- December 2016
- October 2016
- September 2016
- May 2016
- April 2016
- November 2015
- October 2015
- September 2015
- April 2015
- March 2015
- July 2014
- June 2014
- November 2013
- October 2013
- September 2013
- July 2013
- June 2013
- May 2013
- March 2013
- February 2013
- January 2013
- December 2012
- November 2012
- September 2012
- July 2012
- June 2012
- May 2012
- April 2012
- March 2012
- February 2012
- January 2012
- December 2011
- November 2011
- October 2011
- September 2011
- August 2011
- July 2011
- June 2011
Stay In Sync With Us!
More interesting OCD stuff from our archives.
Close to 100 years between these devices always looking for that perfect click track with whatever technology they had available to them at the time.
Both beautiful in their own way.
UREI 964 Digital Metronome
Paquet-Maelzel Mechanical Metronome
Philippe-Nicolas Paquet was born on the 8th Oct 1823. His father was a watch maker and Phillipe followed the family tradition entering the time-keeping industry. He settled in Paris, working for the Maelzel Metronome company and after the death of Johann Nepomuk Mäelzel in 1838 eventually took over the running of the factory, in 1846. Paquet went on to establish a new small machine tools plant and metronome producing plant in Beaumont Sur Oise in 1867 winning many awards at International exhibitions. The factory in Beaumont remained in production until 1983. The two dates shown on the majority of Paquet – Maelzel metronome trade labels are 1815-1846, the former being the year the patent was first approved by Maelzel and 1846 being the year that Paquet took over the helm.
Verified 100% correct operation.
We only had 24 hours with these new toys from Roland. This post is primarily focused on timing and external sync performance but I had a few other observations. They sound great. Really. I’ve given up on the 1:1 audio comparisons with the originals because they are just so close. Being over 30 years old and analogue (hats and cymbals excluded on the TR-909) virtually every TR-909 and TB-303 sound slightly different anyway. In a mix or over a PA it’s never going to be an issue. These are light years ahead of the MC-303/MC-505/MC-909 period of the late 90′s thankfully. Most of the originals need periodic, qualified maintenance and crucial spare parts are getting scarce. At roughly $700 street price for the new versions I’d leave the TR-909 and TB-303 at home under glass ready for the retirement fund. Personally I do miss the individual outs on the TR-09 and the knobs are a little too small and crowded for live, precision tweaking in the dark but I understand fitting it into the ‘Boutique’ physical format was always going to be a squeeze. The TB-03 on the other hand is a little larger than the original and the layout has plenty of familiar real estate.
OK – timing…..note these are both Rev 1.0 OS machines.
TR-09: All Steps - Continuous Rimshot – Single Pattern – 120 BPM
Internal Sync : Precisely 42 samples (0.87ms) push-pull between each consecutive step – think of it as sample-accurate light swing. This seems a common ‘feature’ in many newer drum machines and sequencers.
[Comparison: Original TR-909 = Random 214 samples (4.46ms) push-pull between steps]
External Sync (5 Pin MIDI) – Sync-Gen IILS : Precisely 96 samples (2.00ms) push-pull between each consecutive step. Still sample accurate but heavier swing under external MIDI Clock.
Offset to DAW Grid = 160 samples (3.33ms) behind.
[Comparison: Original TR-909 = Random 374 samples (7.79ms) push-pull between steps]
TB-03: All Steps – Continuous Notes – Single Pattern – 120 BPM
Internal Sync : Very close to sample accurate – only 2 samples (0.04ms) push-pull between each consecutive step.
External Sync (5 Pin MIDI) – Sync-Gen IILS : Precisely 22 samples (0.46ms) push-pull between each consecutive step. Sample-accurate light swing under external MIDI Clock.
Offset to DAW Grid = 95 samples (1.98ms) behind.
Note: Both machines can sync externally to DAW generated USB MIDI Clock, however, with all the possible influencing factors that can affect stability I left this alone for the time being.
We’re still going through the old archives as we tidy things up for anyone who finds this stuff useful. Though not directly related to external hardware tempo-synchronisation it is still a critical aspect of music production with the DAW now the central hub of 99% of all studios – large commercial facilities and bedroom studios alike. Feel free to email if any of this doesn’t ring true as it goes a while back.
The following test case assumes use of an all-in-one prosumer audio interface as a pseudo input mixer/monitoring system rather than an outboard analogue mixing console or an alternative all-analogue signal path for zero latency monitoring.
Here is a real-world example of a latency issue when recording anything in real-time against a DAW generated click track or against previously recorded material as a fresh take or as an overdub. Our DAW audio click track guide was routed directly to a Focusrite Saffire 40 headphone output panned hard left. For the purpose of this test we used a direct clone of that same audio click track but routed to a direct analogue output of the Focusrite Saffire 40 then, on to an external analogue mixer (zero additional latency) and back to an input on the Focusrite Saffire 40 for recording back into the DAW. If we take a very common production scenario where a vocalist wants to monitor their performance through various DAW plugin processing (compression/EQ etc) we must route the armed destination DAW audio track (set to input monitor) to the Focusrite Saffire 40 headphone output. We then pan the destination audio track output hard right in the headphones so we can check the offset between the DAW reference playback guide track (what the performer is singing against) and the live processed take they are monitoring in real-time.
What the vocalist (or any other performer) experiences is a fixed delay between their own natural physical feel and phrasing as they sing/play and their ‘foldback’ processed audio coming back to the headphones. Minimal latency values will result in odd phase/tonal shifts. Larger latency values will produce flamming and in the more extreme cases a slap echo. In either case the less enjoyable the recording experience will be and the end result will most likely be an average take.
As a direct comparison this is a quote from John Klett’s paper - Delay in Large Format Digital Music Consoles:-
“Put yourself in the position of vocalist. You are very good and want a cue mix with your voice in it so you can hear yourself as you sound. You need to hear how you sound fairly accurately so can moderate your tonality.
Typically, in an all-analog path, it takes between 2 and 8 microseconds for a signal to make the trip from a microphone, through a preamp, analog dynamics and EQ to a hole in the patch bay that I would then patch to an input on a console.”
All analogue maximum monitoring latency = 8 microseconds
All digital monitoring latency (DAW@ 48kHz/512 Buffer) = 26 milliseconds
Monitoring delay increased by a factor of 3250 times.
I found these numbers which expand on the previous post – Latency and Feel. I haven’t revisited these for some time but from memory they are accurate. Let me know if anyone thinks they are off the mark. First image shows the waveform offset between the DAW source track audio and the bounced audio after taking a round trip through the audio interface DAC > ADC running at 48kHz with a buffer of 1024 samples and onto a new DAW audio track. A simple application of this would be sending a vocal from a DAW track out to an external effect unit (analogue filter/compressor etc) and then back to an input on the audio interface for recording to a new DAW track. If you are cue/headphone monitoring through the DAW with plugin processing via your audio interface you will have similar issues to contend with. The second image shows the various latency values at a variety of DAW sample rate/buffer settings.
Another one from the archives. Since the release of the MPC-60 in 1988 there have always been rumours about some internal magic feel programmed into the heart of all MPCs thus making them ‘feel’ more musical than playing back real-time MIDI in a DAW environment. The man himself sets the record straight.
Just going back through the archives to extract some useful information for those interested. This relates specifically to digital recording consoles with regards to cue monitoring issues in a digital environment. Track forward to 2016 and the exact same issues still apply to the modern DAW platform with real-time external digital signal processing, active plugin delay compensation, audio interface latency and buffer size, external sequencer tempo-synchronisation, real-time MIDI event delay and external MIDI hardware processing latency.