TOPIC:

dreite wrote:

Oh for crying out loud.  Are you in some sort of anal mode today?
If we're really going to get technical about it even "ms" is not correct.  The proper abbreviation would actually be "msec".
 

No, it's ms not msec (going by the SI system). I only commented as you bizarrely chose to repeat the obvious typo.

Are you saying there is a way to use REW to measure latency using USB loopback though? Please explain if you know how to  .

@ultrasonic 

yeah in REW preferences, I selected the flex as the soundcard and as the “mic”  was easy….

was also able to measure how flat the flex is and wow, it’s literally a flat line, no wiggles at all….

i was just messing around with it…… 

ive been using a creative sound card that’s flat with a soundcard cal file for measurements……. 
I might use a flex now….. and it’s 96k…. Gives me a reason to go for the umik2 I suppose ;-$ 

denver8me4dinner wrote:

@ultrasonic 

yeah in REW preferences, I selected the flex as the soundcard and as the “mic”  was easy….

Yes, I'm familiar with this option and I've used it on my miniDSP SHD to measure what the applied Dirac Live filters look like. My question above regarding timing reference referred to me doing exactly this. Have another read of what I wrote, as I don't think your test is telling you what you think it is. I think a different approach is needed to measure latency. 

@ultrasonic 

aah yes , I agree, however, just as a rough means , I tried it and the ir was a fraction of a ms (or msec) , it would be impossible for the ir to be less then a ms if the latency was more then a ms. Ya follow me…. Unless an fir was loaded. (And I measured all filters off, just a flat 20-20 sweep) 

Because the IR peak is HF biased…. Idk , but I know , your right.  



on a side note I’ve also analyzed Dirac filters that way… interesting stuff…. What did you gather…. Almost always a partial inverted response? Seems it inverted everything except a few deep notches…. But I also noticed some thing left alone that I couldn’t tell as to why… I sorta gave up , I couldn’t make sense what it was trying to do with the phase… in comparison to some dips being left alone…. It almost seemed as if it left the dips alone for sonic attributions. i.e proper spacial imaging in the soundstage….. I could never quite understand that…. 

denver8me4dinner wrote:

@ultrasonic 

aah yes , I agree, however, just as a rough means , I tried it and the ir was a fraction of a ms (or msec) , it would be impossible for the ir to be less then a ms if the latency was more then a ms. Ya follow me…. Unless an fir was loaded. (And I measured all filters off, just a flat 20-20 sweep) 
 

I still think you're missing the point here and not seeing the true latency. If you're not going to answer my question about timing reference then try setting say a 10 ms delay on both channels and repeat your test. You should see a 10 ms latency but my bet is you won't...

Ultrasonic post=59471

denver8me4dinner wrote:

@ultrasonic 

aah yes , I agree, however, just as a rough means , I tried it and the ir was a fraction of a ms (or msec) , it would be impossible for the ir to be less then a ms if the latency was more then a ms. Ya follow me…. Unless an fir was loaded. (And I measured all filters off, just a flat 20-20 sweep) 

 

I still think you're missing the point here and not seeing the true latency. If you're not going to answer my question about timing reference then try setting say a 10 ms delay on both channels and repeat your test. You should see a 10 ms latency but my bet is you won't...

I don't think he understands this at all.    Anyone measuring less than 3ms latency is not measuring it correctly.

Regards your earlier query........I don't use REW for this type of thing.  I'm old school and use my oscilloscopic device on my test bench.  Very straightforward and much easier for end-to-end characterizations like this.

Dave.

Dave and ultrasonic, 

so ….. I think I actually do understand this correctly, and I don’t care for the condescending chatter…. 

so let’s ask why?  10ms are you kidding!!! 

let’s be clear we’re talking the same device. 
a NON Dirac flex? With no firs? 

your high on crack if it has 10ms latency. 


and your 3ms latency, really , idk that actual latency as I just pushed an ir through it ….. seems like a ton of you ask me….. but what do I know….. nothing apparently.  Maybe 3ms with LF filters and eq and boost and all kinds of time variant filters…..  all crossovers and peq on bypass….. how about .3ms ….. just saying 

i also have an old skool scope, an old Agilent from 2004. Maybe I’m just gonna have to bust that out and see. 


all of which sounds pretty fucking boring if you ask me, whatever happened to just measuring your impulse? that pretty much counts for any latency the processor acoustical even on mic latency.  If your syncing with video, same argument, you’re going to end up using the FIR or DIrec. So who really honestly cares about the latency of the device….. or somebody trying to nitpick this device and complain about minidsp again…. Seems to me from someone that uses about 20 different types of DSP on a regular basis, I have never not even one time found a need to know the latency of a circuit board… boring!!!!!! 

He suggested 10ms as an arbitrary point of reference for your testing procedure.  Obviously.
You're high on crack if you can't read/understand the previous postings.  I already told you what the latency is.  Re-read.
It's also obvious at this point you're just trolling and don't have a clue what you're talking about.

Out.

Dave.

Dave , 

don’t be jealous…. 
Be nice. 

i did re read several times, and set it had 10 ms! 

im “out” also…. You guys are boring trying to measure processor latency. 


how do you get this bothers me when people post stuff up to try and say negative things about an Mini DSP…, when minis are the best….  Complaining about latency…. Really 

So I was wrong , 

i used the analog in and analog out on flex as a pass through to a loop back on ch1 and ch2 timing reference (as a loose term just a comparison) as direct loop back , got 3.6ms….. 

assuming my soundcard added .5ms the 3.1ms is probably correct 

I was wrong…… sorry. 

You know, when I post stuff like this it's because I've actually tested it.

Dave.

What’s the latency using SPDF digital in => DSP incl EQ & xover => analog out?

The 2x4HD????  It's a tad higher than 3 milliseconds.  (It's been mentioned numerous times in this thread.) 

Dave.

Nope, Flex. And it’s not clear at all from above. I’m interested neither in ADC latency since SPDIF input nor as with all processing bypassed but with full use of EQ (all bands on all channels) and crossover (LR4 on all channels). 

This for studio use (tracking) where even 5ms is getting annoying and unusable.  

NiToNi wrote:

Nope, Flex. And it’s not clear at all from above. I’m interested neither in ADC latency since SPDIF input nor as with all processing bypassed but with full use of EQ (all bands on all channels) and crossover (LR4 on all channels). 

This for studio use (tracking) where even 5ms is getting annoying and unusable.  

isolate from musicians, and for mastering synch your daw  gear…… 

I mean 5ms is pretty standard in dsp if I remember right….


maybe I’m not thinking properly