USB422 Latency and Yosemite III - Solved

Today I posted on the USB-422 web page signed low latency drivers for Yosemite.

The certification from Apple arrived when I was on holiday and it took a while before I could update the software. I checked the driver on the Yosemite macbook air that I purchased for testing. I certified the driver on a separate computer so that I could be sure that the driver was certified for all computers.

The hardest part of the installation is to be sure that the Low Latency driver is being used. Please read the MacLow installation notes! You can use the Latency check program that I wrote to be sure.

You can see the two versions installed below.

 Please let me know if you have any problems!

Yosemite and Latency II - More Problems

Last week I had a customer report more problems with Yosemite so I tried to replicate my last results - and failed! I then re-installed Yosemite from scratch and still failed!

To help make tests I wrote a small test program which measures the total time to send and receive a message.
 http://www.colinbroad.com/cbsoft/usb422/latency-mac.zip
Note: It is the maximum time that is of interest.

On investigation I discovered with the help of others that Yosemite requires signed drivers, if you rename the kext files or edit the plist file it disables the signing and the driver will no longer install.

I am now waiting for certification from Apple which will take 7 - 10 days assuming that they approve my application. If I get certification I will build a certified driver.

There is one alternative if you need it now, you can run a uncertified driver by disabling the gatekeeper which will then allow uncertified drivers to run. Wolfgang at SMM has confirmed this works.  Note: this allows any uncertified program to run with out confirmation.

TMC-1 User Interface

You may have noticed the change in the description below the title. I have worked on the TMC-1-Penta design since the beginning of the year and I am now happy to release it on our customers. In the process I have updated the User Interface for both versions of the TMC-1 (XMon and Penta) a number of times.

I have to thank Owen Curtin at the Bridge Soundstage for his feedback, Owen was the first person to order a TMC-1-Xmon about 6 months ago. Although I have not always implemented his suggestions they made me look at his problems and come up with my own solutions. Owen also came up with solutions to two of the limitations of the X-Mon - Monitoring the Studio Loudspeakers in the control room and playing back into the Studio from a different source to the Control Room.

I hope that my existing users will find the time and patience to upgrade to the new user interface and will like the changes that I have made.  If there is anything that I have removed that you want back or anything else that you need, let me know and I will try and work out a way.

Unless I have an urgent request or bug fix I hope that this will be the last User Interface upgrade for a while.

The picture below shows the new Meters Page with the Sum source indication, user defined names and Meter layout, you can also see one of the new user key functions [SLS A/B].
Note: I disconnected XMon to stop the Meters.

Yosemite and USB-422 Latency

Having updated Mavericks for USB Latency I was still getting a few emails from customers about Yosemite.  Because I use my macbook as my main machine running Mac, Windows and DOS and do all my work on this computer I was reluctant to upgrade just in case of problems.

Whilst in the USA for NAB I bought a new macbook with Yosemite pre-installed I can now use this machine for trying out new programs. I followed the posted instructions and disabled the Apple driver. You can use the 'ls' command to check that the driver has been disabled (Renamed). I then went to the FTDI web site where I discovered that FTDI had posted a new driver for the Mac that very same week! Version 2.3 2015-04-15. I installed the new FTDI driver, used KextDrop to add the low latency driver and restarted the computer. Everything worked OK and the port was discovered correctly.

I had to wait until I got back to my office to measure the latency, as you can see below it measures at 1mSec s expected. The Yellow trace is the command from the Mac via the USB422 and the Blue trace is the reply from our RM-6 synchroniser. They are both set up to send a message immediately after receiving a message. The USB runs a 1mSec poll so that there is approximately a 1mSec delay on the Mac. The embedded software reacts a lot faster as it is not limited by the USB Poll.

TMC-1 Mechanical Switch Noise in Talkback

I received a report that the mechanical switch noise on the TMC-1 when switching was annoying. Initially I assumed that the problem was switch On (easy to fix with a short delay) but the problem was that the noise from the switch release was too loud.

 I have tried to reduce the mechanical coupling on the panel by using a grommet and thin wires. The pictures show how to cut the grommet and glue to the front panel. The grommet Internal Diameter is 10mm and fixing hole is 12mm . We are supplying future units with this modification.

This reduces the noise but does not completely remove it. There are two alternative approaches you can try.

1) Use an external microphone and connect directly to the X-Mon or Digital system. You can then use the switches on the TMC-1.

2) Use external switches to Enable/Disable the internal talkback on the TMC-1. I have just added Int T/B 1 and Int T/B 2 as options to all the User GP inputs on the TMC-1.

Note this is not a problem with the External talkback microphone and switches.

Let me know how you get on, I am always available and welcome ideas to improve products. 

TMC-1-Penta and Dolby Atmos 7.1.4 and 9.1.2

Today I revised the specification of the TMC-1-Penta to cope with 12 channel Stems. To increase the channel count from 10 to 12 it was necessary to remove one of the stereo cue outputs. The revised specification for the mixer is as follows.

Stem/Input Summing - 12
Four Stereo Cues  - 8
Mono/Stereo main mix - 2
7.1/5.1 Downmix - 8
Stereo Downmix - 2
Total = 32

Here is a revised block diagram of the monitor system

TMC-1-Penta Details

The TMC-1-Penta monitoring is based on a enhanced  TMC-1-Xmon, the changes have involved an increased channel count and removing some limitations of the Digidesign XMon hardware. The block diagram below shows the current design of the system. I expect that this will be modified to respond to users requirements.The main features at present are as follows

• Up to 64 channels on the main monitor path
• Alternate Output (8 Channel)
• Mini Output (8 Channel)
• Five 10 channel wide alternate inputs/stems
• Input and output trims
• Two Listen back Inputs
• Two Talkback Inputs
• Two Talkback output feeds
• Five Cue Outputs
• Three stereo cue inputs
• Stereo Downmix Output
• 8 Channel Downmix - 7.1, 5.1
• Meter Outputs

The complete Digital I/O and Monitoring Solution

CB Electronics (www.colinbroad.com) in cooperation with NTP Technology (www.ntp.dk) have developed a digital version of the TMC-1 monitor controller TMC-1-Penta which will be launched at NAB 2015. The TMC-1-Penta communicates with a Penta 720 or DAD AX32 over ethernet using its native protocol either directly or over a network. The software is based on the XMON version of the TMC-1 with a number of enhancements including the ability to control up to 48 channels on its main input/output and an 8 channel wide Stem Mixer. The flexibility of the NTP Digital interface allows the incorporation of extra features that are not available on Avid’s XMON.

The DAD AX32/NTP Penta combines direct connection to Pro  Tools HD/HDX/HD Native systems with superb A2D/D2A converters,  Madi, AES and SPI. The combination of the DAD AX32 / NTP Penta with the TMC-1-Penta provides a cost effective digital I/O and monitoring solution for any studio.

The block diagram below shows how to use the TMC-1 and NTP Penta 720 or DAD AX32 for digital monitoring and DAW I/O. The Penta 720/AX32 provides all the I/O routing with 16 AES I/O channels, Madi and direct connection with Pro tools as standard. Also available as optional cards are Analog I/O, Dante, SDI and other interfaces providing up to 976 input channels and 976 output channels.

§

Visit NTP Technology on the Sanken/Plus 24 stand  at NAB - C4051 Hall C.

Mac Latency, TMC-1, TMC-2 Penta,TFT Remote Control

I had not realised that so much time has passed since my last post, good intentions does not get it done!

Mac OS X Mavericks and Yosemite Latency

Apple have added their own FTDI driver which has a 16milli-Second Latency, this must be disabled before the modified low latency FTDI driver can be used. See the USB422 product page for details.
http://www.colinbroad.com/cbsoft/usb422/usb422.html

TMC-1

We have now shipped over 10 TMC-1's and they have been received well. Almost every time I demonstrate it the clients give me new idea's. I have incorporated as many as possible in the latest TMC-1 software that I posted last week.  There is a prototype hardware for the TMC-1-S6 mounted in an S6 console at the BVE next week.

TMC-2-Penta

The TMC-2-Penta is a digital version of the TMC-1 that drives the NTP Penta 720 via ethernet. I am working on this at the moment and should have beta units at the end of March. The functionality will be similar to the XMon but more flexible with the main inputs and outputs up to 48 channels wide and where XMon has 2x 7.1 inputs and 4 stereo inputs the Penta version will have 1x 48 channel input and 5x 7.1 inputs.  The TMC-1 can be upgraded to the TMC-2 by adding the ethernet board.

TFT Remote Control

After testing on the prototype at Twickenham we supplied 7 TFT Remote controls (See the earlier post Tuesday, 9 September 2014). One for a DFC console(Twickenham), two for a Harrison MPC4 console (Lenfilm), four for Lawo consoles (ORF). 

TMT Cologne - TMC-1 Analog/Digital Monitoring

The TMC-1 controller for the XMon is now shipping. We will have a S6 version early next year.

Whilst at the Tonmeister Tagung we agreed with Digital Audio Denmark to produce a version of the TMC-1 which provided multi-channel digital monitoring using their Digital interfaces. One advantage of the digital version is that the number of channels is only limited by the DAW so that formats like 7.1+4, 10.2,, 22.1 or larger are all possible.

I still believe that the PD-1 provides the best stem monitoring control by keeping the stems within the DAW (Currently only on Protools or Pyramix but open to others).  One of the major advantages of this approach is that the track and stem names are read from the DAW and do not need to be entered again on the mixer.  We will include this function in the digital monitoring solution so that track and stem names are copied from the DAW to the DAD network.

The PD-1 is currently limited to 8 stems each with only 8 tracks. The next s/w update will increase this to 16 stems each with 16 tracks over two pages. Let me know if you need more!

Film Archive - TC-5, BS-1 Biphase Generator - Lock to Wordclock

It may be news that we still make the BS-1 bi-phase generator, however there is a small archive market that we still supply.

Archiving often requires the transfer of many small segments of audio, it is inefficient to transfer these individually. Multiple segments can be spliced together using an editing table. The TC-5 can be connected to the bi-phase output of the editing table and be used to mark the timecode or feet and frames position of each join.

The BS-1 controls the film machine to transfer the film accurately to audio files with timecode. It is important that the film speed is locked to the audio wordclock or the join locations will not be accurate. In the past video syncs were used to lock wordclock, timecode, and film speed to video syncs.

Many audio transfer setups do not have a video sync reference, only wordclock. We have modified the External reference input on the BS-1 to provide a 24/48Hz reference from 48/96KHz wordclock. This modification is available as an upgrade on existing BS-1's or MC-1's.

PD-1 and Protools 11.2.1

Bill Ritter from Electric Entertainment reported some problems when using the PD-1 with Protools !!.2.1which were not present on Protools 10 as follows

After going into record and then dropping out of record there is no output when switching to Input!

As I am in LA after the AES I went to investigate. The protools screen indicated that the PD-1 was switching the tracks to input but the only way to hear the input was to switch between playback and input on protools.

After some investigation I found that the following sequence of commands from the PD-1 will solve the problem

1) Switch to record
2) Switch to input monitor
3) Switch out of record
4) Switch back to playback monitor

A unexpected benifit of this command sequence was that some monitor glitching on Protools was reduced.

I will add a option to the PD-1 software for this with a variable delay between steps 3) and 4), when available I will update both the PD-1 web page and this blog.

It has been over two years since the last PD-1 software update, if there are any other requests now would be a good time to make them as it will take some time to get back into thr software and test it.

I did get one request at the AES show - to switch any channel not in Record Safe to Input Monitor when stopped, I can implement this as a option at the same time.

Optimising XMon Performance

This week I have been setting up our new Monitor Remote (Now called the TMC-1) on the XMon.  I have heard some bad reports about XMon so I wanted to be sure that I knew how to optimise the setup. 

XMon cannot be considered in isolation but as part of a monitor system that consists of the XMon, Power amps and Speakers.

The main gain section of XMon uses digital gain control to provide a gain range of +31dB to -95db.  As can be seen on the chart below to optimise the signal to noise it is important that XMon is run at between 0dB and +10dB gain for normal levels.

The Speaker trim section of the XMon provides a gain range of 0dB to -63.5dB in 0.5db steps. As can be seen in the chart below the signal to noise remains the same for the first 4dB of attenuation but then the noise floor levels out so that the signal to noise degrades as attenuation is increased.

The factory default settings of the TMC-1 XMon controller are as follows 

Main Gain Section +4dB gain.

Speaker Trim -4dB

Overall gain 0dB

Speaker trim range is restricted to +/- 4dB

Input makeup gain +20dB to -10dB

The default setting provides 5dB of analogue headroom above the digital clip point when set in this mode (XMon clips at approximately +25dBm).

New TFT Monitor Remote for XMon

I have now returned from IBC where we were showing our new TFT based remotes. The new TFT Monitor Remote generated some interest.

Designed originally as a remote for XMon CB plan to interface to other monitor systems including our own digital unit. There are 4 easily accessible main screens arranged in pairs and include: Input/output selection, Meters/Solo, Cue selection and gain, Talkback/Listen Gain.

This unit has the same profile as the PD-1 and sits conveniently on or may be sunk into a desk.

Using an embedded processor the unit communicates directly with XMon via Midi. and controls the full functionality of XMon.  No other computer is required. The remote is fitted with a talk-back microphone and headphone jack and uses the same 15 pin connector as XMon. A GPI port is provided for Hardware Mute, foot switch input and PFL/AFL switching.

The USB connector provides power and allows for software updates. The USB Midi connected to Protools or any MTC source can be used  for Auto-Talkback (or even a Timecode display).

The 4.3" TFT display is fitted with a touch screen in case required for the future, the XMon interface does not use touch.

For the future

we are designing a digital audio monitor system that will use the same remote, We are also looking at other applications using the same technology, one suggestion is as a digital microphone remote.

The photograph below is of the prototype unit and a selection of un-retouched displays.

.

New TFT Remote Panel with LCD Keys for DFC Console

As promised in the last Blog here is the first picture of the new remote panel as suppled to Twickenham studios.

The main features of this panel are as follows

• TFT Display - able to display more system information.
• 21 programable keys - only the keyboard, Shift, Recall, Store are not programable.
• The lower 18 keys are LCD keys which display the key function.
• Jog Wheel, four LCD keys may be replaced by a Jog Wheel.
• Two Serial Ports and a USB serial port.
• Two or more panels may be connected in a daisy chain.
• Direct connection to CBServer via the USB port.
• Sony 9 pin Input for control from desk
• Software updates and keyboard programming via USB port.
• USB Powered  

Fully compatible with existing systems this panel is currently available as a remote to the RM-6 and existing systems as shown in the updated Twickenham block diagram (previous blog). Below is a system diagram using two DFC-TFT remotes and a RM-6.

Atmos, Twickenham and Multi-Machine Track Arming enhancements to the Sony Protocol

It took a few try's to get the system at Twickenham working properly with two 64 track recorders, but after 1 week it seems to be working well. As part of the installation I installed the prototype of our new TFT based remote control panels - see the next blog entry for a full description

If two 64 track recorders becomes normal then it will be necessary to reconfigure the communications between the console and the machines/machine control system. The DFC has 6 sony P2 ports that will support 48 tracks per port but a number of consoles only have one port.

The standard Sony protocol can only control one recorder, here are three possible solutions to track arming multiple machines from a single sony protocol port

1) Implement our xmc protocol,   a number of years ago Harrison implemented this which allows them to track arm multiple record machines. The advantages - no modification required to our systems and frame accurate timecode without video syncs, disadvantage - a major change to the console communications s/w.

2) Map the tracks over multiple ports, the maximum possible is 112 which can be mapped over multiple tracks. The advantage - no modification required to the controlling s/w, disadvantage absolute maximum number of tracks is 112, not all serial ports implement the maximum (eg. DFC 48 maximum),.

3) Enhance the Sony protocol to support multiple recorders. The advantage - can support up to 112 tracks per machine, disadvantage - The controlling s/w must be modified to support the new commands. Most of the required commands are present in the RM-6 to support the our multi-machine protocol (xmc) a subset of these can be added to the Sony protocol.

What are we doing about multiple record machines?

1) The next release of the RM-6 will include all the above solutions. Once implemented the commands will be available for any manufacture to use.

2) I am currently updating the serial control system for one mixer manufacture to increase the number of record channels to 96,  at the same time I will add the commands to support multiple record machines.

Recording up to 192 tracks on two Protools from a DFC

As noted in a earlier blog the DFC can only arm 48 tracks on each serial port, this two ports are required to arm up 96 tracks on a single machine. In order to record up to 96 tracks on two machines 4 ports are required!

The diagram below shows how this is done and the menu settings required to achieve this at Twickenham Studios.  Note that the menu entries are for the latest version of the X4 and RM- software which were updated this week. Contact me for further details if you need them

Diagram updated to include a second remote, our new DFC-TFT as requested by Twickenham.

Detective work in Paris

Some months ago Daniel (our French representative) complained that every morning his SR-4 crashed when turned on, when reset or restarted it worked without problem. My test unit in the UK however did not suffer from the same problem even with identical software and setup.

A visit to Paris found a bug reading the EEPROM that caused the unit to crash if it could not interface with the EEPROM there remained the problem of why it was not reading the EEPROM.

This week Daniel swapped the chip used to read and write to the EEPROM, this cured the problem, but even when the old chip was put back the problem was gone!

The moral of this story is that if you are having strange problems on old equipment (the SR-4 4 was probably 15 years old) try removing any IC's on sockets and cleaning them!  

Advance Information on a new remote for the RM-6

Bellow is the TFT-422 the first in a new series of remotes that we are developing this year, This uses a TFT Display, with combination of  LCD keys and normal keys.

In developing a new remote for the RM-6 we have added two features to our remote controls, as follows

1) TFT display, our new interface can drive 3.5", 4.3" and 5" displays, the TFT display is the same as used on most smart phones and gives a bright clearly legible display capable of showing more information than the existing 2x40 LCD display.. One of the more obvious differences in designing a panel for LCD or TFT displays is the shape, this required changing the panel and the layout of the display. The display on the above panel is a 3.5" TFT.

2) LCD Keys, allowing macros to be changed easily, allowing individual users to define there own key sets. In using LCD Keys you have to take in consideration that they are bigger and deeper that our standard switches  and cost 15x as much. Because this is a free standing unit it would be impossible to use LCD Switches at the front as they are too deep.

We are also planning a new built in panel with all LCD keys except for the Numeric keyboard with an option of extra keys or Jog Wheel

Digitizing Audio from Film using Wordclock as a reference

When transferring audio it is important for the audio to run at the correct speed, positional lock with the moving image a separate procedure often using a punch frame or clapper at the beginning of the reel . If the speed is not correct the sound and image will drift apart, the longer the reel the further they can drift.

We were asked to devise  system to make a simple locked transfer from film to Digital Audio, the customer also wanted to control the film from the DAW. We recommended  our BS-1 Bi phase generator with Sony 9 pin control input. To simplify the system added  a wordclock reference input to the BS-1 removing the need to a video sync generator.

If the workstation does not have a Sony 9 pin output but has Midi Machine Control then our P2MMC-usb can be used to control the BS-1.