info would be greatly appreciated.

Peace

Alright, so the stuttering thing seems to be gone, mostly. It's when I backspin, and the needle doesn't read the signal right, it has nothing to do with Serato, but the needle. It's all good now :).


Peace

I also applied more XP tweaks, which seemed to help a bit.

Rane Sereato Scratch Live is ace stuff :)

Peace

Hey .hz, just a thought but I'm guessing that the stuter you're experiencing (on a spin back) might be caused by the change in velocity on the control record causing the needle to slightly lift and you therefore loose some of the signal. Make sure that when you put your hand on the plater to perform the wind/spin back, that you don't bring your hand down too fast, this causes the opposing side of the record to lift and then causes the neddle to kinda bounce as the record spins back. Also check that your records are not at all warped or concaved.

peace

yeah, i agree with what your saying,, but if the latency is higher, its gonna take you longer to get used to the catch up movements your body has to make to compensate, the higher the latency, the earlier you have to start compensating,,ya feel me?
and obviously if teh latency is lower,, this comes easier

Hey, thank you very much Gnosis :). Absolutely LOVE to learn about this stuff! I was trying different techniques of backspinning, and it definately has to do with the needle reading the groove correctly.

Couldn't agree more with j cue on this one, it just doesn't feel natural to compensate for latency. You strait up shouldn't have to.

With Seratos minimum latency being so excellently low, you won't have to compensate for any noticable delay :-).


Peace

Ok, now, re-re-re-scratch that, I honestly don't notice much of a difference, if at all, when using SSL on a mac. I keep switching back and forth, and they feel identical. I think it is all in your head j cue, because anything under 10ms is undetectable by the human brain, and the latency of SSL on lowest USB buffer setting while using it on a mac is 8ms.

There really is no difference in feeling, atleast for me :).

Gnosis, did you try SSL on a mac yet? If so, what's your take?


Peace

how many times are you gonna change your mind? i was being sarcastic when i said "no latency? really?" it obviously doesn't feel like vinyl, it doesn't sound like vinyl , it doesn't feel like vinyl. latency or no latency. it's a f@c%ing computer playing mp3s. but it's damn close to vinyl. if you want to sound better as a dj, just practice more, and stop worring about latency. DMC doesn't allow you to use a laptop in competition last time i checked. the point being , if you need to do extra technical shit and have it feel exactly like records, use real records. if you wanna do extra technical shit with a laptop, practice on the laptop and accept that it has a different feel.
once you've djed long enough with serato you'll get used to it. and they're always making improvements to the software and to computers and their processing power.
who knows maybe the guys at serato will get the buffer setting to negative 2 years and then you'll be able to play records before they're even made. you'll be able to dj as if you've practiced for 2 more years...

Hey dick, don't come at me with your dumb ass atitude. DJs wouldn't be using it if it didn't give them the vinyl feel they wanted. It replicates vinyl almost perfectly. I have been practicing scratcing for nearly 5 years, and I really can't tell a difference, unless you really look for it. This whole discussion is about latency, so that's what were going to debate about. Got it?

If you have a problem with it, go piss to your diary about it.

Peace

Simple reaction time is the time it takes to react to stimuli. The average human's reaction time falls somewhere between 200 and 270 milliseconds, although athletes and others who train themselves can achieve reaction times approaching 150 milliseconds.

humanbenchmark.com

u don't even know what a millisecond is

Facts:

* There is latency.
* There are people who can feel the latency. There are people who can not feel any latency.
* It sounds different than vinyl.
* There are people who can hear the difference. There are people who can not hear any difference.

No need to call names.

cappinkirk, don't assume something about someone you don't know. I know what a milisecond is buddy. Look up the facts about the human ear taking roughly 10ms to actualy hear the sound once it's generated. Human reaction time is a WHOLE different thing.

I agree completely with those facts nik. Sounding different than vinyl is a relative term, though it does differ a little bit, but if the quality of the sound file your using is high enough, and depending what genre of music your playing, it sounds nearly identical.

There is latency, but it's virtually impossible to feel it when set at an USB buffer of 1 or 2.

Fact:

If someone acts like a dick, there will be a reaction, with low latency ;).


Peace

Gnosis, you still there man?


Peace

hehe lighten up buddy

Done


Peace

Hz, no offense, but you need to freshen up your facts...

1st. The complete system latency is not 1 or 2, when the buffer is set at 1 or 2.

2nd. You are right when you say "the human ear taking roughly 10ms to actualy hear the sound once it's generated." but this is something different here. There is the latency between what you actually do with the hand and what you perceive through your ears. Anyway, you should read this thread which explains the problem with the latency... scratchlive.net <- click:

that is a VERY slippery statement...

first of all you are referring to using the outer ring of the record to get the .5" movement (with this same movement the inner ring would have only moved less than .125" for example) & few dj's skratch with the outer ring of the record.

also if 50 cm = 1 sec. then i millisecond (1,000th of a second), then 1 ms = .05mm, not .5 (huge difference in the math)

correct?

no u r correct i changed cm to mm. .05 cm = .5mm

Steve W is correct there

(eats crow)

mmmmmm tasty

It doesn't make a difference.

50cm/1s =
50cm/1s*(1000/1000) =
(50cm/1000)/(1s/1000) =
(50cm/1000)/(1ms) =
0.05cm / 1ms =
0.5mm / 1ms

do the same math with the inner diameter of a record! you will notice that the results will be much lower. The diameter should be figured from the center if you're going to do that equation in my opinion. it does make a difference.

Let's assume you're doing fast chirps (= 4 chirps per beat) at 120 BPM. That's really fast scratching.

1 beat = 60 sec/min / 120 BPM = 0.5 sec per beat

4 chirps per beat => 0.5 s / 4 = 0.125 sec per chirp

Fwd/bwd movement each take 0.125 sec / 2 = 0.0625 sec = 62.5 ms

BOTTOMLINE (you don't have to read the calculation):

1 ms is only 1.6 % of the record movement (in either direction) when you do really fast chirps.

That's not very much and I doubt that is noticeable.

When I compare both the minimum and maximum latency setting in SSL which is a difference of 20 ms if I remember it correctly I can feel a difference though.

But for many people (including me) even buffer at 20 ms is fine.

36" on the outside but only 14.9 on the inner ring. that's 40 cm per rotation, about 16cm per second (1/3 of 50cm) each ms then = .16mm

oops i should have said 24cm per second, 1/2 of Steve W's math

.24mm=1ms

not trying to be a h8r, just trying to be accurate.

This is my personal opinion and NOT fact but I think the USB issue goes back to the early days of Mac OS X.

Although Apple was the first to really incorporate and PUSH USB with the CRT iMacs with no floppy drive around 2000 (when all PC's still had a floppy and USB was optional), they were also working on competing technology called Firewire. It has been rumored that Apple intentionally crippled USB to help advance Firewire Technology.

I think that legacy cripple code, although improved is still built into the Mac OS so today's USB doesn't perform as well as expected. Compare booting and mounting USB drives on a Mac .vs Firewire drives.

To answer the OP, the problem lies in the Mac OS USB architecture. Something we obviously don't have much control over.

I better edit that known issue, we were just checking you guys were on the ball :-P

cappincirk, I didn't mean it does not make a difference (Though I said it ;) )... what I meant was, yes the travelled distance on the outer ring is higher than on the inner ring, but the amount of audio passed (what matters) during that travelled distance is the same.

word just trying to push the truth

Wow, I love this thread! Pretty awesome to learn about this technology. Anyways, SSL feels just like the PC. While there is latency, it's low enough to not be noticable to make the feel differ from vinyl.

Nik, I know that when you set the USB buffer to Xms of latency, that doesn't equal the total output latency. I beleive it's the time it takes to communicate with the SL1 box and back. True, no?

To the Serato team: You guys say in your statement about the USB buffer (located in the 'Known Issues' forum) that you will have the USB buffer down to 1ms in future updates. Is this still going to happen?

Doesn't really make a difference, though would be another "abnormality" ironed out :).

Peace

Hm, I don't think so. But I am sure Josh can explain it better than I could :)

So Nik, tell me, what's your preffered DVS choice? I know you tested TS and Torq, and own SSL, if you were to be able to obtain one of those for free (legally ;)), which one would you pick if you didn't already own any of those, but still retain the the experience you already have of them?


Peace

TS sounds promising - but they ain't ready yet. Today I've read about a (sort of) bug which reverses playback while playing if the needle has dust.

Torq is a playing toy IMHO.

SSL lacks a lot of features, esp. the crate+library organization needs a looot of work. Some of the crate+library functions which I miss, I coded an own application for it (ScratchTools free, BTW) so it is not biggi for me anymore. SSL is the easiest to use app, and so far it has proven to be the most stable one. I had bad experiences with former NI/Final Scratch software, Torq is just very ugly to use IMHO - no way I will use this thing during a club gig.

Besides this I am using the TTM57 mixer. A lot of promised functions are still missing - but the concepts of the mixer are great, the integration with SSL just dope.

So personally - SSL is my choice. You, yourself, should try all three applications and check which one suits your needs best.

BTW, I did not only test all three, I still own Torq, TS, FS2 (FS1 has been sold) and SSL. ;)

DJ Revolution for example used to be a SSL users and now demonstrates and promotes Torq. I can't tell whether it's just for the money or whether he truely is convinced that Torq is the better product. Guess we'll never find out ;)

I was a former FS2 user before switching over to SSL, and am going to definately stick with SSL for my DVS system. It is the best at what it does. Period. Solid, simple, stable, that's why I love it.

I think the crate/library system is one of the best out there, what improvements are you looking for in it?


Peace

Check the feature suggestions area, there are lots of features which need to be added.

Wow! Good to see some interest in this topic.

Thanks for geting back to me Josh. I'm glad I finally have some real information to pass on to Apple. As I said, I have been speaking with a rep who's card reads, "National Professional Development Manager". This guy was shocked to see me using XP on my Macbook Pro and has offered to look into it. If the issue of lowering the buffer is in the hands of Apple, I would like to take this opportunity to let him/them know about it.

By the way Josh, in helping me pass on the best possible information, can you please describe the relationship between the buffer setting in scratch live and actual system latency. I guess what I'd like to know is... If I were to put the buffer setting to 1 in XP, would I achieve the same 'system latency' as if I were running OSX at the same minimum setting of 1? Nik39 has encouraged me to believe that this is not the case becasue Macs have less system latency by design... Is he right?

I'm am looking forward to testing the latency using pro tools to compare the latency at best settings in XP and OSX but as I said, I have no studio for a while because I'm moving. The sooner I can pass on this letter to Apple the better... I just need confirmation that there is actually an issue. Do macs run Scratch Live with the same overall latency as Windows?

Thanks

Wow, this is absolutely terrific news! Eager to hear what Josh has to say. Either way, it doesn't really matter becuase anything under 10ms is going to feel the same.

Running SSL on a Mac with the USB buffer set to two results in an overall latency of 8ms, so it feels exactly the same as running it on Windows with the USB buffer set to one.

Still very interested in how the system latency being lower in macs affects the performance of SSL.

Peace

Just curious... did you read the linked threads?

this is somewhat related to the circumference discussion - are the inner rings of the record of lower quality sonically than the outer rings (like real records)? how is groove curvature handled with serato? does anyone have any facts on this?

Some of it. What I meant was having the USB buffer setting on 1 on Windows and 2 on the Mac will feel same.

I wouldn't imagine so because SSL reading the same "Noisemap" signal, albeit slight differences to tell the position of the needle. They may have made the signal a little stronger in the inner part of the record to compensate for the groove being wound a little tighter, if that's even how it works, which I imagine it is.

Peace

Well, back to using SSL on the Windows machine. It's also convenient because it's a laptop, but SSL just seems to run smoother on Windows, at least this Toshiba laptop compared to my dual G5.

I tested the latency setting on 2 in Windows, and again, was damn near impossible to tell a difference compared to 1, but I keep it on 1 anyways :)

I hope the Serato team and Apple get the USB buffer situation worked out.

Josh, any further input on this?

Peace

Still there Josh?

Please, buffer and latency, how would you describe the relationship?

Back on the Mac, stickin' with the Mac.


Peace

Well, back to the vastly stripped down Windows XP lappy, and will probably stick with that until I get a Mac lappy.

What the hell happened to this thread anywas? Someone take it out back with a 12 guage?

Ok, so mods, any info on the whole Mac USB vs Windows USB buffer whoobadazzle?


Peace

In my own opinion, my MAC is still faster and tighter with Serato and a buffer setting at 1 than my VAIO was with it set at zero. I don't know what the significance is but I dont think there is a direct correlation between USB buffer settings as far as directly comparing MAC and PC operating systems to each other.

Still lovin' the MAC and I won't go back!!!

Haha, I'm with ya on that one brutha. Swichted back to a mac, a lappy :D. I'm definately, finally, fro' Izzle, stickin' w/ the 'tosh. OS X all the way!


Peace

Well, there went the OS X lappy. It was actually a cracked OS X version running on unsupported hardware, which was pure suckage. OS X is a beast on supported hardware. Unsupported - not so much.

Yup, at the moment, using Windows. When I get the chance, and I have the dough, I will probably opt for a Mac lappy.

Well, so any news on the whole USB thing? It seems like it's pretty much dead. After switching back and forth so much, I must say I can honestly not tell a different between the two.



Peace

Well, back to the mac, I think I will stick with mac for sure now. Everything is just so much smoother to use.



Peace