Relative Time Tool

[Chips Davis]

Many years ago I asked Richard Heyser to implement "Relative Time" into TDS (Time Delay Spectrometry). This gives the ability with an ETC (Energy Time Curve) to figure out time (distances) from the direct, any reflection, or accumulative energy events or energy returns. By blocking the mic from different directions you can determine the direction of the reflection(s) and the time to the reflection with "Relative Time". You can set any reference time point and insert one or multiple points referenced back to the reference point.

Is the Relative Time tool possible with EQW?

I have most of the major measurement programs. However, I am very impressed with this program. I have to thank my client Steve Johnson for showing me REW

Thank you,

Chips Davis Acoustical Designs.
925-518-0388 cell

 

[John Mulcahy]

Hi Chips, how is the time for the direct path set/determined? That is tricky for the measurement software as there are unknown buffering delays in the measurement path, using a loopback connection as a timing reference can overcome that if the path from the soundcard output to the speaker is analog or known to have negligible delay, otherwise that direct path delay would likely be something the user would have to set by entering the distance from source to mic. Without that only the delta between the direct path and the reflection is readily extracted.

 

[Chips Davis]

Hi Chips, how is the time for the direct path set/determined? That is tricky for the measurement software as there are unknown buffering delays in the measurement path, using a loopback connection as a timing reference can overcome that if the path from the soundcard output to the speaker is analog or known to have negligible delay, otherwise that direct path delay would likely be something the user would have to set by entering the distance from source to mic. Without that only the delta between the direct path and the reflection is readily extracted.

 

[Chips Davis]

Sorry this has taken me some time to replay. "Realitive Time".
I'm glad that I have the time to get more into REW and learn more of the great program you have produced.
Relative time and 0 time- Zero time is the time starting with the referenced output from the generator. That time includes the group delay of the DSP and the amp including the latency of the speaker. The time travel from that point (zero time output from the speaker) to the microphone position is the time of travel to the arrival of the IR or ETC at the measurement mic. Relative time starts at that point. See attached.

While viewing the ETC you can see the reflections returning back to the mic position. My early work in near field reflections and minimizing destructive comb filtering using TDS has been documented. To find the offending reflections you set the IR arrival to that time and set the curser to the reflection(s). This gives give you the distance to the surface that is causing the problem. Blocking the mic at different points around mic with a hand of cardboard, reveals the location of the reflection and you know the distance to it you can treat it. I am attaching a slide from the paper I presented at the ASA last November to show the timings.
“Energy Density in Diffusion, Quantifying Aural Quality through Testing and Listening
Side Diffusion and Optimizing the Mixing Environment”

If the reflection is concentrated and above and out of the Hass Effect ( FAY-HALL Effect) parameters, it may be heard as an echo. Near field reflections integrate with the direct and will change the timbre and localization in space.


I don't know if I have a problem with how I'm deriving the reference channel, I am getting negative ms at the arrival of the IR. I am using the ASIO drivers and have set up the preferences for the reference and the measurement channel.

 

[Chips Davis]

Oh, if anyone would like a link to my ASA paper I could post the link.
Chips

 

[John Mulcahy]

REW offers two timing references. "Loopback" uses an input on the soundcard as the reference, typically driven by another output on the card but could also be driven from an output on a piece of equipment further down the chain to the speaker. "Acoustic" uses another speaker to produce a timing signal, the reference is then when that signal arrives at the microphone (mainly for use with USB mics where a loopback connection is not possible). In both cases when using the V5.20 beta versions a timing offset can be specified. Often it would correspond to the flight time from speaker to mic, though it can be set to include any delays between the reference and signal arrival at the mic using the Estimate IR Delay button in the Impulse graph controls. The reference plus the offset define where t=0 appears on the impulse response graph. There are also impulse graph controls to manually shift the t=0 point by a specific time, distance or number of samples or place t=0 at the current cursor position.

Relative times can be measured between any points on the graph by pressing the Ctrl key and clicking and dragging the right mouse button, as used in the example image below. Using the initial peak as the reference point allows the additional time (and hence distance travelled) of any particular reflection relative to the direct arrival to be shown. Does any of that provide the pointers to extract the info you need from the data?