Never used CEntrance, but does it do anything different than what Sonar's External Insert provides?
 
i.e. can't I simply measure RTL by connecting one output to one input and then have the External Insert test and report RTL as number of samples?
 
http://centrance.com/downloads/ltu/

Yes, External Insert should report the same number as CEntrance; it's just a bit more involved to set up (and to describe for someone who hasn't ever used it).

What puzzles me, though, is that External Insert reports different numbers than RTL listed in audio preferences. 
 
I guess External Insert latency is truly measured with a pulse and RTL listed in prefs is reported by the ASIO driver???
 
While I could image that these RTL numbers in audio prefs are not absolutely accurate, they seem to be OK for the Roland interfaces (which do spot on recording latency correction based on the same numbers as listed RTL), yet External Insert reports something completely different.
 
Likewise for a 2nd interface I have tested and compared: RTL seem to be off by a constant offset (which is fine as I adjust Sonar settings to compensate), but External Insert is reporting a totally different latency / offset.
 
 

External insert sends a pulse and measures the latency when it receives it. Centrance is something different since it doesn't know Sonars buffer size etc

Rob[atSound-Rehab]
Never used CEntrance, but does it do anything different than what Sonar's External Insert provides?
 
i.e. can't I simply measure RTL by connecting one output to one input and then have the External Insert test and report RTL as number of samples?
 
http://centrance.com/downloads/ltu/

that url gives me a 404, even after i have to accept a misconfigured ssl certificate...

pwalpwal

Rob[atSound-Rehab]
Never used CEntrance, but does it do anything different than what Sonar's External Insert provides?
 
i.e. can't I simply measure RTL by connecting one output to one input and then have the External Insert test and report RTL as number of samples?
 
http://centrance.com/downloads/ltu/

that url gives me a 404, even after i have to accept a misconfigured ssl certificate...

works fine here ... but just google "centrance latency test utility"; it should get you there ...

Noel Borthwick [Cakewalk]
External insert sends a pulse and measures the latency when it receives it. Centrance is something different since it doesn't know Sonars buffer size etc

They both measure by pinging, but it seems External Insert doesn't include both input and output buffers in the total reported latency. For example, with an ASIO buffer of 32 samples (each way), External Insert reports 32 samples less than CEntrance. I thought I recalled it used to report the total measured RTL - same as CEntrance - but maybe not.

Many interfaces have a built-in "safety buffer" or other processing time that isn't reported by their ASIO driver. Very few manufacturers give any indication of how long it takes their product to process audio or whether the driver's reported latency takes account of that.

tlw
Many interfaces have a built-in "safety buffer" or other processing time that isn't reported by their ASIO driver. Very few manufacturers give any indication of how long it takes their product to process audio or whether the driver's reported latency takes account of that.

I did test this with a MOTU AVB as well which gives you full control over the host buffer and the safety buffer. Yet I still don't understand the numbers reported by the External Insert ...
 
 
Host Buffer:   32    Safety Offset: 32   RTL:  128 Samples  1.33 ms -- External Insert:  179 Samples -- External Insert - RTL: 51
Host Buffer:   64    Safety Offset: 16   RTL:  160 Samples  1.67 ms -- External Insert:  183 Samples -- External Insert - RTL: 23
Host Buffer:   64    Safety Offset: 32   RTL:  192 Samples  2.00 ms -- External Insert:  211 Samples -- External Insert - RTL: 19
Host Buffer:   64    Safety Offset: 64   RTL:  256 Samples  2.67 ms -- External Insert:  275 Samples -- External Insert - RTL: 19

Host Buffer:  128    Safety Offset: 16   RTL:  288 Samples  3.00 ms -- External Insert:  243 Samples -- External Insert - RTL: -45
Host Buffer:  128    Safety Offset: 24   RTL:  304 Samples  3.17 ms -- External Insert:  259 Samples -- External Insert - RTL: -45
Host Buffer:  128    Safety Offset: 32   RTL:  320 Samples  3.33 ms -- External Insert:  275 Samples -- External Insert - RTL: -45
Host Buffer:  128    Safety Offset: 48   RTL:  352 Samples  3.67 ms -- External Insert:  307 Samples -- External Insert - RTL: -45
Host Buffer:  128    Safety Offset: 64   RTL:  384 Samples  4.00 ms -- External Insert:  339 Samples -- External Insert - RTL: -45
Host Buffer:  128    Safety Offset: 128  RTL:  512 Samples  5.33 ms -- External Insert:  467 Samples -- External Insert - RTL: -45

Host Buffer:  256    Safety Offset: 16   RTL:  544 Samples  5.67 ms -- External Insert:  371 Samples -- External Insert - RTL: -173
Host Buffer:  256    Safety Offset: 32   RTL:  576 Samples  6.00 ms -- External Insert:  409 Samples -- External Insert - RTL: -167
Host Buffer:  256    Safety Offset: 64   RTL:  640 Samples  6.67 ms -- External Insert:  471 Samples -- External Insert - RTL: -169
Host Buffer:  256    Safety Offset: 128  RTL:  768 Samples  8.00 ms -- External Insert:  596 Samples -- External Insert - RTL: -172
Host Buffer:  256    Safety Offset: 256  RTL: 1024 Samples 10.67 ms -- External Insert:  851 Samples -- External Insert - RTL: -173

Host Buffer:  512    Safety Offset: 32   RTL: 1088 Samples 11.33 ms -- External Insert:  655 Samples -- External Insert - RTL: -433
Host Buffer:  512    Safety Offset: 64   RTL: 1152 Samples 12.00 ms -- External Insert:  723 Samples -- External Insert - RTL: -429
Host Buffer:  512    Safety Offset: 128  RTL: 1280 Samples 13.33 ms -- External Insert:  852 Samples -- External Insert - RTL: -428
Host Buffer:  512    Safety Offset: 256  RTL: 1536 Samples 16.00 ms -- External Insert: 1107 Samples -- External Insert - RTL: -429
Host Buffer:  512    Safety Offset: 512  RTL: 2048 Samples 21.33 ms -- External Insert: 1619 Samples -- External Insert - RTL: -429

Host Buffer: 1024    Safety Offset: 32   RTL: 2112 Samples 22.00 ms -- External Insert: 1171 Samples -- External Insert - RTL: -941
Host Buffer: 1024    Safety Offset: 64   RTL: 2176 Samples 22.67 ms -- External Insert: 1235 Samples -- External Insert - RTL: -941
Host Buffer: 1024    Safety Offset: 128  RTL: 2304 Samples 24.00 ms -- External Insert: 1364 Samples -- External Insert - RTL: -940

I'm not following your table very well, and it's not clear what the concern is. If what you're seeing differs from the following, please explain:
 
'Hardware' latency (interface bus plus firmware/driver plus A/D/A conversion) = CEntrance measured - 2x ASIO buffer
 
This hardware latency is fixed, and does not vary with buffer size. I would lump the 'safety' buffer in with the firmware part of this 'hardware' latency.
 
SONAR's 'reported' RTL in preferences will typically be smaller than the CEntrance measured value at any given buffer size by anywhere from 1 to 50 samples because the driver under-reports it; this 'hidden' or 'unreported' latency will need to be entered as the 'Manual Offset' in Sync and Caching to get perfect record latency compensation with automatic compensation enabled.
 
External Insert on an analog loopback path with no delay introduced by external hardware = CEntrance measured - 1x ASIO buffer.
 
I don't know why EI under-reports the total 'delay', but its compensation will be correct which is all that really matters.