I have tried tape measurement and ATM from Helix. My new 3-way system sounded pretty decent but I knew something was off. Today I did the time alignment with HolmImpulse like you did. HOLY SHIT, I did not know that it makes such a difference. I listened to Metallica and got goosebumps! Thank you man. The Center, Left Center etc. songs from you worked spot on at the first try.
This is gold! Did it in my car last night, thought my time alignment was good before, but after this the speakers just seems to dissappear lol that's the best way I can describe it.
Hey, is there anyway you could help me with this? I'm having some issues getting consistent measurements after locking in time. It seems to jump around a couple milliseconds everytime I make a measurement even if I make 0 changes. Impulse always looks the same, it just shifts in time
Hello sir, Just came across your channel and have subscribed. Greatly appreciate your videos on this subject. Would love to see a video tutorial of where the rca's from the Focus rite goes in the chain if using dsp to amplifiers. Thanks again. Staying tuned.
I guess you could use "acoustic timing reference" in REW like someone pointed out. But I see that you will not know where you are in time just that if you measure two speakers you can see the time difference between them if you use the closest one as "acoustic timing reference". At least is how I think it works. 30 years ago I built a kit that could use a PC (386 in those days was what I could afford) that sent out a DIRAC pulse and here you also had a mic and an electrical loopback (IMP software and hardware). I chucked this system a few years ago since it was outdated compared to more modern things. But I guess it had some pros compared to REW. HOLMImpulse seems like a new version of this. I can agree what others say: "HolmImpulse for time alignment, REW for phase.".
The acoustic timing reference is just a short chirp before the sweep. You need a tweeter for that chirp as it's located around 4kHz. I would always recommend a two channel interface though as it's more reliable and not that much more money than a UMIK-1. Be prepared to have your XLR microphone calibrated though as the cheaper ones usually don't come with calibration files, let alone a 90° calibration. That said... Nerijus is using a real loopback and could've used REW for simplicity. With impulse response measurements you can measure phase and the impulse response just like HOLMImpulse is doing.
Many thanks for the helpful sharing and I am planning to try myself for this. Want to ask, in this method, we only use just one of the output ('L' or 'R') from the audio interface instead of connecting it to both L & R?
@@RAW-CAt thanks for replying. What happen if we connect one output to DSP and the other as loop back separately? Sorry ... Am trying to understand... (am not professional into this).
REW has the capability to take an impulse response and calculate the diff vs a reference but I never managed to make it “sound” good that way. Did you try that method instead? Do you think this one sounds better? I appreciate you sharing your experience, thanks so much!
Not that its readily audible but did you check absolute polarity of the impulse in the loopback? Often I have seen it starting negative. Due to reversal in the audio interface hardware and/or software combination. Usually there is an inverse button in the software to remedy that systematic error if it occurs in a measurement loop. Fixing it at the source we make sure whatever polarity we see downstream its due to the rest of gear/wiring.
In "device amd signal" tab you can specify the input you are going to use for the mic and I believe, the loopback will be automaticaly taken from the other channel.
Would something like a Behringer U-CONTROL UCA202 work? This is one of the things I have yet to integrate into my tuning procedure. I have the Dayton EMM-6 as well, so I know I would need to get an XLR-to-RCA cable.
@@RAW-CAt I currently use a 48V phantom power supply now. I assumed that I would probably still need it with the Behringer. Thank you for the info and thank you for taking the time to show us your experimentation and results in car audio.
regarding Dayton emm6; on their product specifications page, they mentioned that it is ideal for audio measurements. Also, they are coming with a calibration file. Could you please mention a few xlr mics which I can used for tuning and phase accurate measurements ?
One question, I tried to use the Umik-1 and as a loopback take one of the DSP outputs connected to the microphone input of my laptop (3.5mm), using ASIO. The problem is that over time the delay between the microphone measurement and the reference increases. Does anyone have an idea why that happens?
Good work.👌👌👌 A dump question: I read somewhere that if we use minidsp umic instead of Behringer or Dayton, the audio interface(the soundcard) is not required. Is it true..?
You can use an acoustic timing reference, but it's a bit more work to get reliable data. And you definitely have to remeasure more often due to the clock drift that occurs with USB-microphones. So if possible spend some money on a real interface and XLR-microphone. As Nerijus has shown it's not that expensive. It only gets expensive if you need more inputs/outputs (for measuring multiple impulse responses at once and do automatic averaging) and/or better calibrated microphones.
hi i want to ask i buy at ecm8000 microphone and cheap sound card i getting wrong delay time even i try adjust delay on dsp still no change on holimpulse software iam i doing wrong??sound got mono and loop back button i dun know what the fuction is for.. and use coxial speaker i set to full range hi to low pass it is correct?
@@RAW-CAt thanks sir i want ask how to get the 1st start raise peak at 0ms? do i have to set on manual on offset at at hol impulse setting? and is my speaker is coxial i set hi low pass on dsp is 100 to 20k crossover point and in hol impulse i am also have to set it at that numbers right?
I've been trying for 2 days but I couldn't succeed. I use m-audio mobile preusb sound card and ecm8000 microphone, I wonder if my sound card is not suitable? 2 channel sound card, 1st channel input is connected to the microphone, will I connect the 2nd channel output to the DSP and 2nd channel input? Or the 2nd channel output is dsp and the 1st channel's input (will I connect it parallel to the microphone) I would appreciate if you could help me, thank you
Your videos are literally invaluable for newbs like myself! QQ, if one doesn't have a sub, just an active 3 way system (T/MR/MB in each door) + center and rear MR, what speaker do you recommend starting with first to fix zero?
@@RAW-CAt Just noticed it looks like you're using a signal from the HolmImpulse software. Are you therefore connecting a USB from the laptop into the car itself and using the laptop to send the car the signal?
Actually, I'm still struggling. My Match Up 10 DSP/Amp combo has only a single Line out and an optical in, does not have RCA's. Is there no way to do this with this particular amp/dsp combo?
Just to piggyback on this question. If I'm using 2 set of rcs's from my headunit, 1 pair for LR and another for sub, into the dsp. Do I need to switch the rcas when measuring the sub since they come in on 5 and 6 on the Helix?I
I am confused. I have a Helix DSP and I use an optical single as my source, would this still work? Meaning do I just use the low level RCA's on the DSP and swap it around every measurement? or is there a better way?
@@RAW-CAt Thanks for the response, do you have one low level RCA routed to all the outputs so you don't have to physically swap it around when doing this?
@RAW-Cat tried this method today with decent results. Thing is the ir of subwoofer is way flatter than the midbasses. In this video the impulses are more or less the same. Length and size wise. Do you might have an idea of what I am doing wrong. Same for the phase. Octave below and above while measuring but the graph ends at the xover points.
I would imagine that your reference driver for time 0 is not selected correctly. You need to find the furthest driver in your system. Is it not the subwoofer?
@@RAW-CAt it’s close between right midbass and sub but I think sub is furthest (left drive car) time zero is set on sub. I rewatched the video and saw you using a dB offset for the sub so I think I missed that the first time. Still doesn’t explain the phase stopping at the xovers. Even not having it 100% it outshines the atm by a lot!
I'm bit confused. Why do you have stereo RCA connected to mono output? Can you not use one output as loopback and the second as output to DSP? Thank you
Due to clock differences between outputs of the audio interface, best is to use the same signal for both, DSP and loopback. I used stereo RCAs just hecause I had that cable and the routing in the DSP was done with a pair of inputs. Only one RCA is enough.
Is there any particular reason why you use HOLMImpulse for time alignment but REW for everything else? Can you not use REW for time alignment too? Is HOLMImpulse better?
There is absolutely no reason. Homl was the first software that I tried, learned and liked it. It gave me very good first impressions. Same like Apple vs Android, Android was the first I tried and stuck with it🤷🏻♂️
@@RAW-CAt if I were to try and purchase this DIY interconnect, what would I call it in my search? Struggling to find one online. RCA with y-splitter 6.3mm is the best I've come up with.
I have a question. Every video I’ve watched on timing using IR, O is the middle of the first positive/negative peak and not the point at that it rises. I also googled HOLMimpusle users guide and that’s the way it’s done in the guide, is there a reason you do it differently ?
The point of the impulse you align to doesn't matter as long as you are measuring the same type of drivers. It is extremely difficult to align a subwoofer with much smaller midbass. Probably those guides that you were looking at are trying to align home stereo towers with a sub. If you align midbass and subwoofer impulses based on the first peak, then you will be a few cycles off.
3:07 it seems that you forgot the delay in the amp. Do not know how much that is though. Maybe need loopback from the speaker terminals. But for this I guess it does not matter since it will be the same for all measurements.
If you're doing phase alignment you definitely want to get rid of the system delay to getter better phase readings. Just like you would do with Smaart or software similar to Smaart. You can do that with REW, too. Just press the button "Estimate IR delay" in the Controls of the view "SPL & Phase" in REW. Redo the first measurement then again and all following measurements will have the system delay removed from the measurements, so that only the difference between drivers are shown.
I noticed on your Focusrite, you split the cable coming from one of the outputs…any reason you didn’t just use a separate 1/4” cable, to come out of the other output, to go back to the input?
@@RAW-CAtjust did it today, and seems to have worked GREAT! Only thing I found odd…was when I went to do my Tweeter signals (same frequency range you used), I had to lower the “Number of Taps” from 501 to 150. If I didn’t, the sound would happen SO fast, and HOLMImpulse would error out.
Me again.....by chance can you make a written tutorial in how to set up the holm impulse with the microphone and soundcard, also step by step guide in how to set the whole time alignment as well please.
Because if you set up the TA based on the first peak, the subwoofer and tweeters will not be in time. The beginning of the impulse or the rise of the graph is when the sound starts. The peak is at different times for different frequencies as the wavelengths are very different.
@@RAW-CAt i m tired with measurements too much. I Use pink noise to put bass into Middle.. and its ok. bass on the front... and also align sub to middle speakers
@@RAW-CAtthank you I really missed first few minutes off audio. Cuz I was at work.. I have we watched this video again and I get the concept now. The low back is for impulse ? Inman video that you used umik with rew you mentioned about the forecast rite. Is it referred to this video or you have focusrite with rew video also. I am trying to find that video but to be honest Searching for that video. Thanks
Can you make a tutorial on how to set the phase alignment using the holm impulse as well please? I tried your technique with my widebands and midbass and i'm in love....not even detecting the speakers anymore....its all psychoacoustic now
I am using my UMIK-1 and struggling with REW and the "The Acoustic Time Reference" at the moment. I think I have managed to get the TA quite good but I dont really understand how to get the phase right with help of REW. You seems to have good knowlege about how get everything right so my question for you is if you maybe can make a video about phase aligment in REW or if you have another easy way to get the phase correct? Maybe someone else have any ideas how to do it in another way? I am not fancy of buying more measuring equipment. I also have an ECM measurment microphone and an IRIG Pre that maybe can be used togehter with an extern soundcard. I was just thinkning but Is it not possible just to use one RCA/output from the interface to the DSP and the other output as a loop back or even one output to an AUX input of your HU instead? Do you really need an left and right channel for measuring TA? You can route all the channels in the dsp to just that specific input channel.
With the UMIK-1 you need to make the two measurement sweeps of two drivers in very close time proximity to minimise the clock drift from the two clock sources (soundcard-output of computer and soundcard input of UMIK-1). To achieve this route right channel to the reference driver (f.e. the right tweeter) and left channel to all other drivers. In REW when measuring then set "Output" to "L" and "Ref output" to "R". If measuring the reference channel, f.e. when comparing right to left tweeter, set both output and ref output to R. Then go to view "SPL & Phase" and press "Controls" and then "Estimate IR delay" to get rid of the initial delay. Then remeasure the reference channel again, then reset "Output" to "L" and measure the second tweeter immediately afterwards. For midranges, midbass and sub you just left the output on "L", but for every measurement of pairs reset the "Timing offset" to zero, measure first driver, estimate IR delay, remeasure the same driver and then measure the 2nd driver. You see, it's a bit more complicated with USB-microphones, but it's doable within some margin of error due to clock drift. When comparing the final two measurements of driver pairs you can see the needed time delay in the comment section of the measurement, f.e. you have a comment after measuring like: "Delay 0,0007 ms (0,23 mm, 0,01 in)" for the first driver and "Delay -0,0046 ms (-1,6 mm, -0,06 in)" for the second driver. In this case the 2nd driver is arriving early compared to the first driver and needs to be delayed. In this example this might not be possible as the delay would be around 0.005ms, which most DSPs cannot do due to their limited resolution (f.e. Mosconi PICOs only do 0.00 resolution, so lower delay than 0.01 cannot be done). When having aligned driver pairs you can continue for tweet to mids by measuring driver pairs at once and then aligning the pairs to each other. For aligning midrange+tweeters to midbass or front to subs i would recommend using the alignment tool. Measurements will be done as described above, just play all midrange and tweeters for first measurement and midbass drivers alone for second measurement. Then go to view "All SPL" and click controls, then click "Alignment tool" and check the two new measurements in the dialog. Mark the acoustic crossover and click "Level phase at cursor". Then press "Align phase at cursor". You can also virtually flip phase on one of the measurements to see if you get a better phase alignment in the lower half of the background or less delay for one of the measurements. F.e. without flipping the phase of the sub you might get a result of +11ms delay for the sub and after flipping the phase of the sub you get +0.1ms delay for the sub and better aligment of the phase traces in the background window. In this case use the lower delay and flip the sub. You always want to use the least amount of delay that is needed to get the job done.
@@Cathul I must just thank you for your well explained answer. I followed your guide and are almost done with my 4-way setup. Just the tweeters and the sub that I have to fine adjust a little bit more. A big difference compared to what I had before.