Impressions, bugs and feature suggestion (1.2.1)

Certainly not a bad idea, but then again should you really HAVE to read a book to figure out this kind of thing?
---
As long as this thread has taken a turn in this direction, I have a related question: it seems to me if you manually set a beat grid but don't lock it that Traktor may have its way with the BPM and/or grid the next time that track gets loaded. Is this consistent with what others are seeing? As best I can figure, it's probably related to the autogrid function which, as far as I can tell doesn't automatically grid my tracks anyways, so maybe should just be turned off completely...

 

>>Traktor may have its way with the BPM and/or grid the next tim

What do you mean? is Traktor moving the beatmarker or changing the BPM the next time you load the track?

 

As far as I can tell, yes it is doing one or both. Eventually, I'm sure I'll be able to tell you what the behavior is exactly, but so far, I only realized it in the middle of a gig so I didn't have time to do anything but fix the problem...

 

That sound strange, but interesting. Would like to hear more when you had time checking it....

 

Just wanted to throw my 2 cents in here.

TPRO QUANT ON: Findings and Opinion

I've been sticking with T3 Version 3.22 for a variety of reasons, but found myself "back on the fence" in terms of eval (sound quality / work flow) with the latest version of TPro.

So I took a popular tech house track in WAV format and played / recorded it "flat" in TPro (no EQ, no FX, 0db levels, no KEYLOCK, no QUANT etc) and then played it (inverted phase) together with the orig file in Ableton L8 (no warping). To my surprise / delight: they perfectly phase canceled each other (no music got through). Thus, TPro is indeed "bit transparent" (with no EQ, no FX, 0db levels, no KEY LOCK, no QUANT).

Since Hot Cues (on-the-beat) is such an important part of TPro (versus T3 not having this feature) I then tried the same test (playing / recording same track) with TPro QUANT ON.

Well, when run (inverted phase) against the original WAV track - TPro with QUANT does not phase cancel against the orig track, and you can hear probably 70-80% of the track coming through (meaning there are differences between the orig source and resultant QUANT ON tracks).

I did make sure that the start points of the tracks were the same and even played with Ableton track delay in fractions of millisecs (ms) to make sure that wasn't the issue.

Thus my thinking is that for NI to give us the great feature of QUANT (and allow on-the-beat hot cues, loops, etc): there is a "slicing" (or time slicing) of the source audio signal. While technically I understand this may need to be done (as well as to deal with keylock, pitch and sync demands) my preliminary tests are showing me that TPro with QUANT ON is changing / "degrading ?" the orig audio.

Now maybe I am an "old timer" being 45 years old and was an analog recording engineer during the '80s. But there used to be something called "soul" / "warmth" in the sonic quality of music. That's still my goal. And hey, I'm currently into BreakCore, DnB and DubStep so I'm not talking old 60's/70's RnB and Soul music - I'm talking about the sonic warmth of the sound.

Take Tim Exile's recent NI FX release The Finger and play through the demos. While they / he has done a great job with glitch, mangling, etc -- I can't get it out of my head that NI (this generation) is more into what they can do technically (mathematically / electronically) then any concerns about maintaining sonic quality / warmth.

I originally thought that QUANT's algorithm would be a way to calculate (only) where the beat points are (for hot cues) in TPro, but now I think the algorithm is time slicing and something (audio) is getting lost in the process ?

Not trying to create a flame. I am happy that NI continues to improve / work on TPro. I know that rainerh and gmint_Tf are strong supporters of NI putting out the best possible Traktor product, so I wanted to add my opinion / findings here.

Any thoughts / concerns (by others) on sonic quality issues being related / caused by TPro QUANT ON ?

I think the sonic changes by QUANT are so subtle that most people don't hear the difference. But running the tests with an inverted phase copy with the orig source track in a DAW shows that there is something significant going on when QUANT is on.

 

@iwaiwa

You put quite a bit of effort into that analysis. Your research is very interesting. I'll be interested to see what, if anything NI has to say regarding your hypothesis...

After my gigs this weekend, I have a couple of new issues that I've come across:

1) I thought I observed this last weekend, but I chalked it up to something (that I'll mention later) in my keymap, however weekend, I'm 99% certain that it was independent of that issue. Basically, I loaded a track into a non-playing, non-synced, non-master deck and the tempo of the song was radically (20-30%) shifted. As far as I can recall, it only happened once, but that's enough. I don't know if anyone else has seen this.

2) There is an issue with the hotkey assignment for Sync. Without going too much into the reasons why I did this, I basically setup my map so the same hotkey that loads a track into a deck will simultaneously turn off Sync (Set to Value "0). Unfortunately, it seems that sometimes this command actually ACTIVATES the sync function before turning it off (kind of like a quick toggle). That's a problem because if the other deck is set to master, it will cause the track I just loaded to jump to the tempo of the Master track--i.e. exactly what I don't want to happen.

3) I'm probably going to post this to my Master/Slave thread because I consider it another HUGE problem with that system, but for completeness, I'll mention it here too... As far as I can tell, some (if not all) of the effects are quantized or otherwise synced to the beatgrid of a track. [This is just from observation, so correct me if I'm mistaken]. However, as we all know, not every track can be beat gridded. Now, there are obviously some ways around that but here is the real issue: let's say that I've put a beat grid on a track and it's accurate at the beginning of the track but because of drift, it's not accurate at the end. Usually, that might not be a problem because I'll just use the accurate grid at the beginning to set my loop (I might set another grid at the end for a loop as well). But, for example let's say that the grid is not accurate at the end of the track. Now, when I go to activate an effect (like the gater in my case) the effect is totally off because it's tied to the inaccurate grid. That sucks.

I've never been one of the people who has asked for flexible beat grids (for various reasons), but I really feel that what NI has done is to try to be like Ableton, but without a KEY piece of the always-in-sync piece of the puzzle--flexible grids. When you put the constraint of a fixed tempo grid on this system, there are just so many areas where things fall apart, ESPECIALLY when it comes to older or imperfect tracks. I hate to say it, but this really to the point where it feels like playing anything other than electronic music on Traktor means fighting the software. Which, if you think about it is really stupid. I mean, it is a true novice that can't mix a couple of perfect-tempo 128 BPM tracks together without screwing up. It's a little harder to mix something that has significant variations in the beat. But, as it stands Traktor makes it EASIER to do the simple task, but HARDER to do the more difficult task. To me, that's backwards.
---
Oh, and here's one feature suggestion:

- A hotkey/MIDI command that will sort cues so that they are chronologically located on a track.

I always find it odd (although understandable) that cue #2 can be prior (on the track) to cue #1. However, it would be nice if there were a way to fix that other than going back and deleting and resetting cues...

 

I thought I'd expand this suggestion a bit, since I just noticed last week (and I don't know how it has managed evade me thus far) that - interestingly - it seems like Traktor already does this "checking" while fetching an iTunes playlist... the items that already exist in my collection have a small symbol indicating it.

Why then is it that it can not show the "Traktor-data" (key, analyzed-status, cover image, etc.) along with the data that is being fetched from iTunes?

You know... just saying

 

i've seen it but couldnt reproduce it reliably so far. but i'm into it.

 

Thanks for the confirmation. It's good to know someone else can confirm that it's happened to them as well...

 

Another weekend, more feedback--

1) Audio Glitches -- Both my brother (on TPro for the first time this weekend) and I experienced audio glitches this weekend. I can recall at least two and my brother indicated that he had approximately one per hour. I *thought* I had heard glitches before, but I wasn't sure until this weekend. We're both on very recent MacBook Pros with 4 Gigs of RAM running on Mackie D.4 mixers. I think it goes without saying that I'm disappointed that this issue is back again. I suspect that this is probably related to the new deck caching mechanism or the new time stretching algorithm, but whatever the case, it's obviously a problem.

2) FX -- Again, I'm going to reiterate what I set out in my previous post. I think it is (again) a huge oversight that the FX units are tied to the beatgrid of a song. Here again, when the grid is on, the FX sound great, but when the grid is not perfect, the effects can sound like complete trash. NI, please, at least give us the OPTION to just have the effect START when it is turned on (like in T3) not when Traktor thinks it should start. PLEASE!

 

2) Aren't the FX tied to the master tempo rather than the beatgrid? (Semantics probably) Would tap tempo help for tracks with an imperfect beatgrid?

 

Yes and no. I gave a more lengthly description of the problem in my thread entitled Master/Slave, but basically the issue (so far as I can ascertain) is this:

In 3.4, let's say you turned on the delay. You hit the on button and the delay starts filling up from the time that you hit the button. The long and short is that the delay essentially "starts" from the time that you turn it on.

--Note: I'm pretty certain about the foregoing. As for the following, this is my best guess. Unless someone from NI confirms or denies, this is as good as it gets.

On the other hand, Pro tries to be smart. What it does is it uses the beat grid to determine when the effect should start (i.e. on the beat). However, if the grid is off, the effect is off too. Unfortunately, this makes the effects a real crap shoot especially when it comes to older songs. Which, in turn, sucks because transitions between older songs is where I personally find the effects to be most useful.

Again, it's another instance where I want to at least have the option to turn off the automatic....

 

FX are tied to the Master Clock.

 

Which is the set using the beatgrid as a reference no?

 

I am with everybody on these ones: Cueing issues, back suing issues, sound glitches, loss of control on decks, etc. I really like the performance of version 1.2.1 but this software is just too darn buggy. i am downgrading to version 1.1 until NI fixes these issues.

 

yes and no.
you could use INT mode and dial in a tempo which equals the real tempo of the track.

 

Yes, but even then once you sync to the master clock, you still get the same problem with the FX...

 

I'd just like to add another subjective confirmation of that.

I was mucking around on the weekend started doing some proper a/b tests of the keylock and it's really warbly even when tracks are sped up by only a couple of bpm.

Keylock settings:

Hiq, scratch.

 

Another great "improvement"!

 

Time into a track

why cant I see the time into a track I can only see the time off a track. On the older version you could see both?