What is the time between snare hits?

[Dabbler]

In a song that is 140 bpm 4/4 signature.

How is the time between snare hits calculated?

GIven, normal snare hits on 2 and 4

If I take 140/60 I get 2.33 bps.

Divide that by 4 for one bar and I get 0.583 seconds per beat?

I'm trying to use math to set the A&R on a compressor but my brain ain't working.

[orthodox]

In a song that is 140 bpm 4/4 signature.

How is the time between snare hits calculated?

GIven, normal snare hits on 2 and 4

If I take 140/60 I get 2.33 bps.

Divide that by 4 for one bar and I get 0.583 seconds per beat?

I'm trying to use math to set the A&R on a compressor but my brain ain't working.

You need seconds per beat, that is 60 sec / 140 bpm = 0.428 sec/beat.
2 beats will be: 0.428 sec/beat * 2 beats = 0.857 sec

[Dabbler]

Ah of course.

So when the first snare hit occurs, the next one is 0.857 seconds behind it.

857 milliseconds.

I think I can work with this.

Thanks orthodox.

[Creativemind]

2 x 0.428 = 0.856, not 0.857?

[Jagwah]

Here dude:

[Raveshaper]

Here dude:

"Aaaaggghh! Make it stop!"

Great series.

[Dabbler]

Here dude:

Woof!

[Last Alternative]

Damn! That video tripped me out. Creepy stuff lol.

[Jagwah]

Here dude:

"Aaaaggghh! Make it stop!"

Great series.

Absolutely! The songs get stuck in my head too D:

[ravisoni]

In a song that is 140 bpm 4/4 signature.

How is the time between snare hits calculated?

GIven, normal snare hits on 2 and 4

If I take 140/60 I get 2.33 bps.

Divide that by 4 for one bar and I get 0.583 seconds per beat?

I'm trying to use math to set the A&R on a compressor but my brain ain't working.

Your 2.33 bps was correct, too. Just take the reciprocal (1/2.33) and you get the same answer as others have gotten for seconds per beat.

[selig]

In a song that is 140 bpm 4/4 signature.

How is the time between snare hits calculated?

GIven, normal snare hits on 2 and 4

If I take 140/60 I get 2.33 bps.

Divide that by 4 for one bar and I get 0.583 seconds per beat?

I'm trying to use math to set the A&R on a compressor but my brain ain't working.

It's important to remember that compressors don't work like envelopes - the attack and release time are expressed as a RATE (which is a Ratio), not a TIME.
Rate can only be determined if you know TIME and DISTANCE. For the SSL Channel compressor (as one example) the attack is listed as 3ms per 20 dB (like how your measure speed in MPH or KPH). This means that when you have 10 dB reduction, the attack is 1.5 ms but if on the next "hit" you have 20 dB reduction the attack will take twice as long at 3ms.

What does this mean?
It means that you can't go by the numbers on a compressor like you can on an envelope unless you know the ms per dB RATE of the attack/release. You may get lucky and use an amount of gain reduction that corresponds to the front panel values. But if you don't get the expected results after all these calculations, this is why!

[orthodox]

In a song that is 140 bpm 4/4 signature.

How is the time between snare hits calculated?

GIven, normal snare hits on 2 and 4

If I take 140/60 I get 2.33 bps.

Divide that by 4 for one bar and I get 0.583 seconds per beat?

I'm trying to use math to set the A&R on a compressor but my brain ain't working.

It's important to remember that compressors don't work like envelopes - the attack and release time are expressed as a RATE (which is a Ratio), not a TIME.
Rate can only be determined if you know TIME and DISTANCE. For the SSL Channel compressor (as one example) the attack is listed as 3ms per 20 dB (like how your measure speed in MPH or KPH). This means that when you have 10 dB reduction, the attack is 1.5 ms but if on the next "hit" you have 20 dB reduction the attack will take twice as long at 3ms.

What does this mean?
It means that you can't go by the numbers on a compressor like you can on an envelope unless you know the ms per dB RATE of the attack/release. You may get lucky and use an amount of gain reduction that corresponds to the front panel values. But if you don't get the expected results after all these calculations, this is why!

The attack and release times surely correspond to the rates of the respective processes. But the dBs mentioned in expressions of attack and release speeds like "3ms per 20dB" are not those conventional dBs of gain reduction, they are rather notional. The gain change process in attack or release is not linear in dBs and it does not complete in a finite time.

If the compressor attack speed is 3ms per 20dB, the gain reduction of 20dB is not reached in 3ms. It is rather like this: the first half of the difference between the initial and target gain reduction levels is made in 6/20*3 = 0.9ms, a half of the remaining half - in another 0.9ms and so on. In 3ms the gain reduction will be -14.5dB.

[normen]

The attack and release times surely correspond to the speeds of the respective processes. But the dBs mentioned in expressions of attack and release speeds like "3ms per 20dB" are not those conventional dBs of gain reduction, they are rather notional. The gain change process in attack or release is not linear in dBs and it does not complete in a finite time.

If the compressor attack speed is 3ms per 20dB, the gain reduction of 20dB is not reached in 3ms. It is rather like this: the first half of the difference between the initial and target gain reduction levels is made in 6/20*3 = 0.9ms, a half of the remaining half - in another 0.9ms and so on. In 3ms the gain reduction will be -14.5dB.

Without talking about a specific compressor none of these values are correct. The curve, timing, dependence on program input all vary from compressor to compressor which is why people don't use just one compressor for everything. Even if you test a compressor with short white noise peaks and continuous white noise that doesn't tell you everything about how it will react to different input material. You HAVE TO use your ears when setting up compressors.

[orthodox]

You HAVE TO use your ears when setting up compressors.

Totally agree.

Without talking about a specific compressor none of these values are correct.

I forgot to say, I was talking about MClass, Mix Channel and Master Section compressors, they share the same algorithm.

[normen]

I forgot to say, I was talking about MClass, Mix Channel and Master Section compressors, they share the same algorithm.

No they don't give the same results. You state that as if it was a fact but unless you decompiled Reason you simply don't know that. What you can say is that with your test setup you got the same results maybe.

[selig]

In a song that is 140 bpm 4/4 signature.

How is the time between snare hits calculated?

GIven, normal snare hits on 2 and 4

If I take 140/60 I get 2.33 bps.

Divide that by 4 for one bar and I get 0.583 seconds per beat?

I'm trying to use math to set the A&R on a compressor but my brain ain't working.

It's important to remember that compressors don't work like envelopes - the attack and release time are expressed as a RATE (which is a Ratio), not a TIME.
Rate can only be determined if you know TIME and DISTANCE. For the SSL Channel compressor (as one example) the attack is listed as 3ms per 20 dB (like how your measure speed in MPH or KPH). This means that when you have 10 dB reduction, the attack is 1.5 ms but if on the next "hit" you have 20 dB reduction the attack will take twice as long at 3ms.

What does this mean?
It means that you can't go by the numbers on a compressor like you can on an envelope unless you know the ms per dB RATE of the attack/release. You may get lucky and use an amount of gain reduction that corresponds to the front panel values. But if you don't get the expected results after all these calculations, this is why!

The attack and release times surely correspond to the rates of the respective processes. But the dBs mentioned in expressions of attack and release speeds like "3ms per 20dB" are not those conventional dBs of gain reduction, they are rather notional. The gain change process in attack or release is not linear in dBs and it does not complete in a finite time.

If the compressor attack speed is 3ms per 20dB, the gain reduction of 20dB is not reached in 3ms. It is rather like this: the first half of the difference between the initial and target gain reduction levels is made in 6/20*3 = 0.9ms, a half of the remaining half - in another 0.9ms and so on. In 3ms the gain reduction will be -14.5dB.

You've missed the point of my comment - attack TIME depends on gain reduction amount, nothing more. I'm not sure how you can claim to know the specific response curve of a compressor's time constants, or imply they all work the same…I was speaking in generalities, you are being VERY specific - where do you get your numbers?

[selig]

You HAVE TO use your ears when setting up compressors.

Totally agree.

Without talking about a specific compressor none of these values are correct.

I forgot to say, I was talking about MClass, Mix Channel and Master Section compressors, they share the same algorithm.

Where did you hear this?

[orthodox]

You HAVE TO use your ears when setting up compressors.

Totally agree.

Without talking about a specific compressor none of these values are correct.

I forgot to say, I was talking about MClass, Mix Channel and Master Section compressors, they share the same algorithm.

Where did you hear this?

I did not, I analysed their output and interpolated the gain change curve. I have the formula that fits to their behavior with precision of 10^-6.

[normen]

I did not, I analysed their output and interpolated the gain change curve. I have the formula that fits to their behavior with precision of 10^-6.

Correction: You have the formula that fits to the behavior of the compressors to your input signal.

[orthodox]

I did not, I analysed their output and interpolated the gain change curve. I have the formula that fits to their behavior with precision of 10^-6.

Correction: You have the formula that fits to the behavior of the compressors to your input signal.

Yes, sure. The rest is just a speculation based on common sense. However, I am pretty sure the compressors work the way I expect.
I'm going to write a topic about that in Tutorials, where everyone will be able to disprove my findings.

[selig]

You HAVE TO use your ears when setting up compressors.

Totally agree.

Without talking about a specific compressor none of these values are correct.

I forgot to say, I was talking about MClass, Mix Channel and Master Section compressors, they share the same algorithm.

Where did you hear this?

I did not, I analysed their output and interpolated the gain change curve. I have the formula that fits to their behavior with precision of 10^-6.

While it's possible you are correct and they DO share the same algorithm, it's also possible they do not. So the best you can say is the measured results were similar, as it is possible to get the same results from different algorithms when measuring specific parameters.

Unless you measure every possible parameter, then your results can only prove that the parameters that were measured gave very similar results, right?

I have to ask how you got such precise measurements, as I've only been able to get rough measurements when comparing compression curves (and I'd love to improve my comparison testing!).

[selig]

I did not, I analysed their output and interpolated the gain change curve. I have the formula that fits to their behavior with precision of 10^-6.

Correction: You have the formula that fits to the behavior of the compressors to your input signal.

Yes, sure. The rest is just a speculation based on common sense. However, I am pretty sure the compressors work the way I expect.
I'm going to write a topic about that in Tutorials, where everyone will be able to disprove my findings.

I hope you meant to write "where everyone may dispute my findings", because if everyone disproves your findings than it suggests your findings are not valid, right?

[orthodox]

I have to ask how you got such precise measurements, as I've only been able to get rough measurements when comparing compression curves (and I'd love to improve my comparison testing!).

I prepared series of audio clips (32 bit float audio files at 192kHz) with exact sample values I generated. Then I ran them through the compressors in various modes and exported the output as 24 bit int audio files. The output files were analyzed in Wolfram Mathematica as arrays of values, and every sample was taken into account with its exact value.

I hope you meant to write "where everyone may dispute my findings", because if everyone disproves your findings than it suggests your findings are not valid, right?

Yes, I would not like to suggest one way or the other in advance.

[selig]

I have to ask how you got such precise measurements, as I've only been able to get rough measurements when comparing compression curves (and I'd love to improve my comparison testing!).

I prepared series of audio clips (32 bit float audio files at 192kHz) with exact sample values I generated. Then I ran them through the compressors in various modes and exported the output as 24 bit int audio files. The output files were analyzed in Wolfram Mathematica as arrays of values, and every sample was taken into account with its exact value.

Very interesting - maybe it's time to get Mathematica…
Next question:
How did you get the various compressors to output the same level, or set them to the same settings?

[orthodox]

Next question:
How did you get the various compressors to output the same level, or set them to the same settings?

I analyzed them without any notion of what they mean by their settings. The results were independent for each compressor. The output level was measured.