Hey folks, please help me with getting Inguz/DRC/ setup on my SBT!!

[Jeff Flowerday]

Yup it works! And sounds horrible!

I must have not done such a great job on the correction filter. I used the Inguz method with DRC. Is audiolense worth the cheese?

I'm intrigued but also skeptical about something messing with the audio before. It looks like audiolense is the proper way to do it, but it isn't cheap.

There isn't much else I can do with my room, so the last option would be DRC, but I need to hear some rave reviews.

[Phil Leigh]

Yup it works! And sounds horrible!

I must have not done such a great job on the correction filter. I used the Inguz method with DRC. Is audiolense worth the cheese?

What microphone did you use? You need to use a GOOD flat mic, preferably one that is calibrated.
Can you describe "horrible"?
Is the sound distorted? - can you check in the log.txt file to see if it mentions "clipping" in the log messages.

[Phil Leigh]

Yup it works! And sounds horrible!

I must have not done such a great job on the correction filter. I used the Inguz method with DRC. Is audiolense worth the cheese?

Yes Audiolense is worth every penny. You also need a very good, flat microphone and a good sound card or USB audio interface. I've never used DRC so can't help with that.

One thing to bear in mind is that most systems require the overall level to be reduced quite significantly in order to avoid clipping on corrected peaks. This is all handled by inguzdsp.exe and its config file... But you will notice that you will need to increase the level of your amp/preamp to compensate.
Also, note that most people do not like a " totally flat" response, favouring instead a gradual roll off in the upper frequencies.

Regardless of the above, the limiting factor is the quality and accuracy of the convolution wav file fed into Inguzdsp.exe. This hinges on the quality of mic and sound card and on the software used to make the measurements and generate the correction file. There is no free lunch here unfortunately. You need GOOD mics, a GOOD sound card and GOOD measurement software. Everything else is free!

[evdplancke]

What microphone did you use? You need to use a GOOD flat mic, preferably one that is calibrated.
Can you describe "horrible"?
Is the sound distorted? - can you check in the log.txt file to see if it mentions "clipping" in the log messages.

Same experience so I can answer for myself. Was using Behringer ECM8000 as recommended by Inguz site best price/quality, connected to Behringer mixing table for Phantom power and level adjustements. Using Olympus LS11 on line in as digital recorder in 44,1/24 and converted to 44,1/16 by Audacity. Processed with DRC.

Result: no noticeable improvements/deterioration on orchestral music but horrible cavernal sound on voices with a lot of metallical echo and back layed voices.. No clipping in log.txt.

What's wrong? Will maybe try 96/24 sweep recording to see if there is a difference.

Don't want to pay for Audiolense without garantied results. Is it possible to import in Audiolense sweep files recorded by external recorders in order to avoid using a bad PC sound card?

[adamdea]

I'm intrigued but also skeptical about something messing with the audio before. It looks like audiolense is the proper way to do it, but it isn't cheap.

There isn't much else I can do with my room, so the last option would be DRC, but I need to hear some rave reviews.

There is already plenty of material available on the benefits of room correction.

If you want to hear rave reviews I would stick to foo: you will always find plenty of rave reviews for that.

I found that audiolense made a very distinct improvement on my system. For my own part i got quite a bit of benefit just using the inguz equaliser and an SPL meter from maplins. But using a measring mic and audiolense was better.

But it's obvious that the benefits of any implementation depend on 1) what problems you have to start with 2) how well you use the room correction.

I am suspicious of anyone who simply says "I've tried it and it doesn't work". You have to analyse out the problems and decide which ones are worth correcting and which aren't: I have a null at between 30-40hz in my room- not amount of boosting will help it. If you tell the software to try it will probably mess everything up. But there are peaks and smaller dips in response which can be addressed. If you try one approach and don;t like the sound, you have to rethink the strategy- maybe you are being too ambitious. AFAIK most users do not aim for a target curve with an entirely flat frequency response but just aim to get it nearer

There is a cognitive dissonance for audiophiles because it involves systematically altering the sound; audiophiles gernally seem to prefer the illusion that they are purely reproducing something even when they are (unsystematically) tinkering with frequency and/or phase eg with cables, valves, Nos dacs.


IMHO anyone who says it doesn't work simpliciter either a) has a perfect room b) doesn't know how to work the software or c) is wrong

I have only tried frequency domain correction and have not moved onto time domain because that is a lot more complex and there are more tradeoffs. But I will probably give that a go in the future.

[evdplancke]

Another dummy question: there are several places where I can adjust recording level in my setup: recorder rec gain, mixer main level, mixer Mic input gain, mixer Mic line level. Where is the best place to ajust recording level in order to avoid clipping but still minimise noise?

[naamanf]

Same experience so I can answer for myself. Was using Behringer ECM8000 as recommended by Inguz site best price/quality, connected to Behringer mixing table for Phantom power and level adjustements. Using Olympus LS11 on line in as digital recorder in 44,1/24 and converted to 44,1/16 by Audacity. Processed with DRC.

Result: no noticeable improvements/deterioration on orchestral music but horrible cavernal sound on voices with a lot of metallical echo and back layed voices.. No clipping in log.txt.

What's wrong? Will maybe try 96/24 sweep recording to see if there is a difference.

Don't want to pay for Audiolense without garantied results. Is it possible to import in Audiolense sweep files recorded by external recorders in order to avoid using a bad PC sound card?

Same setup but I was using a Profire 610. I also had the same results. I do have clipping in my log file so I am sure that's one of the problems.

[Phil Leigh]

Same setup but I was using a Profire 610. I also had the same results. I do have clipping in my log file so I am sure that's one of the problems.

To reduce the number of clipping messages in log.txt, change the gain setting in inguzdsp.exe.config:

<?xml version="1.0" encoding="utf-8" ?>
<configuration>
<appSettings>
<add key="dither" value="1" />
<add key="partitions" value="2" />
<add key="gain" value="-16" />
</appSettings>
</configuration>

Mine is set to -16... as you can see above. This eliminates 95%+ of clipping in my system with my correction file. I'm not bothered about the odd clipped sample here and there, they are inaudible.

However, if you are hearing "cavern" noises that implies something has gone (very) wrong with the measurement process. This is where Audiolense starts to pay for itself because:

1) it will tell you/show you if the measurement has worked properly (mic setup, gain, recording levels etc) and exactly what the results are in terms of measured frequence response

2) it will let you draw a target curve for the correction - this curve can be anything you want it to be. You can also (if you want to) restrict the range of frequencies that are subject to correction. Some people (not me) prefer to only correct the lower frequencies - say below 500Hz or even lower.


The sort of room correction we are discussing here is categorically not "plug and play". It takes time, practice and attention to detail plus first-rate equipment and software to get good results.

[Phil Leigh]

The behringer mic is "OK". It's HF response is not very flat. Its calibration files have also proven somewhat unreliable. However, it should give "reasonable" results.

It is VERY IMPORTANT that all measurement takes place in a quiet room, free from extraneous noise (e.g. traffic, clocks ticking etc) and should use the average of many sweeps (another benefit of the Audiolense software). You need to use a good, vibration-proof mic stand to hold the mic properly.

[naamanf]

The behringer mic is "OK". It's HF response is not very flat. Its calibration files have also proven somewhat unreliable. However, it should give "reasonable" results.

It is VERY IMPORTANT that all measurement takes place in a quiet room, free from extraneous noise (e.g. traffic, clocks ticking etc) and should use the average of many sweeps (another benefit of the Audiolense software). You need to use a good, vibration-proof mic stand to hold the mic properly.

Any suggestions on a mic then that doesn't break the bank. I just happened to have everything already on hand and don't want to invest a large amount if the results are not worth the squeeze.