I actually do understand what beta releases are for, thank you. I also understand what alphas and stable releases are for.
I’m not sure that accusing your customers of a lack of understanding is helpful, especially if you dont know their background and experience.

I was listing the bugs that are being commonly experienced in a “stable” release which was unusable unless you used the “bleeding edge” betas…

I completely understood that was how you had chosen to implement it. I just thought that it was farcical to claim that being able to exit it without turning off the machine is a “feature” and not a bug, especially as you have now added that feature. If it was the correct design decision and you are happy with it, why add the feature to not have to power cycle? Or are all Polyend bugs just features that the user doesnt understand?

Accusing your customers of being ridiculous is very revealing, along with your previous comments. Thank you for illuminating your attitude towards your customers who are unable to use advertised features.

I have acknowledged your problem (multiple times) and told you we are working towards addressing it (multiple times). I’m just rolling my eyes at your growing temper tantrum.

7 Likes

Again, insulting your customers is rather unedifying. I genuinely wanted to have a discussion (and you’ve avoided my questions) - just thrown insults and called me names. It’s not a great business practice usually… I’m genuinely not angry and having a “temper tantrum” - I’m asking you questions and explaining the problems with the Tracker. You’ve started calling names… it’s very illuminating about the attitude of Polyend to its customers. I actually came close to not buying it after seeing some of the comments to customers on the fb group from the mods - they were pretty insulting. It seems the attitude persists here.

hey y’all – actually took delivery of a tracker the other week, stoked to dig in!

in the meantime though, it seems like a break could be helpful. i’m setting this thread to slow mode for the next 24 hours. this means that each user can only post once every 4 hours. please remember that all public threads are crawled by search engines and indexed – a direct message to work something out can go a long way.

14 Likes

who let the elektronauts into this thread ? smh.

9 Likes

For what it’s worth, I’ve had a Tracker since July, and I’ve been astounded at the feverish support activity and rapid responsiveness of Polyend throughout. I don’t think I’ve seen that level of support and iteration on a similar product, well, ever. By my count, there’s been around 35 beta firmware releases in 7 months, most of them addressing high-priority bugs and also adding significant new features. That averages out to one beta release per week. I didn’t feel like I had a crippled product when it first arrived, either - I encountered some bugs and things I would’ve liked to see improved, but no show-stoppers or lost music, and I could certainly be productive with it. Granted, I don’t use stem export (song export works great) or some other features people think are deal-breakers, so maybe I’m lucky to an extent. Anyway, kudos to the team for what they’ve done so far. It’s always a source of excitement when I get a notification about a new Tracker beta.

9 Likes

Thank you @RPLKTR and @klingklangmatze for answering my questions, and @naxuu for providing your perspective!

I cannot find any local-to-Chicago shops which stock the PT and I’m not inclined to mail-order one; we have to support our local synth stores in these times. However, if/when I can get my hands on one, I will do a deep dive on all the concerns presented here and do my best to represent the unit I get accurately; please feel free to ask questions.

4 Likes

Thank you @RPLKTR for participating in this thread. It’s great to have someone from the Polyend team sharing information in public like this!

My Tracker is the best music device I’ve bought in a long time. Your hard work is very much appreciated.

10 Likes

@RPLKTR Can I ask some questions about the clock sync please?
How does the delay compensation work on external clock? I am a little confused. Is there a reason it is a global manual setting?
Is there a reason why this isnt a fixed value? As it is a value the user sets for the whole project, does that imply the offset is always fixed? If so, is there a reason it isnt fixed/automatic? If not, does that mean that the drift/lag/offset of the audio from the clock varies per step ?

Thanks a lot for answering this question. I know lots of people have struggled keeping it in sync (I certainly have and was just confused as to why the user needs to set the offset themselves)
Thanks.

Again - I really like the tracker and want to use it - it’s a great idea - I’m just struggling with all the bugs and instability.

picked up a tracker from my local shop in sf today. deleted all the default samples and projects and dove in after spending some time with the loopop demo. it’s a pure delight thus far. i’ll wire it up to my midi garbage heap tomorrow and see what’s up, but from my first few hours it’s a delight in the purest sense. the workflow is much less obtuse than i’d anticipated.

4 Likes

This is how I understand it:

The delay compensation means that depending on what you’re clocking the Tracker with, there may be an offset in the playback between the master and slave device. To put it other way, there’s a positive or negative delay between receiving a MIDI clock tick and the master device actually outputting what should be played on that specific tick.

The delay is dependent on the audio / MIDI latency and synchronization of the other device(s), so it cannot be a fixed value in the receiving end. Only you can hear whether the devices you’re syncing start the playback in the same phase / at the exact same moment or not, and the amount of positive or negative latency compensation you need can vary from master clock device to another. As long as you’re using the same MIDI clock master and the sending nor the receiving end isn’t completely broken, the clock compensation setting should always be kept the same once you set it once.

Separately from this, there was an issue in the Tracker firmware that seems to be marked fixed in one of the 1.3.0 betas, and hence in the final 1.3.0 as well: [1.3.0b5] Polyend Tracker does not sync to midi clock properly · Issue #559 · polyend/TrackerIssues · GitHub - so the MIDI clock kept in sync as long as the clock sync compensation was not used, but drifted (or glitched?) when the clock compensation was set to something else than 0. So it simply didn’t keep in sync when the compensation was used, which wasn’t something that could be corrected with adjusting the compensation, but some bug in the Tracker’s internal sync code.

So if your issue sounds like the one just described and you still have it, either you have some version of the firmware active that is not the final 1.3.0, or the said issue hasn’t been completely fixed in the final 1.3.0 even it has been marked so.

Hope this helps (and hope I understood your problem correctly in the first place).

Thanks, @kbra, your explanation is very good. Let me expand on it.

External clock synchronization

Manual external clock compensation is needed because the master clock might also have non-zero latency. Additionally, in hardware-only setups users often resort to MIDI daisy-chaining which adds additional latency.

This is not a Polyend Tracker-specific problem and the solution isn’t atypical either. Look at how Renoise solves that same problem:

It’s also global and manual.

Tracker’s status: Importantly, there are no known issues with our implementation of this. External clock synchronization is considered a solved problem.

Internal clock synchronization

If you’re using Tracker as the master clock, essentially the question we’re after is:

If you have “C-4 M01” on one track and “C-4 01” on another track during the same step, do they sound at the same time?

The answer here depends on many factors which is why we don’t have this yet in Polyend Tracker. At this point you need to compensate latency on the MIDI receiver end. On that same picture above you see that the “MIDI Clock Master” section in Renoise also has an Offset slider. That slider isn’t all that you need to consider.

  • If the MIDI out that you’re sending to is a synth with zero latency, the “C-4 M01” note will sound sooner than “C-4 01” because Tracker’s audio engine has some internal latency due to its audio buffer it needs for its synthesis engines, reverb, master limiter, and our hardware codec’s “bass boost” and “space enhancer”.
  • However, that synth might either have latency of its own, or it can be chained to a pedal or other processor that adds some latency of its own. So calculating this latency automatically isn’t possible to figure out.
  • Better yet, the question arises: how are you listening to your external synth sequenced by Tracker? Are both Tracker and your external synth mixed by some other mixer or DAW? Or is the external synth’s sound engine routed back to Tracker via its “Line In” audio input? In the latter case, the sound from the synth is additionally shifted by Tracker’s audio buffer.
  • Finally, the kicker: what if you have more than one device connected to MIDI? “C-4 M01” (MIDI channel 1) might have different latency from “C-4 M02” (MIDI channel 2).

Again, is this a Polyend Tracker-specific problem? Nope. Let’s see how Renoise solves this. Apart from the “Offset” slider we’ve seen already, it’s got “automatic plugin delay compensation”. It can do this because VST plugins report their latency to the VST host. Tracker works with hardware and this kind of reporting is impossible.

On top of this, Renoise has an instrument-specific MIDI OUT latency setting:

That’s not all. To deal with the situation of audio from the external MIDI synth being routed back to Renoise, it’s got a special mode in its #Line Input device:

Tracker’s status: I hope you can appreciate that this is a lot of functionality to be implemented. I’m working on this personally, this issue is tracked as #570.

Live clock drift

Realistically, no two hardware devices clock the same in a live setting. This is why MIDI doesn’t just send “start” and “stop” messages but also sends clock with a high PPQN. This is how you keep multiple devices in sync. Because MIDI is a serial protocol with relatively low resolution, it still has some jitter. Jitter means that the timing between two CLOCK messages is never exactly the same. In most circumstances it doesn’t matter but if you care about phasing between multiple devices, there are special devices which are specifically designed to output rock-solid clock, usually through an audio port and not MIDI. Those devices cost more than the Tracker.

Tracker’s status: Tracker’s current internal live clocking can be +/- 0.05% off from the declared BPM. This is within norm for hardware sequencers, we won’t be doing anything about this.

Song export clock drift

Tracker’s offline song rendering also can be off by +/- 0.05% from the declared BPM. We don’t consider this a deal breaker as all DAWs currently allow for audio stretching to match whatever BPM you need, if the 0.05% difference is unacceptable for you. Alternatively, you can export live with AUDIO OUT clocking Tracker externally from your DAW, making the timing match exactly what you need.

Tracker’s status: Unlike the live clock drift, we feel like we can do better than +/- 0.05%. Due to the audio engine’s implementation this is not a trivial improvement to make. We make no promises as to when this will be addressed.

Stems

To reiterate, we are fixing performance of stem export as well as issues with stems alignment between each other. This is tracked as #697. Stems have the same clock drift as song export (see above). The drift is identical between all stems so they stay in sync with one another when laid out on a multitrack audio editor.


I hope this clears things up. Cheers!

11 Likes

Thanks a lot for the detailed explanation. I have used delay compensation in DAWs and have done some coding of MIDI on embedded devices so have some awareness of clock issues (jitter, precision, lag, etc), hence the reason for the questions. I can think of a number of ways you may have implemented it. I have read your explanation about four times and am still not quite sure about the answer to one of my questions:

(for the sake of argument, let’s assume the clock is constant with no jitter and is precise and accurate, I’m just asking if the time from the step trigger/rise of the clock pulse to the audio generation is constant and independent of the processing - if it’s not, how large is it - eg mean and standard deviation would be great - just because I am aware that the current latency is reported to be 15ms from the clock but I’m keen to know how constant this is)

I’m also confused about your comment on “daisy-chaining” hardware MIDI devices causing latency - do you mean using MIDI thru ? (I always thought something like a 6n138 would only add a small number of microseconds to the latency eg typically 2 micro seconds ish - you’d have to chain 500 devices to get a 1ms latency would you not ??? Is that one of the issues for the tracker? I admit this is a gross simplification and am ignoring the latency that is in the order of nanoseconds in the 74HC14…).

Very happy to admit I am wrong - for hardware chaining, I haven’t seen or heard significant latency from chained MIDI thru devices. I’d love to learn what the latency is for midi thru on hardware from your perspective and if it is an issue for the Tracker - eg erica synths quote a latency of 60 nanoseconds for their hardware midi thru box… are they incorrect ?

Usually “daisy chaining” in MIDI-land seems to refer specifically to connecting something to some device’s MIDI input, then connecting the MIDI output of that device to next one’s input, et cetera. So for each device, add the time it takes for that device to merge (or just relay) and send the messages forward.

I suppose traditionally this has been done either because a lot of (cheaper) synths didn’t and still don’t have a thru port, or for the sake of being able to have several control sources controlling several sound sources.

Using proper hardware MIDI thrus (as long as every device in the chain has them) does mitigate this issue, so you’re correct with your calculation.

3 Likes

Thanks a lot for that - didnt realise people use midi in and out that much !!! Pretty much all of my hardware synths and keyboards have midi thru or I use a midi thru box for the odd one that doesnt (a midi thru mult is so easy to build)

2 Likes

I’m other words it’s better to use Tracker on its own unless you make Ambient music so timing doesn’t have to be perfect.

2 Likes

Ha ha - that made me laugh - and the noise just adds character ! (Really wanted to use stems to avoid the noise)

Erk, really? I appreciate the hard work going into fixing the various problems, and have a lot of goodwill for Polyend, but this kind of statement isn’t particularly edifying. It’s not unreasonable for customers to expect core features to work at purchase date.

I too would like to know if the clock drift is expected to be consistent across the stems (as in, out at a consistent rate across the project), or if this is going to cause problems when chopping out particular sections, due to an inconsistency in the variation across the whole track.

1 Like

Just to add my .02, I have had a tracker since June. Even if it had never been updated, it arrived then as a complete, inspiring instrument and I have been astounded at the level of support, regular flow of betas (like every few days), and engagement with the community. I’ve never seen anything like it in 25 years of buying gear.

I have never messed around with exporting stems - no interest - and I have not noticed any sort of inappropriate noise. I record line-in audio and am an ascended master of proper gain staging when doing so.

Features I have politely suggested and asked for and received with concierge quickness:

  • option to turn off anti-aliasing
  • full pitch effect added to effect lanes (for pitching sliced beats)
  • auto bounce selection back to note lane after changing instrument lane
  • ext sync latency compensation
  • live sample external audio while running sequencer
  • bit depth added as lane effect
    and more
14 Likes

FWIW I love the Tracker. I’ve said elsewhere that the design is beautiful and I love composing with it. I am glad that they are using GH to track issues and are responsive. I get that an issue like this can be tough to deal with, and isn’t a deal breaker for everybody.

That said, after waiting months for a fix for something that is a core feature, it does seem a bit unfair to tell people that it’s essentially their fault for not doing extended research on what features are broken in current firmware and what isn’t. I reported the export issue on day one that I got the device, and was told that exports would be fixed - but now it seems like that expectation is being walked back. If I had known that there was a chance that stem exports were going to be substantially broken and perhaps not even completely fixed at some point, I wouldn’t have bought the Tracker, or at least been able to return it within the first month of ownership. As it is, I’m fine to wait a bit for the problems to be ironed out, but don’t really appreciate what seems to be finger pointing, as if it’s the consumer’s fault for expecting that a device which costs a fair chunk of cash would work to a certain standard. That’s a surefire way to kill off any goodwill that folks might have.

There was a similar situation with the Digitakt, which was advertised as ‘Overbridge ready’, which then apparently meant it would work with Overbridge, but some version of Overbridge that didn’t exist yet - and then wasn’t relased until years later. The attitude of Elektron at that point was extremely poor, telling customers that it was their fault for not interpreting their advertising properly - and it left a sour taste with many folks. I really hope this doesn’t become the same kind of scenario, because the Tracker has so much potential, and I appreciate the hard work that has gone into it, and how responsive the Polyend folks have been on GitHub.

TL;DR - I get that releasing a product like this is tough. It’s complex and requires juggling a whole bunch of demands and expectations. I’m happy to wait on further development, but please don’t start blaming folks for buying a device on the premise that core features should at least be usable.

2 Likes